title:
author:
publisher:
isbn10 | asin:
print isbn13:
ebook isbn13:
language:
subject
publication date:
lcc:
ddc:
subject:
Corrosion Engineering Handbook Corrosion
Technology (New York, N.Y.) ; 11
Schweitzer, Philip A.
CRC Press
0824797094
9780824797096
9780585347233
English
Corrosion and anti-corrosives--Handbooks,
manuals, etc.
1996
TA462.C6545 1996eb
620.1/1223
Corrosion and anti-corrosives--Handbooks,
manuals, etc.
Corrosion Engineering Handbook
CORROSION TECHNOLOGY
Editor
Philip A. Schweitzer, P.E.
Consultant
Fallston, Maryland
1. Corrosion and Corrosion Protection Handbook: Second Edition,
Revised and Expanded, edited by Philip A. Schweitzer
2. Corrosion Resistant Coatings Technology, Ichiro Suzuki
3. Corrosion Resistance of Elastomers, Philip A. Schweitzer
4. Corrosion Resistance Tables: Metals, Nonmetals, Coatings,
Mortars, Plastics, Elastomers and Linings, and Fabrics: Third
Edition, Revised and Expanded (Parts A and B), Philip A. Schweitzer
5. Corrosion-Resistant Piping Systems, Philip A. Schweitzer
6. Corrosion Resistance of Zinc and Zinc Alloys, Frank C. Porter
7. Corrosion of Ceramics, Ronald A. McCauley
8. Corrosion Mechanisms in Theory and Practice, edited by P.
Marcus and J. Oudar
9. Corrosion Resistance of Stainless Steels, C. P. Dillon
10. Corrosion Resistance Tables: Metals, Nonmetals, Coatings,
Mortars, Plastics, Elastomers and Linings, and Fabrics: Fourth
Edition, Revised and Expanded (Parts A, B, and C), Philip A.
Schweitzer
11. Corrosion Engineering Handbook, edited by Philip A. Schweitzer
ADDITIONAL VOLUMES IN PREPARATION
Page i
Corrosion Engineering Handbook
Edited By
Philip A. Schweitzer, P. E.
Consultant
Fallston, Maryland
Page ii
Library of Congress Cataloging-in-Publication Data
Corrosion engineering handbook / edited by Philip A. Schweitzer.
p. cm.(Corrosion technology; 11)
Includes index.
ISBN 0-8247-9709-4 (hardcover: acid-free paper)
1. Corrosion and anticorrosivesHandbooks, manuals, etc. I. Schweitzer,
Philip A. II. Series: Corrosion technology (New York, N.Y.); 11.
TA462.C6545 1996
620.1'1223dc20
96-18884
CIP
The publisher offers discounts on this book when ordered in bulk
quantities. For more information, write to Special Sales/Professional
Marketing at the address below.
This book is printed on acid-free paper.
Copyright © 1996 by Marcel Dekker, Inc. All Rights Reserved.
Neither this book nor any part may be reproduced or transmitted in any
form or by any means, electronic or mechanical, including
photocopying, microfilming, and recording, or by any information
storage and retrieval system, without permission in writing from the
publisher.
Marcel Dekker, Inc.
270 Madison Avenue, New York, New York 10016
Current printing (last digit):
10 9 8 7 6 5 4 3 2 1
PRINTED IN THE UNITED STATES OF AMERICA
Page iii
Preface
Corrosion is both costly and dangerous. Billions of dollars are spent
annually for the replacement of corroded structures, machinery, and
components, including metal roofing, condenser tubes, pipelines, and
many other items. In addition to replacement costs are those
associated with preventive maintenance to prevent corrosion,
inspections, and the upkeep of cathodically protected structures and
pipelines. Indirect costs of corrosion result from shutdown, loss of
efficiency, and product contamination or loss.
While the actual replacement cost of an item may not be high, the loss
of production resulting from the need to shut down an operation to
permit the replacement may amount to hundreds of dollars per hour.
When a tank or pipeline develops a leak, product is lost. If the leak
goes undetected for a period of time, the value of the lost product
could be considerable. In addition, contamination can result from the
leaking material, requiring cleanup, and this can be quite expensive.
When corrosion takes place corrosion products build up, resulting in
reduced flow in pipelines and reduced efficiency of heat transfer in
heat exchangers. Both conditions increase operating costs. Corrosion
products may also be detrimental to the quality of the product being
handled, making it necessary to discard valuable materials.
Premature failure of bridges or structures due to corrosion can also
result
Page iv
in human injury or even loss of life. Failures of operating equipment
resulting from corrosion can have the same disastrous results.
When all of these factors are considered it becomes obvious why the
potential problem of corrosion should be considered during the early
design stages of any project and why it is necessary to constantly
monitor the integrity of structures, bridges, machinery, and equipment
to prevent premature failures.
In order to cope with the potential problems of corrosion it is
necessary to understand
1. Mechanisms of corrosion
2. Corrosion resistant properties of various materials
3. Proper fabrication and installation techniques
4. Methods to prevent or control corrosion
5. Corrosion testing techniques
6. Corrosion monitoring techniques
It is the intention of this book to provide a source of information on
these six items for designers, engineers, and maintenance and
operating personnel. This knowledge is essential when dealing with
the problems of corrosion.
PHILIP A. SCHWEITZER, P.E.
Page v
Contents
Preface
iii
Contributors
ix
1. Fundamentals of Metallic Corrosion
Paul K. Whitcraft
1
2. Atmospheric Corrosion
Philip A. Schweitzer
23
3. The Corrosion of Carbon and Low-Alloy Steels
Gary N. Kirby
35
4. Corrosion of Stainless Steels
Paul K. Whitcraft
53
5. Corrosion of Nickel and High-Nickel Alloys
Philip A. Schweitzer
79
Page vi
6. Corrosion of Copper and Copper Alloys
Philip A. Schweitzer
89
7. Corrosion of Aluminum and Aluminum Alloys
Bernard W. Lifka
99
8. Corrosion of Titanium
Philip A. Schweitzer
157
9. Corrosion of Tantalum
John B. Lambert
165
10. Corrosion of Zirconium
Te-Lin Yau
195
11. Corrosion Resistance of Cast Alloys
James L. Gossett
253
12. Mechanisms of Chemical Attack, Corrosion
Resistance, and Failure of Plastic Materials
Philip A. Schweitzer
297
13. Corrosion of Thermoset Plastics
Dirk L. Pletcher
347
14. Chemical Attack and Failure of Elastomers
Philip A. Schweitzer
385
15. Corrosion Resistance of Specific Elastomers
Philip A. Schweitzer
395
16. Aqueous Corrosion of Advanced Ceramics
Eugene L. Liening and James M. Macki
419
17. Chemical-Resistant Mortars, Grouts, and Monolithic 459
Surfacings
Augustus A. Boova
18. Glass Linings
Donald H. De Clerck
489
19. Cathodic Protection
Philip A. Schweitzer
545
Page vii
20. Corrosion Inhibitors
Philip A. Schweitzer
555
21. Painting for Protection
Walter M. McMahon
563
22. Liquid-Applied Linings
Philip A. Schweitzer
581
23. Sheet Linings
Philip A. Schweitzer
589
24. Stress Corrosion Testing
Russell D. Kane
607
25. Corrosion Monitoring
Allan Perkins
623
26. Selecting Materials of Construction
Robert B. Puyear and David A. Hansen
653
27. Designing to Prevent Corrosion
V. Mitchell Liss
677
Index
733
Page ix
Contributors
Augustus A. Boova Vice President of Operations, Atlas Minerals and
Chemicals, Inc., Mertztown, Pennsylvania (retired)
Donald H. De Clerck Consultant, Rush, New York
James L. Gossett Technical Consultant, Materials, Department of
Research and Engineering, Fisher Controls, Marshalltown, Iowa
David A. Hansen Director, Metallurgy, Mechanical Department, Fluor
Daniel, Inc., Sugar Land, Texas
Russell D. Kane President, CLI International, Inc., Houston, Texas
Gary N. Kirby President, Kirby Corrosion Control, Inc., Brielle, New
Jersey
John B. Lambert Consultant, Lake Forest, Illinois
Eugene L. Liening Senior Materials Associate, Department of
Materials Engineering, Michigan Division, The Dow Chemical
Company, Midland, Michigan
Bernard W. Lifka Technical Consultant, New Kensington,
Pennsylvania (retired)
Page x
V. Mitchell Liss Engineering Consultant, La Grange, Texas
James M. Macki Associate Director, Materials Technology Institute of
the Chemical Process Industries, Inc., St. Louis, Missouri
Walter M. McMahon Technical Director, Southern Coatings, La
Habra Heights, California (retired)
Allan Perkins Applications Engineering Manager, Marketing
Department, Rohrback Cosasco Systems, Sante Fe Springs, California
Dirk L. Pletcher Senior Research Engineer, Composite Materials
Research, Zimmer, Warsaw, Indiana
Robert B. Puyear Consultant, Chesterfield, Missouri
Philip A. Schweitzer Consultant, Fallston, Maryland
Paul K. Whitcraft Director, Quality Assurance/Engineering
Department, Rolled Alloys, Inc., Temperance, Michigan
Te-Lin Yau Senior Corrosion Engineer, Teledyne Wah Chang,
Albany, Oregon
Page 1
1
Fundamentals of Metallic Corrosion
Paul K. Whitcraft
Rolled Alloys, Inc.
Temperance, Michigan
Corrosion is all around us. It shows itself in rust-stained structures that
are falling down with age. It shows itself in the beautiful green patina
associated with copper sculptures and roofs. Corrosion processes cost
billions of dollars each year in maintenance and repair. Corrosion is
relentless and patient, and inevitable. But in some ways corrosion is
also predictable and useful. The use of controlled corrosion processes
is routinely selected as the most cost-effective solution to catastrophic
corrosion. Galvanized steel construction materials have served well
since their invention in the early 1800s. The standard lead-zinc
storage batteries in most of our automobiles are in effect controlled
and reversible corrosion cells.
Corrosion is the degradation of a material's properties or mass over
time due to environmental effects. It is the natural tendency of a
material's compositional elements to return to their most
thermodynamically stable state. For most metallic materials this
means the formation of oxides or sulfides, or other basic metallic
compounds generally considered to be ores. Fortunately, the rate at
which most of these processes progress is slow enough to provide
useful building materials. Only inert atmospheres and vacuums can be
considered free of corrosion for most metallic materials.
Page 2
These corrosion processes, as with all of the physical world, follow
the basic laws of thermodynamics. Corrosion is an electrochemical
process. Under controlled conditions it can be measured, repeated, and
predicted. Since it is governed by reactions on an atomic level,
however, corrosion processes can act on isolated regions, uniform
surface areas, or result in subsurface microscopic damage. Complicate
these forms of corrosion with further subdivisions, add just basic
environmental variables such as pH, temperature, and stress, and the
predictability of corrosion begins to suffer rapidly. Figure 1 illustrates
the basics of this complexity.
Figure 1
Corrosion environments.
Page 3
I
Forms of Corrosion
The effect of corrosion on a metallic surface can take many forms.
Identifying these forms can assist in understanding the corrosion
process and offer insight into its control.
A
Uniform Corrosion
The simplest form of corrosion is ''uniform" or "general" corrosion.
Uniform corrosion is an even rate of metal loss over the exposed
surface. It is one of the most easily measured and predictable forms of
corrosion. Many references exist which report average or typical rates
of corrosion for various metals in common media.
Since the corrosion is so uniform, corrosion rates for materials are
often expressed in terms of metal thickness loss per unit of time. One
common expression is "mils per year" or, sometimes, millimeters per
year. Because of its predictability, low rates of corrosion are often
tolerated and catastrophic failures are rare if planned inspection and
monitoring is implemented. For most chemical process and structures,
general corrosion rates of less than 3 mils per year (mpy) are
considered acceptable. Rates between 2 and 20 mpy are routinely
considered useful engineering materials for the given environment. In
severe environments, materials exhibiting high general corrosion rates
of between 20 and 50 mpy might be economically justifiable.
Materials which exhibit rates of general corrosion beyond this are
usually unacceptable. It should be remembered that not only does the
metal loss need to be considered, but where that metal is going must
also be considered. Contamination of product, even at low
concentrations, can be more costly than the replacement of the
corroded component.
Uniform corrosion is generally thought of in terms of metal loss due
to chemical attack or dissolution of the metallic component into
metallic ions. In high-temperature situations, uniform metal loss is
more commonly preceded by its combination with another element
rather than its oxidation to a metallic ion. Combination with oxygen to
form metallic oxides, or scale, results in the loss of the material in its
useful engineering form, which ultimately flakes off to return to
nature.
B
Galvanic Corrosion
When two different metallic materials are electrically connected and
placed in a conductive solution, an electrical potential will exist. This
potential difference will provide a stronger driving force for the
dissolution of the less noble (more
Page 4
electrically negative) material. It will also reduce the tendency for the
more noble material to dissolve.
While the relative differences in potential will change from one
environment to another, they remain fundamentally the same since the
potential is related to the energy required to oxidize them to metal
ions in the given environment. The significance of this becomes more
apparent when a variety of materials are listed in order of their
electrical potential in a familiar environment such as that shown in
Table 1. Notice that the precious metals of gold and platinum are at
the high potential (more noble or cathodic) end of the series, while
zinc and magnesium are at the low potential (less noble or anodic)
end.
The practical implications of this concept constantly surround us. It is
this principle that forms the scientific basis for using a material such
as zinc to "sacrificially" protect a stainless steel drive shaft on a
pleasure boat. It supplies the logic for the use of galvanized steel.
Galvanic corrosion is often experienced by owners of older homes
where more modern copper water tubing is connected to the older
existing carbon steel water lines.
Page 5
C
Erosion Corrosion
An increased rate of corrosion attack which is attributable to the
movement of a corrodent over a surface is recognized as erosion
corrosion. The movement of the corrodent can be associated with
mechanical wear. The increase in localized corrosion due to the
erosion process is usually related to the removal or damage of the
protective surface film. The mechanism is usually identified by
localized corrosion which exhibits a pattern that follows the flow of
the corrodent.
Fretting corrosion is a specialized form of erosion corrosion where
two metal surfaces are in contact and experience very slight relative
motion causing damage to one or both surfaces. Again, in the
presence of a corrodent, the movement causes mechanical damage of
the protective film leading to localized corrosion. The corrosion
usually takes the form of pitting-type attack.
A second subset of erosion corrosion is the case of cavitation. A type
of corrosion familiar to pump impellers, this form of attack is caused
by the formation and collapse of tiny vapor bubbles near a metallic
surface in the presence of a corrodent. The protective surface film is
again damaged, in this case by the high pressures caused by the
collapse of the bubbles.
D
Pitting Corrosion
Pitting corrosion is in itself a corrosion mechanism, but it is also a
form of corrosion often associated with other types of corrosion
mechanisms. It is characterized by a highly localized loss of metal. In
the extreme case, it appears as a deep, tiny hole in an otherwise
unaffected surface. Figure 2 illustrates the nature of pitting and other
forms of corrosion.
The initiation of a pit is associated with the breakdown of the
protective film on the metal surface. In cases where pit depths
increase rapidly, the environment is usually such that no repair or
repassivation of the protective layer can be accomplished. For other
instances where many shallow pits form, the environment is usually
one where repassivation of the damaged film can be made but
initiation of new sites is occurring on a regular basis.
The localized nature of pitting attack can be associated with
component geometry, the mechanics of the corrosion process, or with
imperfections in the material itself. The growth of pits, once initiated,
is closely related to another corrosion mechanism, i.e., crevice
corrosion.
E
Crevice Corrosion
Crevice corrosion occurs in some environments because the nature of
the environment within the crevice will become more aggressive with
time. Movement
Page 6
Figure 2
Forms of corrosion.
of the corrodent within a crevice is slow or nonexistent. Over time,
small changes due to minor localized corrosion may become
magnified because they are not constantly being replenished or
minimized by the bulk solution.
As a result of a slow initial rate of corrosion the pH of the crevice
environment may become more acidic or detrimental ion species may
concentrate. As a result of the low flow condition the crevice region
may become depleted of oxygen or preclude the replacement of
reacted inhibitors.
F
Selective Leaching
Selective leaching is the process whereby a specific element is
removed from an alloy due to an electrochemical interaction with the
environment. Dezincification of brass alloys is the most familiar
example of this type of corrosion. It occurs most commonly when
exposed to soft waters and can be accelerated by high carbon dioxide
concentrations and the presence of chloride ions.
The result of this corrosion is that of leaving a porous and usually
brittle shadow of the original component. Other alloy systems are
susceptible to this form of corrosion. Examples include the selective
loss of aluminum in aluminum-copper alloys and the loss of iron in
cast iron-carbon steels.
G
Intergranular Corrosion
As the name suggests, this particular corrosion mechanism attacks
those sites where individual grains within a metallic material touch
each other. These
Page 7
Figure 3
Photomicrograph of intergranular sulfidation attack in a high-nickel alloy
(alloy 600). This type of attack can occur in high-nickel alloys in the
presence of sulfur-bearing compounds under reducing conditions at
temperatures above about 1150°F. The intergranular attack by sulfidation
is often preceded by carburization, which also takes the form of
intergranular attack. The etchant was glyceregia (×100).
boundaries are natural regions of higher energy due to the greater
frequency of dislocations of atoms from the natural order of the
material's structure. Figure 3 is an example of intergranular attack
caused by high-temperature sulfidation in a nickel base alloy. In
addition, these regions also tend to act as sites for the formation of
secondary phases, which are essentially small islands within the
matrix that have a chemical composition different from the alloy
itself.
Depending on the corrodent and the alloy system, corrosion attack
may initiate at these locations due to preferential attack of the
secondary phase itself, or attack the surrounding matrix which was
locally dealloyed in forming the secondary phase. Either mechanism
will result in the metallic surface being etched along the grain
boundaries. As the attack progresses, individual grains are separated
from the matrix and the surface layer becomes porous. In severe cases
the surface texture will be grainy or powdery leading to more rapid
metal loss.
Page 8
Figure 4
Chloride stress corrosion cracking found in a S30400 stainless
steel stack operating around 150°F. The upper photomicrograph
illustrates the extensive branching associated with this type of
corrosion (×25). The transgranular nature of this corrosion
mechanism is shown in the lower photomicrograph (×100). The
etchant was electrolytic oxalic acid.
Page 9
H
Stress Corrosion Cracking
The mechanism of stress corrosion cracking (SCC) is specific to
certain alloys (or alloy systems) in specific environments. It is
characterized by one or more crack fronts which have developed as a
result of a combination of the particular corrodent and tensile stresses.
Depending on the alloy system and corrodent combination, the
cracking can be intergranular or transgranular. The rate of crack
propagation can vary greatly and is affected by stress levels,
temperature, and concentration of the corrodent. In some severe
combinations, such as type 304 stainless steel in a boiling magnesium
chloride solution, extensive cracking can be generated in a matter of
hours. Fortunately, in most industrial applications the progress of SCC
is much slower. However, because of the nature of the cracking it is
difficult to detect until extensive corrosion has already developed,
which can lead to unexpected catastrophic failure. Such an example of
this crack mechanism is shown in Fig. 4. Alloy system and corrodent
combinations which are known to exhibit SCC are fairly well
documented and should be considered in initial design stages.
Apart from the SCC mechanism, stress can assist in other corrosion
processes. Since this stress-assisted corrosion is related to tensile
stresses it is logical to expect that it will also accelerate the simple
mechanical fatigue process. Corrosion fatigue is often difficult to
differentiate from simple mechanical fatigue but is recognized as a
factor when the environment has been judged to have accelerated the
normal fatigue process. Such systems can also have the effect of
lowering the endurance limit such that fatigue will take place at a
stress level at which, without the environmental effect, fatigue failures
would not be expected.
II
Corrosion Processes
A
Electrochemical Nature of Corrosion
Corrosion by its simplest definition is the process of a material
returning to the natural thermodynamic state. For most metallic
materials this means the formation of the oxides or sulfides from
which they originally started when they were taken from the earth
before being refined into useful engineering materials.
These changes are electrochemical reactions which follow the laws of
thermodynamics. Understanding the interaction of materials with their
environment now takes on the added dimensions of chemistry and
electricity. These concepts help explain why corrosion processes are
time-and temperature-dependent. They also establish that corrosion
reactions, or rates, are affected by ion and corrodent concentrations.
They also explain why some reactions are reversible, or controllable,
while others are not. One of the most basic corrosion reactions
involves the oxidation of a pure metal when exposed to a strong acid.
A familiar
Page 10
case is that of placing pure iron in hydrochloric acid. The resulting
chemical reaction is obvious with the solution beginning to bubble
violently. The chemical reaction can be expressed as follows:
We can see the result of this reaction by the gradual disappearance of
the iron and the hydrogen bubbles rising rapidly to the surface. On an
electrochemical level, there is also an exchange of electrons taking
place.
The iron has been converted to an iron ion by giving up two electrons
(oxidation) which were picked up by the hydrogen ions. By gaining
electrons the hydrogen ion was reduced and formed hydrogen gas.
Note that the chlorine atom does not enter into the reaction itself. The
transfer of electrons is taking place on the metal's surface. Those
locations where electrons are being given up are identified as anodes.
The sites where electrons are being absorbed are denoted as cathodes.
A difference in electrical potential exists between these two areas and
a complete electrical circuit is developed. Negatively charged
electrons flow in the direction of anode to cathode and positively
charged hydrogen ions in the solution move toward the cathode to
complete the circuit. The faster the dissolution of the metal (rate of
corrosion), the higher the current flow. The sites of the anodes and
cathodes can change locations on the surface. In fact, this is exactly
what happens when general corrosion takes place, with the anodic
areas moving uniformly over the metal's surface.
Anodic reactions in metallic corrosion are relatively simple. In fact,
the reactions are always such that the metal is oxidized to a higher
valence state. During general corrosion, this will result in the
formation of metallic ions of all the alloying elements. Metals which
are capable of exhibiting multiple valence states may go through
several stages of oxidation during the corrosion process.
Cathodic reactions are more difficult to predict but can be categorized
into one of five different types of reduction reactions.
Page 11
B
Cell Potentials
Understanding electrochemical behavior and the possible reactions
can help in predicting the possibility and extent of corrosion. A
reaction will only occur if there is a negative free energy change
(DG). For electrochemical reactions the free energy change is
calculated from:
where n is the number of electrons, F is Faraday's constant, and E is
the cell potential.
Therefore, for a given reaction to take place the cell potential must be
positive. The cell potential is taken as the difference between the two
half-cell reactions, the one at the cathode minus the one at the anode.
The half-cell potential exists because of the difference in the neutral
state compared to the oxidized state, such as Fe/Fe2+, or, at the
cathode, the difference between the neutral state and the reduced state,
as in H+/H2. These reductionoxidation (redox) potentials are
measured relative to a standard half-cell potential. The chart shown in
Table 2 lists potentials relative to the H+/H2, which is set as zero.
Table 2 Standard
OxidationReduction Potentials,
25°C, Volts vs. Hydrogen
Electrode
Au « Au3+ + 3e
1.498
O2 + 4H+ + 4e « 2H2O 1.229
Pt « Pt2+ + 2e
1.2
Ag « Ag+ + e
0.799
Fe3+ + e « Fe2+
0.771
O2 + 2H2O + 4e « 4OH 0.401
Cu « Cu2+ + 2e
0.337
0.000
2H+ + 2e « H2
Pb « Pb2+ + 2e
-0.126
Ni « Ni2+ + 2e
-0.250
Fe « Fe2+ + 2e
-0.440
Cr « Cr3+ + 3e
-0.744
Zn « Zn2+ + 2e
-0.763
Al « Al3+ + 3e
-1.662
Mg « Mg2+ + 2e
-2.363
Source: A. J. de Bethune and
N. A. S. Loud, Standard
Aqueous Electrode Potentials
and Temperature Coefficients
at 25°C, Clifford A. Hampel,
Skokie, IL, 1964.
Page 12
Looking at the example of iron corroding freely in acid, the cell
potential is calculated to be
Since the cell potential is positive, the reaction can take place. The
larger this potential difference, the greater the driving force for the
reaction. Whether corrosion does occur and at what rate is dependent
on other factors. In order for corrosion to occur, there must be a
current flow and a completed circuit which is then governed by Ohm's
law: I = E/R. The cell potential calculated here represents the peak
value for the case of two independent reactions. If the resistance were
infinite, the cell potential would remain as calculated but there would
be no corrosion at all. If the resistance of the circuit was zero, the
potentials of each half-cell would approach the other while the rate of
corrosion would be infinite.
C
Polarization
At an intermediate resistance in the circuit, some current begins to
flow and the potentials of both half-cell reactions move slightly
toward each other. This change in potential is called polarization. The
resistance in the circuit is dependent on a number of factors including
the resistivity of the media, surface films, and the metal itself. The
relationships between the polarization reactions at each half-cell are
represented in Fig 5. The intersection of the two polarization lines
(curves) closely approximates the corrosion current and the combined
cell potentials for the freely corroding situation.
Once the corrosion current is determined, the corrosion current
density can be calculated by determining the surface area. Using
Faraday's laws, a corrosion rate in terms of metal loss per unit time
can be determined. However, polarization data can be more useful
than just estimating corrosion rates. The extent of polarization can
help predict the type and severity of corrosion. As polarization
increases, corrosion decreases. Polarization may be preferential to
either the cathodic or anodic reactions. Understanding the influence of
environmental changes on polarization can offer insight to controlling
corrosion. For example, in the ironhydrochloric acid example,
hydrogen gas formation at the cathode can actually slow the reaction
(increased circuit resistance) by blocking access of hydrogen ions to
the cathode site. This results in cathodic polarization and lowers the
current flow and corrosion rate. If oxygen is bubbled through the
solution the hydrogen is removed more rapidly by combining to form
water and the corrosion rate increases significantly. Although this is
an oversimplified view of the effects
Page 13
Figure 5
Polarization of iron in acid.
of oxygen, it does indicate that the degree of polarization can be
affected by changes in the environment, either natural or induced.
There are three basic causes of polarization, termed activation,
concentration, and potential drop. Potential drop is the change in
voltage associated with effects of the environment and the circuit
between the anode and cathode sites. It includes the effects of the
resistivity of the media, surface films, corrosion products, etc.
Activation polarization is due to a rate-controlling step within the
corrosion reaction(s) at either the cathode or anode sites. An example
of this can be seen with the H+/H2 conversion reaction. The first step
of this process, 2H+ + 2e ® 2H, takes place at a rapid pace. The
second part of this reaction, 2H ® H2, occurs more slowly and can
become a rate-controlling factor.
Concentration polarization is the effect resulting from the excess of a
species which impedes the corrosion process, or with the depletion of
a species critical to the progression of the corrosion process. The
earlier case with an excess concentration of hydrogen gas impeding
the rate of reaction is an example of concentration polarization. While
in this case it occurred at the cathode, it can also develop at the anode.
Page 14
D
Oxygen Concentration Cells
The oxygen reduction reaction which occurs in neutral or basic
solutions, O2 + 2H2O + 4e ® 4OH-, plays a significant role in many
corrosion processes. It not only contributes to sustaining a cathodic
reaction but can induce one. This occurs when substantial differences
in dissolved oxygen content exist at one area on the metal surface
relative to another. The natural tendency is to equal concentrations
and the means of achieving this by corrosion is to lower the oxygen
concentration at the region where it is highest. The oxygen reduction
reaction accomplishes this, but in the process the area where this
occurs becomes cathodic to the lower oxygen concentration region.
Because of the current flow created by this action, corrosion will
occur at the anodic or low oxygen concentration site.
E
Metal Ion Concentration Cells
In a similar fashion, metal ion concentration cells can also develop
and fuel the corrosion process. This situation arises when a significant
difference in metal ion concentration exists over a metal surface.
Again, the tendency is to reach equilibrium in ion concentration, and
in a corrosive environment this is managed by putting more metal ions
into solution at the low-concentration area. This area becomes the
anode and the current flow generated by this process can result in
plating out metal ions at the cathodic or high metal ion concentration
region.
III
Measuring Polarization
While polarization always leads to lower rates of corrosion,
identifying the effects of the environment on polarization of the
corrosion circuit is useful in predicting corrosion behavior.
Measurement of the corrosion current while the corrosion potential is
varied is possible with the apparatus shown in the Fig. 6.
Again, turning to the example of iron corroding in a hydrochloric acid
solution, if the iron sample is maintained at the natural corrosion
potential of -0.2 V, no current will flow through the auxiliary
electrode. The plot of this data point in the study would equate to that
of A or C in Fig 7. As the potential is raised, anodically polarized, the
current flow will increase and curve AB will approximate the
behavior of the true anodic polarization curve. Alternatively, if the
potential were lowered below -0.2 V, the current measurements would
result in the curve CD and approximate the nature of the cathodic
polarization curve. By using the straight line portion, or Tafel regions,
of these curves, an approximation of the corrosion current can be
made.
Most often it is the anodic polarization behavior that is useful in
understanding alloy systems in various environments. Anodic
polarization tests can be conducted with relatively simple equipment
and the scans themselves can be done
Page 15
Figure 6
Anodic polarization measurement apparatus.
in a short time. They are extremely useful in studying the activepassive behavior that many materials exhibit. As the name suggests,
these materials can exhibit both a highly corrosion-resistant behavior
and that of a material that corrodes actively, while in the same
corrodent. Metals that commonly exhibit this type of behavior include
iron, titanium, aluminum, chromium, and nickel. Alloys of these
materials are also subject to this type of behavior.
Active-passive behavior is dependent on the materialcorrodent
combination and is a function of the anodic or cathodic polarization
effects which occur in that specific combination. In most situations
where active-passive behavior occurs, there is a thin layer at the metal
surface that is more resistant to the environment than the underlying
metal. In stainless steels this layer is composed of various chormium
and/or nickel oxides which exhibit substantially different
electrochemical characteristics than the underlying alloy. If this
resistant, or passive, layer is damaged while in an aggressive
environment, active corrosion of the freshly exposed surface will
occur. The damage to this layer can be either mechanical or
electrochemical in nature.
The behavior of iron in nitric acid underscores the importance of
recognizing the nature of passivity. Iron is resistant to corrosion in
nitric acid at concentrations around 70%. Once passivated under these
conditions it can also
Page 16
Figure 7
Anodic and cathodic polarization curves.
exhibit low rates of corrosion as the nitric acid is diluted. However, if
this passive film is disturbed, rapid corrosion will begin and
repassivation will not be possible until the nitric acid concentration is
raised to a sufficient level.
A
Anodic Polarization
Active-passive behavior is schematically represented by the anodic
polarization curve shown in Fig. 8. Starting at the base of the plot, the
curve starts out with a gradually increasing current as expected.
However, at point A, there is a dramatic polarizing effect which drops
the current to a point where corrosion is essentially halted. As the
potential is increased further, there is little change in current flow until
the next critical stage, B, where a breakdown of the passive film
occurs and the corrosion current again begins to rise.
Even with an established anodic polarization behavior, the
performance of a material can vary greatly with relatively minor
changes in the corrodent. This is also illustrated in Fig 9. Frame 1
illustrates the case where the anodic and cathodic polarization curves
intersect much as in materials with no active-passive behavior. The
anode is actively corroding at a high but predictable rate.
Page 17
Figure 8
Anodic polarization curve for material exhibiting active-passive
behavior.
Frame 2 represents the condition often found perplexing when using
materials that exhibit active-passive behavior. With relatively minor
changes within the system, the corrosion current could be very low
when the material is in the passive state or very high when active
corrosion begins.
Frame 3 typifies the condition sought after when using materials in
the passive state. In this example the cathodic polarization curve
intersects only in the passive region, resulting in a stable and low
corrosion current. This type of system can tolerate moderate upset
conditions without the onset of accelerated corrosion.
Figure 9
Schematic representation of a material with active-passive behavior in different
corrosive environments.
Page 18
Figure 10
Effects of environment and alloy content on anodic polarization behavior.
The anodic polarization technique is also useful in studying the effects
of variations in the environment and the benefits of alloy additions. As
illustrated in Fig. 10, temperature increases cause a shift of the curve
to higher currents. Increasing chromium contents in steel expands the
passive region significantly; adding molybdenum raises the potential
required for the initiation of pitting-type attack. The presence of
chloride or other strong oxidizing ions will shrink the passive region.
IV
Other Factors Affecting Corrosion
As was just noted, temperature can have a significant influence on the
corrosion process. This is not surprising since it is an electrochemical
reaction and reaction rates do increase with increasing temperature.
There are additional influences on corrosion other than the corrodent
itself.
The relative velocities between the component and the media can have
a direct effect on the corrosion rate. In some instances, increasing the
velocity of the corrodent over the surface of the metal will increase
the corrosion rate. When concentration polarization occurs, the
increased velocity of the media will
Page 19
disperse the concentrating species. However, with passive materials,
increasing the velocity can actually result in lower corrosion rates.
This occurs since the increased velocity shifts the cathodic
polarization curve such that it no longer intersects the anodic
polarization curve in the active corrosion region as shown in Fig 11.
The surface finish of the component also has an impact on the mode
and severity of the corrosion that can occur. Rough surfaces or tight
crevices can facilitate the formation of concentration cells. Surface
cleanliness can also be an issue with deposits or films acting as
initiation sites. Biological growths can behave as deposits or change
the underlying surface chemistry to promote corrosion.
Other variations within the metal surface on a microscopic level
influence the corrosion process. Microstructural differences such as
secondary phases or grain orientation will affect the way corrosion
manifests itself. For corrosive environments where grain boundaries
are attacked the grain size of the material plays a significant role in
how rapidly the material's properties deteriorate. Chemistry variations
in the matrix of weld deposits are also factors.
Radiation can have an effect on a material's mechanical properties.
The effect on metallic materials is very gradual and not very
pronounced. Stresses,
Figure 11
Increased corrodent velocity can shift the cathodic polarization
curve such that passive behavior can be induced.
Page 20
either residual or applied, impact the mode of corrosion and lower the
energy needed for corrosion to begin. Stress is a requirement for stress
corrosion cracking or corrosion fatigue, but can also influence the rate
of general corrosion.
Finally, time is a factor in determining the severity of corrosion.
Corrosion rates are expressed using a time dimension. Some corrosion
processes are violent and rapid while most are so slow as to be
imperceptible on a day-to-day basis. Equipment is planned to provide
a useful service life. A chief goal in understanding corrosion is the
proper selection of materials, equipment, processes, or controls to
optimize our natural and financial resources.
V
Corrosion Basics
This chapter has provided a foundation for the recognition and
comprehension of corrosion processes. The basic principles outlined
here can be applied to identified corrosion problems and provide
solutions or alternatives. Corrosion control in many forms and
approaches is founded on these concepts.
The principle of cathodic or ''sacrificial" protection is founded in the
natural potential differentials between different metals. Zinc anodes
are intentionally placed in electrical contact with steel structures so
that as they corrode the steel is protected. In other systems a current
may be applied to the structure to be protected so as to cause the
current to flow to an artificial anode.
For similar reasons, it is desirable to build process systems from the
same materials. In systems where contact of dissimilar metals cannot
be avoided, it is helpful to have the less noble material possess the
largest surface area. By doing so the corrosion current that is
generated is distributed over a much greater area and slows the overall
rate of penetration. In many such systems it is also possible to
electrically insulate one alloy network from the other.
Anodic protection finds its basis in the understanding of activepassive behavior. By increasing the potential of the component to be
protected, it moves from an actively corroding situation to one where
passivity can be induced. Such techniques can be quite cost-effective
but must be applied under well controlled operating conditions since
slight over-or underprotection can lead to accelerated rates of
corrosion.
The types and varieties of inhibition systems are diverse but also
derive their fundamental logic from the principles reviewed here.
Inhibitors slow corrosion by increasing polarization at either the
anodic or cathodic reactions, or by increasing the electrical resistance
of the media.
The corrosion engineer can play a major role in system design,
material selection, process, or environmental control and remediation.
The focus of these efforts should not necessarily be the complete
elimination or avoidance of corrosion but rather in choosing the most
cost-effective means or corrosion control and abatement.
Page 21
References
1. A. Brasunas (ed.), NACE Basic Corrosion Course, National
Association of Corrosion Engineers, Houston, 1970.
2. M. Fontana and N. Greene, Corrosion Engineering, McGraw-Hill,
New York, 1967.
3. L. Darken and R. Gurry, Physical Chemistry of Metals, McGrawHill, New York, 1953.
4. A. Sedriks, Corrosion of Stainless Steels, John Wiley and Sons,
New York, 1979.
5. M. Henthorne, Corrosion, Causes and Control, Chemical
Engineering Magazine, New York, 19711972.
6. P. Schweitzer, Corrosion and Corrosion Protection Handbook,
Marcel Dekker, New York, 1979.
Page 23
2
Atmospheric Corrosion
Philip A. Schweitzer
Fallston, Maryland
Atmospheric corrosion, though not a separate form of corrosion, has
received considerable attention because of the staggering associated
costs which result. With the large number of outdoor structures such
as buildings, fences, bridges, towers, automobiles, ships, and
innumberable other applications exposed to the atmospheric
environment, there is no wonder that so much attention has been given
to the subject.
Atmospheric corrosion is a complicated electrochemical process
taking place in corrosion cells consisting of base metal, metallic
corrosion products, surface electrolyte, and the atmosphere. Many
variables influence the corrosion characteristics of an atmosphere.
Relative humidity, temperature, sulfur dioxide content, hydrogen
sulfide content, chloride content, amount of rainfall, dust, and even
position of the exposed metal exhibit marked influence on corrosion
behavior. Geographic location is also a factor.
Because this is an electrochemical process, an electrolyte must be
present on the surface of the metal for corrosion to occur. In the
absence of moisture, which is the common electrolyte associated with
atmospheric corrosion, metals corrode at a negligible rate. For
example, carbon steel parts left in the desert remain bright and tarnish
free over long periods. Also, in climates where the air temperature is
below the freezing point of water or of aqueous condensation on
Page 24
the metal surface, rusting is negligible because ice is a poor conductor
and does not function effectively as an electrolyte.
Atmospheric corrosion depends not only on the moisture content
present but on the dust content and the presence of other impurities in
the air, all of which have an effect on the condensation of moisture on
the metal surface and the resulting corrosiveness. Air temperature can
also be a factor.
I
Atmospheric Types
Since corrosion rates are affected by local conditions, atmospheres are
generally divided into the following major categories:
Rural
Industrial
Marine
Additional subdivisions such as, urban, arctic, and tropical (wet or
dry) can also be included. However, of main concern are the three
major categories.
For all practical purposes, the more rural the area, with little or no
heavy manufacturing operations, or with very dry climatic conditions,
the less will be the problem of atmospheric corrosion.
In an industrial atmosphere, all types of contamination by sulfur in the
form of sulfur dioxide or hydrogen sulfide are important. The burning
of fossil fuels generates a large amount of sulfur dioxide, which is
converted to sulfuric and sulfurous acid in the presence of moisture.
Combustion of these fossil fuels and hazardous waste products should
produce only carbon dioxide, water vapor, and inert gas as
combustion products. This is seldom the case. Depending on the
impurities contained in the fossil fuel, the chemical composition of the
hazardous waste materials incinerated, and the combustion conditions
encountered, a multitude of other compounds may be formed.
In addition to the most common contaminants previously mentioned,
pollutants such as hydrogen chloride, chlorine, hydrogen fluoride, and
hydrogen bromide are produced as combustion products from the
burning of chemical wastes. When organophosphorous compounds
are incinerated, corrosive phosphorous compounds are produced.
Chlorides are also a product of municipal incinerators.
Road traffic and energy production lead to the formation of NOx
which may be oxidized to HNO3. This reaction has a very low rate;
therefore in the vicinity of the emission source the contents of HNO3
and nitrates are very low. The antipollution regulations that have been
enacted do not prevent the escape into the atmosphere of quantities of
these materials sufficient to prevent corrosion problems. The
corrosivity of an industrial atmosphere diminishes with increasing
distance from the city.
Page 25
Marine environments are subject to chloride attack resulting from the
deposition of fine droplets of crystals formed by evaporation of spray
that has been carried by the wind from the sea. The quantity of
chloride deposition from marine environment is directly proportional
to the distance from the shore. The closer to the shore, the greater the
deposition and corrosive effect. The atmospheric test station at Kure
Beach, North Carolina shows that steels exposed 80 feet from the
ocean corrode 1015 times faster than steels exposed 800 feet from the
ocean.
In addition to these general air contaminants, there may also be
specific pollutants found in a localized area. These may be emitted
from a manufacturing operation on a continuous or spasmodic basis
and can result in a much more serious corrosion problem than that
caused by the presence of general atmospheric pollutants.
Because of these varying conditions, a material that is resistant to
atmospheric corrosion in one area may not be satisfactory in another.
For example, galvanized iron is perfectly suitable for application in
rural atmospheres, but it is not suitable when exposed to industrial
atmospheres.
II
Factors Affecting Atmospheric Corrosion
As previously described, atmospheric corrosion is an electrochemical
process and as such depends on the presence of an electrolyte. The
usual electrolyte associated with atmospheric corrosion is water
resulting from rain, dew, fog, melting snow, or high humidity. Since
an electrolyte is not always present, atmospheric corrosion is
considered a discontinuous process. Corrosion only takes place during
the time of wetness.
A
Time of Wetness
The term "time of wetness" refers to the length of time during which
the metal surface is covered by a film of water that renders significant
atmospheric corrosion possible. The time of wetness is dependent on
local climatic conditions such as frequency of rain, fog, and dew;
temperature of the metal surface; temperature of air; relative humidity
of the atmosphere; wind speed; and hours of sunshine.
The time of wetness can be determined either by meteorological
measurements of temperature and relative humidity or by
electrochemical cells. The time of wetness determined by
meteorological measurements may not necessarily be the same as the
actual time of wetness because wetness is influenced by the type of
metal, pollution of the atmosphere, presence of corrosion products,
and degree of coverage against rain. However, the results from these
measurements
Page 26
usually show a good correlation with corrosion data from field tests
under ordinary outdoor conditions.
B
Adsorption Layers
The adsorption of water on the metal surface may be the result of the
relative humidity of the atmosphere, of the chemical and physical
properties of the corrosion products, of the properties of materials
deposited from the air, or a combination of all three. Industrial
atmospheres contain suspended particles of carbon, carbon
compounds, metal oxides, sulfuric acid, sodium chloride, and
ammonium sulfate. When these substances combine with moisture or
when because of their hygroscopic nature they form an electrolyte on
the surface, corrosion is initiated.
When hygroscopic salts which are deposited or formed by corrosion
absorb moisture from the atmosphere, the metal surface may become
wetted. Such absorption occurs above a certain relative humidity,
called the critical relative humidity, which corresponds to the vapor
pressure above a saturated solution of the salt present. The amount of
water on the surface has a direct effect on the corrosion rate. The more
water present, the greater the corrosion rate.
C
Phase Layers
Phase layers are the result of the formation of dew by condensation on
a cold metallic surface, precipitation in the form of rain or fog, and
wet or melting snow. The rate of corrosion will be dependent on the
concentration and nature of the corrodents in the electrolyte, which
will vary depending on the deposition rates, frequency of wetting,
drying conditions, and degree of rain protection provided.
If the surface is wetted after a long dry spell during which there has
been a large accumulation of surface contamination, the corrosion rate
will be greater than that for a smaller amount accumulated during a
shorter dry period. Corrosion will also be affected by the quantity of
electrolyte present.
Dew is an important source of atmospheric corrosionmore so than
rainand particularly under sheltered conditions. Dew forms when the
temperature of the metal surface falls below the dew point of the
atmosphere. This can occur outdoors during the night when the
surface temperature of the metal is lowered as a result of radiant heat
transfer between the metal and the sky. It is also common for dew to
form during the early morning hours when the air temperature rises
faster than the metal temperature. Dew may also form when metal
products are brought into warm storage after cold shipment.
Under sheltered conditions dew is an important cause of corrosion.
The high corrosivity of dew is a result of several factors.
Page 27
1. Relatively speaking, the concentration of contaminants in dew are
higher than in rain water, which leads to lower pH values. Heavily
industrialized areas have reported pH values of dew in the range of 3
and lower.
2. The washing effect, which occurs with rain, is usually slight or
negligible.
With little or no runoff, the pollutants remain in the electrolyte and
continue their corrosive action. As the dew dries, these contaminants
remain on the surface to repeat their corrosive activity with
subsequent dew formation.
Depending on the conditions, rain can either increase or decrease the
effects of atmospheric corrosion. Corrosive action is caused by rain
when a phase layer of moisture is formed on the metal surface. This
activity is increased when the rain washes corrosive promoters such as
H+ and
from the air (acid rain). Rain has the ability to decrease
corrosive action on the surface of the metal as a result of washing
away the pollutants that have been deposited during the preceding dry
spell.
Whether the rain will increase or decrease the corrosive action is
dependent on the ratio of deposition between the dew and wet
contaminants. When the dry period deposition of pollutants is greater
than the wet period deposition of sulfur compounds, the washing
effect of the rain will dominate and the corrosive action will be
decreased.
In areas where the air is less heavily polluted, the corrosive action of
the rain will assume a much greater importance because it will
increase the corrosion rate.
High concentrations of sulfate and nitrate and high acidity will be
found in areas having an appreciable amount of air pollution. The pH
of fog water has been found to be in the range of 2.24.0 in highly
contaminated areas. This leads to increased corrosivity.
D
Dust
On a weight basis in many locations, dust is the primary air
contaminant. When in contact with metallic surfaces and combined
with moisture, dust can promote corrosion by forming galvanic or
differential aeration cells that, because of their hygroscopic nature,
form an electrolyte on the surface. This is particularly true if the dust
contains water-soluble particles or particles on which sulfuric acid is
absorbed. Dust-free air therefore is less likely to cause corrosion.
E
Temperature
During long-term exposure in a temperate climatic zone, temperature
appears to have little or no effect on the corrosion rate. The overall
effect of temperature on
Page 28
the corrosion rate is complex. As the temperature increases the rate of
corrosive attack is increased as the result of an increase in the rate of
electrochemical and chemical reactions as well as the diffusion rate.
Consequently, under constant humidity conditions, a temperature
increase will promote corrosion.
By the same token, an increase in temperature can cause a decrease in
the corrosion rate by causing a more rapid evaporation of the surface
moisture film created by rain or dew. This reduces the time of
wetness, which in turn reduces the corrosion rate. In addition, as the
temperature increases, the solubility of oxygen and other corrosive
gases in the electrolyte film is decreased.
When the air temperature falls below 32°F/0°C the electrolyte film
may freeze. When freezing occurs there is a pronounced decrease in
the corrosion rate, which is illustrated by the low corrosion in
subarctic or arctic regions.
In general, temperature is a factor influencing corrosion rates, but it is
of importance only under extreme conditions.
III
Specific Atmospheric Corrodents
The electrolyte film on the surface will contain various materials
deposited from the atmosphere or originating from the corroding
metal. The composition of the electrolyte is often the factor which
determines the rate of corrosion.
The primary contaminants in the air that lead to atmospheric corrosion
are SOx, NOx, chlorides, and oxygen.
SOx
Sulfur dioxide, which results from the burning of fossil fuels (such as
coal and oil) and the combustion products from the incineration of
organic and hazardous wastes, is the most important corrosive
contaminant found in industrial atmospheres. Most of the sulfur
derived from the burning of fossil fuels is emitted in the form of
gaseous SO2. Once in the atmosphere their physical and chemical
state undergoes change. The sulfur dioxide is oxidized on moist
particles or in droplets of water to sulfuric acid:
The sulfuric acid can be partially neutralized, particularly with
ammonia resulting from the biological decomposition of organic
matter. This neutralization forms particles containing (NH4)2SO4 and
forms of acid ammonium sulfate such as NH4HSO4 and
(NH4)3H(SO4)2. Atmospheric corrosion results from the deposition of
these various materials on metallic surfaces. Deposition of these sulfur
compounds is accomplished by:
Page 29
1. Dry deposition
a. Absorption of sulfur dioxide gas on metal surfaces
b. Impaction of sulfate particles
2. Wet deposition
a. Removal of gas from the atmosphere by precipitation in the form
of rain or fog
The primary cause of atmospheric corrosion is the dry deposition of
sulfur dioxide on metallic surfaces. This type of corrosion is usually
confined to areas having a large population, many structures, and
severe pollution. Therefore the atmospheric corrosion caused by
sulfur pollutants is usually restricted to an area close to the source.
NOx
Nitrogen oxide emissions originate from combustion processes other
than those emitting SOx. Road traffic and energy production are the
primary sources. Most of the nitrogen oxides are emitted as NO in
combustion processes. In the atmosphere oxidation to NO2 takes place
successfully according to
As the pollutant moves further from the source it is further oxidized
by the influence of ozone:
Near the emission source nitrogen dioxide is considered to be the
primary pollutant. The NO2/NO ratio in the atmosphere varies with
time and distance from the source. Allowed enough time the NOx may
be further oxidized according to the reaction
Since this reaction occurs at a very slow rate, the amounts of HNO3
and nitrates in the vicinity of the source are very low.
C
Chlorides
In marine environments chloride deposition is in the form of droplets
or crystals formed by evaporation of spray that has been carried by the
wind from the sea. As distance from the shore increases, this
deposition decreases as the droplets and crystals are filtered off when
the wind passes through vegetation or when the particles settle by
gravity.
Page 30
Gaseous HCl is a combustion product derived from the burning of
coal and municipal incinerators. This gaseous HCl is very soluble in
water and forms hydrochloric acid, which is extremely corrosive.
D
Oxygen
Oxygen is a natural constituent of air and is readily absorbed from the
air into the water film on the metal surface, which may be considered
saturated, thus promoting any oxidation reactions.
E
Hydrogen Sulfide
Trace amounts of hydrogen sulfide are present in some contaminated
atmospheres. This can cause the tarnishing of silver and copper by the
formation of tarnish films.
IV
Effects on Metals Used for Outdoor Applications
Carbon steel is the most widely used metal for outdoor applications
although large quantities of zinc, aluminum, copper, and nickelbearing alloys are also used. Metals customarily used for outdoor
installations will be discussed.
A
Carbon Steel
Except in a dry, clean atmosphere carbon steel does not have the
ability to form a protective coating as some other metals do. In such
an atmosphere a thick oxide film forms which prevents further
oxidation. Solid particles on the surface are responsible for the start of
corrosion. This settled airborne dust promotes corrosion by absorbing
SO2 and water vapor from the air. Even greater corrosive effects
result when particles of hygroscopic salts, such as sulfates or
chlorides, settle on the surface and form a corrosive electrolyte.
To protect the surface of unalloyed carbon steel, an additional surface
protection must be applied. This protection usually takes the form of
an antirust paint or other type of paint protection formulated for
resistance against a specific type of contaminant known to be present
in the area. On occasion, plastic or metallic coatings are used.
B
Weathering Steels
Weathering steels are steels to which small amounts of copper,
chromium, nickel, phosphorus, silicon, manganese, or various
combinations thereof have been
Page 31
added. This results in a low-alloy carbon steel which has improved
corrosion resistance in rural areas or in areas exhibiting relatively low
pollution levels. Factors which affect the corrosion resistance of these
steels are
Climatic conditions
Pollution levels
Degree of sheltering from the atmosphere
Specific composition of the steel
Exposure to most atmospheres results in a corrosion rate which
becomes stabilized in 35 years. Over this period a protective film or
patina, dark brown to violet in color, forms. This patina is a tightly
adhering rust formation on the surface of the steel which cannot be
wiped off. Since the formation of this film is dependent on the
pollution in the air, in rural areas where there may be little or no
pollution a longer time may be required to form this film. In areas that
have a high pollution level of SO2, loose rust particles are formed
with a much higher corrosion rate. This film of loose particles offers
little or no protection against continued corrosion.
When chlorides are present, such as in a marine environment, the
protective film will not be formed. Under these conditions corrosion
rates of the weathering steels are equivalent to those of unalloyed
carbon steel.
In order to form the patina a series of wet and dry periods is required.
If the steel is installed in such a manner as to be sheltered from the
rain, the dark patina does not form. Instead a rust lighter in color
forms which provides the same resistance. The corrosion rate of the
weathering steels will be the same as the corrosion rate of unalloyed
steel when it is continuously exposed to wetness, such as in water or
soil.
Since the patina formed has a pleasant appearance, the weathering
steels can be used without the application of any protective coating of
antirust paint, zinc, or aluminum. This is particularly true in urban or
rural areas.
In order to receive the maximum benefit from the weathering steels,
consideration must be given to the design. The design should
eliminate all possible areas where water, dirt, and corrosion products
can accumulate. When pockets are present the time of wetness
increases, which leads to the development of corrosive conditions.
The design should make maximum use of exposure to the weather.
Sheltering from rain should be avoided.
While the protective film is forming, rusting will proceed at a
relatively high rate, during which time rusty water is produced. This
rusty water may stain masonry, pavements, and the like.
Consequently, steps should be taken to prevent detrimental staining
effects, such as coloring the masonry brown, so that any staining will
not be obvious.
Page 32
C
Zinc
Galvanized steel (zinc coating of steel) is used primarily in rural or
urban atmospheres for protection from atmospheric corrosion.
Galvanizing will also resist corrosion in marine atmospheres
providing saltwater spray does not come into direct contact. In areas
where SO2 is present in any appreciable quantity, galvanized surfaces
will be attacked.
D
Aluminum
Except for aluminum alloys that contain copper as a major alloying
ingredient, these alloys have a high resistance to weathering in most
atmospheres. When exposed to air, the surface of the aluminum
becomes covered with an amorphous oxide film that provides
protection against atmospheric corrosion, particularly that caused by
SO2.
The shiny metal appearance of aluminum gradually disappears and
becomes rough when exposed to SO2. A gray patina of corrosion
products forms on the surface. If aesthetics are a consideration, the
original surface luster can be retained by anodizing. This anodic
oxidation strengthens the oxide coating and improves its protective
properties.
It is important that the design utilizing aluminum eliminate rainsheltered pockets on which dust and other pollutants may collect. The
formation of the protective film will be disturbed and corrosion
accelerated by the presence of these pollutants.
E
Copper
When exposed to the atmosphere over long periods of time copper
will form a coloration on the surface known as patina, which in reality
is a corrosion product that acts as a protective film against further
corrosion. The length of time required to form the patina depends on
the atmospheres because the color is due to the formation of copper
hydroxide compounds. Initially the patina has a dark color, which
gradually turns green. In urban or industrial atmospheres the
compound is a mixture of copper/hydroxide/sulfate and in marine
atmospheres a mixture of copper/hydroxide/chloride. It takes
approximately 7 years for these compounds to form. When exposed to
clean or rural atmospheres tens or hundreds of years may be required
to form the patina.
The corrosion resistance of copper is the result of the formation of this
patina or protective film. Copper roofs are still in existence on many
castles and monumental buildings that are hundreds of centuries old.
Page 33
F
Nickel 200
When exposed to the atmosphere a thin corrosion film (usually a
sulfate) forms dulling the surface. The rate of corrosion is extremely
slow but will increase as the SO2 content of the atmosphere increases.
When exposed to marine or rural atmospheres the corrosion rate is
very low.
G
Monel Alloy 400
The corrosion of Monel is negligible in all types of atmospheres.
When exposed to rain a thin graygreen patina forms. In sulfurous
atmospheres, a smooth brown adherent film forms.
H
Inconel Alloy 600
In rural atmospheres Inconel alloy 600 will remain bright for many
years. When exposed to sulfur-bearing atmospheres a slight tarnish is
likely to develop. It is desirable to expose this alloy to atmospheres
where the beneficial effects of rain in washing the surface and sun and
wind in drying can be utilized. It is not recommended to design on the
basis of a sheltered exposure.
Page 35
3
The Corrosion of Carbon and Low-Alloy Steels
Gary N. Kirby
Kirby Corrosion Control, Inc.
Brielle, New Jersey
I
Overview
This chapter will cover the corrosion of steels in a wide range of
corrosive environments. Both overall corrosion and localized forms of
corrosion will be included. In each catergory, the behavior of carbon
steel will be compared with that of low-alloy steels.
A
Metallurgical Definitions
As the name implies, a carbon steel owes its properties chiefly to the
presence of carbon, without substantial amounts of other alloying
elements. However, manganese is present to improve notch toughness
at low temperatures. The steels discussed in this chapter contain less
than 0.35% carbon in order to make them weldable.
Low-alloy steels are of two types:
1. Weathering steels, which contain small additions of copper,
chromium, and nickel to form a more adherent oxide during
atmospheric exposures. An example is U.S. Steel's Cor-Ten steel.
2. Hardenable steels, which offer higher strength and hardness after
proper heat treatment and which contain additions of chromium or
Page 36
molybdenum and possibly nickel. Common examples include 4130
and 4340 steels.
B
Overall or General Corrosion
Carbon and low-alloy steels are primarily affected by overall or
general corrosion, also known as wastage. The natural occurrence of
iron is in the form of various oxides, which we use as ores for refining
to make steel. Steels in atmospheric service therefore tend to return to
their oxide form, by the process we call rusting.
The corrosion of steel is the most common form of corrosion the
general public sees, but it is very complex, having over a dozen
variables determining the rate of corrosion, as will be covered below.
Of course, the most common corrosive solvent is water, in everything
from dilute solutions to concentrated acids and salt solutions, but
organic systems may cause serious corrosion too, as will also be
explained below.
C
Localized Corrosion
Localized corrosion involves discrete attack on a metal's surface, with
the surrounding alloy surface virtually unaffected.
1
Stress Corrosion Cracking
Stress corrosion cracking (SCC) is a form of environmentally assisted
cracking caused by tensile stresses (residual or operational) plus
certain chemical species, resulting in fine, branched cracking
characteristic of SCC. SCC is more common at elevated temperatures.
Chemicals that cause SCC in steels include alkalies (e.g., NaOH),
nitrates, dry ammonia, hydrogen sulfide, and hydrogen gas. One big
advantage of plain carbon steel is that SCC can be prevented by stress
relieving after fabrication, without harmful precipitation, such as the
chromium carbide precipitation that occurs during stress relief of
stainless steels and that sensitizes the alloys to severe corrosion in
areas depleted of chromium. Of course, with some large or complex
steel shapes, stress relief may not be feasible. For high-strength, heattreatable alloys, the heat treatment can be selected to lower the alloy's
susceptibility to SCC.
2
Pitting
Immersed steels may pit under low-flow or stagnant conditions. The
pits are generally shallow as, for example, in seawater, where the
pitting rate is 545 mpy (mils per year, where one mil = 0.001 in. or
25.4 µm), while the overall corrosion rate in the aerated splash zone is
as high as 17 mpy (430 µm/year) [1].
Page 37
3
Hydrogen Effects
Under various conditions, hydrogen gas can embrittle, blister, crack,
or decarburize steels, the effect being generally more pronounced the
higher the strength of the steel.
4
Corrosion Fatigue
Prolonged cyclic loading of steels can induce fatigue failures. If
conditions are corrosive, the fatigue problem is worse, sometimes
involving corrosion deposits accumulating in the cracks to concentrate
the cycling stresses. About 10 years ago, corrosion fatigue failures in
boiler feedwater deaerators around the United States caused several
deaths [2].
D
Microbiologically Influenced Corrosion (MIC)
There are many bacteria that accumulate in water systems, then grow
on metal surfaces in colonies and change the localized chemistry in
the colonies to highly corrosive conditions. Stagnant hydrotest water
is a frequent cause of this corrosion.
II
Corrosion Data
A
Overall Corrosion in Aqueous Systems
In dilute water solutions, the most important variable is acidity or
solution pH. Figure 1 [3] shows the effect of pH on the corrosion of
steel at 22°C (72°F) and 40°C (104°F). This diagram is suitable for
water flowing at a moderate flow rate.
Figure 1
Effect of pH on the corrosion of carbon steel.
(From Ref. 3.)
Page 38
There is a range of pH from 5.5 to 10 where the corrosion rate is
constant at about 1012 mpy (250305 µm/year). In this range there is
an alkaline solution of saturated ferrous hydroxide covering the steel's
surface, this hydroxide solution having a constant pH of about 9.5 [4].
The rate-determining reaction in this corrosive range is the diffusion
of oxygen through the ferrous hydroxide film to feed the
electrochemical cathodic reduction of the oxygen to hydroxyl ion.
Thus dissolved oxygen is another key variable in aqueous corrosion.
At lower pH values the cathodic reaction changes to the relatively
rapid reduction of hydrogen ions in the acidic solution to produce
hydrogen gas bubbles. Different acids have different values of pH
where the onset of this rapid corrosion occurs. As shown, carbonic
acid (dissolved carbon dioxide) initiates it at pH 5.5. Hydrochloric
acid starts it at pH 4. The effect is dramatic; at pH 2.7, the corrosion
rate reaches 80 mpy (2 mm/year).
Under stagnant conditions, the overall corrosion rate is lower. The
author has tested carbon steel in about 40 waters, including tap water,
well waters, seawater, and chemical waste-waters at ambient
temperatures, and the corrosion rate was about 35 mpy (75125
µm/year) for all samples. However, stagnant conditions are generally
to be avoided since they are exactly where the various forms of
localized corrosion become serious, including pitting, oxygen
concentration cells, and microbiologically influenced corrosion
(MIC). These localized corrosions penetrate faster than overall or
general corrosion.
Figure 2 shows more effects on carbon steel from varying solution pH
[5]. This shows that for water alone (in this case, distilled water),
lowering the pH from 4.5 to 3.7 raised the corrosion rate in this test,
from less than 2 mpy (0.05 mm/year) up to over 16 mpy (0.4
mm/year). But notice that this pH effect is not apparent for the 0.5 M
sodium sulfate solution, so that some strong salt solutions may not be
strongly pH-dependent in their corrosiveness, at least over narrow pH
ranges.
Figure 2
Corrosion rates of steel in oxygen-free water containing
carbon dioxide at 25°C (77°F), with and without 0.5 M
sodium sulfate.
(From Ref. 5.)
Page 39
The next important variable to consider is fluid flow rate. Figure 3
shows the effect of flow on the corrosion rate of steel from stagnant to
8 feet/s. Note that as the flow rate rises from zero, the corrosion rate
increases to a maximum around 1 or 2 feet/s. This increase comes
from an increase in the oxygen supplied for the oxygen reduction
process occurring on the cathodic areas of the steel. Higher flow rates
then supply enough oxygen so that adsorbed oxygen and the ferrous
hydroxide layer can cover the entire steel surface, a complete level of
passivation. At 68 feet/s (1.82.4 m/s), which is a common range of
flow rates in the chemical industry, the corrosion rate settles at 1015
mpy (250380 µm/year). Figure 3 also shows the effect of surface
roughness of the steel, another variable affecting corrosion.
With higher flow rates, the corrosion rate increases up to around 40
feet/s (12 m/s), where the attack changes to erosion-corrosion, which
means that any protective oxide or adsorbed layer is stripped away
and bare steel is open to accelerated attack. Turbulence has a similar
effect. Figures 4 and 5 show the effects of increasing flow velocity for
distilled water and seawater. At 39 feet/s, the corrosion rate in
distilled water at 50°C (122°F) exceeds 200 mpy (5 mm/year).
With corrosion, as with any other chemical reaction, temperature
plays a major role. Figure 1 shows the increase in corrosion from
increasing temperature. In neutral or alkaline waters, however, the
temperature effect is more complicated. In an open system, higher
temperatures will drive off oxygen, eventually to very low levels.
Since oxygen provides the cathodic reaction in the corrosion process,
if there is no oxygen there will be little corrosion. Figure 6 shows this
effect, with the corrosion beginning to decrease around 80°C (176°F)
and becoming very low above 100°C (212°F).
Figure 3
Effect of water flow velocity on the corrosion of steel. Increased
oxygen leads first to higher corrosion, then to oxygen
passivation that lowers corrosion.
(From Ref. 23.)
Page 40
Figure 4
Effect of pH of pure water on erosion corrosion of steel at
50°C at flow velocity of 39 feet/s.
(From Ref. 24.)
The behavior of weathering low-alloy steels in aqueous corrosion tests
and applications is unpredictable. In 1953 early tests results on
weathering steels containing copper, chromium, phosphorus, and
nickel showed superior corrosion resistance in immersion tests in river
water at a pH of 3.54.0 [6]. After 4 years immersion, for example,
Cor-Ten steel showed an average corrosion rate of 2.7 mpy (69
µm/year), with a maximum pit depth of 14 mils (360 µm). For
Figure 5
Effects of seawater velocity on corrosion of steel (ambient
temperature).
(From Ref. 25.)
Page 41
Figure 6
Effect of temperature on corrosion of iron in water containing
dissolved oxygen.
(From Refs. 26 and 27.)
comparison, carbon steel corroded at an average of 4.1 mpy (104
µm/year) and pitted to a maximum of 21 mils (530 µm).
However, in the laboratory tests described above, involving carbon
dioxide and sodium sulfate, low-alloy steels performed worse than
carbon steel. Three alloy steels containing phosphorus, copper,
chromium, and nickel in various permutations corroded severely at
4060 mpy (1.01.5 mm/year) in 0.5 M sodium sulfate containing 100
ppm oxygen with carbon dioxide bubbling through at 25°C (77°F).
For comparison, four carbon steels corroded at rates of only 2 mpy
(50 µm/year).
Such inconsistencies explain why one power plant the author visited
in Pennsylvania had a high regard for weathering steels in a wide
variety of services, while a power plant in Florida had disappointing
results with a weathering steel in a water immersion service. One
solution is to test weathering steels in prospective immersion services,
using National Association of Corrosion Engineers (NACE)
guidelines [7].
Table 1 shows corrosion data for several low-alloy steels compared
with carbon steels [8]. The table includes results for solutions
containing 200 ppm propionic acid and 200 psig carbon dioxide at
130°F (54°C). Again as in Fig 1, carbonic acid can be more
aggressive than other acids, in this instance, exceeding an order of
magnitude in some cases.
In Table 1, with the exception of some of the Cr-Mo steels in the
propionic acid tests, the alloy steels are generally more resistant than
the carbon steels, especially the 9% chormium steel. This is
reasonable since steels become ''stainless" at the 1113% chromium
level. Furthermore, the higher corrosion
Page 42
Table 1 Overall Corrosion Rate, 130°F (54°C)
(During Indicated Exposure Time), mils/year
Carbonic acid Propionic acid
(200 psig)
(200 ppm)
Alloy
7 Days 70 Days 7 Days 70 Days
Carbon steels
API J-55
62
5.8
1.8
4.6
API H-40
60
9.1
1.1
3.1
API N-80
77
5.5
1.5
4.7
Alloy Steels
2.25 Cr-1 Mo
59
3.0
3.1
3.3
5 Cr-0.5 Mo
46
4.2
4.2
3.3
9 Cr-1 Mo
2
0.2
0.5
0.6
3.5 Ni
28
3.1
1.5
1.5
5 Ni
26
2.1
1.3
1.6
9 Ni
27
1.2
1.0
1.7
rates for the other Cr-Mo steels in the propionic acid solution are still
in a usable range for many applications.
The large decrease in corrosion rates in carbonic acid between 7 and
70 days is attributed to the eventual formation of protective surface
films, which may be fragile under certain fluid flow conditions.
1
Atmospheric Corrosion
Atmospheric corrosion is a factor in most applications of alloy steels,
presenting a wide variety of problems in various industries [9]. Figure
7 summarizes test results in industrial or semiindustrial environments,
comparing plain carbon steel with structural copper steel and highstrength, low-alloy (HSLA) steels. In such environments the alloy
steels are clearly better than plain carbon steel. Table 2 lists average
reductions in thickness for various steels in several environments [9].
Again, the alloy steels show superiority.
B
Localized Corrosion in Aqueous Systems
1
Stress Corrosion Cracking
In 1979 the author was called in to study the cracking of the top girth
weld in a vertical high-pressure caustic fusion autoclave in the
dyestuff industry. The crack
Page 43
Figure 7
Atmospheric corrosion in a semi-industrial or industrial environment.
(From Ref. 9.)
was detected during inspection and estimated by ultrasonic testing as
1 in. deep into the vessel wall, which was over 4 in. thick. A white
powdery film covered this weld and the surrounding plate material.
The vessel was made from ASME SA-515 (grade 70) pressure vessel
plate steel and saw service in caustic soda and hydrated lime at
temperatures up to 250°C (480°F). The cause of the cracking was
caustic stress corrosion cracking (SCC) induced by vaporous NaOH
deposition and residual weld stresses plus bending stresses from a
vessel support ring. The vessel was repaired by gouging out the
cracked material and rewelding with a low-hydrogen filler metal and
then stress-relieving in place. Eventually the vessel was replaced with
a nickel-clad steel autoclave, nickel being substantially immune to
caustic SCC.
Figure 8a shows a temperature-concentration plot for collected
industrial service conditions where caustic SCC has been observed
[10]. This has led to guidelines [11] for stress-relieving carbon steel
welds in caustic service (Fig. 8b), the purpose being to avoid cracking
such as in the autoclave described above.
The susceptibility of low-alloy steels to SCC depends on their strength
level. General tensile strength levels for alloy steels for resistance to
SCC in steels such as AISI 4130 and 4340 can be given as [12]
1. High SCC resistance: tensile strength below 180,000 psi (1240
MPa)
2. Moderate SCC resistance: tensile strength 180,000200,000 psi
(12401380) MPa)
3. Low SCC resistance: tensile strength over 200,000 psi (1380 MPa)
Table 2 Corrosion of Various Steels in Various Environments
Average reduction in thickness
Carbon A242 (K11510) A588 (K11430) Crsteel
Cu-P Steel
V-Cu Steel
Environment
Exposure mils µm
mils
µm
mils
time (yr)
Urban industrial
3.5
3.3 84
1.3
33
1.8
7.5
4.1 104
1.5
38
2.1
Rural
3.5
2.0 51
1.1
28
1.4
7.5
3.0 76
1.3
33
1.5
Severe marine (25 m/80 ft
0.5
7.2 183
2.2
56
3.8
from ocean)
2.0
36.0 914
3.3
84
12.2
3.5
57.0 1448
28.7
5.0
Destroyed 19.4
493
38.8
Chloralkali plant
0.5
4.1 104
2.4
61
2.7
2.0
18.8 478
5.7
145
7.4
Sulfur plant
0.5
15.5 394
7.4
188
9.4
2.0
43.3 1100 20.4
518
32.4
Sulfuric acid plant
0.5
3.3 84
1.8
46
1.9
2.0
8.9 226
3.0
76
3.3
Chlorinated hydrocarbon
0.5
5.4 137
1.8
46
1.8
plant
2.0
44.1 1120 4.1
104
4.6
Hydrochloric acid plant
0.5
12.3 312
5.8
147
7.1
2.0
49.8 1265 25.2
640
31.6
Page 45
Figure 8
SCC of carbon steel in caustic soda (NaOH)
at elevated temperatures. (a) Industrial data.
(From Ref. 10.)
(b) NaOH service graph to avoid SCC of
carbon steel.
(From Ref. 11.)
Note that the carbon-silicon steel that cracked in the caustic fusion
vessel described above had a nominal tensile strength of 70,000 psi
(483 MPa)far below these guidelines for higher alloy steels. For such
plate steels, stress relief is still an option to prevent SCC, with
guidelines available in the ASME Boiler and Pressure Vessel Code.
Chemical species that induce SCC in carbon or low-alloy steels, even
at low concentrations, include [12]
Hydroxides, gaseous hydrogen
Gaseous chlorine, HCl, and HBr
Hydrogen sulfide gas, Mns, and MnSe inclusions in alloy
Aqueous nitrate solutions
As, Sb, and Bi ions in aqueous solutions
Carbon monoxidecarbon dioxidewater gas mixtures
Many of these chemical systems will crack a steel at room
temperatures. Table 3 lists some cracking test results at 40°C (104°F)
[13].
Another chemical that causes SCC in steels is anhydrous ammonia
[14,15]. Alloys affected include carbon steel in storage tanks and
ASTM A517 quenched and tempered steel in motor vehicle cargo
tanks. Various grades of A517 steel contain small amounts of Cr, Ni,
Mo, B, V, Ti, Zr, and Cu. This cracking can be alleviated by (1)
adding 0.2% water to the ammonia, (2) eliminating air contamination
from ammonia systems, and (3) stress-relieving tanks or fabricating
with hot-formed or stress-relieved heads.
Page 46
Table 3 SCC of Mild (Carbon) Steel and
Low-Alloy Steel U Bends in CO/CO2
Mixture at 40°C (104°F)
Test 1 Test 2 Test 3
Test condition
79.9
79.9
79.9
CO (%):
20.1
20.1
20.1
CO2 (%):
1.0
1.0
Air (kg/cm2):
16
16
21
Total P
(kg/cm2):
1
2
Time/weeks: 2
Results
Liquid phase
No
Crack Crack
Mild steel
crack
Crack Crack
1 Cr-0.5 Mo
Crack Crack
2.25 Cr-1 Mo No
crack
Crack
5 Cr-0.5 Mo
No
9 Cr-1 Mo
crack
Vapor phase
Crack No
Mild Steel
crack
Crack No
1 Cr-0.5 Mo
crack
2.25 Cr-1 Mo Crack No
crack
2
Pitting
In the corrosive systems cited above involving 200 psi carbon dioxide
at 130°F (54°C), the test steels also showed considerable pitting.
Table 4 gives the distribution and maximum depth of pits after 14
consecutive 7-day exposures, with coupon cleaning between the 7-day
periods [16]. The authors postulate that the frequent cleaning probably
decreased the maximum pit depth and increased the number of pits
per unit area.
The Corrosion Data Survey (Metals Section, 6th ed.), published by
NACE in 1985, lists only five chemicals besides carbonic acid as
causing pitting in carbon steel (Table 5). This is from a sample of
around 900 chemicals, perhaps suggesting that the shallow pitting
characteristic of carbon steels is not usually a major problem distinct
from overall corrosion. However, the corrosion fatigue of deaerators
cited above [2] frequently involved deep pitting that initiated cracking
throughout the vessels. The inclusion in the NACE data of two
amines, which can be corrosion inhibitors, suggests that pitting in
steels can sometimes be a matter
Page 47
Table 4 Number and Depth of Pits in Most Severely Pitted cm2
200 psi Carbon Dioxide at 130°F
No. of pits/cm2 at listed depth for
various depths of pits (mils or
0.001 in.)
Maximum pit
Alloy
0.63.0 3.16.06.19.09.112.12.115. depth (mils)
Carbon
steels
API J-55
2 pits 17
10
2
0
12
API H-40
9
4
14
5
3
15
API N-80
32
8
0
0
0
7
Alloy steels
2.25 Cr-1
1
2
0
0
0
4
Mo
0
0
0
0
0
5 Cr-0.5 Mo 0
0
0
0
0
9 Cr-1 Mo
3.5 Nickel
9
28
10
1
1
15
5 Nickel
9
20
6
0
0
17
9 Nickel
14
13
5
0
0
15
Weathering
0
6
5
10
4
14
steel
of incomplete corrosion inhibition at inhibitor concentrations that are
inadequate to coat the entire surface. Thus, the inhibitor's stable,
protective surface film may be incomplete, leaving certain discrete
anodic unprotected areas severely prone to pitting. This pitting is
driven by the large surrounding cathodic or protected areas. Such
pitting is known to occur with anodic inhibitors such as chromates,
where inadequate dosages leave unprotected anodes susceptible to
pitting.
One more environment that pits steel is soil, which obviously affects
buried
Table 5 Chemicals Causing Pitting in Steels
Allylamine
Antimony trichloride
Carbonic acidcarbon dioxide
Epichlorohydrin
Methallyiamine
Nickel nitrate
Note: No entry for seawater.
Source: NACE Corrosion Data Survey, 6th ed.,
Metals Section, 1985.
Page 48
pipelines. In one study [9] of 10 carbon and alloy steels containing Cr,
Ni, Cu, and Mo and exposed to a variety of soils for 13 years, the
conclusion was that factors such as soil pH, resistivity, and degree of
aeration have more influence on the severity of corrosion than the
alloy content of the steel. In any case, protective coatings and cathodic
protection are the best means of reducing the corrosion of buried
pipelines.
3
Hydrogen Damage
The body-centered crystal structures of carbon and low-alloy steels
are susceptible to four types of hydrogen damage, two of which are
low-temperature processes and two high-temperature:
Low-temperature
Hydrogen blistering
Hydrogen embrittlement
High-temperature
Decarburization
Hydrogen attack
The diffusion of hydrogen through steels to harm mechanical
properties involves atomic or nascent hydrogen since molecular
hydrogen cannot diffuse through metals [17]. Common sources of
atomic hydrogen include corrosion (including the acid pickling of
steel), misapplied cathodic protection, high-temperature moist
atmospheres, electroplating, and welding.
Hydrogen Blistering
During some acid services, such as the acid pickling of steels,
hydrogen atoms may penetrate the crystal lattice and collect in
fissures or cavities in the steel. These atoms then combine into
hydrogen gas molecules, eventually reaching pressures of several
hundred thousand atmospheres and forming blisters on the steel's
surface. In petroleum process streams this problem is promoted by socalled hydrogen evolution poisons such as sulfides, arsenic
compounds, cyanides, and phosphorus-containing ions. In closed
systems like pickling operations, chemical inhibitors are added to the
acid to reduce the hydrogen penetration.
Hydrogen Embrittlement
Another harmful effect of hydrogen penetration of steel is
embrittlement, which is a more complicated metallurgical effect
possibly involving the interaction of hydrogen atoms with the tip of an
advancing crack. For low-alloy steels the alloys are most susceptible
in their highest strength levels. Alloys containing nickel or
molybdenum are less susceptible.
If hydrogen is initially present in a steel, for example, from
electroplating, the hydrogen can be baked out. In fact, this
embrittlement decreases with increasing service temperature,
especially above 150°F (65°C). Generally, hydrogen embrittlement is
not usually a problem in steels with yield strengths below about 1000
MPa (150 ksi), but if hydrofluoric acid or hydrogen sulfide is present,
the yield strength must be below 550 MPa (80 ksi) for good
resistance. Welding
Page 49
conditions should be dry and low-hydrogen filler metal should be used
to minimize hydrogen embrittlement.
Decarburization
The hardness and strength of a steel depends on its carbon content. A
loss of carbon (decarburization) lowers the tensile strength of steels
and can be caused by moist hydrogen at high pressures and high
temperatures. Figure 9 shows the Nelson diagram [18] depicting the
limits of service conditions for carbon and alloy steels in hydrogen
services.
Hydrogen Attack
High-temperature hydrogen attack refers to a reaction between
hydrogen and a component of the alloy [19]. For example, in steels
hydrogen reacts with iron carbide at high temperatures to form
methane gas, according to the following reaction:
Because methane cannot diffuse out of steel, it accumulates and
causes fissuring and blistering, thereby decreasing alloy strength and
ductility. Alloy steels containing chromium and molybdenum are
beneficial in such services because the carbides formed by the
alloying elements are more stable than iron carbide and therefore
resist hydrogen attack. It is noteworthy that water vapor and carbon
dioxide at high temperatures can also decarburize steels [19].
4
Corrosion Fatigue
As the name implies, corrosion fatigue is affected by both the severity
of corrosive conditions and mechanical, cyclical stress factors. Stress
raisers such as
Figure 9
Schematic of hydrogen damage for low-alloy steels in hydrogen service.
Decarburization and hydrogen attack above alloy lines.
(From Ref. 18.)
Page 50
notches, holes, weld defects, or corrosion pits can initiate fatigue
cracks, and a corrosive environment can reduce crack initiation time
[20]. For many materials the stress range required to cause fatigue
failure diminishes progressively with increasing time and with the
number of cycles of applied stress. In the case of the carbon steel
deaerators cited previously [2], major observations about the cracking,
which occurred over 827 years of service, included the following:
1. Deep corrosion pits along the welds initiated many cracks.
2. Residual weld stresses and localized stresses (e.g., bending
stresses) contributed to cracking.
3. Stress-induced cracking also occurred.
4. The worst cracks were located in circumferential and head-to-shell
welds in horizontal vessels.
5. Stress relief and designing to reduce localized stresses were
supported.
In many cases NDE was used to find the cracks, by ultrasonic testing
and wet, fluorescent magnetic particle testing.
C
Microbiologically Influenced Corrosion
Under certain conditions, bacterial colonies change the chemistry of
an alloy's surface and induce rapid corrosion. One common example
involves sulfatereducing bacteria (SRB), which produce acidic
hydrogen sulfide, which is highly corrosive to steel. Then other
bacteria may act on the hydrogen sulfide and produce sulfuric acid
[21,22]. Other instances of microbiologically influenced corrosion
(MIC) report corrosion due to the formation of acetic and formic acids
[22]. These acids may be highly concentrated and cause pitting and
rapid failure in the "slime," nodules, or tubercles where the bacteria
live.
MIC is generally associated with stagnant or low-flow aqueous
systems over a range of pH values from 110.5, at temperatures of 0°C
(32°F) to 100°C (212°F). There are 5060 bacteria species believed to
be associated with MIC, both aerobic and anaerobic species. The
corrosive attack can be rapid, often occurring within weeks of
introducing the bacteria. One solution that suffered MIC involved
dissolved polymeric organic materials and ammonium phosphate
compounds, so the affected solutions can be quite varied.
There are various treatments used to prevent or alleviate MIC,
including adding biocides such as ozone or hydrogen peroxide to the
water. This, however, will be ineffectual if the bacteria have
previously formed protective nodules as their habitat. These nodules
must be mechanically removed in order to kill the bacteria.
Bacteria must always be assumed to be present in untreated water, so
it is obvious that untreated hydrotest water should be removed from a
system or vessel
Page 51
as soon as possible. If this is not possible, keep the water flowing at
velocities over 5 feet/s (1.5 m/s) since bacteria require low-flow
conditions to colonize.
D
Organic Corrosives
There are four categories of organic compounds that can be corrosive
to steels:
1. Organic acids, such as acetic or formic acid.
2. Compounds that hydrolyze to produce acids. This includes
chlorinated hydrocarbons such as carbon tetrachloride or
trichloroethane, which react with water to produce hydrochloric acid.
Another such compound is dimethyl sulfate, which hydrolyzes to
make sulfuric acid.
3. Chelating agents, which take up or combine with transition
elements. The author once tested steel coupons in an aqueous solution
containing 5000 ppm of an ammonium salt of
ethylenediaminetetraacetic acid (EDTA) at a pH of 8.5 at 204°C
(400°F) and 200 psig. The corrosion rate was 110 mpy (2.8 mm/year),
which explained the disappearance of steel packing in a packed
column. Remember, sometimes the organic product molecule can be a
chelating agent!
4. Inorganic corrosives dissolved and dissociated in organic solvents.
This may include such combinations as hydrochloric acid dissolved in
methanol or sulfuric acid dissolved in dimethylformamide. Other
candidates include chlorine, bromine, or iodine dissolved in methanol.
Test if in doubt, especially if the consequences of failure are great.
References
1. Guidelines for Selection of Marine Materials, International Nickel
Company, New York, 1971, p. 4.
2. R. J. Franco and G. M. Buchheim, Case histories of deaerator
failure analysis, Mat. Perform. 25(10):9 (1986).
3. G. W. Whitman, R. P. Russell, and V. J. Altieri, Effect of
hydrogen-ion concentration on the submerged corrosion of steel,
Indust. Eng. Chem. 16, No. 7, 1924, p. 665.
4. H. H. Uhlig and R. W. Revie, Corrosion and Corrosion Control,
3rd Ed., John Wiley and Sons, New York, 1985, p. 96.
5. G. Schmitt and B. Rothmann, Corrosion of unalloyed and low
alloyed steels in carbonic acid solutions, Carbon Dioxide Corrosion
in Oil and Gas Production, (NACE Task Group T-1-3, ed.) NACE,
Houston, 1984, p. 167.
6. C. P. Larrabee, Corrosion resistance of high-strength low-alloy
steels as influenced by composition and environment,
CorrosionNACE, 9:259 (1953).
7. TM-01-69-76, Standard Test Method, Laboratory Corrosion
Testing of Metals for the Process Industries, Revised 1976, NACE,
Houston.
8. D. W. DeBerry and W. S. Clark, Corrosion due to use of carbon
dioxide for enhanced
Page 52
oil recovery, Carbon Dioxide Corrosion in Oil and Gas Production
(NACE Task Group T-1-3, ed.), NACE, Houston, 1984, p. 22.
9. T. G. Oakwood, Corrosion of alloy steels, Metals Handbook, 9th
ed., Vol. 13, Corrosion, ASM International, Metals Park, OH, 1987,
pp. 531546.
10. H. W. Schmidt, P. J. Gegner, G. Heinemann, C. F. Pogacar, and E.
H. Wyche, Stress corrosion cracking in alkaline solutions, Corrosion,
Vol. 7, NACE, 1951, p. 295.
11. Corrosion Data Survey, 6th ed., Metals Section, NACE, Houston,
1985, p. 176.
12. J. D. Wood, Environmental assisted cracking, Industrial
Corrosion (Course Notes), Center for Professional Advancement,
East Brunswick, NJ, 1994, Section N, pp. 8, 11.
13. M. Kowaka and S. Nagata, Stress corrosion cracking of mild and
low alloy steels in CO-CO2-H2O environments, Carbon Dioxide
Corrosion in Oil and Gas Production (NACE Task Group T-1-3, ed.),
NACE, Houston, 1984, pp. 228, 229.
14. A. W. Loginow, Detection and diagnosis of stress corrosion
cracking in ammonia tanks, Mat. Perform., 15(6):33 (1976).
15. A. W. Loginow, A review of stress corrosion cracking of steel in
liquified ammonia service, Mat. Perform., 26(12):18 (1986).
16. D. W. DeBerry and W. S. Clark, Ref. 8, p. 34.
17. M. G. Fontana, Hydrogen damage, Corrosion Engineering, 3rd
ed., McGraw-Hill, New York, 1986, p. 143.
18. Steels for Hydrogen Service at Elevated Temperatures and
Pressures in Petroleum Refineries and Petrochemical Plants, API
Publication 941, 4th ed., American Petroleum Institute, Washington,
D C, 1990.
19. M. G. Fontana, Decarburization and hydrogen attack, Corrosion
Engineering, 3rd ed., McGraw-Hill, New York, 1986, pp. 529534.
20. D. O. Sprowls, Evaluation of corrosion fatigue, Metals Handbook,
9th ed., Vol. 13, Corrosion, ASM International, Metals Park, OH,
1987, p. 291.
21. J. D. Wood, Microbial corrosion, Industrial Corrosion (Course
Notes), Center for Professional Advancement, East Brunswick, NJ,
1994, Section Q, p. 1.
22. D. H. Pope and J. G. Stoecker, Microbiologically influenced
corrosion, Process Industries Corrosion: The Theory and Practice (B.
J. Moniz and W. I. Pollack, eds.), NACE, Houston, 1986, p. 227.
23. R. P. Russell, E. L. Chappell, and A. White, Effect of velocity on
corrosion of steel under water, Indust. Eng. Chem., p. 67, Jan. 1927.
24. M. G. Fontana, Erosion corrosion, Corrosion Engineering, 3rd
ed., McGraw-Hill, New York, 1986, pp. 91108.
25. Guidelines for Selection of Marine Materials, International Nickel
Company, New York, 1971, p. 6.
26. F. Speller, Corrosion: Causes and Prevention, McGraw-Hill, New
York, 1951, p. 168.
27. H. H. Uhlig and R. W. Revie, Iron and steel, Corrosion and
Corrosion Control, John Wiley and Sons, New York, 1985, pp.
90122.
Page 53
4
Corrosion of Stainless Steels
Paul K. Whitcraft
Rolled Alloys, Inc.
Temperance, Michigan
Annual stainless steel consumption in United States is approaching 2
million metric tons. It is used in a wide variety of household items in
addition to being a significant factor in industrial process equipment.
Consumption in automobiles has continued to increase and now
averages over 30 pounds per vehicle. Uses for these materials can be
found in nearly every industry.
Worldwide production of stainless steel exceeds 12.5 million metric
tons. Stainless steel production in the early 1900s was zero when the
first trials of adding chromium to mild steel were conducted. The
original efforts in this area were presumably based on the observation
that chromium-plated steel parts were highly corrosion-resistant. The
end result was the introduction of the ferritic family of stainless steels.
The first documentation of the development of this class of steel
began to appear in the 1920s. The first American Society for Testing
and Materials (ASTM) specifications for stainless steels were
published in 1935.
Today there are seven basic families of stainless steels with
compositions that contain 1133% chromium, 038% nickel, and 07%
molydbenum as the major alloying elements. These families are
Ferritic
Austenitic
Precipitation hardenable
Page 54
Superferritic
Martensitic
Duplex (ferritic-austenitic)
Superaustenitic
I
Families of Stainless Steel
A
Ferritic
Chromium is a metal that readily forms an oxide which is transparent
and happens to be extremely resistant to further degradation. As a
further benefit to alloying with steel, it is less noble than iron and thus
tends to form its oxide first. Increasing the chromium content in steel
gradually above about 2% improves mild atmospheric corrosion
resistance steadily up to a level of about 12% where corrosion is
essentially arrested. For exposure to mild, wet environments the
addition of about 11% chromium is sufficient to prevent ''rusting" of
steel components, hence the term "stainless."
Ferritic stainless steels are magnetic, have body-centered cubic atomic
structures, and possess mechanical properties similar to those of
carbon steel, though less ductile. Continued additions of chromium
will also continue to improve corrosion resistance in more severe
environments. Chromium additions are particularly beneficial in terms
of resistance in oxidizing environments, at both moderate and elevated
temperatures. Addition of chromium is the most cost-effective means
of increasing the corrosion resistance of steel, with chromium costing
less than a dollar per pound. Chromium contents in the ferritic
stainless alloys top out around 28%. These materials are historically
known as "400" series stainless as they were identified with numbers
beginning with 400 when the American Institute for Iron and Steel
(AISI) had the authority to designate alloy compositions. Alloy
identification is now formally handled by the Unified Numbering
System (UNS) whereby stainless alloy identification numbers
generally begin with "S" followed by a five-digit number. Most of the
old AISI designations were retained as the first three digits of the
UNS number such that the old Stainless 405, a basic 12% Cr balance
iron material, is designated UNS S40500.
Table 1 lists some of the more common ferritic stainless steels. Type
430 stainless (UNS S43000) is alloyed with 18% chromium. This is
the next logical step in alloy additions since steady improvement in
corrosion resistance is obtained by increasing the chromium content
over 11%. Incremental additions over 18% become less effective,
particularly for aqueous corrosion. Beyond this, the highest practical
level of chromium content in iron is afforded by type 446
Page 55
Table 1 Selected Ferritic Stainless Steels
CTEa µin./in.
Alloy C Cr Ni
Mo
S405000.0612.0
S409000.0611.0
S430000.0716.5
S430350.0518.0
S443000.1221.0
S446000.0725.5
Room
temp.
yield
Room temp.
Alloy (KSI)
tensile (KSI)
S40500 40
70
S40900 35
60
S43000 45
74
S43900 40
70
S44300 50
90
S44600
50
70
Density
N
Other
bb #/in.3
6.8
600 0.279
Ti-0.4
6.7
600 0.280
6.6
600 0.278
0.03 Ti-0.7
6.6
600 0.280
Cu-1.0
6.7
600 0.277
6.3
600 0.270
Elevated
temp.
Room temp. Toughness ft-lb. strength (KSI
elong. %
@ °F
@ °F)c
30
75 @ RT
30
28
30
22
2 @ 1200°
CRP
25
25 @ RT
2 @ 1200°
CRP
Values are approximate.
aCoefficient of thermal expansion for range of 72°1200° F.
bMagnetic permeability.
cCRP is stress required to produce 1% creep strain in 10,000 hours.
(UNS S44600). This stainless alloy is used primarily for hightemperature oxidation resistance.
1
Stabilization
Examination of the compositions of stainless types 409 and 439
introduce an additional approach to improving corrosion resistance. It
also underscores the importance of carbon in stainless alloys. The role
of carbon as an alloy addition to steel is primarily that of increasing
strength. Increasing carbon content, since it is an interstitial element,
pins the movement of atoms within the matrix resulting in higher
stresses required to cause deformation. This is also a factor in stainless
steels, but increasing carbon content can have a deleterious effect on
corrosion resistance.
During melting and high-temperature working operations, the carbon
content in stainless steel is generally in solid solution, i.e., uniformly
distributed within the steel matrix analogous to sugar in solution in
warm water. As the steel cools from a temperature of around 1600°F,
there is a preference for the formation of a chromium carbide
compound, which precipitates preferentially at grain
Page 56
boundaries. This is somewhat analogous to the sugar water solution as
the stage where cooling has caused the sugar to crystallize and
precipitate to the bottom of the container. The solubility limit of
carbon in austenitic steel is illustrated in Fig. 1. The solubility of
carbon in ferrite is slightly higher.
Chromium carbides in themselves do not suffer from poor corrosion
resistance. The detrimental effect is in the fact that chromium is
depleted from the surrounding matrix. In fact, the chromium depletion
can be so severe as to lower the chromium content locally to below
the 11% content considered to be the minimum for stainless steel. In
actuality, any depletion can be significant if the environment is severe
enough to cause the depleted zone to become anodic to the matrix. In
high-temperature service, even where the component is used at a
temperature which will cause chromium carbide precipitation, grain
boundary chromium depletion is usually not a concern. Due to
diffusion of chromium from within the grain toward the grain
boundary, chromium depletion at elevated temperatures is short-lived.
One way to avoid the precipitation of chromium carbides is to force
the precipitation of another carbide first. Two elements, titanium and
niobium (columbium) are particularly effective. Titanium will tie up
carbon in the ratio of about five times its weight. Niobium is more
efficient tying up about 15
Figure 1
Solubility of carbon in austenite.
Page 57
times its own weight. In both types 409 and 439, titanium is used as
the stabilizer. In other alloys, such as some of "superferritic"
materials, both elements are used since in higher concentrations each
element can produce detrimental side effects.
Ferritic stainless steels offer useful resistance to mild atmospheric
corrosion and most freshwaters. They will corrode with exposure to
seawater atmospheres. These alloys are also useful in hightemperature situations, with 446 exhibiting useful oxidation (scaling)
resistance through about 2100°F. Ferritic materials which contain
more than about 18% chromium are also susceptible to an
embrittlement phenomenon when exposed to temperatures in the
range of 600°1100°F. This is due to the formation of a secondary
phase and is termed 885° embrittlement after the temperature which
causes the most rapid formation. These materials are not brittle in this
temperature range but lose ductility when cooled to room temperature.
B
Martensitic
Within a certain range of compositions based on ferritic stainless
steels, as indicated by the adjacent diagram (Fig. 2) developed by
Schaeffler, martensitic structures can be developed. These alloys are
hardenable because of the phase transformation from body-centered
cubic to body-centered tetragonal. As with the low-alloy steels, this
transformation is thermally controlled.
Figure 2
Schaeffler constitutional diagram for stainless steels.
Page 58
The corrosion resistance of the martensitic steels is again dependent
solely on chromium, and since the carbon contents are generally
higher than the ferritic alloys, they are less corrosion-resistant.
Nevertheless, the combination of useful corrosion resistance in mild
environments coupled with high strengths makes the martensitic
family of materials a significant portion of the stainless realm. Alloys
in this family are shown in Table 2. Type 410 (S41000) stainless,
which is hardenable to about Rockwell C 32, is a popular choice for
many components requiring a combination of corrosion resistance,
strength, and toughness. Type 440C (S44000), which is the highest
carbon version of this 18% chromium stainless, is hardenable to
Rockwell C 60.
Figure 3 illustrates the general relationship between the ferritic and
martensitic families of stainless. In addition, two other families of
magnetic stainless steels are represented in the figure and will be
discussed in detail later in this chapter.
Table 2 Selected Martensitic Stainless Steels
Alloy
C
S41000 0.1
S41600 0.12
S42000 0.22
S43100 0.11
S44002 0.65
S44004 1.0
Cr
12.0
12.0
13.0
16.0
17.0
17.0
Ni
Mo N Other
S-0.2
1.5
CTEa
µin./in.
6.5
6.5
5.7
6.8
5.6
5.6
bb
750
750
750
800
800
800
Density
#/in.3
0.280
0.276
0.279
0.280
0.277
0.275
Room temp. Room Room
yield (KSI) temp. temp.
Elevated
Hardened
tensileelong. Toughness temp. strength
Alloy condition
(KSI) % ft-lb. @ °F (KSI @ °F)c
S410001850°OQ +
155
188 17
49 @ RT
5 @ 1100°
700°
CRP
S416001800°OQ +
136
171 14
20 @ RT
700°
S420001900°OQ +
215
250
8
15 @ RT 96 @ 1000°
400°
CRP
S431001800°OQ +
163
202 16
25 @ RT
3 @ 1200°
700°
CRP
S440021900°OQ +
240
260
5
8 @ RT
600°
S440041900°OQ +
275
285
2
5 @ RT
600°
Values are approximate.
aCoefficient of thermal expansion for range of 72°1200° F.
bMagnetic permeability.
cCRP is stress required to produce 1% creep strain in 10,000 hours.
Page 59
Figure 3
Magnetic stainless steels.
C
Austenitic
This family of stainless accounts for the widest usage of all the
stainless steels. These materials are nonmagnetic, have face-centered
cubic structures, and possess mechanical properties similar to those of
the mild steels, but with better formability. The AISI designation
system identified the most common of these alloys with numbers
beginning with 300 and resulted in the term 300 series stainless.
The relationship between alloying elements and alloy types illustrated
in the Schaeffler diagram (Fig. 2) is an important concept in
understanding stainless steels. It has been established that certain
elements, specifically chromium, molybdenum, and silicon, are ferrite
formers. Aluminum and niobium are also ferrite formers, although
their effect is dependent on the alloy system. There are also elements
which tend to promote the formation of austenite. The most often used
are nickel, manganese, carbon, and nitrogen.
Examination of the Schaeffler diagram offers insight into the reason
for the composition of type 304, the cornerstone of austenitic alloy
family. Once the corrosion resistance plateau of 18% chromium is
reached, the addition of about 8% nickel is required to cause a
transition from ferritic to austenitic. The primary
Page 60
benefit of this alloy addition is to achieve the austenitic structure
which, relative to the ferritics, is very tough, formable, and weldable.
The added benefit, of course, is the improved corrosion resistance to
mild corrodents. This includes adequate resistance to most foods, a
wide range of organic chemicals, mild inorganic chemicals, and most
natural environmental corrosion.
Nickel is used judiciously as an alloying element since its cost is
substantially higher than chromium, averaging $2.50$3.00 per pound
over the last decade. However, type 304 is balanced near the
austenite-ferrite boundary for another reason. Compositions similar to
type 304 that can form no ferrite when solidifying after welding are
prone to cracking during solidification and are more difficult to hotwork. As a result, adding more nickel to the 188 composition offers
little benefit from a corrosion standpoint and would be detrimental in
other regards.
The next major step in alloying additions comes from the metal
molybdenum. This element also provides excellent corrosion
resistance in oxidizing environments, particularly in aqueous
corrosion. It participates in strengthening the passive film which
forms on the stainless steel surface along with chromium and nickel.
A significant benefit is realized with the addition of only about 2%
molybdenum. Added directly to the 188 composition the alloy would
contain too much ferrite, so it must be rebalanced. The resulting
chemistry is roughly 16% chromium, 10% nickel, and 2%
molybdenum.
Anodic polarization studies, such as the examples in Fig. 4, can be
useful in understanding the benefits of different alloy additions.
Chromium significantly increases the area of passivity as indicated in
the adjacent diagram. It lowers the potential required for the onset of
passivity and raises the pitting potential. Further increase in chromium
also shifts the current in the passive region to lower levels.
Molybdenum has the particular benefit of raising the pitting potential.
It is about three times more efficient at this than chromium.
Austenitic alloys also make use of the concept of stabilization.
Stainless types 321 and 347 are versions of type 304 stabilized with
titanium and niobium, respectively. The austenitic family of stainless
also prompted another approach to avoiding the effects of chromium
carbide precipitation. Since the amount of chromium which
precipitated was proportional to the carbon content, lowering the
carbon could prevent sensitization. As shown in Fig. 1, maintaining
the carbon content to below about 0.035%, vs. the usual 0.08%
maximum, will avoid the precipitation of harmful levels of chromium
carbide. This discovery along with improvements in melting
technology resulted in the development of the low-carbon version of
many of these alloys. When first introduced, extralow carbon (ELC)
grades required premiums on pricing due to higher production costs.
This differential has essentially disappeared in the face of modern
argon-oxygen decarburization (AOD) furnaces.
AOD furnaces, utilized as a final refining stage in melting, are
designed to
Page 61
Figure 4
Effects of environment and alloy content on anodic polarization behavior.
permit the bubbling of the molten steel with oxygen, which facilitates
the removal of carbon and sulfur. During this process the exposed
surface of the melt is protected with an inert argon atmosphere. This
arrangement also permits bubbling with nitrogen gas, which will
dissolve as atomic nitrogen into the steel. Nitrogen acts in a fashion
similar to carbon by pinning slip planes, thus leading to higher
strength materials.
Modern melting technology is also responsible for another trend in
stainless metallurgy. At one time the permissible chemistry ranges for
alloying elements needed to be broad to accommodate inhomogeneity
in electric furnace melts, chemical analysis variations, and raw
material quality. For example, the chromium range for type 304 was
18.020.0, and still heats were occasionally missed. With current
technology it is possible to maintain ±3s limits on chromium to 0.5%
or better. The result is that alloys are currently being produced with
0.50%0.75% less of an alloying element than they were just 15 years
ago.
Chemistries and properties of the 300 series alloys are listed in Table
3. An overview of these and other nonmagnetic stainless steel families
is shown in Fig. 5. Further improvements in general, localized, and
high-temperature corrosion resistance are gained by additions of
chromium, nickel, molybdenum, or
Page 62
Table 3 Selected Austenitic Stainless Steels
Alloy
C
S20100 0.08
Cr
16.5
S20910 0.05
22.0
S21904 0.025 21.0
S30300 0.06
18.5
S30400 0.05
18.5
S30403 0.025 18.2
S32100 0.05
17.5
S34700 0.05
17.5
S31603 0.025 16.5
S31726 0.025 17.5
S31008 0.05
24.5
S30615 0.2
18.0
S30815 0.07
21.0
Room temp.
Alloy yield (KSI)
S20100
45
S20910
65
Density
Ni
Mo N Other CTEa µin./in. bb #/in.3
5.8
0.2 Mn-5.8
10.3
< 0.283
1.02
12.5
0.25Mn-5.0
10.5
< 0.285
1.02
6.0
0.3 Mn-9.0
10.9
< 0.283
1.02
8.5
0.04S-0.2
10.4
< 0.285
1.02
8.5
0.04
10.4
< 0.285
1.02
8.6
0.04
10.4
< 0.285
1.02
9.5
0.05Ti-0.6
10.4
< 0.284
1.02
9.5
0.05Cb-0.6
10.4
< 0.285
1.02
10.5 2.3 0.04
10.3
< 0.287
1.02
14.0 4.3 0.14
10.1
< 0.287
1.02
19.5
0.04
9.8
< 0.289
1.02
14.0
Si-3.8
10.1
< 0.274
1.02
11.0
0.18Si-1.8,
10.1
< 0.282
Ce-0.05
1.02
Elevated temp.
Room temp. Room temp. Toughness strength (KSI
tensile (KSI) elong. % ft-lb. @ °F
@ °F)c
95
45
115 @
-100°
118
40
115 @
34 @ 1500°
-100°
STYS
S21603
50
98
50
S30300
35
90
50
S30400
45
95
45
S30403
28
75
55
S32100
35
85
55
S34700
37
88
50
S31603
28
75
55
S31726
S31008
60
43
100
92
40
45
S30615
46
108
55
S30815
55
105
55
213 @
26 @ 1300°
-110°
STYS
120 @
2 @ 1400°
-300°
CRP
120 @
2 @ 1400°
-300°
CRP
80 @ -300° 1 @ 1400°
CRP
110 @
2 @ 1400°
-300°
CRP
110 @
2 @ 1400°
-300°
CRP
110 @
3 @ 1400°
-300°
CRP
90 @ -300°
60 @ -300° 3 @ 1400°
CRP
95 @ RT
4 @ 1400°
CRP
110 @ RT
5 @ 1400°
CRP
Values are approximate.
aCoefficient of thermal expansion for range of 72°1200° F.
bMagnetic permeability.
cCRP is stress required to produce 1% creep strain in 10,000 hours. STYS
represents a short-time tensile yield strength.
Page 63
Figure 5
Austenitic stainless steels.
other more minor alloying elements. These modifications have led to
other austenitic alloys such as type 310, used primarily for hightemperature (above 1100°F) applications due to oxidation and
sulfidation resistance. More recent austenitic alloys include materials
such as S30815 (253MA®) and S30615 (RA85H®). S30815 is a
highly oxidation-resistant material with exceptional elevated
temperature mechanical properties. The oxidation resistance is a result
of the 22% chromium content combined with a small cerium addition
which helps form a tightly adherent scale. The strength is enhanced by
a nitrogen alloy addition, without a deleterious effect on corrosion
resistance.
Even with alloying additions such as molybdenum to improve
localized corrosion resistance to halogens, the workhorse 304 and 316
alloys are susceptible to chloride stress corrosion cracking (SCR).
This cracking mechanism manifests itself as branched, generally
transgranular cracks that are so fine as to be virtually undetectable
until it has progressed to catastrophic proportions. This mode of
failure can occur when the austenitic alloy is under stress in the
presence of halogen ions at temperatures above about 120°F. Studies
by Copsen, summarized in Fig. 6, underscored the benefit of very low
nickel contents, such as the ferritic stainless steels, or nickel levels in
excess of about 20%. In fact, the nickel
Page 64
Figure 6
Chloride stress cracking vs. nickel content.
contents in these two alloys are in the range which tend to crack most
quickly in chloride-bearing environments.
Another group of austenitic alloys is based on the substitution of
manganese for nickel. Manganese has about half the austenitizing
power of nickel. This approach was first used during and shortly after
World War II in response to nickel shortages. Stainless type 201 was
developed as a substitute for type 304 stainless. By adding about 4%
manganese and 0.2% nitrogen, the nickel content could be lowered to
about 5%. Although the strength of this alloy is higher than that of
type 304, its corrosion resistance is inferior.
Other alloys in this line have been developed. These include Nitronic
40 (S21900) and Nitronic 50 (S20910). The corrosion resistance of
S20910 exceeds that of type 316 stainless with the additional benefit
of higher mechanical properties.
D
Duplex
Stainless alloys that contain roughly equal amounts of austenite and
ferrite are termed duplex stainless. This family of alloys grew out of
one basic material originally identified as type 329. They are balanced
to contain relatively high chromium contents, with only enough nickel
and austenitizers to develop about 50% austenite. A partial list of
these alloys is presented in Table 4.
These alloys offer several useful advantages. First, their general
corrosion resistance is typically slightly above that of 316 in most
media. In addition, since
Page 65
Table 4 Selected Duplex (Austenitic-Ferritic) Stainless Steels
CTEa
Density
Alloy
C
Cr
Ni
Mo
N Other
µin./in. bb #/in.3
S32900 0.05 26.5
4.5
1.5
6.4 < 0.280
100
S31803 0.02 22.0
5.5
3.0 0.18
7.2 < 0.281
100
S32550 0.03 25.5
5.5
3.5 0.2 Cu-2.0
6.5 < 0.282
100
S32760 0.02 25.0
6.5
3.5 0.25Cu-0.7,
7.2 < 0.280
W-0.7
100
S32750 0.02 25.0
7.0
4.0 0.25
7.5 < 0.280
100
Elevated
temp.
strength
Room temp. Room temp. Room temp. Toughness (KSI @
Alloy
yield (KSI) tensile (KSI)
elong. % ft-lb. @ °F °F)c
S32900
82
100
31
90
57 @ 700°
STYS
S31803
70
98
25
180
40 @ 600°
STYS
S32550
98
125
30
140
76 @ 600°
STYS
S32760
90
120
28
160
S32750
90
120
28
160
55 @ 600°
STYS
Values are approximate.
aCoefficient of thermal expansion for range of 72°400° F.
bMagnetic permeability.
cSTYS represents a short-time tensile yield strength.
the nickel content is held low, they offer very good resistance to
chloride stress corrosion cracking. In combination with good
corrosion resistance, duplex stainless alloys offer higher strengths
than those typically found with austenitic steels.
Although their corrosion resistance is good, the boundary between
acceptable and poor performance is sharper than with austenitic
materials. As a result, they should not be used under conditions that
operate close to the limits of their acceptability. Although more
formable than the ferritic alloys, they are not as ductile as the
austenitic family of alloys. Welding requires more care than with the
austenitic alloys due to a greater tendency to compositional
segregation and sensitivity to weld heat input.
Due to the high chromium contents, duplex alloys are sensitive to
885°F embrittlement. This generally limits their usage to 600°F
maximum for pressure vessels. Due to the presence of nickel,
chromium, and molybdenum they are also susceptible to the formation
of s phase. This is a brittle phase which forms islands in the matrix
and will affect mechanical properties and corrosion resistance due to
alloy depletion. The s phase forms in the temperature range of 1100°F
to around 1600°F, and most rapidly at about 1450°F. The deleterious
effects of s phase formation are not obvious at the elevated
temperature but can become a factor at room temperature. The
formation of s phase in these alloys is sufficiently
Page 66
rapid to have an effect on properties due to slow (air) cooling after
anneal. A measurable effect as the result of exposure in this
temperature range due to welding has been demonstrated.
A reduction of properties due to s phase formation is also possible in
some austenitic alloys, particularly those with higher chromium and
nickel contents such as type 310. However, such effects are significant
only after thousands of hours of exposure at elevated temperature.
Formation of s phase in type 316 is also possible after long elevated
temperature exposure, but the volume fraction of s phase formed is so
low as to barely affect mechanical properties.
Duplex alloys can be a cost-effective choice for solving particular
corrosion and engineering design problems. The details of their
application, fabrication, and operation should be carefully considered
before they are utilized.
E
Precipitation Hardenable
This family of stainless alloys utilizes a thermal treatment to
intentionally precipitate phases which cause a strengthening of the
alloy. The precipitating phase is generated through an alloy addition
of one or more of niobium, titanium, copper, molybdenum, or
aluminum. The metallurgy is such that the material can be solutiontreated, i.e., all alloying elements are in solid solution and the material
is in its annealed or softest state. In this condition the material can be
machined, formed, and welded in the desired configuration. After
fabrication the unit is exposed to an elevated temperature cycle
(aging) which precipitates the desired phases to cause an increase in
mechanical properties.
Precipitation hardenable (PH) stainless steels are themselves divided
into three alloy types: martensitic, austenitic, and semiaustenitic. An
illustration of the relationship between these alloys is presented in Fig.
7. Chemical analyses and aged mechanical properties are listed in
Table 5. The martensitic and austenitic PH stainless steels are directly
hardened by thermal treatment. The semiaustenitic steels are supplied
as an unstable austenite, which is the workable condition and must be
transformed to martensite before aging.
As a class these alloys offer high mechanical properties, although not
as high as martensitic low-alloy steels, in combination with very
useful corrosion resistance. On average, their general corrosion
resistance is below that of type 304 stainless. The corrosion resistance
of the PH 157 Mo and A-286 alloys approaches that of type 316. The
martensitic and semiaustenitic PH grades are resistant to chloride
cracking. These materials are susceptible to hydrogen embrittlement.
F
Superferritic
The ability of the ferritic alloys to resist chloride stress corrosion
cracking is one of their most useful features in terms of corrosion
resistance. During the 1970s
Page 67
Figure 7
Precipitation hardening stainless.
developmental efforts were directed at producing ferritic materials
that could also exhibit a high level of general and localized pitting
resistance as well.
The first commercially significant alloy to meet this expectation was
an alloy containing 26% chromium and 1% molybdenum. In order to
obtain the desired corrosion resistance and acceptable fabrication
characteristics, the material had to have very low interstitial element
contents. To achieve these levels the material was electron
beamrefined under a vacuum and introduced as E-Brite Alloy. Carbon
plus nitrogen contents were maintained at levels below 0.020%.
E-Brite Alloy (S44627), partly because of its high level of corrosion
resistance for a ferritic material and partly because it is located so far
into the ferritic zone on the Schaeffler diagram, was termed a
superferritic. This alloy's usage grew for a period of several years
until its benefits for the construction of pressure vessels were
overshadowed by the difficult nature of fabrication and a concern over
toughness. Due to the very low level of interstitial elements, it was
prone to absorbing these elements during welding processes. Increases
in oxygen plus nitrogen to levels much over 100 ppm resulted in poor
toughness. Even without these effects, the alloy could exhibit a
ductile-to-brittle transition (DBTT) around room temperature. Other
ferritic alloys were also developed, some of which are shown in the
adjacent table. Stainless 444 (S44400) was originally
Page 68
Table 5 Selected Precipitation Hardening Stainless Steels
Alloy
C
Martensitic
S17400 0.04
S15500 0.04
S45000 0.03
CTEa
µin./in.
Density
bb #/in.3
Cr Ni Mo
N
Other
16.2 4.2
15.0 4.5
15.0 6.0 0.7
3.6
3.5
1.5
Cb-0.25
Cb-0.30
Cb-0.75
6.5
6.5
6.5
A1-1.1
6.5
A1-1.1
A1-1.1
N-0.1
6.5
6.6
7.2
0.282
0.282
0.282
9.9
1.01 0.286
S13800 0.04 13.0 8.0 2.3
Semiaustenitic
S17700 0.05 17.0 7.0
S15700 0.05 15.0 7.0 2.5
S35000 0.08 16.5 4.5 3.0
Austenitic
S66286 0.05 15.0 25.01.3
Ti-2.1, V0.3
Room
Room
temp.
temp.
Hardened yield Room temp.
elong.
Alloy condition (KSI) tensile (KSI)
%
Martensitic
S174001900°OQ+ 183
198
15
900°
S155001900°OQ+ 185
200
14
900°
S450001900°OQ+ 188
196
14
900°
S138001700°OQ+ 210
225
12
950°
Semiaustenitic
S177001750°SC+ 220
235
6
950°
S157001750°SC+ 225
240
6
950°
S350001900°SC+ 162
198
15
850°
90
90
>
50
70
0.282
0.282
0.280
0.280
Elevated
temp.
Toughness
strength
ft-lb. @ °F (KSI @ °F)c
16 @ RT
23 @ 900°
CRP
15 @ RT
40 @ RT
76 @ 1050°
STYS
30 @ RT
<15 @ RT
<15 @ RT
14 @ RT
35 @ 1000°
SR
85 @ 1000°
STYS
Austenitic
S662861800°OQ+ 95
145
24
64 @ RT
25 @ 1200°
1325°
SR
Values are approximate.
aCoefficient of thermal expansion for range of 72°1200° F.
bMagnetic permeability.
cCRP is stress required to produce 1% creep strain in 10,000 hours. SR
represents the stress required to produce stress rupture. STYS represents a shorttime tensile yield strength.
Page 69
introduced as 18Cr-2Mo and was stabilized with both titanium and
niobium. This material was less sensitive to contamination with
interstitial elements, but still exhibited relatively high DBTT values.
Characteristics of superferritic stainless steels are shown in Table 6.
Materials such as Sea-Cure (S43635) and 29-4C Alloy (S44735)
represent the most recent developments in superferritic materials. These
alloys do exhibit excellent localized corrosion resistance. Although the
superferritic materials alloyed with some nickel have improved
mechanical toughness and are less sensitive to contamination from
interstitial elements, their availability is still limited to thicknesses
below about 0.200 in. This is related to the formation of embrittling
phases during cooling from annealing temperatures. Section thicknesses
over these levels cannot be cooled sufficiently fast to avoid a loss of
toughness.
G
Superaustenitic
During the 1970s and into the 1980s much attention was focused on a
family of stainless alloys which came to be identified as
superaustenitic. The foundation for the development of this class of
materials was in the development
Table 6 Selected Superferritic Stainless Steels
Alloy
S44400
S44627
S44660
S44800
Alloy
S44400
N
C
Cr
Ni
Mo
0.02 18.2
2.2
0.02
0.002 26.0
1.0
0.010
0.02 26.0 2.5
3.0
0.025
0.005 29.0 2.2
4.0
0.01
Room temp. Room temp. Room temp.
yeild (KSI) tensile (KSI) elong. %
50
80
25
Other
Ti + Cb - 0.4
Ti + Cb - 0.5
Toughness
ft.Ib. @ °Fc
CTEa Density
µin./in. bb #/in.3
6.1
0.280
5.5
0.280
5.8
0.280
5.2
0.277
Elevated temp.
strength (KSI @
°F)d
S44627
60
75
25
5 @ 1200° STYS
S44660
80
95
30
S44800
75
90
25
Values are approximate.
a Coefficient of thermal expansion for range of 72°200° F.
b Magnetic permeability.
c Toughness in these alloys is highly dependent on section thickness and
temperature. Ductile-to-brittle transition temperatures near room temperature are
possible.
d STYS represents a short-time tensile yield strength.
Page 70
of Carpenter No.20 stainless, introduced in 1951. Consisting of 28%
nickel and 19% chromium with additions of molybdenum and copper,
this alloy was first produced as a cast material. Developing the
process to produce this material as a wrought product and later
refinements in chemistry ultimately resulted in the introduction of
20Cb-3 stainless in 1965. Superaustenitic materials are tabulated in
Table 7.
20Cb-3 stainless became popular in the chemical process industry as
an intermediate step between type 316 stainless and the more highly
alloyed nickel base materials. In particular, it was a cost-effective way
to combat chloride SCC. This form of cracking, illustrated in Fig 8, is
particularly difficult to combat by
Table 7 Selected Superaustenitic Stainless Steels
Alloy
C
Cr Ni
N08020 0.02 19.5 33.0
Mo
2.2
N
Other
Cu-3.2
Si-1.2
CTEa
µin./in.
8.9
N08330 0.05 19.5
35.0
N08367 0.02 20.5
25.0
N08800 0.08 19.5
32.0
N08825 0.02 20.0
38.5
3.0
N08904 0.02 20.0
25.0
4.5
S31254 0.02 20.0
18.0
6.1 0.2 Cu-0.7
9.4
S31654 0.02 24.0
22
10.0
S35315 0.05 25.0
35.0
7.3 0.5 Cu-0.5,
Mn-3.0
0.15 Si-1.8, Ce0.05
6.1 0.22
9.5
9.5
Al-0.4, Ti0.4
Cu-2.0, Ti0.8
Cu-1.5
9.6
9.1
9.4
9.5
Density
bb #/in.3
< 0.292
1.02
< 0.287
1.02
< 0.291
1.02
< 0.287
1.02
< 0.294
1.02
< 0.289
1.02
< 0.289
1.02
< 0.289
1.02
< 0.285
1.02
Room temp. Room temp.
yield (KSI) tensile (KSI)
Alloy
N08020
48
90
N08330
37
86
N08367
55
110
N08800
36
85
N08825
N08904
S31254
44
36
44
100
85
94
S31654
62
108
S35315
46
103
Room Toughness ft-ib. Elevated
temp.
@ °F
temp.
elong. %
strength
(KSI @
°F)c
45
145 @ -300° 1.5 @
1300° CRP
48
240 @ RT 5.3 @
1300° CRP
50
85 @ -300° 22 @ 900°
STYS
45
105 @ RT 5.5 @
1300° CRP
43
70 @ -300°
40
125 @ RT
35
88 @ RT
23 @ 750°
STYS
40
130 @ RT 43 @ 750°
(STYS)
48
142 @ RT 4.3 @
1400°
Values are approximate
a Coefficient of thermal expansion for range of 72°1200° F.
b Magnetic permeability.
c CRP is stress required to produce 1% creep strain in 10,000 hours. STYS
represents a short-time tensile yield strength
Page 71
Figure 8
Chloride stress corrosion cracking found in a S30400 stainless
steel stack operating around 150°F. The upper photomicrograph
illustrates the extensive branching associated with this type of
corrosion (×25). The transgranular nature of this corrosion mechanism
is shown in the lower photomicrograph (×100). Superaustenitic alloys
containing in excess of about 24% nickel are virtually immune
to this type of corrosion in most industrial environments. The
etchant was electrolytic oxalic acid.
Page 72
means other than alloy selection. Because of the high nickel content of
20Cb-3 stainless, it received a nickel base alloy UNS designation as
UNS N08020. However, since the major constituent is iron, it is truly
a stainless steel. The superaustenitic term is derived from the fact this
composition plots high above the austenite-ferrite boundary on the
Schaeffler diagram. Unlike the 300 series stainless alloys, there is no
chance of developing ferrite in this material.
In a similar time frame, another superaustenitic alloy was introduced
based on the wrought version of the heat-resistant cast alloy, HT. This
alloy, identified as RA330 stainless, contains about 35% nickel and
20% chromium with an addition of silicon. This superaustenitic
stainless also was assigned a nickel base UNS number (N08330).
N08330 offers excellent oxidation and carburization resistance in
combination with good elevated temperature mechanical properties.
Other superaustenitic stainless alloys with long histories include
Inconel 825 (N08825) and Inconel 800 (N08800), which have
similarities with N08020 and N08330, respectively. The driving force
for the development of newer superaustenitic stainless materials lay
primarily in the desire for alloys with better resistance to localized
corrosion. While alloys N08020 and N08825 exhibit good general
corrosion resistance to strong acids, their pitting resistance is only
slightly better than that of type 316L. Their performance in seawater
or brackish water is marginal at best.
The main approach to improving the pitting and crevice corrosion
resistance of the basic 35% nickel, 19% chromium, 2% molybdenum
alloy was to increase the molybdenum content. Among the first of the
newer alloys introduced was 904L (UNS N08904), which boosted the
molybdenum content to 4% and reduced the nickel content to 25%.
The reduction in nickel content was beneficial as a cost-saving factor,
with minimal loss of general corrosion resistance and sufficient
resistance to chloride SCC.
The next progression was to raise the molybdenum content to a higher
level, 6%, and offset the tendency for the formation of s phase by the
alloying addition of nitrogen. This concept was introduced with two
alloys, 254SMO® (UNS S31254) and AL-6XN® (UNS N08367). The
major benefit of the addition of nitrogen was the ability to produce
these alloys in heavy product sections such as plate, bar, and forgings.
An additional benefit was derived from alloying with nitrogen in
terms of increased pitting resistance. A significant amount of work by
a large body of researchers has demonstrated a relationship between
pitting or crevice corrosion resistance and alloy content which is
approximated by
where increasing values indicate increased resistance. A value in
excess of approximately 33 is considered necessary for pitting and
crevice resistance to ambient seawater.
Page 73
II
Cast Stainless Steel
Thus far the discussion has been devoted to examining the different
families of stainless steel metallurgy. The alloys discussed were
wrought materials, i.e., materials which are hot-worked following
their being cast into ingots. The practice of hot-working steels
improves the uniformity of their chemical, mechanical, and corrosionresistant properties. These materials are suited for fabrication by
bending and welding.
Cast stainless steels can be divided into the same families as the
wrought materials, except for the superferritics. Castings offer the
particular advantage of being able to obtain complex shapes without
extensive fabrication or machining. Cast alloys usually cost less per
pound than the wrought counterpart since the hot-working operations
are avoided. Cast stainless steels can also have chemistry
modifications to enhance properties that would otherwise render them
unworkable as a wrought product. Heat-resistant cast alloys, such as
HK, usually have high carbon and silicon contents which improve
elevated temperature strength considerably, but at the expense of
room temperature toughness. The cast structure of such materials is
also less resistant to thermal fatigue than the wrought material.
While the compositions of the basic austenitic cast alloys are very
similar to the wrought versions, the cast versions usually contain
significant amounts of delta ferrite. As in the solidification of weld
metal, ferrite is beneficial in reducing the tendency for the material to
form cracks during solidification. The ferrite content in CF-8M can
approach 20% and can readily attract a magnet. While high ferrite
contents are often not of concern, the ferrite can be attacked
preferentially in some environments such as urea, nitric acid, and
hydrochloric acid.
The existence of significant amounts of ferrite is one form of
segregation that can be encountered in cast stainless alloys. Since
cooling rates are generally slow for cast components, other secondary
phases can form. Chromium carbide precipitation is a particular
concern for many of these materials and under most circumstances the
casting should be solution-annealed prior to being placed in corrosive
service. A general list of cast stainless alloys is given in Table 8.
III
Welding
From an engineering standpoint, the ability to weld stainless steels
with relative ease is a major advantage to their usefulness. Weld
deposits, since they are cast structures, are subject to discussion
regarding corrosion resistance similar to the cast materials. The
chemistry of a weld deposit is likely to exhibit segregation and,
depending on the alloy and the welding technique employed, may
develop deleterious secondary phases in either the weld or heataffected zone.
Two ways to address this concern have already been discussed. These
Page 74
Table 8 Selected Cast Stainles Steel Alloys
Alloy C
CF-8 0.06
CF-3 0.02
CF0.06
8M
CF0.02
3M
CN0.06
7M
CN0.02
3MN
CD0.03
4MCu
HH
0.45
HK
0.45
HT 0.5
HX 0.5
Cr
18.5
18.0
18.5
Ni
8.5
8.5
9.5
Mo
2.2
CTEa
µin./in.
10.4
10.4
10.3
18.0
9.5
2.2
10.3
<1.5 0.287
19.5
28.0
2.2
9.2
<1.1 0.292
21.0
25.0
6.2
0.22
9.9
<1.1 0.291
26.0
5.0
2.0
0.1
7.0
<100 0.282
25.0
25.0
18.0
17.0
12.0
12.0
35.0
65.0
Room temp.
Alloy yield (KSI)
CF-8
37
CF-3
36
CF8M
CF3M
CN7M
CN3MN
CD-
42
40
32
55
81
N
Other
Cu-3.2
Cu-3.0
13.0
12.9
11.9
12.2
Density
bb #/in.3
<1.5 0.285
<1.5 0.285
<1.5 0.287
1.5 0.279
<1.1 0.286
1.5 0.286
2.0 0.294
Elevated
temp.
Room temp. Room temp. Toughness ft- strenght
tensile (KSI) elong. %
lb.@ °F
(KSI @ °F)c
77
50
80 @ RT
2 @ 1400°
CRP
77
80
80 @ RT
1 @ 1400°
CRP
77
50
80 @ RT
2 @ 1400°
CRP
80
50
80 @ RT
2 @ 1400°
CRP
69
45
85 @ RT
2 @ 1400°
CRP
107
40
85 @ RT
22 @ 900°
STYS
108
25
40 @ RT
60 @ 700°
4MCu
HH
50
85
25
25 @ RT
HK
50
75
17
20 @ RT
HT
40
70
10
4 @ RT
HX
36
85
9
4 @ RT
STYS
3.0 @ 1400°
CRP
10.2 @
1400° CRP
8.0 @ 1400°
CRP
6.4 @ 1400°
CRP
Values are approximate.
aCoefficient of thermal expansion for range of 72°F to 1400°F.
bMagnetic permeability.
cCRP is stress required to produce 1% creep strain in 10,000 hours. STYS
represent a short-time tensile yield strength.
involve the reduction of carbon content to low levels and the use of
stabilizers to prevent chromium depletion. Either of these methods is
typically used for components which are assembled in the field
because subjecting the fabricated unit to an annealing treatment is
neither practical nor desirable in most instances.
In many cases, a small decrease in the corrosion resistance of
weldments is tolerable. When the environment is particularly severe
for the alloy being used, the weld may be attacked preferentially. This
condition can be exaggerated by the area effects of the more noble
base metal compared to the small weld zone. An
Page 75
alternate approach for field welding is to select a higher alloy welding
consumable so the weld deposit is more noble than the base metal.
Preferential attack can also occur in the heat-affected zone of the base
metal. This is typical of weldments made in standard type 304 where
the carbon content will lead to chromium carbide precipitation. Of
course this condition cannot be avoided by using a different filler
metal and the only remedy is a postweld anneal.
Aside from the actual weld deposit chemistry, welding technique can
have an influence on corrosion resistance. First and foremost, the area
to be joined should be clean and free of dirt and grease. Carbonaceous
materials will contaminate the weld deposit and deplete chromium
from the alloy. For a similar reason, carbon arc gouging or cutting
should be avoided. Contamination from other metals should also be
avoided. Although free iron will essentially be melted into the weld
deposit unnoticed, rust can affect weldability. At best it can lead to
lack of fusion or porosity and at worst it may act as a preferential site
for the onset of corrosion. Joint preparation should be accomplished
using properly sharpened tooling and wire brushing should be
performed using stainless steel brushes.
Low-melting-point metals are of particular concern. Molten copper,
zinc, or aluminium will attack the grain boundaries of austenitic alloys
preferentially. Copper alloy clamps or fixtures used to hold work
while welding have been known to leave smears of metal which have
subsequently caused cracking. Zinc from galvanized steel or paint
primers has also been known to contaminate weld joints.
Full-penetration weld joints should also be made. This is a good
practice from a strength and fatigue resistance standpoint but is also a
factor in avoiding corrosion. Unfused joints are sites likely to trap
corrodents and corrosion products increasing the likelihood that
oxygen or metal ion concentration cells are developed. This will
generally mean that joints will have to be beveled if the thicknesses to
be joined are in excess of 3/16 in. Beveling joints also ensures that
adequate filler metal will penetrate to the root in those instances
where overalloying is desired.
Finally, the surface finish of the weld area should be similar to that of
the base metal. While a slightly higher roughness is usually
unavoidable unless welding is followed by grinding to blend in the
weld, minimizing roughness in the weld zone can be beneficial. Weld
spatter should be removed by grinding. The weld slag from covered
electrodes, which prevents oxidation of the metal during
solidification, should be completely removed prior to making a second
weld pass or placing the weld in service. Slag deposits on the surface
will act as crevices in corrosive service. Removal of heat tints on the
surface from bare wire or autogenous welding processes is preferred.
In severe service these areas may be attacked preferentially.
Page 76
IV
Passivation
Stainless steels offer useful resistance because they tend to exhibit
passive corrosion behavior as a result of the formation of protective
oxide films on the exposed surfaces. Under normal circumstances,
stainless steels will readily form this protective layer immediately on
exposure to oxygen. When this protective film is violated or fails to
form, active corrosion can occur. Some fabrication processes can
impede the reformation of this passive layer and to ensure that it is
formed stainless steels are subjected to ''passivation" treatments.
The most common passivation treatments involve exposing the metal
to an oxidizing acid. Nitric and nitric-hydrofluoric acid mixtures
represent the predominant usage in stainless steel production. The
nitric-hydrofluoric acid mixtures are more aggressive and are typically
used to remove the oxide scales formed during thermal treatment.
This "pickling" process provides two benefits. First it removes the
oxide scale and passivates the underlying metal surface. Second, due
to its aggressive nature, the process will remove any chromiumdepleted layer that may have formed as a result of the scale formation.
For passivation treatments other than scale removal following thermal
treatment, less aggressive acid solutions are usually employed. The
primary purpose of these treatments is to remove contaminants that
may be on the component's surface and could prevent the formation of
the oxide layer locally. The most common contaminant is imbedded
or free iron particles from forming or machining tools. A dilute (10%)
solution of nitric acid is effective to removing free iron. For ferritic,
martensitic, or precipitation hardening grades a nitric acid solution
inhibited with sodium dichromate is used so as not to attack the
stainless too aggressively. For the more resistant stainless alloys
phosphoric acid at 1% concentration and nitric acid at 20%
concentrations are also effective. Other commercially available
chelating agents can be employed.
References
1. A. Brasunas, NACE Basic Corrosion Course, National Association
of Corrosion Engineers, Houston, 1970.
2. C. Dillon, Corrosion Control in the Chemical Process Industries,
McGraw-Hill, New York, 1986.
3. M. Fontana and N. Greene, Corrosion Engineering, McGraw-Hill,
New York, 1967.
4. M. Henthorne, Corrosion: Causes and Control, Chemical
Engineering Magazine, New York, 19711972.
5. H. McGannon, The Making, Shaping and Treating of Steel, U.S.
Steel Corporation, Pittsburgh, 1971.
6. Mechanical and Physical Properties of the Austenitic ChromiumNickel Stainless Steels at Elevated Temperatures, International Nickel
Company, New York, 1963.
7. Metals Handbook, 10th ed., ASM International, Metals Park, OH,
1990.
Page 77
8. P. Schweitzer, Corrosion and Corrosion Protection Handbook,
Marcel Dekker, New York, 1979.
9. A. Sedriks, Corrosion of Stainless Steels, John Wiley and Sons,
New York, 1979.
10. Steel Castings Handbook, Steel Founder's Society of America,
Rocky River, OH, 1981.
11. Technical Data Sheet, 20Cb-3 Stainless Steel, Carpenter
Technology Corporation, Reading, PA, 1987.
12. Technical Data Sheet, AL-6XN Alloy, Allegheny-Ludlum
Corporation, Pittsburgh, 1991.
13. Technical Data Sheet, E-Brite Alloy, Allegheny-Ludlum
Corporation, Pittsburgh, 1981.
14. Technical Data Sheet, SEA-CURE Stainless, Colt Industries, East
Troy, WI, 1978.
Page 79
5
Corrosion of Nickel and High-Nickel Alloys
Philip A. Schweitzer
Faliston, Maryland
The nickel-based alloys exhibit the widest range of applications of any
series of alloys. This is the result of the ability to nickel to be
metallurgically compatible with a variety of alloying elements such as
chromium, copper, and molybdenum. In general, nickel-based alloys
contain more alloying elements and are more corrosion-resistant than
the iron-based alloys.
In the electrochemical series nickel is nobler than iron but more active
than copper. Reducing environments, such as dilute sulfuric acid, find
nickel more corrosion-resistant than iron but not as resistant as copper
or nickel-copper alloys. The nickel-molybdenum alloys are more
corrosion-resistant to reducing environments than nickel or nickelcopper alloys.
While nickel can form a passive film in some environments, it is not a
particularly stable film; therefore nickel cannot generally be used in
oxidizing media, such as nitric acid. When alloyed with chromium a
much improved stable passive film results producing a greater
corrosion resistance to a variety of oxidizing environments. However,
these alloys are subject to attack in media containing chloride or other
halides, especially if oxidizing agents are present. Corrosion will be in
the form of pitting. The corrosion resistance can be improved by
adding molybdenum and tungsten.
The high-nickel-containing alloys, both nickel-based alloys and
austen-
Page 80
itic stainless steels containing high nickel content, are often used in
applications requiring stress corrosion cracking resistance to chloridecontaining environments.
Nickel and the high-nickel alloys will be treated individually with
their specific advantages and disadvantages discussed. Only those
alloys which find application in the corrosion resistance field will be
covered.
I
Nickel
There are two basic pure nickel alloys, each containing a minimum of
99% of nickel: alloy 200 and alloy 201. Alloy 201 is the low-carbon
version of alloy 200. Alloy 200 is subject to the formation of a grainboundary graphitic phase, which reduces ductility tremendously.
Consequently, nickel alloy 200 is limited to a maximum operating
temperature of 600°F (315°C). For application above this temperature
alloy 201 should be used.
The corrosion resistance of alloys 200 and 201 are the same. They
exhibit outstanding resistance to hot alkalies, particularly caustic soda.
Excellent resistance is shown at all concentrations at temperatures up
to and including the molten state. Below 50% the corrosion rates are
negligible, usually being less than 0.2 mil/year (mpy) even in boiling
solutions. As concentrations and temperatures increase, corrosion
rates increase very slowly. Impurities in the caustic, such as chlorates
and hypochlorites, will determine the corrosion rate.
Nickel is not subject to stress corrosion cracking in any of the chloride
salts and it exhibits excellent general resistance to nonoxidizing
halides. Oxidizing acid chlorides such as ferric, cupric, and mercuric
are very corrosive and should be avoided.
Nickel 201 also finds application in the handling of hot, dry chlorine
and hydrogen chloride gas on a continuous basis up to 1000°F
(540°C). The resistance is attributed to the formation of a nickel
chloride film. Dry flourine and bromine can be handled in the same
manner. The resistance will decrease when moisture is present.
Nickel exhibits excellent resistance to most organic acids, particularly
fatty acids such as stearic and oleic, if aeration is not high.
Nickel is not attacked by anhydrous ammonia or ammonium
hydroxide in concentrations of 1% of less. Stronger concentrations
cause rapid attack.
Nickel also finds application in the handling of food and synthetic
fibers because of its ability to maintain product purity. The presence
of nickel ions is not detrimental to the flavor of food products and it is
nontoxic. Unlike iron and copper, nickel will not discolor organic
chemicals such as phenol and viscose rayon.
Page 81
II
Nickel-Copper Alloys
Nickel and copper are completely soluble in each other. This has
resulted in a series of alloys, but we will only be dealing with those
alloys on the nickel-rich side.
A
Alloy 400
In 1905 the first nickel alloy containing approximately one-third
copper and two-thirds nickel was produced. This was known as
Monel* alloy 400. The present equivalent, alloy 400, remains one of
the most widely used nickel alloys.
Nickel-copper alloys offer somewhat higher strength than unalloyed
nickel without sacrificing ductility. The thermal conductivity of alloy
400, though somewhat lower than that of nickel, is significantly
higher than that of nickel alloys containing substantial amounts of
chromium or iron.
Alloy 400 has many of the characteristics of chemically pure nickel
with improvements in certain areas of corrosion resistance over that of
pure nickel. The general corrosion resistance of alloy 400 in the
monoxidizing acids such as sulfuric, hydrochloric, and phosphoric is
improved over that of pure nickel. However, there is no improvement
over nickel when in contact with oxidizing media such as nitric acid,
ferric chloride, chromic acid, wet chlorine, sulfur dioxide, or
ammonia.
One of the major areas of application is in the handling of water,
including brackish and seawaters. Under high velocity conditions it
gives excellent service. In stagnant seawater alloy 400 is subject to
pitting, but at a much lower rate than nickel. The absence of chloride
stress corrosion cracking is also a factor in the selection of this alloy
for these applications.
Alloy 400 has excellent resistance to hydrofluoric acid solutions at
various concentrations and temperatures. It is subject to stress
corrosion cracking in moist, aerated hydrofluoric or hydrofluorosilic
acid vapor. If completely immersed in the acid cracking is unlikely.
Alloy 400 undergoes negligible corrosion in all types of atmospheres.
Indoor exposure produces a very light tarnish that is easily removed
by occasional wiping. Outdoor surfaces that are exposed to rain
produce a thin graygreen patina. In sulfurous atmospheres a smooth,
brown adherent film forms.
B
Alloy 405
Sulfur is added to this alloy to improve machinability. The corrosion
resistance of alloy and alloy 405 are essentially the same.
*Monel 400 is the trademark of International Nickel.
Page 82
C
Alloy K-500
This is an age-hardenable alloy that has the advantage of increased
strength and hardness while retaining the excellent corrosion resistant
properties of alloy 400. Strength is maintained up to approximately
1200°F (649°C), and the alloy is strong, tough, and ductile at
temperatures as low as -423°F (-233°C).
III
Nickel-Molybdenum Alloys
There is one major alloy in this seriesalloy B-2, which is a low carbon
and silicon (0.02%, 0.08% maximum) version of alloy B. This alloy is
unique because it does not contain chromium. Molybdenum, the
primary alloying ingredient, provides significant corrosion resistance
to reducing environments.
Since this alloy was originally developed to handle hydrochloric acid
it is only logical that this is a major area of application. Alloy B-2 is
capable of handling all concentrations of hydrochloric acid in the
temperature range 158212°F (70100°C) and of handling wet hydrogen
chloride gas.
It also exhibits excellent resistance to pure sulfuric acid at all
concentrations and temperatures below 60% acid and good resistance
to 212°F (100°C) above 60% acid.
The high molybdenum content provides corrosion resistance of alloy
B-2 in many nonoxidizing environments among which are
hydrofluoric and phosphoric acids, along with numerous organic acids
such as acetic, formic, and cresylic. It is also resistant to many
chloride-bearing salts (nonoxidizing) such as aluminum chloride,
magnesium chloride, and antimony chloride.
The high molybdenum content also provides protection against pitting
attack in acid chloride environments. Being nickel-rich it is also
resistant to chloride-induced stress corrosion cracking.
Alloy B-2 has extremely poor corrosion resistance in oxidizing
environments. It has virtually no corrosion resistance to oxidizing
acids such as nitric and chromic, or to oxidizing salts, such as ferric
chloride or cupric chloride. Care must be taken if oxidizing salts are
present in reducing acids. Concentrations as low as the ppm range of
oxidizing salts such as ferric chloride, ferric sulfate, or cupric chloride
can accelerate the attack in hydrochloric or sulfuric acids. Even
dissolved oxygen has sufficient oxidizing power to affect the
corrosion rates for alloy B-2 in hydrochloric acid.
Alloy B-2 is not recommended for elevated temperature service
except in specific applications. The high molybdenum content
provides the alloy with excellent mechanical properties at elevated
temperatures, greater than 1650°F (900°C), and consequently has
been used for mechanical components in reducing environments and
vacuum furnaces. The use of alloy B-2 in the temperature range
Page 83
11121562°F (600850°C) is not recommended because of the
formation of the intermetallic phase of NiMO.
IV
Nickel-Chromium-Molybdenum Alloys
The major alloys in this family are Inconel alloy 625, Hastelloys C276, C-4, and C-22. These alloys were produced to provide improved
corrosion resistance in oxidizing environments.
A
Inconel Alloy 625
This alloy finds application where strength and corrosion resistance is
required. It exhibits exceptional fatigue strength and superior strength
and toughness at temperatures ranging from cryogenic to 2000°F
(1093°C). The columbium and tantalum stabilization makes the alloy
suitable for corrosion service in the as-welded condition. It also has
good resistance to chloride stress corrosion cracking.
Alloy 625 has excellent resistance to phosphoric acid solutions,
including commercial grades of acids containing fluorides, sulfates,
and chlorides in the production of superphosphoric acid (72% P2O5).
The alloy also exhibits good resistance to aqueous solutions in a
variety of applications including organic acids, sulfuric and
hydrochloric acid at temperatures below 150°F (65°C). Alloy 625 is
resistant to mixtures of nitric-hydrofluoric where stainless steel loses
its resistance.
B
Hastelloy Alloy C-276
This alloy is a low-carbon and silicon version of Hastelloy alloy C and
can therefore be used in the as-welded condition in most applications.
Alloy C-276 is extremely versatile because it possesses good
resistance to both oxidizing and reducing media, including conditions
with halogen ion contamination. When dealing with acid chloride
salts, the pitting and crevice corrosion resistance of the alloy makes it
an excellent choice.
Alloy 276 has exceptional corrosion resistance to many chemical
process materials, including highly oxidizing neutral and acid
chlorides, solvents, chlorine, formic and acetic acids, and acetic
anhydride. It also resists highly corrosive agents such as wet chlorine
gas, hypochlorites, and chlorine solutions.
Exceptional corrosion resistance is exhibited in the presence of
phosphoric acid at all temperatures below the boiling point of
phosphoric acid, when concentrations are less than 65% by weight.
Corrosion rates of less than 5 mpy were recorded. At concentrations
above 65% by weight and up to 85%, alloy C-276 displays similar
corrosion rates, except at temperatures between 240°F (116°C) and
the boiling point, where corrosion rates may be erratic and may reach
25 mpy.
Page 84
C
Hastelloy Alloy C-4
Although alloy C-276 is resistant to carbide precipitation,
precipitation of the intermediate µ-phase can still occur. Alloy C-4
was developed to reduce this latter precipitation. The composition of
alloys C-4 and C-276 is, with the exception of iron and tungsten,
approximately the same. The general corrosion resistance of the two
alloys are generally the same. In a strongly reducing medium, such as
hydrochloric acid, alloy C-4 has slightly higher rates than C-276,
while in an oxidizing medium the results are reversed.
Alloy C-4 can be subjected to temperatures in the normal sensitizing
range of 10221994°F (5501090°C) for extended periods without
experiencing severe corrosive attack. This temperature exposure can
be the result of welding; thermomechanical processing, such as hotforming or rolling operations, stress relief, or normalizing treatments;
or operation of process equipment in the sensitizing range.
D
Hastelloy Alloy C-22
Alloy C-22 was developed to improve on the shortcomings of alloys
C-276 and C-4. Alloy C-276 is limited in its applications in oxidizing
environments containing low amounts of halides and in environments
containing nitric acid. In addition, the thermal stability of the alloy is
insufficient to permit it to be cast. Although alloy C-4 has a much
higher thermal stability, it does not have satisfactory corrosion
resistance to chloride-containing environments.
Alloy C-22 overcomes these shortcoming. It is not only superior in
oxidizing environments containing nitric acid, but it also has
improved pitting resistance over that of alloy C-276. The general
areas of application of alloy C-22 are the same as for alloy C-276.
V
Nickel-Chromium-Iron Alloys
A
Inconel Alloy 600
Alloy 600 has excellent mechanical properties and a combination of
high strength and good workability. It performs well with
temperatures from cryogenic to 1200°F (649°C).
Alloy 600 is practically free from corrosion by fresh waters, including
the most corrosive of natural waters containing free carbon dioxide,
iron compounds, and dissolved air. It remains free from stress
corrosion cracking even in boiling magnesium chloride.
The alloy exhibits greater resistance to sulfuric acid under oxidizing
conditions than either nickel 200 or alloy 400. The addition of
oxidizing salts to
Page 85
sulfuric acid tends to passivate alloy 600, which makes it suitable for
use with acid mine waters or brass pickling solutions, where alloy 400
cannot be used.
Alloy 600 is not subject to stress corrosion cracking in any of the
chloride salts and has excellent resistance to all nonoxidizing halides.
The alloy has excellent resistance to dry halogens at elevated
temperatures and has been used successfully for chlorination
equipment at temperatures up to 1000°F (538°C).
Alloy 600 has been substituted for alloy 201 in certain hightemperature applications in which sulfur is present because of its
improved resistance. However, alloy 600 is subject to stress corrosion
cracking in high-temperature, high-concentration alkalies. Therefore
the alloy should be stress-relieved before use and the operating
stresses kept to a minimum. Alloy 600 is almost entirely resistant to
attack by solutions of ammonia over the complete range of
temperatures and concentrations.
The usefulness of this alloy is in its high-temperature applications.
Alloy 600 exhibits corrosion resistance in mildly oxidizing aqueous
media but is limited in use in reducing acid solutions.
B
Incoloy Alloy 800
This alloy is used primarily for its oxidation resistance and strength at
elevated temperatures. It is particularly useful for high-temperature
equipment because the alloy does not form the embrittling s phase
after long exposures at 12001600°F (649871°C). High creep and
rupture strengths are other factors that contribute to its performance in
many applications.
At moderate temperatures the general corrosion resistance of alloy
800 is similar to that of the other austenitic nickel-iron-chromium
alloys. As the temperature increases alloy 800 continues to exhibit
good corrosion resistance, while other austenitic alloys are
unsatisfactory for the service.
Alloy 800 has excellent resistance to nitric acid at concentrations up
to about 70%. It is also resistant to a variety of oxidizing salts but not
halide salts. It also has good resistance to organic acids, such as
formic, acetic, and propionic. Alloy 800 is particularly suited for the
handling of hot corrosive gases such as hydrogen sulfide.
This alloy is not widely used for aqueous service. In aqueous
corrosion service the corrosion resistance of alloy 800 falls between
that of 304 and 316 stainless steels. On occasion alloy 800 may be
substituted for the austenitic stainless steels since the stress corrosion
cracking resistance of alloy 800 is somewhat better than that of the
austenitics.
C
Incoloy Alloy 825
The composition of alloy 800 has been modified to produce alloy 825,
which has an improved aqueous corrosion resistance. The higher
nickel content of alloy 825,
Page 86
as compared to alloy 800, makes it resistant to chloride ion stress
corrosion cracking. Additions of molybdenum and copper provide
resistance to pitting and to corrosion in reducing acid environments
such as sulfuric or phosphoric acid solutions. Alloy 825 is resistant to
pure sulfuric acid solutions up to 40% by weight at boiling
temperatures and at all concentrations at a maximum temperature of
150°F (60°C). The presence of oxidizing salts, such as cupric or
ferric, actually reduces the corrosion rates. Alloy 825 has limited use
in hydrochloric or hydrofluoric acids.
The chromium content of alloy 825 provides resistance to a variety of
oxidizing environments such as nitrates, nitric acid solutions, and
oxidizing salts. Alloy 825 has good resistance to stress corrosion
cracking in neutral chloride environments. If localized corrosion is a
problem with 300 series stainless steel, alloy 825 may be substituted.
Alloy 825 also offers excellent resistance to corrosion by seawater.
D
Alloy 800H
Alloy 800H is a controlled carbon version of alloy 800. The carbon
content is maintained between 0.05% and 0.1% to provide the alloy
with better elevated temperature creep and stress rupture properties. It
is used in a variety of high-temperature applications in the refining
and heat treatment industries.
VI
Nickel-Chromium-Iron-Molybdenum Alloys
The addition of molybdenum to the Ni-Cr-Fe alloys provides for an
increase in the corrosion resistance to environments containing
chlorides and to moderately corrosive reducing environments such as
dilute sulfuric and phosphoric acids. The amount of molybdenum that
can be alloyed is limited because of the presence of iron, which leads
to embrittlement resulting from precipitation of intermetallic phases.
The iron present also tends to reduce the cost of these alloys.
A
Alloy G/Hastelloy Alloy G-3
Alloy G-3 is a low-carbon version of alloy G. The corrosion resistance
of alloy G and alloy G-3 are approximately the same, but the thermal
stability of alloy G-3 is greater. These alloys are highly resistant to
pitting and stress corrosion cracking in both acid and alkaline
environments, including hot sulfuric and phosphoric acids,
hydrofluoric and contaminated nitric acids, mixed acids, and sulfate
compounds.
Alloy G has found many applications in pollution control equipment.
Excellent performance has been demonstrated in municipal garbage
incinerator systems, including fans, ducts, and scrubber equipment.
The SO2 scrubber systems for power plants that use water or alkaline
quench have been incorporat-
Page 87
ing alloy G as a major material of construction because of its
resistance to sulfuric acid and conditions during which the chloride
ion can concentrate.
B
Hastelloy Alloy X
Alloy X provides high strength and excellent oxidation resistance at
elevated temperatures. A typical application in the process industries
is the nitric acid catalyst grid supports operating at 1650°F (900°C).
The strength and resistance to warpage and distortion at high
temperature provide outstanding performance.
Alloy X is also used for distributor plates in the manufacture of
magnesium chloride. Other high-temperature applications include
flare nozzles, thermowell protection tubes, expansion bellows, furnace
internal retorts, muffles, and trays.
C
Hastelloy Alloy G-30
This alloy is a modification of alloy G with a higher chromium
content, which gives it a higher resistance to oxidizing environments
than other alloys in this series.
In acid mixtures such as nitric plus hydrofluoric and sulfuric plus
nitric acids, alloy G-30 shows the highest resistance of this class of
alloys.
The alloy is also used in the evaporators of commercial wet process
phosphoric manufacturing systems. In these environments, which are
complex mixtures of phosphoric, sulfuric, and hydrofluoric acids and
contain various oxides, alloy G-30 has shown the lowest corrosion
rate of a number of alloys tested. The corrosion rate for alloy G-30
was 6 mpy as compared to 16 mpy for alloys G-3 and 625.
VII
Alloys for High Temperature Corrosion
Alloys designed to resist high-temperature corrosion are basically
oxidation-resistant materials since all forms of attack at elevated
temperatures are considered to be oxidation. As with aqueous
corrosion a protective oxide film is formed. The rate at which the
metal oxidizes will depend on the stability of the film. If the film is
stable and remains in place the rate will be logarithmic, diminishing
with time.
Cycling temperatures will tend to spall off the surface film, leading to
a stepwise oxidation of the alloy. Changes in the environment can also
have the same effect.
Although all high-temperature corrosion is considered to be oxidation,
there are other terms that are also encountered such as
oxidationreduction, sulfidation, fuel ash corrosion, carburization, and
nitridation, to mention a few.
While many of the high-nickel alloys previously discussed can be
utilized at elevated temperatures, there are some instances where these
materials are not
Page 88
satisfactory. Consequently, other alloys have been developed to
overcome these shortcoming.
A
Haynes Alloy No. 556
The presence of 18% cobalt in this alloy provides greater resistance to
sulfidation than many nickel-based alloys such as alloy X or alloy
800H. In pure oxidation alloy 556 shows good resistance but is
superseded in performance by other alloys such as X and 214.
In chloride-bearing oxidizing environments that alloy shows better
resistance than alloys 800H and X but not as good as alloy 214.
In carburizing environments the alloy is better than 310 stainless steel
and some nickel-based alloys such as X and 617 but not as good as the
aluminum-containing alloys such as 214.
Typical applications include internals of municipal waster incinerators
and refractory anchors in a refinery train-gas-burning unit.
B
Haynes Alloy No. 214
Haynes alloy no. 214 possesses the highest oxidation resistance of any
of the nickel-based alloys to both static and dynamic environments.
Alloy 214 develops a tenacious aluminum oxide layer at the surface.
The aluminum film also provides superior resistance to carburizing
environments containing chlorine and oxygen.
As typical of many high-temperature alloys, this alloy does not
possess good resistance to aqueous chloride solutions. Therefore dew
point conditions must be avoided.
Typical applications of this alloy include mesh belts for supporting
chinaware while being heated in a kiln, strand annealing tubes for
making medical grade stainless wire, and honeycomb seals in turbine
engines.
C
Hastelloy Alloy No. 230
The outstanding feature of Hastelloy alloy no. 230 is its superior
nitridation resistance. This property, with its high creep strength, has
enabled use of the alloy as a catalyst support grid in the manufacture
of nitric acid. It also exhibits good resistance to carburization.
However, the alloy does not possess adequate resistance to sulfidizing
environments.
Reference
1. Philip A. Schweitzer, Corrosion Resistance Tables, Parts AC, 4th
ed., Marcel Dekker, New York, 1995.
Page 89
6
Corrosion of Copper and Copper Alloys
Philip A. Schweitzer
Fallston, Maryland
Copper is a very useful material. It has excellent electrical and thermal
conductivity properties, is malleable and machinable, but has low
mechanical properties. In order to obtain strength the metal must be
cold-worked or alloyed. As a result there are hundreds of copper
alloys. The Copper Development Association, together with the
American Society of Testing and Materials and the Society of
Automotive Engineers, developed a five-digit system to identify these
alloys. This system is part of the unified numbering system for metals
and alloys. The numbers C-10000 through C-79999 denote the
wrought alloys while the cast copper and copper alloys are numbered
C-80000 through C-99999.
I
Coppers
To be classified as copper the compound must contain a minimum of
99.3% copper. Elements such as silver, arsenic, phosphorus,
antimony, tellurium, nickel, cadmium, lead, sulfur, zirconium,
magnesium, boron, and bismuth may be present singly or in
combination.
Copper is noble to hydrogen in electromotive force (emf) series and
thermodynamically stable with no tendency to corrode in water and in
nonoxidizing acids free of dissolved oxygen. With copper and its
alloys the predominant
Page 90
cathodic reaction is the reduction of oxygen to form hydroxide ions.
Therefore the presence of oxygen or other oxidizing agents is
necessary for corrosion to take place. In oxidizing acids or in aerated
solutions of ions which form complexes, e.g., CN-,
, corrosion can
be severe. Copper is also subject to attack by turbulently flowing
solutions, even though the metal may be resistant to the solution in a
stagnant condition. Most of the corrosion products formed on copper
and copper alloys produce adherent, relatively impervious films with
low solubility that provides the corrosion protection.
Copper finds many applications in the handling of seawater and/or
freshwater. Copper pipe was used by the Egyptian pharaoh Cheops to
transport water to the royal bath. Several years ago a remnant of this
pipe was unearthed, still in usuable condition, a testimony to copper's
durability and resistance to corrosion.
The corrosion resistance of copper, when handling freshwater or
seawater, is dependent on the surface oxide film which forms. In order
for corrosion to continue oxygen must diffuse through this film. Highvelocity water will disturb the film while carbonic acid or organic
acids, which are present in some freshwaters or soils, will dissolve the
film. Either situation leads to an appreciably high corrosion rate. If the
water velocity is limited to a maximum of 45 feet/s the film will not
be disturbed.
Sodium and potassium hydroxide solutions can be handled at room
temperature by copper in all concentrations. Copper is not corroded
by perfectly dry ammonia but may be rapidly corroded by moist
ammonia and ammonium hydroxide solutions. Alkaline salts, such as
sodium carbonate, sodium phosphate, and sodium silicate, act like
hydroxides but are less corrosive.
When exposed to the atmosphere over long periods the protective film
which forms is initially dark in color, gradually turning green. This
corrosion product is known as patina. Since the coloration is given by
copper hydroxide products, the length of time required to form this
coloration is dependent on the atmosphere. In a marine atmosphere,
the compound is a mixture of copper/hydroxide/chloride and in an
urban or industrial atmosphere a mixture of copper/hydroxide/sulfate.
Pure copper is immune to stress corrosion cracking. However, alloys
of copper containing more than 15% zinc are particularly subject to
this type of corrosion.
The coppers are resistant to urban, marine, and industrial atmospheres.
For this reason copper is used in many architectural applications such
as building fronts, downspouts, flashing, gutters, roofing, and
screening. In addition to their corrosion resistance, their good thermal
conductivity properties make the coppers ideal for use in solar panels
and related tubing and piping used in solar energy coversion. These
same properties plus their resistance to engine coolants have made the
coppers suitable for use as radiators.
Page 91
Large amounts of copper are also used in the beverage industry,
particularly in the brewing and distilling operations.
In general, the coppers are generally resistant to
1. Seawater
2. Fresh waters, hot or cold
3. Deaerated; hot or cold, dilute sulfuric acid, phosphoric acid, acetic
acid, and other nonoxidizing acids
4. Atmospheric exposure
The coppers are not resistant to
1. Oxidizing acids such as nitric, hot concentrated sulfuric, and
aerated nonoxidizing acids (including carbonic acid)
2. Ammonium hydroxide (plus oxygen). A complex ion
Substituted ammonia compounds (amines) are also corrosive
forms.
3. High-velocity aerated waters and aqueous solutions
4. Oxidizing heavy metal salts (ferric chloride, ferric sulfate, etc.)
5. Hydrogen sulfide, sulfur, and some sulfur compounds
II
High-Copper Alloys
High-copper alloys contain a minimum of 95% copper if wrought and
94% copper if cast. The corrosion resistance of these alloys is
approximately the same as that of the coppers. They find their main
area of application in the electrical and electronics field. Application
in corrosion-resistant service is found when mechanical strength
greater than that of the coppers is required.
III
Copper-Zinc Alloys (Brasses)
Brass alloys which contain more than 15% zinc can be subject to
dealloying (dezincification). This is a type of corrosion in which the
brass dissolves as an alloy and the copper constituent redeposits from
solution onto the surface of the brass as a metal in porous form. The
zinc constituent may be deposited in place of an insoluble compound
or carried away from the brass as a soluble salt. The corrosion can
take place uniformly or be local. Uniform corrosion is more apt to
take place in acid environments while local corrosion is more apt to
take place in alkaline, neutral, or slightly acid environments. The
addition of tin or arsenic will inhibit this form of corrosion.
Conditions of the environment which favor dezincification are high
temperature, stagnant solutions, especially of acid, and porous
inorganic scale formation. Other factors which stimulate the process
are increasing zinc concentrations and the presence of both cuprous
and chloride ions.
Page 92
As the dealloying proceeds a porous layer of pure or almost pure
copper is left behind. This reaction layer is of poor mechanical
strength. The dezincification process on copper-zinc alloys is
therefore very detrimental. These alloys are also subject to stress
corrosion cracking. Moist ammonia in the presence of air will cause
this form of corrosion. The quantity of ammonia present need not be
great as long as the other factors are present.
Relative resistance of the brasses to stress corrosion cracking is as
follows:
Low resistance
Brasses containing > 15% zinc
Brasses containing > 15% zinc and small amounts of lead, tin, or
aluminum
Intermediate resistance
Brasses containing < 15% zinc
If the metal is cold-formed residual stresses may be present which can
also cause stress corrosion cracking. By heating the metal to a
temperature high enough to permit recrystallization the stresses will
be removed. It is also possible to provide a stress relieving anneal at a
lower temperature without substantially changing the mechanical
properties of the cold-worked metal.
Table 1 lists the compositions of some of the brasses which find
application in corrosion engineering. Copper alloys UNS C-44300
through C-44500 are known as admiralty brasses. They are resistant to
dealloying as a result of the presence of tin in the alloy.
Admiralty brass finds application mainly in the handling of seawater
and/or
Table 1 Copper-Zinc Alloys: Maximuma Composition (%)
Copper Alloy UNS No.
Cu
Pb Fe Zn Sn
Others
C-27000
63.068.5 0.10 0.07 Rem
C-28000
59.063.0 0.30 0.07 Rem
C-44300
70.073.0 0.07 0.06 Rem 0.81.2 0.020.10 As
C-44400
70.073.0 0.07 0.06 Rem 0.81.2 0.020.10 Sb
C-44500b
70.073.0 0.07 0.06 Rem 0.81.2
C-46400
59.062.0 0.20 0.10 Rem 0.51.0
C-46500
59.062.0 0.20 0.10 Rem 0.51.0 0.020.10 As
C-46600
59.062.0 0.20 0.10 Rem 0.51.0 0.020.10 Sb
C-46700b
59.062.0 0.20 0.10 Rem 0.51.0
C-68700c
76.079.0 0.07 0.06 Rem
0.020.10 As
aUnless shown as a range.
bAlso contains 0.020.10% P.
cAlso contains 1.82.5% Al.
Page 93
freshwater, particularly in condensers. These brasses are also resistant
to hydrogen sulfide and therefore find applications in petroleum
refineries.
Red brass is an alloy containing 15% zinc. It has basically the same
corrosion resistance of copper but with greater mechanical strength.
Most brass piping and fittings are produced from this alloy.
IV
Copper-Tin Alloys
Copper-tin alloys are known as tin bronzes or phosphor bronzes.
Although tin is the principal alloying ingredient, phosphorus is always
present in small amounts, usually less than 0.5% because of its use as
an oxidizer. Tables 2 and 3 list the principal tin bronzes used for
corrosion engineering.
These alloys are probably the oldest alloys known, having been the
bronzes of the Bronze Age. Even today many of the artifacts produced
during that age are still in existence. Items such as statues, vases,
bells, and swords have survived hundreds of years of exposure to a
wide variety of environments testifying to the corrosion resistance of
these materials.
Alloys which contain more than 5% tin are especially resistant to
impingement attack. In general, the tin bronzes are noted for their
high strength. Their main application is in water service for such items
as valves, valve components, pump casings, and so forth. Because of
their corrosion resistance in stagnant waters they also find wide
application in fire protection systems.
V
Copper-Aluminum Alloys
Copper-aluminum alloys are known as aluminum bronzes and find
application where combinations of strength, corrosion resistance, and
wear resistance are of importance. They are available in both wrought
and cast form. Tables 4 and 5 list the aluminum bronzes used in
corrosion engineering.
Table 2 Wrought Copper-Tin Alloys: Maximuma Composition (%)
Copper Alloy UNS No.
Zn
Cu Pb Fe
Sn
P
C-51000
Rem 0.05 0.10 4.25.8 0.30 0.030.35
C-51100
Rem 0.05 0.10 3.54.9 0.30 0.030.35
C-52100
Rem 0.05 0.10 7.09.0 0.20 0.030.35
C-52400
Rem 0.05 0.10 9.011.0 0.20 0.030.35
C-54400
Rem 3.54.5 0.10 3.54.5 1.54.5 0.010.50
aUnless shown as a range.
Page 94
Table 3 Cast Copper-Tin Alloys: Maximuma,b Composition (%)
Copper Alloy UNS No.
Cu
Sn
Pb
Zn Fe Sb Ni S P
C-90300
86.089.0 7.59.0 0.30
3.05.00.200.201.0 0.050.05
C-90500
86.089.0 9.011.0 0.30
1.03.00.250.201.0 0.050.05
C-92200
86.089.0 5.56.5 1.02.0 3.05.00.250.251.0 0.050.05
C-93700
78.092.0 9.011.0 8.011.0 0.8 0.150.551.0 0.080.15
C-93800
75.079.0 6.37.5 13.016.0 0.8 0.150.8 1.0 0.080.05
C-93900
76.579.5 5.07.0 14.018.0 1.5 0.4 0.500.8 0.081.5
C-94700c
85.090.0 4.56.0 0.10
1.02.50.250.154.56.50.050.05
aUnless shown as a range.
bIn addition to the alloying ingredients shown each alloy also contains 0.0005%
aluminum and 0.0005% silicon.
cMn 0.20%.
Page 95
Table 4 Wrought Copper-Aluminum Alloys: Maximuma Concentration
(%)
Copper Alloy UNS
No.
Cu
Al
Fe
Ni Mn Si
Sn
Others
C-60800
92.594.85.06.5 0.10
0.020.35
As 0.10
Pb
C-61000
90.093.06.08.5 0.50
0.10
0.02 Pb
C-61300
88.692.06.07.5 2.03.00.15 0.100.100.200.500.01 Pb
C-61400
88.092.56.08.0 1.53.5
0.01 Pb
C-61500
89.090.57.78.3
1.82.2
0.015 Pb
C-61800
86.991.08.511.00.51.5
0.10
0.02 Pb
C-62300
82.289.58.511.02.04.01.0 0.5 0.250.60
C-63000
78.085.09.011.02.04.04.05.51.5 0.250.20
C-63200
75.984.58.59.5 3.05.04.05.53.5 0.10
0.02 Pb
aUnless shown as a range.
Page 96
Table 5 Cast Copper-Aluminum Alloys: Maximuma
Concentrations (%)
Copper Alloy UNS
No.
Cub Al
Fe
Ni
Mn
C-95200
86.0 8.59.5 2.54.0
Others
1.0
total
C-95300
86.0 9.011.0 0.81.5
1.0
total
C-95400
83.010.011.5 3.05.0 2.5 0.5
0.5
total
C-95500
79.010.011.5 3.05.0 3.05.53.50
0.5
total
C-95700c
71.0 7.08.5 2.04.0 1.53.011.014.00.5
totale
C-95800c
79.0 8.59.5 3.54.5d4.05.00.81.5 0.5
totalf
aUnless shown as a range.
bMinimum.
cContains maximum of 0.01% Si.
dIron content shall not exceed the nickel content.
eMaximum 0.03% Pb.
fMaximum 0.02% Pb.
Alkalis such as sodium and potassium hydroxides can be handled.
When constantly immersed in concentrations of 145% sodium
hydroxide at temperatures from room to 125°F (52°C) the corrosion
rate will be less than 5 mpy for alloys C-61400 and C-95300. At
175°F (80°C) the corrosion rate will range from 0.1 to 8.1 mpy in
concentrations of 4510%.
Table 6 Chemical Composition of the Cupronickels (%)
Copper Alloy UNS
No.
Cu Ni
Fe
Mn
Others
C-70600
Bal 9.011.0 1.01.8 1.0 Pb 0.05
max max
Zn 1.0 max
C-71500
Bal29.033.0 0.400.7 1.0 Pb 0.05
max max
Zn 1.0 max
C-71900
Bal29.032.0 0.25 0.51.0 Cr 2.63.2
max
Zr 0.080.2
Ti 0.020.08
C-96200a
Bal 9.011.0 1.0 1.8 1.5 Pb 0.03c
max max
C-96400b
Bal28.032.0 0.251.5 1.5 Pb 0.03c
max max
a1.0% max niobium; 0.25% max silicon.
b1.0% max niobium; 0.70% max silicon.
cFor welding grades Pb must not exceed 0.01%.
Page 97
These alloys are resistant to nonoxidizing mineral acids such as
sulfuric and phosphoric. Resistance is controlled by the presence of
oxygen or an oxidizing agent. At room temperature up to 90°F (32°C)
the corrosion rate of alloy C-95400 in sulfuric acid concentrations
ranging from 0.5% to 50% will be less than 5 mpy.
Aluminum bronzes are also resistant to many organic acids such as
acetic, citric, formic, and lactic. When in contact with these materials
it is possible to experience copper pickup in the finished product.
Although the concentration is very low, discoloration of the product
may occur. The color contaminant can easily be removed by a carbon
filtration polishing step.
This alloy also finds many applications in the handling of seawater for
such items as valves, fittings, and pump casings, particularly in
desalination plants aboard naval and commercial vessels.
Aluminum bronzes are not affected by pitting, crevice corrosion, or
stress corrosion.
VI
Copper-Nickel Alloys
Copper-nickel alloys are referred to as cupronickels and are useful in
waters ranging from fresh to brackish and sea. Their biofouling
resistance is excellent. Corrosion rates for alloys C-70600 and C71500 in seawater are approximately 1 mpy. They find their greatest
use in saltwater service where they are used as piping, fittings,
condenser tubes and plates, and pump casting.
Of all the copper alloys, the copper nickels are the most resistant to
stress corrosion cracking in the presence of ammonia and ammonical
solutions and are highly resistant to stress corrosion cracking in
general. Table 6 lists the cupronickels used for corrosion engineering.
Page 99
7
Corrosion of Aluminum and Aluminum Alloys
Bernard W. Lifka
New Kensington, Pennsylvania
I
Passivity of Aluminum and Alloys
Aluminum is the most prevalent metallic element (about 8%) in the
solid portion of the earth's crust; but it always occurs in a combined
form, usually a hydrated oxide, of which bauxite is the principal ore.
Thermodynamically, metallic aluminum is very active and seeks to
return to the natural oxidized state through the process of corrosion.
The activity of aluminum can be appreciated when one considers that
fine aluminum powder undergoing rapid chemical oxidation is the
primary fuel in modern aerospace rockets.
A
Oxide Coating
Aluminum attains high resistance to corrosion in many environments
because of very rapid formation of a thin, compact, and adherent
oxide film over the surface that limits further corrosion. The normal
surface film formed in air is only about 5 nm (50 Å) thick. The film is
thickened by formation at elevated temperatures and when formed in
the presence of water or water vapor. This oxide film is stable
(insoluble) in the pH range of about 49, which includes many
atmospheric and aqueous environments. The film dissolves at lower
and higher pH, with some notable exceptions such as stability in
concentrated nitric acid (pH 1) and in
Page 100
concentrated ammonium hydroxide (pH 13). The high resistance to
nitric acid is noteworthy in that a 1-to 5-min immersion in commercial
strength nitric acid (specific gravity of 1.42) is an approved method
for cleaning of aluminum without excessive removal of the base metal
[1].
The oxide film is not homogeneous and contains weak points.
Breakdown of the oxide film at weak points leads to the onset of
localized corrosion. The oxide film becomes more nonhomogeneous
with increasing alloying content, and on heat-treatable alloys as
opposed to non-heat-treatable alloys.
II
Metallurgy of Aluminum and Alloys
A
Composition
The world aluminum is used generically for both the pure metal and
for the alloys, but almost all commercial products are alloys.
Examples where pure aluminum is used are superelectrical
conductors, capacitor foil, some rotogravure foil, and as a cladding
material on aluminum Alclad sheet.
Commercially, a minimum aluminum content of 99.00% or greater is
considered as ''unalloyed" aluminum. Most unalloyed specifications
range from 99.00% to 99.75% minimum aluminum, but experimental
metal has been produced with ultrapurity as high as 99.9990%.
Elements such as bismuth and titanium are intentionally added to
assist in the smelting process, while such others, as chromium,
manganese, and zirconium are used for grain control during
solidification of large ingots. Elements such as copper, magnesium,
manganese, nickel and zinc are added to attain desired properties such
as strength, formability, stability at elevated temperatures, etc. Some
elements are present as unintentional impurities coming from trace
elements from the ore, from pickup from ceramic furnace linings, or
from the use of scrap metal in recycling. For example, most 2XXX
aluminum-copper alloys permit up to 0.25% zinc as an impurity,
which could occur if some 7XXX aluminum-zinc alloy scrap was
used during a remelt.
The Aluminum Association (AA) publishes a handbook that provides
the nominal composition, the allowable composition range, and other
property data for the most frequently used unalloyed aluminum and
aluminum alloys [2]. This handbook can be obtained at a nominal cost
from the AA and is a must for aluminum alloy designers as it contains
much general information on temper designations, fabrication
techniques, quality control, nomenclature, and standards. The full
listing of composition limits for registered alloys are contained in Ref.
3 for wrought aluminum and wrought alloys, and in Ref. 4 for
aluminum alloys in the form of castings and ingot.
The user should be aware that there are three types of composition
listings in use. First there is the nominal, or target, composition of the
alloy. This is used
Page 101
Table 1 Nominal (Target) Chemical Composition (%) of
Wrought Alloysa
AlloySi Fe Cu
Mn
Mg
Cr Zn
Ti
1160 99.60% minimum aluminum. All other elements £
0.40%.
2024
4.4
0.6
1.5
3004
1.2
1.0
aAluminum and normal impurities constitute remainder.
in discussing generic types of alloys and their uses, etc. Next there are
the alloy limits registered with the AA. These are the specification
limits against which alloys are produced. In these limits, intentional
alloying elements are defined as an allowable range. The usual
impurity elements are listed as the maximum amount permissible, and
rare trace elements are grouped into an each other category. Each
trace element cannot exceed a specified "each" amount, and the total
of all trace elements cannot exceed the slightly higher "total" amount.
Finally, when a particular sample is analyzed, there is the listing of the
elements actually present. Examples of all three types of these
compositions are shown in Tables 13 for unalloyed aluminum (1160),
a heat-treatable alloy (2024), and a non-heat-treatable alloy (3004).
When a customer obtains an analysis of a particular sample, the
intentional elements must be within the prescribed range, but not
necessarily near the midpoint if the range is wide. For example, as
shown in Table 3, copper in 2024 alloy can be skewed to high content
(sample 2) to enhance strength or low content to enhance toughness
(sample 3). When the allowable range is only about 0.5 percentage
point or less, the producer will be targeting for the nominal
composition.
Specified impurity elements have to be at or below the maximum
limit. Individual nonspecified impurities should be less than the
"0.05% each" level
Table 2 Registered Chemical Composition Limits (%) of Wrought
Aluminum Alloysa
Others
Alloy Si Fe
Cu
Mn
Mg Cr Zn Ti Each Total
1160 0.25 0.35 0.05
0.03
0.03
0.05 0.03 0.03
2024 0.50 0.50 3.84.9 0.300.9 1.21.8 0.10 0.25 0.05 0.05 0.15
3004 0.30 0.70 0.25 1.01.5 0.81.3
0.25 0.05 0.05 0.15
aAlloying elements shown as a required range, impurity elements as the
maximum tolerable. Aluminum and trace impurities constitute
remainder.
Page 102
Table 3 Analyses of Particular Aluminum Samples, Weight Percent of
Alloying Elements Actually Presenta
Alloy No. Si
Fe
Cu
Mn Mg
Cr
Ni
Zn
Ti
1160 1 0.080 0.100 0.000 0.000 0.000 0.000 0.000 0.000 0.020
2024 2 0.25 0.32 4.77 0.61 1.77 0.000 0.000 0.025 0.03
2024 3 0.10 0.12 4.40 0.55 1.45 0.000 0.000 0.000 0.02
3004 4 0.30 0.42 0.00 1.25 1.10 0.000 0.000 0.000 0.05
aReminder is aluminum.
with the total less than 0.15%. The customer must realize that, as
shown in Table 3, some impurity elements will be present, but usually
well below the allowed limit. Not all of the allowable impurity
elements will be present in every sample. However, some amount of
iron and silicon is virtually inevitable, except in ultrarefined pure
aluminum. Certain alloys are produced in several purity variants, with
the less pure versions minimizing cost and the higher purity versions
enhancing some property. For example, it is well known that
toughness is enhanced when the iron and silicon levels are both less
than 0.10%.
The important thing is to know that other metallic elements are
present and necessary for desired properties. Many elements combine
with one another and with aluminum to produce intermetallic
compounds that are either soluble or insoluble in the aluminum
matrix. The presence of second-phase particles (sometimes called
constituent particles) is normal and they can be seen and identified by
metallographic examination. Intermetallic particles are present even in
commercial pure aluminum, i.e., 1100 aluminum, and in relatively
dilute, nonheat-treatable alloys such as alloy 3004 (Fig. 1). The more
highly alloyed heat-treatable alloys contain significantly larger
amounts of both soluble and insoluble particles (Fig. 2).
Many intermetallic particles have an electrochemical activity different
from that of pure aluminum or aluminum solid solutions. Examples,
of the electrode potentials of pure (99.95%) aluminum, some
aluminum solid solutions and some constituent particles can be found
in the literature [5]. The referenced examples span about 1 V and the
potential of the pure metal is close to the midpoint. Thus particles can
be an anodic or a cathodic site in the metal, thereby promoting
localized corrosion. Processing methods affect the size, amount, and
distribution of the particles, and this will alter resistance to corrosion.
Alloys will show a variety of microstructural features depending on
the processing route. Experience is needed to assess whether a
particular structure is normal, desirable, and how it affects resistance
to corrosion.
Page 103
Figure 1
Photomicrograph of a cross-section from 3004-H19 sheet showing the
presence of intermetallic particles. Particles are primarily small
Al12(MnFe)3Si dispersoid particles with a few larger Al6(FeMn)
constituent particles (as polished).
(Courtesy of Alcoa Technical Center, Lawrence Johnston.)
B
Cast Alloys
Aluminum can be fabricated but virtually all of the conventional
casting processes, e.g., vacuum and high-pressure die casting,
permanent mold casting, sand, plaster, and investment mold castings.
There is no single commercial designation system for castings, but the
three-digit registration system of the AA is the most widely used [4].
Other systems have been developed by the American Society for
Testing and Materials (ASTM) [6] and the Society of Automotive
Engineers (SAE) [7]. In addition, there are some proprietary
nomenclatures.
The AA system contains designations for pure aluminum ingot, used
as the starting material (pig) for the smelting of cast or wrought
alloys, as well as designations of final alloys used to produce cast
products. The AA designation system, shown in the table at the top of
page 105, is based on the major alloying element(s).
Page 104
Figure 2
(a) Photomicrographs of 7075-T6 in the as-polished condition showing
intermetallic phases present that consist of both insoluble and soluble
particles (as polished).
(b) Same sheet as in (a), but etched to reveal the fine precipitates from
aging and the grain structure. Note the elongated, pancake-shaped grains
(etch: Graff/Sargent's).
(Courtesy of Alcoa Technical Center, Lawrence Johnston.)
Page 105
Series
Alloy system
1XX 99.9% minimum aluminum
2XX Aluminum + copper
3XX Aluminum + silicon +
magnesium
Aluminum + silicon + copper
Aluminum + silicon + copper +
magnesium
4XX Aluminum + silicon
5XX Aluminum + magnesium
7XX Aluminum + zinc
8XX Aluminum + tin
Much less corrosion testing has been done on cast alloys compared
with wrought alloys. However, the corrosion resistance of castings
generally is comparable to that of similar wrought alloys. The grain
structure of castings tends to be equiaxed, which precludes
susceptibility to exfoliation corrosion. Some instances of stress
corrosion cracking (SCC) have occurred with the higher strength
alloys, particularly in the 2XX Al-Cu series. However, the level of
resistance to SCC is more like that of a transverse T-grain structure
rather than the more susceptible short-transverse ST-grain structure,
which will be discussed later for wrought alloys. Some general
remarks follow for each class of cast alloys.
The 1XX series is assigned to pure aluminum. Besides ingot, the only
major commercial use of pure aluminum castings is electrical
conductor parts such as collector rings and bus bars. Because of the
low strength, these products usually are cast with integral steel
stiffeners. Electrical conductors have high resistance to general
corrosion, even when used outdoors, because they operate at slightly
elevated temperatures due to the current flow. This prevents
condensation and keeps the surfaces dry.
2XX, aluminum + copper alloys were the first type of casting alloys to
be used commercially and they still are used appreciably. They
provide medium to high strengths but are somewhat difficult to cast.
These alloys are the least resistant to corrosion and can be susceptible
to SCC in the maximum strength T6 temper. These alloys are
produced in both as-cast (F) tempers and heat-treated tempers (T4
through T7).
3XX, Al-Si-Mg, Al-Si-Cu-Mg provide the best combination of
strength and good corrosion resistance. These alloys are produced in
both as-cast (F) tempers and heat-treated tempers (T5 through T7).
4XX, aluminum-silicon castings are the most prevalent because of
their superior casting characteristics. They provide reasonable good
resistance to corrosion but low to medium strength.
5XX, aluminum-magnesium castings provide the highest resistance to
corrosion, good machinability, and weldability. However, they also
have low to
Page 106
medium strength and are more difficult to cast, generally being
limited to sand castings or simple permanent mold shapes.
7XX, aluminum-zinc castings are used in limited applications. They
are difficult to cast and limited to simple shapes. They have medium
to good resistance to corrosion and high melting points, which makes
them attractive for brazed assemblies. The as-cast F temper will
naturally age to relatively high strength within a few weeks to a month
at room temperature. Alternatively, the same or higher strength can be
attained by heat treatment to T6 or T7 tempers.
8XX, aluminum-tin castings were designed for bearings and bushings
in internal combustion engines. Required properties are the ability to
carry high compressive loads and good fatigue resistance. The typical
environment is internal combustion lubricating oils (plus inherent
contaminants). In this environment, all 8XX alloys have exhibited
resistance to corrosion superior to competitive materials. The alloys
usually are produced in a heat-treated T5 temper. Some T5 castings
are subsequently compressively cold-worked about 4% to provide
improved compressive yield strength.
C
Wrought Ingot Metallurgy Alloys
1
Alloy Designation
All American producers follow the AA designations for wrought
aluminum products [3]. The general system is shown below, with the
primary alloying element listed first, followed by other usual
elements. An 8XXX series is reserved for miscellaneous alloys not
covered by the 1XXX to 7XXX groupings.
Series
Alloy series
1XXX99.9% minimum aluminum
2XXXAl-Cu, Al-Cu-Mg, Al-Cu-Mg-Li, and
Al-Cu-Mg-Si
3XXXAl-Mn and Al-Mn-Mg
4XXXAl-Si
5XXXAl-Mg and Al-Mg-Mn
6XXXAl-Mg-Si, Al-Mg-Si-Mn, and Al-MgSi-Cu
7XXXAl-Zn, Al-Zn-Mg, Al-Zn-Mg-Mn, and
Al-Zn-Mg-Cu
2
General Fabrication
Most wrought products start out as a section from a continuously cast,
rectangular or round ingot. Ingots are mechanically worked into
usable shapes by hot and cold rolling, extruding, or forging. Ingots
have a cellular structure that tends to be equiaxed. In the as-cast
condition, high concentrations of intermetallic particles
Page 107
are present in the interstices of the cellular structure. Ingots are
subsequently homogenized (preheated) to reduce this condition.
Figures 3 and 4 show an ingot structure before and after preheating.
The amount of preheating depends primarily on the alloy and
subsequent hot working method, but characteristics desired in the final
product may require extended preheats or special temperatures.
Rectangular ingots for sheet and plate range from about 20 to 60 cm
thick and can be over 280 cm in length. A typical commercial ingot
can exceed 4500 kg (10,000 pounds) in weight. Another process used
for a more economical production of thin (less than about 75 mm)
sheet and foil is continuous casting of a thin slab (approximately 612
mm), which is immediately rolled to the final sheet or foil thickness.
The size of the starting ingot and the casting method affects the
metallurgical structure and, pending further processing, usually affects
resistance to corrosion. Thin sheet and foil produced from a large-size,
precast ingot has a different structure, and often improved corrosion
resistance, compared to similar gage products produced by the
continuous roll-cast process.
Figure 3
Photomicrograph of an as-cast 6010 alloy ingot showing the large amounts
of both soluble and insoluble particles at interstitial sites in the ingot
dendrite structure. This ingot condition is not suitable for subsequent hot
rolling. Rapid heating to the hot rolling temperature could cause melting of
the soluble phases, leading to internal splitting in the metal (as polished).
1 inch = 100 µm.
(Courtesy of Alcoa Technical Center, Lawrence Johnston.)
Page 108
Figure 4
Photomicrograph of same ingot as in Fig. 3, but after preheating for 16 h at
575°C to dissolve the soluble phases. Only the insoluble phases remain
and the ingot can now be safely hot-rolled (as polished). 1 inch = 100 µm.
(Courtesy of Alcoa Technical Center, Lawrence Johnston.)
Round ingot range in diameter from about 15 to 90 cm. Long lengths
are cast and then cut to an appropriate length (called a billet) required
for a particular extrusion or forged part.
Forged parts can be produced from many types of starting stock.
Billets can come from rectangular ingot, rolled plate, cast or extruded
rounds, and preforged billet. Two general forging processes exist: (1)
hand forgings which produces simple geometrical shapes, and (2)
closed die forging which produces complex shapes of varying
thickness. The precision closed die process can produce die forgings
of limited size with dimensions at or very close to those of the
finished part.
Effect on Grain Morphology
There are three principal ways in which the fabrication can affect the
grain morphology, which is likely to affect resistance to certain types
of corrosion. First the cellular ingot structure is changed to a grain
structure greatly elongated in the principal (longitudinal) direction of
working but also changed in the other grain coordinates. Grain
morphology tends to conform with the physical dimensions of the
final product. Hence rectangular products
Page 109
such as plate and sheet have grains with a long, wide, thin platelet
shape. Such grains have directions designated: longitudinal (L)
(primary direction of working), long-transverse (LT) (width) and
short-transverse (ST) (thickness). Mechanical properties and corrosion
resistance can vary in each of these directions, tending to be highest in
the L direction and lowest in the ST direction. An example of such a
grain structure is shown in Fig. 5.
Products having a cross-section with a low aspect ratio of 2 or less
(such as a bar or a round, square, hexagonal shape, etc.) have
elongated grains, but with nearly the same width and thickness. These
grain directions are designated longitudinal (L) and transverse (T),
making no distinction between the width and thickness dimensions.
An example of such a grain structure is shown in Fig. 6. For most
properties, the T-direction property is similar to that of LT properties.
One important exception is for alloys susceptible to SCC in which
case the
Figure 5
Composite photomicrographs taken at the center of 38-mm plate of 7075-T651
alloy, illustrating the grain structure in the three principal directions. This long,
wide, and thin grain structure is typical of thick, unrecrystallized plate of most
alloys (etch: Keller's).
(Courtesy of Alcoa Technical Center, John J. Liput.)
Page 110
Figure 6
Composite of photomicrographs of 25-mm diameter rolled rod of 7075-T651
alloy. The grains are relatively long in the longitudinal, rolling direction,
but equiaxed in the cross-section (etch: Keller's).
(Courtesy of Alcoa Technical Center, John J. Liput.)
resistance of a T-grain structure is intermediate between that of an LT
and an ST grain structure [8].
In extrusions and die forgings with complex cross-sections, the grain
flow in the cross-section tends to follow the product configuration
(Fig. 7).
Secondly, grains will be either unrecrystallized or recrystallized
depending on the amount and temperature of working. Large amounts
of working and lower temperatures promote recrystallization during
subsequent heating. These factors, plus the rate of heating to
temperature, also control the grain size. From a corrosion standpoint,
the preferred grain structure depends on the type of corro-
Page 111
Figure 7
Photomicrograph of the cross section of one half of a symmetrical, dumbbellshaped, 7050T74 die forging. Note how the flow is affected by the forging contour.
It is possible to position vertical specimens so that the grain flow would be either
diagonal or perpendicular to the specimen axis. Marked differences in grain flow
can affect resistance to SCC and, sometimes, the original tensile properties
(etch: Graff-Sargent).
(Courtesy of Alcoa Technical Center, David A. Lukasak.)
sion one expects to encounter and wants to prevent. Generally, a
large-grain, fully recrystallized grain structure is the least favorable.
Finally, the thermal treatments (duration, temperature, and rate of
temperature change) can drastically alter the amount, size, and
distribution of both soluble and insoluble intermetallic particles.
Depending on the alloy and final temper, this can have no effect, or a
major effect, on the resistance to certain types of corrosion.
D
Strengthening Mechanisms
Pure aluminum is strengthened only by cold working. Aluminum
alloys can be further strengthened by solid solution strengthening, up
to the solid solubility limit of the strengthening element. Heattreatable alloys are appreciably strengthened by solution heat
treatment and by precipitation heat treatment.
1
NonHeat-Treatable Alloys
If an element has appreciable solubility (a few percent) in solid
aluminum at relatively low temperatures (about 100°C or less), then
its addition to aluminum strains the lattice thereby increasing strength.
Elements with an appreciable effect are magnesium, cooper, silicon,
and silver, in that order. Magnesium is the element most used
commercially and additions of up to 3 wt % are added without any
effect on corrosion. An example is shown in Table 4. The O temper
strength increasing from 70 to 195 MPa as percent magnesium is
increased from 0 to
Page 112
Table 4 Examples of Strengthening Aluminum
Alloy Sheet by (1) Solid Solution Hardening
Through Increased Magnesium Content and (2)
Cold Working by Rolling at Temperature of 50°Ca
H34
H38
Nominal O temper temper
temper
wt %
ultimate ultimate ultimate
Alloymagnesium T.S., MPa T.S., MPa T.S., MPa
1100
0.0
70
100
130
5005
0.8
125
160
200
5050
1.4
145
190
220
5252
2.5
195
260
290
aThe O temper receives no cold reduction, the
H34 temper about 38% cold reduction, and the
H38 about 75% cold reduction.
2.5%. Likewise the H34 temper increased from 100 to 260 MPa and
the H38 temper from 130 to 290 MPa. (Compare columns in Table 4
vertically for the effect of solid solution strengthening by the addition
of magnesium.)
When the alloying element remains in solid solution its addition has
little affect on resistance to corrosion. Adding too much of an element
causes it to precipitate out as a separate phase, which generally has an
adverse affect especially if the second phase is segregated to particular
regions rather than being randomly distributed. This can be a problem
with the 3.55.5% magnesium content alloys when used for long times
at temperatures in the range of 80175°C. Exposure at even higher
temperatures is less of a problem because this agglomerates the
precipitate into fewer particles and reduces the adverse effect.
All alloys can be cold-worked to increase strength. This has little
affect on their resistance to corrosion but decreases properties such as
elongation, forming, and toughness. Table 4 shows how the strength
of the four alloys increases when the fully annealed O temper is coldrolled to the H34 and H38 tempers. (Compare horizontal rows in
Table 4 for the effect of strengthening by cold work.) Excessive cold
working can cause banded slip planes in the metal. These planes are in
a higher thermodynamic state and can be more susceptible to
corrosion. Also they contain numerous dislocations that act as sites for
localized precipitation leading to localized corrosion. This mechanism
can cause susceptibility to transgranular exfoliation in 5XXX alloys
containing more than 3% Mg after exposure above 80°C. It also has
caused localized midplane corrosion problems in some thin 3004
sheet produced by the roll caster process.
2
Heat-Treatable Alloys
Solution heat treatment is complex, but basically it involves a eutectic
or peritectic alloy in which the solubility of the alloying element is
much greater at
Page 113
high temperatures (370545°C) than at room or slightly elevated
temperatures. Holding the metal at an elevated temperature above the
Solvus temperature causes the element to dissolve. Rapid cooling
(quenching) to room temperature retains much of the dissolved
alloying element in a supersaturated solid solution, the actual amount
retained depends on the degree of supersaturation, the mobility of the
particular element, and the rate of cooling. Al-Cu alloys were the first
known and still are an extensively used heat-treatable alloy system.
This system requires a cooling rate greater than 560°C/s to retain all
copper in solid solution, which is attainable only in thin sheet or small
parts, such as rivets. However, cooling rates of 170°C/s or faster will
develop substantially improved strengths. However, as the copper
precipitates out of solid solution, it forms intermetallic particles which
create new surfaces. Consequently, these particles tend to form first
along grain boundaries, which results in susceptibility to intergranular
corrosion. Intergranular corrosion and its cause will be discussed later
in more detail. In the as-quenched condition, the temper designation is
T4 when there is no subsequent cold work and T3X for subsequent
cold work; the second digit indicates a specified amount of cold work.
When an as-quenched alloy is held for a sufficient time at an elevated
temperature below the Solvus temperature of the respective alloy
(usually in the range 100200°C), fine precipitates form not only at the
grain boundary, but randomly distributed on lattice vacancies
throughout the grains. This process is called precipitation heat
treatment or artificial aging. It greatly improves mechanical properties
and, if the amount of aging is sufficient, improves corrosion resistance
by eliminating the tendency for localized intergranular corrosion.
The artificial aging time is inversely proportional to the temperature
and can vary greatly with individual alloys, even involving multiple
step treatments. For most 2XXX alloys the introduction of small
amounts of cold work prior to aging increases the local strain and
dislocations around the insoluble intermetallic Al-Mn particles,
greatly accelerating the aging response. Table 5 illustrates how alloy
2024 (nominal 4.4% Cu) can be strengthened from the fully annealed
O temper by (1) solution heat treatment (T42), (2) the addition of cold
work (T31), (3) precipitation heat treatment (T63), and (4) cold work
plus precipitation heat treatment (T81 and T86).
The size, shape, and distribution of the intermetallic particles that
capture the dislocations and act as nucleation sites for precipitation is
determined by high-temperature treatments early in the fabrication
stage. Hence ingot preheating and subsequent fabrication temperatures
can affect the performance of the final product. Thus it is not
surprising to see differences between large ingot metallurgy products
which receive separate preheat (homogenization) treatments,
compared with that of continuous roll-cast products, which are not
homogenized prior to rolling.
Other heat-treatable alloys are the 6XXX aluminum-silicon and the
7XXX
Page 114
Table 5 Examples of Strengthening Aluminum 2024 Alloy
Sheet by (1) Solution Heat Treatment (SHT) to the T42
and T31 Tempers and (2) SHT Plus Precipitation Heat
Treatment (Artificial Aging) to the T62, T81, and T86
Tempersa
Temper
Mode =
O
T42 T31 T62 T81
T86
UTS =
220 425 440 435 460
480
TYS =
95
260 290 345 400
440
aUltimate tensile strength (UTS) and ultimate 2% yield
strength (TYS) mega-pascals.
Note: The T31 and T81 tempers are stretched 13% cold
work, and the T86 temper is stretched 610% cold work
prior to aging. The O temper is used as a baseline
representing no thermal nor mechanical strengthening.
aluminum-zinc systems. Cold work prior to aging has a only a small
affect on the strengths of both 6XXX and 7XXX alloys.
6XXX alloys are generally used in the artificially aged, peak strength,
T6 temper. However, they are frequently supplied in the T4 temper
because they have superior formability in this temper. The user will
subsequently apply artificial aging or have this performed by an
independent heat treater. The room temperature interval between
quenching and artificial aging is not critical and can be as long as 6
months to a year.
Many 6XXX alloys can tolerate a less rapid quench. Consequently,
certain products like extrusions sometimes are fan-quenched directly
from the fabrication temperature and then artificially aged to the T5
temper without a separate SHT step. The advantage is lower
production cost and less warpage during quenching. Slightly lower
strengths of about 2035 MPa (35 ksi) are obtained, but there is no
significant effect on resistance to corrosion.
6XXX alloys are the easiest of the heat-treatable alloys to fabricate,
which reduces their respective cost. Alloys 6061-T6 and 6063-T6 are
the workhorse aluminum alloys for most general applications because
of their moderately high strength, good machinability, and high
resistance to corrosion in natural environments. Alloy 6061-T6 is
slightly higher alloyed and contains a nominal 25% copper. Hence
6061-T6 provides somewhat higher tensile properties, on the order of
6070 MPa (810ksi); but 6063-T6 has notably higher resistance to
pitting corrosion.
7XXX alloys are not used in the as-quenched ''W" temper because
they are not metallurgically stable and gradually increase in strength
by natural aging at room temperature. This process has been followed
on alloys 7075 and 7050 and showed a continual change for more
than 20 years. Other characteristics also are changed by the natural
aging process; consequently, 7XXX products
Page 115
typically are artificially aged to a stable temper within a month or two
after quenching. A minimum 4-day room temperature interval
between quenching and artificial aging often is specified to ensure
attaining a more consistent strength in the final temper.
The peak strength T6 temper provides the highest strength in
aluminum alloys. However, many 7XXX-T6 products can be
susceptible to exfoliation or to SCC if stressed in the T or ST
directions. Consequently, alloys such as 7075, 7050, 7150, and 7055
generally are aged beyond maximum strength to a variety of T7-type
tempers (e.g., T73, T74, T76, and T77) to provide the highest strength
attainable together with a specified degree of resistance to corrosion.
E
Powder Metallurgy Alloys
Powder metallurgy, generally using atomized aluminum flake powder,
subsequently ball-milled with other metal powders, is a method of
introducing higher amounts of an alloying element than can be
attained by molten ingot metallurgy. The intent is either lower
fabrication cost or improved properties, such as better retention of
strength at high temperature.
1
Pressed and Sintered Powder Metallurgy Parts
Aluminum powder metallurgy parts can be produced by pressed,
compacted, and sintered processes, similar to those used for powder
iron parts. The advantage of the aluminum powder parts is low cost,
near-finish dimensions in the compacted part, together with higher
compaction, lower weight, and better corrosion resistance than
comparable iron parts. A number of alloys exist which, for the most
part, use a proprietary numbering system [9].
2
Wrought Powder Metallurgy "P/M" Products
In addition to the conventional powder products, the Aluminum
Company of America developed a process whereby the compacting
and sintering process was used to produce a moderate size round or
rectangular billet. These "powder metallurgy billets" (P/M) were then
fabricated and heat-treated to final products by extrusion and by die
forging. Some experimental plate was rolled. Alloys 7090 and 7091
were developed in an attempt to provide a better combination of
strength and toughness than that provided by conventional ingot
alloys. An Al-Fe-Ce alloy was also developed for maximum retention
of strength at temperatures of 200350°C. The resistance to corrosion
of all three alloys was high.
P/M alloys contain dispersed aluminum oxide through the product
resulting from the oxide layer on the many fine aluminum powder
particles; hence chemical analyses will detect oxygen. P/M alloy
products have not been used extensively, to some extent because of
the increased processing cost, but mainly because the size of the
starting billet developed to date is much smaller than
Page 116
convention ingot and greatly limits the size of parts that can be
produced. Rolled P/M sheet and plate are not commercially available.
III
Selection of Aluminum and Aluminum Alloys for Specific
Applications
The volume of aluminum metal needed affects the alloy selections
available. Metal for applications requiring many thousands of pounds
most likely would be purchased directly from an aluminum producer.
In such cases all alloys and tempers can be considered, and the
endproduct produced at or close to the final dimensions and other
requirements met.
For developmental projects, major suppliers frequently keep a small
inventory of their newest and developmental products that can be
obtained in small quantities for trial and evaluation purposes.
For applications requiring a small amount of metal, the material most
likely will have to be purchased from distributors who warehouse the
most commonly used alloys, products, and gages. Usually a few
alloys of each alloy system are available, and common size gauges of
sheet, plate, rod, and so forth are stocked. The user procures the
nearest gauge available or a sufficiently thick bar or plate stock to
machine the finished part. Independent heat-treating sources are
available, so some distributors stock only the more basic tempers and
apply additional artificial aging as needed. The advantage of using
distributors are quick delivery, "cut-to-size" dimensions, and small
size orders of even a fraction of a square meter in size.
Surprisingly, resistance to corrosion rarely if ever is the primary
deciding factor in the choice of an alloy for a particular application.
Decisions are based primarily on physical properties required to
produce the part and that are needed for the end applicationbe it high
electrical conductivity for bus bar, or high strength and long fatigue
life for a jet fuselage skin, etc. Once physical properties are specified,
procurement cost and availability are important considerations.
Having set the physical property requirements, producers generally
want the most resistant alloy attainable that is consistent with
engineering properties needed and which does not drastically increase
cost. As might be expected, compromises often are necessary and
protective finishes and treatments are considered when less resistant
alloys must be used. Designers should give careful consideration to
durability of coating or treatment. The frequency and ease of
maintenance procedures must be considered for assemblies that are to
provide long service life, as contrasted with disposable and easily
replaced parts. The cost of the total service life often can justify the
increased procurement cost of a more resistant alloy or temper. As
with most metals, corrosion of aluminum frequently initiates within
joints, on internal and faying surfaces and the like, so that periodic
maintenance can be costly or even
Page 117
impossible. The longer the life expectancy and the more critical the
application, the more thought need be given to product design and
alloy selection.
IV
General Resistance to Corrosion
This section discusses the relative performance of the various classes
of wrought aluminum alloy, as was previously done for cast alloys.
The alloys are discussed in order of decreasing resistance to corrosion
rather than by strict numerical sequence.
All of the non heat-treatable alloys are resistant enough to be
effectively usable in many applications without surface protection.
Protection, such as an anodic coating, often is applied both to extend
life in more corrosive environments and to retain a bright, decorative
appearance even in mild atmospheres. Regarding the heat-treatable
alloys, this is true only for the 6XXX alloys. Protective measures
appropriate to the particular application are always recommended for
the high-strength 2XXX and 7XXX alloys.
A
Non-Heat-Treatable Alloys
1
1XXX Pure Aluminum
Pure aluminum (99.00% or purer) is more resistant than any of the
aluminum alloys. Rapid dissolution will occur in highly acidic or
alkaline solutions, but in the oxide stable range of pH 49 aluminum is
subject only to water staining of the surface and to localized pitting
corrosion. High-purity aluminum (99.990% or purer) has resistance to
pitting that is notably superior to that of the commercial purity grades.
Pure aluminum does not incur any of the more drastic forms of
localized corrosion (intergranular, corrosion, exfoliation, SCC) that
will be discussed later.
Chemical grades of aluminum exceeding 99.99% purity, made by the
Hoopes electrolytic process [10], were available as early as 1920.
Newer methods, such as zone refining and preparation from amalgams
or alkyds, can now provide metal purity exceeding 99.9990%.
Commercial aluminum made by the Hall process ranges from 99.00%
to 99.99% purity. Foil is produced in 1100 (99.00% minimum) and
1145 (99.45%) purity metal. Products such as sheet, plate, tubing, and
wire are produced in 1100 (99.00%) or 1160 (99.60% minimum)
purity metal. Cladding on alclad sheet uses 1230 (99.30% minimum)
or 1175 (99.75% minimum) purity metal.
2
3XXX Al-Mn and Al-Mn-Cu Alloys
Seventeen 3XXX alloys are registered with the AA, but the two most
frequently used are 3003 and 3004. The Al-Mn constituent particle
has an electrochemical potential similar to that of aluminum; hence it
is not a significant site for corrosion initiation. Like pure aluminum,
3XXX alloys do not incur any of the more drastic
Page 118
forms of localized corrosion, and pitting corrosion is the principal
type of corrosion encountered. With low copper content (less than
0.05%), 3003 and 3004 alloys are almost as resistant to pitting as pure
aluminum. Increasing copper increases the tendency to pitting, with
the effect becoming notable at about 0.15% copper. This tendency is
accelerated by the presence of the chloride ion. Lowering the iron
content reduces the tendency to pitting corrosion.
Alloy 3003 is produced in most product forms, while 3004 is
marketed as sheet, plate, and extruded tube. The remaining 3XXX
alloys are produced primarily as sheet or foil. The most notable
applications have been cooking utensils, food packaging, building
siding, chemical containers, moisture barrier foil laminates, and
roofing sheet. Virtually all metal beverage cans are made of fully
hardened 3004-H119 alloy, protected internally with a clear coating.
Both 3003 and 3004 sheet and tube can be obtained as an alclad
product, the cladding being alloy 7072. Alclad tubes have frequently
been used to prevent perforation in application such as boilers,
condensers, and evaporators.
3
5XXX Al-Mg and Al-Mg-Mn Alloys
From all the commercial alloys, 5XXX alloys with less than 3% Mg
have the best resistance to pitting corrosion and the lowest rate of pit
propagation, particularly in seawater and aqueous chloridecontaining
solutions. The exception is a relatively few alloys, such as 5017 and
5043, which intentionally have about 0.2% copper added to them.
These lower magnesium content alloys likewise do not incur
intergranular corrosion, exfoliation, and SCC.
The higher alloyed Al-Mg-Mn alloys, containing more than 3% Mg,
can become susceptible to intergranular corrosion, exfoliation, and
SCC, if exposed for long time to elevated temperature. Special
tempers, involving thermal treatments to stabilize the precipitate
microstructure, have been developed that greatly minimize this
possibility.
All 5XXX alloys are available as rolled products, sheet, plate, rod,
and bar. Alloys 5083, 5086, 5154, and 5454 are produced as
extrusions, and alloys 5083 and 5456 as forgings. Cold working of
5XXX alloys becomes significantly more difficult as the magnesium
content increases.
4
4XXX Aluminum-Silicon Alloys
The wrought aluminum-silicon alloys consist for the most part of
specialty products. Examples are alloys 4004, 4343, and 4045,
claddings on brazing sheet, and 4043, welding electrodes and weld
wire. 4XXX alloys have reasonable resistance to corrosion when
exposed by themselves. However, they usually are used in
comparatively small amounts in direct contact with other alloys.
Therefore care must be taken that 4XXX alloys and tempers are
electrochemically compatible with the other alloys and tempers.
Welding applications are a notable case in which large
electrochemical
Page 119
differences can exist between the 4043 weld, the heat-affected zone
(HAZ) of the weldment, and the unaffected parent alloy that was
welded. The worst case is when the HAZ, generally the smallest
volume of metal, is the most anodic region. In this case the HAZ will
undergo localized corrosion. Postweld thermal treatments (called
postweld aging) tend to smooth out chemical differences and
eliminate such localized effects. An example is shown in Figure 8.
This local corrosion is most severe when corrosion is confined to a
"knife line" attack at the interface between the HAZ and the weld
bead.
B
Heat-Treatable Alloys
1
4XXX Aluminum-Silicon Alloy
An exception to the general class of 4XXX alloys is a heat-treatable
die forging alloy, 4032, containing a nominal 12% Si, 1% Mg, 0.9%
Cu, and 0.9% Ni. The alloy is SHT and artificially aged to the T6
temper. Its particular merit is a high-temperature strength, with a low
coefficient of thermal expansion. The alloy
Figure 8
Photograph of two butt-welded panels of 12-mm-thick plate of 5083-H321 and
7005T6E132 alloys immersed for 3 years in seawater at Miami, Florida. The top,
as-welded 7005-T6E132 panel shows localized corrosion in a 1-cm-wide
heat-affected zone adjacent to the weld. This localized corrosion was virtually
eliminated by postweld-aging the bottom panel for 16 h at 150°C (300°F).
(Courtesy of Alcoa Technical Center, John J. Liput.)
Page 120
retains usable yield strength, 170 MPa (25ksi) or higher up through
about 230°C (445°F).
2
6XXX Al-Mg-Si, Al-Mg-Si-Cu, and Al-Mg-Si-Mn Alloys
The 6XXX alloys are strengthened by the stoichiometric precipitate,
Mg2Si, and magnesium and silicon generally are added in
approximate proportions to produce this compound. Most alloys
contain a slight amount of excess silicon to ensure maximum strength.
The 6XXX alloys are the most versatile and extensively used,
particularly as general structural alloys, because they possess a
number of desirable factors:
1. They are fabricated by both the producers and users.
2. They are available in all product forms at moderate cost.
3. They have very high formability in the T4 temper and good
machinability in both T4 and T6 tempers.
4. They provide a range of medium level strengths in the T4 tempers.
(see below) and higher strength in the T6 temper. T6 aging can be
applied by either the producer or the end user. Currently the highest
strength sheet is alloy 6013-T6, having guaranteed minimum LT
strengths (MPa) of 353 UTS and 312 TYS, with typical values of 395
UTS and 365 TYS.
6XXX Alloy Typical Ultimate Tensile and Tensile
Yield Strengths, MPa
T4 Temper
T6 temper
UTS
TYS
UTS
TYS
150350
90170
235400
210365
5. They have a proven record of high resistance to corrosion in most
common environments including industrial and seacoast atmospheres,
fresh water and seawater, and soil.
There are more registered 6XXX alloys than any other series.
However, warehouses may carry only the most frequently used 6005,
6061, and 6063 alloys. New alloys, with improved stamping and
drawing formability, are being developed for automotive applications.
Any alloy eventually used in high volume would most likely be
warehoused by distributors.
Alloys containing less than 0.5% Cu have low quench sensitivity so
that all soluble elements are usually retained in solid solution (except
for very thick sections or very slow quench media, e.g., still air).
Therefore, in the T4 temper these alloys are susceptible only to pitting
corrosion. Resistance to pitting decreases as the copper and iron
contents increase, and the effect is synergistic.
Page 121
For example, a laboratory 6.35-mm plate of 6061-T4 and T6 was
fabricated containing (1) the nominal 0.28% Cu plus a typical level of
0.3% Fe, (2) the maximum allowable 0.4% Cu and 0.7% Fe. Panels
were exposed to the James River estuary near Newport News. The
initial rate of pitting on the less pure panels was three to four times
higher than on the nominal purity plate. Eventually some perforation
occurred on the less pure plate during the 8-year test, whereas the
maximum depth of pitting was about 1.5 mm for the plate with
nominal purity.
At copper levels higher than 0.5% some intergranular corrosion may
occur in T4 material, and intergranular corrosion is typically
encountered in T6 temper products of all 6XXX alloys and
thicknesses even in less severe environments, such as atmospheric
exposure at the seacoast and inland. The depth of intergranular
penetration is self-limiting and confined to the near surface (generally
less than 1 mm in depth).
This intergranular corrosion does not result in susceptibility to
exfoliation or to SCC. Often it can be detected only under
magnification and it is very minor compared with that which occurs in
2XXX and 7XXX alloys. Intergranular corrosion has not been found
to cause significant los of strength. However, the sharpness of the
corrosion tips can create a notch effect that significantly reduces
fatigue life. This has eliminated unprotected 6061-T6 and 6013-T6
sheet from fatigue-limited applications, such as fuselage skin sheet on
pressurized jet liners. Work is in progress to understand the cause of
this intergranular corrosion and, more importantly, to find ways to
prevent it.
There has been no reported occurrence of SCC 6061-T4 or T6 parts in
service. In laboratory studies, susceptibility to SCC was induced only
by very slow quenching from a very high SHT temperature, 565°C,
and then only for the T4 temper.
3
2XXX Al-Cu, Al-Cu-Mg, Al-Cu-Mg-Li, and Al-Cu-Mg-Si Alloys
The 2XXX alloys are the oldest heat-treatable alloys, dating back to
"Duralumin" developed in Germany in 1919 and subsequently
produced in the United States as alloy 2017. The 2XXX designation
applies to alloys having copper as the primary alloying element, with
additional alloying elements being Mg, Si, and, most recently, Li.
Solid solution strengthening is produced by GP zone formation during
solution heat treatment followed by rapid quenching. Additional
strengthening can be achieved by subsequent cold working, artificial
aging, or a combination of the two. Each fabrication method has its
own temper designation (see Table 5).
Strengthening during artificial aging involves the formation of
semicoherent transition precipitates within the grain bodies. Principal
strengthening phases are CuAl2 for alloys with less than 1%
magnesium (e.g., alloys 2014 and 2219), CuMgAl2 for a magnesium
content above 1% (e.g., alloys 2024 and 2034),
Page 122
CuLiAl2 when lithium is present (alloys 2090 and 2091), and Mg2Si
in alloys containing relatively low copper together with similar
amounts of magnesium and silicon (e.g., 2117 and recently 2008). The
first two groups of alloys provide medium strength together with
excellent toughness in the as-quenched, T3, T4 tempers. The Mg2Sicontaining alloys are also used in the T3 and T4 tempers. They
provide less strength but markedly improved formability. Alloy 2117
is used for highly drivable rivets, whereas 2008 is a sheet alloy for
deep-drawn auto body parts.
Lithium-containing alloys are used only in aged tempers to provide
maximum strength.
The major use of 2XXX alloys has been in the as-quench tempers to
provide medium strength along with excellent toughness and
formability. High resistance to corrosion depends on the ability to
quench the products fast enough to retain virtually all soluble
elements in solid solution, so that the primary form of corrosion is
pitting. Slow quench rates, caused either by a slow quench media or
by thick products, results in incoherent precipitation at grain
boundaries. When this happens, T3 and T4 temper products are highly
susceptible to intergranular corrosion, exfoliation and SCC in the ST
and T directions, and even in the LT direction if very highly stressed.
This will be discussed in more depth in the following section on
specific types of corrosion.
Artificial aged tempers provide considerably higher strength but much
lower toughness. They tend to incur more numerous but less deep
pitting than the as-quenched tempers. The peak strength T6 temper
can exhibit slight susceptibility to exfoliation and to SCC. However,
all 2XXX alloy products in T8 type tempers are aged slightly beyond
peak strength, generally by about 20 MPa (3 ksi). These 2XXX-T8X
products have a proven record of high resistance to both exfoliation
and SCC.
4
7XXX Al-Zn, Al-Zn-Mg, Al-Zn-Mg-Mn, and Al-Zn-Mg-Mn-Cu
Alloys
Alloy 7072 (Al + 1% Zn) is the only commercial binary 7XXX alloy.
It is the cladding alloy normally used when a more anodic potential
(i.e., more electronegative) than that of pure aluminum is required,
being typically -0.87 V on the saturated calomel electrode (SCE) scale
vs. -0.74 V for pure aluminum. In addition, there are four, copperfree, specialty cladding alloys7008, 7108, 7011, and 7013which are
even more electronegative and have higher strength than 7072.
For heat-treatable 7XXX alloys, solid solution strengthening is
achieved by GP zone formation during solution heat treatment
followed by rapid quenching to the W temper. Because the W temper
is unstable, all 7XXX alloys are artificially aged to a stable T6-or T7type temper. Principal strengthening phases during aging are the
stoichiometric Eta phase MgZn2, sometimes the Mg3Zn3Al2
Page 123
phase, and several variants of the nonstoichiometric Eta' phase
MgZn2Cu, which exhibit varying degrees of coherency with the
aluminum lattice.
The copper-free 7XXX alloys, with or without manganese, have good
resistance to pitting, medium strength, and are weldable. Alloy 7005,
for example, is frequently used for irrigation tube and bridge
pontoons. Generally these alloys are used in thin sections and afford
good resistance to exfoliation and SCC. Susceptibility to SCC can
occur in thick sections and in thin products with fully recrystallized
grains, particularly when they are highly worked (bent, indented,
pierced, etc.) into the plastic region resulting in permanent residual
stresses.
Alloys containing less than 1% copper are similar to the copper-free
alloys, incurring mild pitting corrosion but providing higher strength.
Two notable exceptions exist as regards resistance to corrosion: (1)
sheet of alloys 7046 and 7146 with an unrecrystallized grain structure
has shown high susceptibility to exfoliation in outdoor atmospheres
and (2) with prolonged exposure, alloy 7079-T6 in all product forms
is highly susceptible to SCC in the ST and T directions even in very
mild environments such as an indoor atmosphere. These products no
longer are produced and their use is not recommended.
7XXX alloys containing 13% copper provide the highest strengths
available in wrought aluminum alloys. Early alloys such as 7075 used
chromium as the grain refiner, but this causes quench sensitivity and
markedly reduces strength beginning at a thickness of about 5 cm.
Newer alloys use zirconium and can provide high strength in
thicknesses up to about 17 cm.
If rapid quench conditions are achieved (about 130220°C/s) the alloys
incur pitting corrosion. Slower quench cooling rates cause incoherent
precipitates at the grain boundary and susceptibility to intergranular
corrosion.
7XXX alloys usually have a directional grain structure (see Figs. 5
and 6). If they are exposed to severe environments (e.g., seacoast
atmosphere) in the peak strength T6 temper they can be very
susceptible to exfoliation and to SCC if a sustained stress is applied in
the ST and T grain directions.
To counteract this susceptibility, the industry has developed a variety
of tempers aged beyond peak strength to provide the best high
strength together with a specified resistance to corrosion. Verification
of resistance to exfoliation usually involves 7 days exposure as per
Ref. 11, while determination of resistance to SCC involves 20 or 30
days exposure as per Ref. 12.
The first such temper was 7075-T73, which was required to exhibit no
exfoliation after a 7-day test and pass a 30-day SCC test at an ST
stress of 285 MPa (42 ksi). However, the necessary overaging caused
about a 12% decrease in strength from peak. After more than 12 years
of trouble-free service, design engineers decided that the inherent
degree of resistance to SCC was more than was needed. Intermediate
T7-type tempers were then developed, some of which are shown
below. Close control of the multiple aging steps at two or more
temperatures is needed, sometimes involving other processing steps.
The purpose
Page 124
is to retain as much of the coherent strengthening phases as possible
and still produce sufficient randomly distributed precipitates to
achieve the specified corrosion resistance. Some of the tempers are
proprietary and certain ones can be applied only by a mill producer.
Other tempers can be applied by the customer. For example, 7075-T6
products can be converted to either the T73 or T76 tempers simply by
the addition or the appropriate second step age.
Typical
LT-YS, Exfoliation
AlloyTemper MPa
requirement
7075 T73
405
No
exfoliation
7075 T76
445
EB rating or
less
7050
T74
410
7150
T76
470
No
exfoliation
EB or less
7150
T77
505
EB or less
7055
T77
550
EB or less
SCC-ST
test stress,
MPa
285
135 to 170
(pending
product)
240
120 to 145
(pending
product)
120 to 145
(pending
product)
@ 102
Nondestructive measurement of electrical conductivity is a convenient
method of assessing the degree of aging, hence many T7 type tempers
are required to meet both a minimum strength and a minimum
electrical conductivity. The electrical conductivity usually is
expressed in % IACS (International Annealed Copper Scale). For
example, alloy 7075-T73 must have an electrical conductivity of at
least 38% IACS and be no more than 81 MPa above the guaranteed
long-transverse yield strength. Alternatively, 7075-T73 can exceed the
minimum guaranteed strengths by any amount, provided the electrical
conductivity is 40% IACS or higher.
V
Specific Types of Corrosion
A
Uniform Dissolution
The protective oxide film on aluminum and its alloys readily dissolves
in strong acids and alkalis that are outside the stable pH range of
about 49. When this happens the aluminum dissolves more or less
uniformly over the entire surface. Commercial etchants are available
for both cleaning and chemically sizing aluminum parts. Typical
solutions for this purpose are aqueous solutions of sodium hydroxide
and sulfuric/chromic or phosphoric acid solutions.
Page 125
Dissolution is most uniform in pure aluminum, then in dilute alloys
and the non-heat-treatable alloys. Highly alloyed, heat-treatable alloys
often show some surface roughness, especially when thick crosssections are etched because variable dissolution rates result from
through-thickness variations in solid solution concentration of the
alloying elements and in the distribution of constituent particles.
Pourbaix potentialpH diagrams can be used to predict the regions of
oxide stability and of uniform corrosion in various solutions [13].
However, the rate of dissolution has to be determined experimentally
and depends on the particular chemical solution, temperature, and
possible presence of trace ions.
Table 6 shows examples of variations in the room temperature
corrosion rate of 1100-H14 aluminum sheet in several solutions.
Additional tests in 80 wt % sulfuric acid showed raising the
temperature to 50°C increased the corrosion rate by a factor of about
4.5. At either temperature, the corrosion rate was further accelerated
by the addition of a 1025 ppm of chloride ions (added as hydrochloric
acid).
B
Pitting Corrosion
The oxide film is never completely uniform and is heterogeneous on
heat treatable alloys. Microcracks in the oxide occur from rapid
temperature excursions or from the metal being stressed. This leads to
local film breakdown and the initiation of localized pitting corrosion.
Pitting corrosion occurs in all environments and on all alloys. Pure
aluminum is the most resistant, followed by the 5XXX and then
Table 6 Effect of Chemical
Solutions at Similar pH Values on
the Mass Loss of 1100-H14
Aluminum Sheet, µ/year
pH
Solution
Metal
loss
36
100
178
214
605
1430
125
2.0
Acetic acid
2.0
Nitric acid
2.0
Sulfuric acid
2.0 Hydrochloric acid
2.0
Phosphoric acid
2.0 Hydrofluoric acid
10.5
Ammonium
hydroxide
10.5 Sodium carbonate
1180
10.5 Sodium hydroxide
1600
Source: Alcoa Technical Center,
Alcoa Center, PA.
Page 126
the 3XXX alloys. Alloys containing copper are the least resistant to
pitting corrosion.
Pits begin somewhat hemispherical in shape, but the corrosion follows
the grain morphology. Therefore pitting in aluminum is not
necessarily a smooth attack and pits can be significant stress risers. An
example of pitting corrosion is shown in Fig. 9.
The chloride ion is known to facilitate breakdown of the aluminum
oxide. Aluminum chloride (AlCl3) usually is present in the solution
within pits and the concentration increases as corrosion progresses or
during drying in environments that are alternatively wet and dry. A
saturated AlCl3 solution has a pH of about 3.5, so the bottom of
corrosion pits and cracks often will not repassivate and stop
corroding, as long as oxygen and the corrosive electrolyte still can
migrate to the bottom.
In most environments, including industrial and seacoast atmospheres
and
Figure 9
Photomicrograph of pitting corrosion in 2024-T3 sheet exposed to seacoast
atmosphere at Point Judith, RI. The maximum depth is about 100 µm. Pits have
relatively similar depth and width, but follow the grain structure and are not a
smooth hemisphere. The pit shown has sufficient roughness to be a potential
stress riser (as polished).
(Courtesy of Alcoa Technical Center, Edward L. Colvin).
Page 127
immersion in seawater, the rate of pitting rapidly diminishes and
becomes self-limiting. Figure 10A illustrates this self-limiting effect
of pitting corrosion. This plot provides a reasonable estimate of the
rate at which perforation would occur in aluminum. Note, however,
that the loss in strength caused by the pitting corrosion occurs at a
different rate (Fig. 10B). The rationale behind this difference follows.
Pits can initiate relatively quickly and grow to a limiting depth, at
which mass transport no longer provides sufficient oxygen and the
corroding species. At this point, further penetration of that pit is
stifled. A few isolated deep pits have a small effect on the reduction in
cross-section, so that the initial reduction in strength and load carrying
ability is less pronounced than is the depth of penetration. However,
new corrosion pits initiate at other sites and corrosion continues, but at
a reduced rate. Eventually a significant reduction in cross-section
occurs and the effect on strength is noticeable. Without any
protection, new sites of corrosion continually occur, hence the selfstopping effect on loss in strength is less abrupt than on the depth of
perforation (compare parts A and B in Fig. 10). Likewise corrosion on
a freely exposed object, like a highway road sign, occurs on both
surfaces at approximately equal rates and the effect is additive. For
many structures, corrosion occurs only on the outer, exposed surface.
For tubes and containers, corrosion will probably occur at different
rates on the inner and outer surfaces.
The effect on corrosion fatigue is still different and more complicated.
A few isolated deep pits can be very significant if they occur at the
location of high fatigue stress. In such a case, pits act as local stress
risers and can greatly reduce the number of cycles required to initiate
a fatigue crack. Mild general pitting over the entire surface is unlikely
to produce such an effect. However, once initiated, a fatigue crack
will propagate more rapidly in metal that has been weakened by
overall, general corrosion. Thus the designer needs to consider not
only the rate at which corrosion occurs but how the corrosion might
affect the critical design properties.
Finally, although pitting is regarded as the least damaging form of
corrosion, microscopic studies of corrosion in situ have shown that the
more severe modes of corrosion, such as SCC, frequently initiate and
grow from pits.
C
Intergranular Corrosion
Intergranular corrosion (IGC; sometimes called intercrystalline
corrosion) occurs to some extent in all heat treated products and is
often related to copper depleted regions, or anodic precipitates at the
grain boundary region. IGC commonly occurs in 2XXX-T3 and T4
alloys that are not quenched rapidly enough to keep all of the solute
elements in solid solution. IGC also is prevalent in the peak strength
T6 tempers, especially in 6XXX-T6 products exposed to atmospheres.
Corrosion is limited to the immediate grain boundary region and may
not
Page 128
Figure 10
Plots of the maximum depth of pitting corrosion, on 100 and 3003
alloy sheet, and of the percent loss in strength resulting from
exposure to seacoast atmosphere at Point Judith, RI. Both curves
show the self-stopping nature of pitting corrosion of aluminum
but at different rates of change. The rationale behind the
difference is discussed on page 127.
(Courtesy of Alcoa Technical Center, Marsha Egbert.)
Page 129
be visible to the naked eye. IGC will penetrate more quickly than
pitting corrosion, but it too reaches a self-limiting depth due to limited
transport of oxygen and corroding species down the narrow corrosion
path. When depth of penetration ceases, IGC begins to spread laterally
over the entire surface, as opposed to pitting corrosion that often
remains confined to discrete sites. Both types of corrosion have an
adverse effect of fatigue life, but the sharper tips of IGC are
considerably greater stress risers than corrosion pits. Figure 11 shows
the narrow, highly localized nature of intergranular corrosion.
A simple laboratory test to determine the susceptibility to IGC of
heat-treatable alloys is contained in Ref. 14. However, care must be
taken in relating the results obtained to the performance to be
expected in other environments [15].
1
Mechanism of Intergranular Corrosion in 2XXX Alloys
Intergranular corrosion in 2XXX alloys essentially is galvanic
corrosion of the very narrow, anodic copperdepleted regions at the
grain margins, driven by the relatively larger cathodic area of the
copper-rich grain matrix.
In the 2XXX system, corrosion potential measurements [16] were
conducted on high-purity aluminum and various binary Al-Cu alloys
up to and beyond the limit of solid solubility of 5.65% copper. Also it
was possible to
Figure 11
Photomicrograph of narrow intergranular corrosion (IGC) in 6013-T6 sheet
exposure to seacoast atmosphere at Point Judith, RI. Corrosion is confined
to the grain boundary and can be a very sharp stress riser (etch: Keller's).
(Courtesy of Alcoa Technical Center, Edward L. Colvin.)
Page 130
produce large particles of the stoichiometric precipitates CuAl3 and
CuMgAl3 so that their corrosion potentials could be measured. Figure
12 is a plot of the corrosion potential of the various materials as a
function of copper content showing that significant potential
differences of as much as 0.15 V can exist between pure aluminum and
Al + 4% (or more) dissolved copper.
When 2XXX alloys are slowly quenched, large, incoherent, copper-rich
precipitates form at the grain boundary, creating a copper-depleted
zone in the grain margins. Figure 13 is a schematic presentation of this
condition. The depleted zone is very narrow, typically 0.1 m or less in
width. This results in the most adverse galvanic situation of a small
anode and a relatively large cathode. Thin foils of 2024 prepared from
sheet given a slow quench to a T42-type temper were immersed for a
few seconds in the solution given in [14] and then examined by
transmission electron microscopy [17]. Corrosion was confined to the
copper-depleted zone and proceeded rapidly down this zone. After the
sheet was aged to the peak strength, corrosion still initiated at the
boundary, but in the form of a
Figure 12
Plot of corrosion potentials of pure aluminum and of binary aluminum plus copper
alloys, plus the two stoichiometric precipitates. The binary alloys were fully
solution-heat-treated and quenched as rapidly as possible to retain the maximum
amount of copper in solid solution. Note that the addition of copper raises the
corrosion potential of pure aluminum by about 0.14 V.
(Courtesy of Alcoa Technical Center, Marsha Egbert.)
Page 131
Figure 13
Schematic of grain boundary region in a 2XXX alloy. Precipitation of the very high
copper content precipitates on the boundary causes a copper-depleted zone on
either side of the boundary. The difference in electrochemical potentials of the
copper-depleted zone and the copper-rich matrix form a strong galvanic cell with a
potential difference of about 0.12 V. Furthermore, the anodic copper-depleted zone
is small in area compared with the area of the cathodic grain matrix, resulting in a
high driving force for rapid intergranular corrosion.
(Courtesy of Alcoa Technical Center, Edward L. Colvin.)
Page 132
string of circular pits that gradually progressed outward into the
copper-rich grain, resulting in a much slower rate of penetration.
Further aging past peak strength (as in 2XXX-T8 type tempers)
produced many fine incoherent precipitates throughout the grains,
each being surrounded by a ring of copper-depleted aluminum. At this
point the grain boundary no longer is a preferred site for corrosion,
and corrosion reverts to general pitting. The electron
photomicrographs of Ref. 17 appear on p. 232 of Ref. 5 and in Ref 18
where this subject is discussed in detail on pp. 478486.
2
Mechanism of Intergranular Corrosion in 7XXX Alloys
The mechanism of intergranular corrosion in 7XXX alloys is believed
to be analogous to that in 2XXX alloys, but it has not been studied in
as great detail because most of the precipitate phases cannot be grown
to a sufficient size to be analyzed. However, addition of zinc and the
dissolution of the MgZn2 phase both shift the potential of these alloys
in the anodic direction, so that a potential difference of as much as
0.24 V could exist between pure aluminum and the alloys. Again,
precipitation occurs first at grain boundaries causing a depleted zone
which creates a galvanic cell between the narrow depleted zone and
the zinc/magnesium-rich grains. This condition exists in both the asquenched W temper and the peak strength T6 temper. Overaging to
the various T7 tempers produces semicoherent or incoherent
precipitates throughout the grains and reduces the driving force for
localized intergranular corrosion. This is discussed in more extent on
pp. 489498 of Ref. 18.
D
Exfoliation Corrosion
When intergranular corrosion occurs in a product with a highly
direction grain structure, of the type shown in Fig. 5, it propagates
internally, parallel to the surface of the metal. The entrapped corrosion
product is about five times as voluminous as the metal consumed.
This produces an internal stress that splits off the overlying layers of
metal. Hence the name ''exfoliation." Figure 14 shows exfoliation of a
susceptible plate alloy exposed to a seacoast atmosphere.
Exfoliation is a very deleterious form of corrosion because the
splitting off of uncorroded metal rapidly reduces load carrying ability.
The splitting action continually exposes film free metal, so the rate of
corrosion is not self-limiting. Exfoliation generally proceeds at a
nearly linear rate.
Exfoliation requires elongated (pancake-shaped) grains, a susceptible
grain boundary condition, and a relatively severe environment. The
most damaging natural environments are those with high chloride ion
content, such as deicing salts, and a seacoast atmosphere. The
presence or absence of an applied stress has no significant effect.
Coatings can delay exfoliation but the best procedure is use of a
resistant temper.
Figure 14
Photograph of an exfoliated panel from 12-mm-thick 7075T651 plate after 6 years of exposure to a seacoast
atmosphere. The exfoliated layers on the left side have swelled the specimen to three times the
thickness and pulled the aluminum mounting bolt out of the wooden test rack.
(Courtesy of Alcoa Technical Center, John J. Liput.)
Page 134
E
Stress Corrosion Cracking (SCC)
SCC can occur in the high-strength 2XXX and 7XXX alloys, and
5XXX alloys with more than 3% magnesium. In aluminum alloys,
SCC characteristically is intergranular in nature and occurs most
readily in products having long, in-line grain boundaries and when a
sustained tension stress acts perpendicular to these boundaries (i.e.,
stressed in the ST direction). Compressive surface stresses can reduce
susceptibility to SCC. Detailed discussion of SCC can be found in
Refs. 5, 8, and 9.
SCC is insidious because it can occur in relatively mild environments
(condensed moisture indoors can be sufficient). Figure 15 shows a part
that failed by SCC with little visible corrosion. Metallographic
examination of a cross-section through a secondary crack in this part
(Fig. 16) showed the intergranular crack path that is characteristic of
SCC in aluminum alloys.
SCC requires an incubation period that can range from a few hours to
many years depending on the environment. Once cracking initiates,
propagation is rapid and the rate of propagation tends to be constant.
In thick parts, cracking may arrest because of gradual loss of stress at
the crack tip as the crack grows. However, complete failure frequently
occurs because SCC also leads to failure by other modes, i.e., tensile
overload or fatigue cracking. The best protective measure is use of
resistance alloys and tempers. Coatings are not effective because no
coating is perfect, and coating damage is likely. Thus there is high
probability of sites at which SCC can initiate and grow.
F
Filiform Corrosion
Filiform corrosion occurs on the surface of aluminum, underneath a
pliable coating. It results in the blistering of the coating in a "worm
track" pattern. Initiation occurs at a break in the coating, such as a
fastener edge or a scratch. The worm tracks slow down with time but
appear to propagate most quickly on
Figure 15
Photograph of an extruded and drawn 7075-T651 tube with a press-fit plug at the
right end (plug has been partially withdrawn). Circumferential tension stress from
the interference fit caused SCC resulting in a through-thickness crack about 4 cm
in length, plus several secondary cracks.
(Courtesy of Alcoa Technical Center, John J. Liput.)
Page 135
Figure 16
Photomicrograph of a section through one of the secondary cracks in the
failed part shown in Fig. 15. The intergranular nature of the crack is clearly
shown. Intergranular cracking is characteristic of SCC in aluminum alloys.
1 inch = 250 µm.
(Courtesy of Alcoa Technical Center, John J. Liput.)
copper-containing alloys. Preparation of the metal surface for coating,
surface cleanliness, coating flexibility, thickness, and adherence all
are important factors that determine whether this form of corrosion
will occur.
Filiform corrosion always is shallow in depth and causes loss of
product integrity only when it occurs on thin sheet (about 0.05 mm or
thinner) and foil (defined as £ 0.15 mm in thickness), e.g., food
containers, or a foil moisture barrier on insulation board or on foillaminated paper packaging.
On thicker painted sheet, as for aircraft or automobiles, filiform
corrosion is primarily a cosmetic problem, but it causes loss of paint
adhesion and can act as a site for initiation of pitting or other forms of
corrosion.
G
Crevice Corrosion
Crevice corrosion is not truly a separate form of corrosion. Rather it is
corrosion that initiates in the narrow opening between two fastened
parts. Crevices aggravate corrosion because they cause a chemical
environment different from that on freely exposed surfaces. Crevices
exclude oxygen, retain moisture, collect pollutants, and concentrate
corrosion products. When crevices cannot be eliminated, proper
design to promote drainage and sealing of edges to prevent ingress of
moisture is the best protection. The automotive industry has reported
that a "hot wax dip" can effectively seal cervices. Faying surfaces that
will form a crevice are often painted prior to assembly.
Page 136
H
Poultice Corrosion
Poultice corrosion is not a separate form of corrosion, but rather early
initiation and aggravation of corrosion occurring beneath an
hygroscopic attachment or insert. This could be lamination of paper,
cloth, or wood to a single layer of aluminum or a multilayered
laminate. An unintentional poultice is the gradual collection of
hygroscopic particulate matter (e.g., soil) on ledges and the like.
Design prevention measures would be the use of laminate material
that does not absorb moisture and the sealing of edges. Periodic
cleaning and drying are good preventative measures.
In this regard, depending on the species, freshly cut wood contains
over 50% moisture and organic acids that can be quite corrosive [19].
Properly dried (kilned) wood is much more compatible. Wood treated
against disease and insects may contain chemicals that can leach out
and be corrosive. Some aerosol-type fungicides required for
international shipment in wood containers can be particularly
corrosive.
I
Galvanic (Cathodic) Corrosion
All corrosion is electrochemical in nature, involving dissolution of
metal into metallic ions with the release of electrons at the anodic site,
and one of several possible chemical reactions at the cathodic site to
use up those electrons.
However, galvanic corrosion implies accelerated corrosion resulting
from electrical contact of two dissimilar metals or alloys. It is
important to remember that various aluminum alloys have sufficiently
different corrosion potentials (by as much as 0.4 V in extreme cases)
to cause strong galvanic cells when in contact with each other. A list
of some corrosion potentials can be found in Ref. 5 (pp. 210 and 212)
and Ref. 9 (pp. 257 and 258).
Of the normal construction metals, only magnesium and zinc are
anodic to aluminum. Aluminum and its alloys will corrode
preferentially to copper, graphite, iron, steel, titanium, and many other
metals. When possible, the best protective measure is to break
electrical contact between the junction of the two materials with a
suitable insulator. For example, a polyvinyl coupling may be used to
connect aluminum and steel pipe.
Galvanic corrosion accelerates corrosion but generally does not
change the type or morphology of the inherent corrosion. For
example, it will not cause susceptibility to SCC in an alloy that is
normally resistant to SCC. The more anodic metal (alloy) will corrode
preferentially to protect the cathodic metal (alloy), but the rate of
corrosion usually is controlled by the cathodic (less negative) metal
(alloy). Therefore, the worst situation is a large cathode area
contacting a small anode, e.g., anodic fasteners in a metallic structure.
Since the anodic material is the one most prone to corrode, designers
and engineers often want to coat or protect it. However,the proper
protective measure is to coat or
Page 137
protect the cathodic metal to isolate it from the anodic material. This
is explained further in the following section. Coating or protection of
both the anodic and cathodic materials is permissible and often used.
1
Mechanism of Galvanic Corrosion
Every metal immersed in a solution will corrode at a definite
potential, with a fixed rate of current flow. These values can be
determined by means of polarization measurements [20]. Schematic
examples of anodic and cathodic polarization curves, and the resulting
corrosion potential, Ecorr, and corrosion current, icorr, are shown in
Fig. 17. If two different metals are exposed in the same electrolyte but
are not in electrical contact, each will corrode at its respective rate, as
shown in Fig. 18a.
Galvanic corrosion occurs when two metals with different corrosion
potentials are in the same electrolyte and in electrical contact with
each other. Electrical connection can be by actual physical contact or
through connection by a separate
Figure 17
Polarization behavior of a single alloy is a specific solution. Current is plotted
as a function of applied potential for both the anodic and cathodic reactions.
Linear extrapolation to the point of intersection yields a potential (Ecorr) where
the anodic and cathodic currents (icorr) are equal and the rate of oxidation
(corrosion) equals the rate of reduction (i.e., consumption of electrons).
(Courtesy of Alcoa Technical Center, Marsha Egbert.)
Page 138
Figure 18
Logarithmic plots of current density in two dissimilar metals, A and C, in a
single electrolyte. (a) The two metals are not in electrical contact with one
another and each metal has its own corrosion rate.
(b) The two metals are in electrical contact. A new corrosion rate exists that
essentially parallels that of the uncoupled metal C; but the mixed potential,
E - AC, is intermediate between the uncoupled potentials, E - A and E - C.
The more passive metal A now corrodes at a slower rate, i'corrA, while the
more active metal C corrodes at a faster rate, i'corrC.
(Courtesy of Alcoa Technical Center, Marsha Egbert.)
Page 139
conductor (electrical wire, fastener, automobile chassis, etc.). A
complete electrical circuit is required, with electrons moving through
the two connected metal and ions flowing through the electrolyte.
The more stable (cathodic) metal polarizes the less stable (anodi)
metal to a new intermediate potential. At this new potential the anodic
metal corrodes more rapidly (Fig. 18b). At that same potential,
corrosion of the more stable cathodic material is reduced and, pending
conditions, may be entirely prevented. Reference (21) discusses
development of a galvanic chart to predict the effect one metal (alloy)
may have on another.
Oxidation of metal and production of electrons primarily occurs on
the anodic metal. Electrons flow by ionic transport through the
electrolyte and are consumed on the more stable metal by cathodic
reactions that do not involve the consumption of metal. Typical
cathodic reactions (equations shown below) are the evolution of
hydrogen; the reduction of oxygen and hydrogen ions to water in
acidic solutions; or the reduction of oxygen and water to hydroxyl
ions in neutral or basic solutions; and, finally, metal deposition,
"plating out," of a more cathodic metal ion from the electrolyte. For
example, some of the copper in aluminum alloys will go into solution,
and then may plate back out on the aluminum, causing another small,
local galvanic cell.
Following are typical cathodic reactions that consume electrons:
Rates of corrosion reactions depend on the potential difference, the
conductivity of the electrolyte, temperature, the relative size of the
cathode/anode area, and whether the electrolyte is quiescent or
aerated. An example of the effect of area ratio and aeration is shown
in Fig. 19. The total rate of current flow and the consequent metal
dissolution is determined primarily by the cathodic reactions; hence
the need to minimize the cathode surface area and, if possible,
eliminate it entirely from the electrical circuit. Reference 22 gives
guidelines for prevention of galvanic corrosion. It has not been
published but is available on request.
VI
Corrosion Prevention Measures
A
Design
The first place to prevent corrosion is proper design of a product.
Corrections and improvements made during the design stage are far
less expensive than subse-
Page 140
Figure 19
Effects of aeration and of cathode-to-anode ratio on the rate of galvanic
corrosion of 1008 steel coupled to 6111 aluminum. Test environment is 3.5%
sodium chloride.
(Courtesy of Alcoa Technical Center, Marsha Egbert. Data from Edward
L. Colvin.)
quent changes, repairs, and stop-gap procedures made on a faulty
product. If an existent part is being replaced or improved, a good
place to start design is to determine why the prior material was
inadequate or failed, and what the possible misapplications were. As
early as 1941, Mears and Brown [23] summarized 18 mechanisms or
factors which could cause electrochemical potential differences on the
surface of the metal and be sites for initiation of corrosion. These
causes are shown in Table 7. No single book or course can provide
solutions to all possible corrosion problems. Intelligent, creative
planning is required and many suppliers will assist in assessing
designs. Design measure to prevent corrosion should include the
following:
1. Selection of the most resistant aluminum alloy and the appropriate
temper consistent with the required engineering properties,
availability, and cost. Temper selection is particularly important to
prevent exfoliation and to provide the degree of resistance to SCC
needed for a particular application.
Page 141
Table 7 Causes of Corrosion Sites (Listed
Alphabetically)
1. Complex cells
2. Contact with dissimilar metalsa
3. Differences in shape
4. Differential aerationa
5. Differential agitation
6. Differential concentration or
composition of solutiona
7. Differential grain size
8. Differential heating
9. Differential illumination
10.Differential preexposure to air or
oxygen
11.Differential strain
12.Differential thermal treatmenta
13.Externally applied potentials
14.Grain boundaries
15.Impurities in the metal
16.Local scrathches or abrasionsa
17.Orientation of grain
18.Surface roughness
aParticularly likely to determine a site of
corrosion.
Source: Ref. 23.
2. Fabrication and assembly techniques should be carefully planned to
minimize problem areas such as crevices, dissimilar metal contact,
sharp edges that will not take a good coating, and assembly methods
that induce high, locked-in stresses. Consideration should be given to
favorable designs that promote complete drainage and drying, plus
ease of periodic cleaning and maintenance. Faying surfaces should be
sealed, or else the bottoms should be opened to drain. Tubing should
not be "bundled" as this provides locations for debris to accumulate,
which cannot be readily removed.
3. Compatible fasteners are needed because fasteners usually are in
electrical contact with the joined parts. When aluminum fasteners are
not suitable, stainless steel, cadmium-plated steel, galvanized, or
aluminized fasteners are good choices for use with aluminum and
alloys.
4. Piping, beams, etc., of dissimilar metals, particularly steel, should
not be installed over or against aluminum structures in such a manner
that their corrosion products will run off onto the aluminum structure.
Ferric or ferrous ions in the corrosion products are likely to plate out
on the aluminum and cause galvanic corrosion sites.
Page 142
B
Maintenance
Aluminum metal should be cleaned regularly. The polishing of
airliners is a prime example of the benefit of cleaning, but merely
hosing off accumulated dirt and pollutants will appreciably extend
life.
Any corrosion that does occur should be removed by mechanical
cleaning (brushing and the like) and sanding the area smooth. This
should be done before corrosion propagates deep into the metal.
Usually, these ''blended" areas are then given a hand-applied
conversion repair coating. Chemical cleaners can be used to remove
corrosion products, provided the size of the part permits this and the
parts can be removed, cleaned, rinsed, and reinstalled. In situ chemical
cleaning is rarely practical because of the likelihood of retention of
some of the chemical cleaner, which in turn could become a cause of
corrosion problems.
As indicated by Table 7, local scratches are a prime site for initiation
of corrosion. Consequently, paint films and other coatings should be
maintained and repaired as needed to prevent moisture penetration.
C
Tempers
For aluminum and the lower strength, strain-hardened 3XXX and
5XXX alloys, corrosion resistance of a particular alloy does not differ
appreciably with the temper used. For 5XXX alloys containing
magnesium of 3.5% and higher, a stabilized temper, such as H343 of
H116 and others, frequently is specified to prevent exfoliation and
SCC problems in the event that elevated service temperatures are
encountered.
Most heat-treatable alloys are used in artificially aged tempers that
have been developed to minimize susceptibility to exfoliation and
SCC. Examples are 2XXX-T8X, 6XXX-T6, and 7XXX-T7X tempers.
A comparison of the relative resistance of 2XXX and 7XXX alloys to
SCC in all three grain directions is given in Table 2 of Ref. 24.
There are two notable exceptions to the previous paragraph. The main
one is appreciable use of 2024 alloy sheet in T3 and T4 tempers
because of its formability and toughness. In thin gauges,
approximately 2.5 mm and thinner, 2024 sheet can be quenched
rapidly enough to retain copper in solution and only pitting corrosion
is encountered. Various protective measures are effective against
pitting corrosion.
Thicker 2024-T3 sheet and T351 plate corrode by IGC and can incur
severe exfoliation and SCC in many environments. Exfoliation is most
prone in product thickness of about 640 mm. For SCC, the thickness
only needs to be sufficient so that sustained tension stress can be
present in the ST direction.
The other exception is that many 7075 alloy products are warehoused
in the T6 temper because this is the most readily available highstrength alloy/temper.
Page 143
Alloy 7075-T6 products have provided useful service in many
applications, but once again the thicker products are prone to IGC,
leading to exfoliation and SCC.
The designer needs to ensure that the temper selected is adequate for
the intended purpose. This may require actual testing. The conducting
of corrosion tests is beyond the scope of this chapter, but the reader is
cautioned that accelerated, laboratory corrosion tests are only an
approximation. They are most often used to ensure lot quality and
consistency and do not necessarily imply adequate performance in
service. Information gained from accelerated tests needs to be
combined with service experience in real environments. Ideally, this
service experience should include any protective measures or
aggravating conditions that are expected.
D
Cathodic Protection (Sacrificial Anodes)
Much was said previously about adverse effects of galvanic corrosion
resulting from the contact of two dissimilar metals. This principle can
be used to provide protection to a structure by intentionally making it
the cathode. The glossary of corrosion terms developed by the
International Association of Corrosion Engineers (NACE) defines
cathodic protection as "reduction or elimination of corrosion by
making the metal a cathode by means of an impressed direct current
or by the attachment of sacrificial anodes (usually magnesium, zinc or
aluminum)."
Buried aluminum pipelines have been protected by magnesium
anodes, and some structures and ships have been protected by anodes
or direct current rectifiers. Proper design of a protective system is
complex to ensure that the entire structure is protected and that no
portion is "overprotected," which can result in local accumulation of
hydroxyl ions leading to caustic attack. The services of a professional
cathodic protection company are recommended. The reader is referred
to Ref. 25 for a detailed discussion of cathodic protection.
1
Anodic Coatings
An everyday use of cathodic protection is in galvanized and
aluminized steel products, such as galvanized containers and
fasteners. Although they are not common marketplace items,
aluminum alloy products likewise can be dipped or sprayed with an
anodic coating of zinc, pure aluminum, or a more anodic aluminum
alloy.
2
Alclad Coating
Alclad sheet was developed initially to provide good corrosion
resistance to the high strength sheet used so profusely in aircraft
applications. A thin plate of pure aluminum is placed on the top and
bottom of the alloy ingot. This "sandwich" is first roll-bonded to form
a metallurgical bond between the cladding and core. The
Page 144
clad ingot then is rolled to the finished gage sheet or plate. Clad
tubing can be made by the extrusion process. The two most common
claddings are pure aluminum on most 2XXX alloys and 7072
(aluminum + 1% zinc) on 7XXX alloys and some 2XXX-T8 products.
Cladding thickness depends on the core alloy and final product gage.
Cladding thicknesses per side are nominally 1.5, 2.5, 4, 5, and 10% of
the final composite thickness (see table 6.1 of Ref. 2).
Initially the cladding provides a barrier layer of more resistant
material than the core. Its real benefit occurs when the clad surface is
intentionally or accidentally penetrated by sawing, drilling,
inadvertent scratches and dings, or when the natural pitting corrosion
of the cladding finally reaches the core. At this time the
electrochemical nature of the cladding comes into play. Further
corrosion spreads laterally, confining itself to the cladding rather than
penetrating into the core. Cladding is one of the few surface protection
methods that provides effective protection even when a break or flaw
exists in the coating. Unpublished data on 7008 Alclad 7075-T651
plate [26], in which grooves of various widths were machined through
the cladding showed that protection was provided to void of about
1. 35 mm wide in industrial and in seacoast atmospheres
2. 6-mm-wide in immersion in 3.5% NaCl solution, and
3. as wide as 25 mm for continuous immersion in the very good
electrolyte of 3.5% + H2O2 listed in Ref. 14
E
Barrier Coatings
There are two types of barrier coatings. One type involves a thin
surface layer of more resistant metal. For example, there is a
proprietary process for roll-bonding stainless steel to non-heattreatable aluminum alloys. Aluminum can also be chrome-plated by
conventional plating methods or can be coated with a variety of
metals by vapor deposit or ion deposition, etc. These latter methods
are more expensive and have size limitations. Metal barrier coatings
differ from an Alclad coating, in that they normally do not provide
any electrochemical protection at a break in the coating.
The more conventional barrier coatings are conversion coatings and
organic paint films. These provide an envelope to exclude the
environment. They are most effective when the entire part can be
completely coated. Recent development in electrostatic coating
procedures by the automotive industry are a prime example of this
protection. Some protection is lost if the coating is damaged by a
scratch or chip. Newer pain films are more resistant to undercutting
and delamination by corrosion. The ability of a coating to resist
delamination by corrosion is evaluated by scoring through the paint
coating, prior to exposure, or by causing chip damage in a "gravelprojecting" machine using a specified exposure time and size of
gravel [27].
Page 145
Corrosion evaluations of coatings have long recognized water, either
in the liquid or vapor stage, and the presence of chloride ions plus
other acid producing ions as the primary factors that degrade coatings.
Recently, more attention has been paid to the degradation of coatings
by ultraviolet (UV) light. Outdoor exposure and laboratory spray
cabinets have been the usual test methods, followed by visual and
metallographic examination. Currently there is considerable interest in
the use of potentiodynamic polarization techniques as a means of
rapidly assessing the durability of metal that has been painted,
anodized, or given various polymeric surface treatments [28, 29].
1
Anodizing
Anodizing is a commercial surface treatment unique to aluminum.
The aluminum object is immersed as the anode in an acidic solution
(usually chromic, sulfuric, or phosphoric acid) and a controlled, direct
current is applied. Oxidation of the surface produces a hard, porous
film of aluminum oxide that is 10003000 times thicker than the
natural oxide, film thickness being on the order of 825 µm. The
columnar, porous oxide is sealed by immersion in boiling water
(sometimes with acetate salts added) which imparts impermeability to
the film. Before sealing, the porous film can be dyed to various colors,
and special electrolytes produce inherent coloration.
Anodized surfaces provide a shinier, more wear-resistant surface than
the natural oxide and they are easier to keep clean. Eventually they
will roughen or wear through from use. The principal attributes of
anodizing are its decorative appearance and ease of cleaning.
Unfortunately, its pH response is similar to the natural oxide, and
anodized films can be readily damaged by alkaline building materials
such as concrete or mortar.
Anodizing also is an excellent surface preparation for subsequent
painting, but usually less expensive chemical treatments are used for
this purpose.
2
Paints
Aluminum is painted for both decorative and protective reasons. The
surface should be treated for good pain adhesion. This involves
thorough cleaning plus an etch to roughen the oxide so that the paint
bonds to it. Examples of surface treatments used are in Ref. 30. Such
standards are developed over time and provide proven techniques.
However, for major applications recommendations should be obtained
from commercial paint and chemical suppliers, who can provide
information on the newest, state-of-the-art products and procedures.
Cleaning is followed by a primer paint and then most often a top coat,
such as an epoxy or urethane. Some new one-step, self-priming
products are available. When a very high gloss is desired, a final
coating of clear lacquer is applied.
The final coating often is tailored to the application. Examples are
high gloss on auto bodies and self-cleaning paint on residential siding.
Optimum
Page 146
procedures for both surface preparation and painting of aluminum
often differ from those for steel, particularly for electrostatic painting.
Compromises have to be made when painting a multimetal (material)
product, e.g., an auto body, and designers may be limited to using
existent paint line conditions.
Maximum protection depends on maintaining an unbroken paint
envelope, and repairs should be made when needed. This depends
greatly on the application and life expectancy desired. For example,
painted jet airliners are stripped of their coating and completely
repainted on a regular basis. Automobiles are repainted as needed,
usually for appearance purposes. Dents and scratches in residential
siding are rarely even repaired, whereas rain-carrying systems (gutters
and down spouts) often are less expensive to replace than to repair and
repaint.
Antifouling paints to prevent growth of algae, barnacles, and other sea
organisms must be tailored to use on aluminum. The common
antifouling paints for steels are not suited for use with aluminum
because they contain leachable heavy metals, such as arsenic, copper,
and lead, that can plate out on the aluminum and cause severe local
corrosion.
3
Polymer Coatings
Clear polymer coatings are used to provide protection while retaining
a glossy metallic appearance. All beverage and food containers are
coated for prolonged shelf life and to prevent contamination of the
food product. A hole-free coating is required. These coatings can be
color-tinted to identify that the metal, has indeed, been coated, or to
color-code the type of coating applied. These coatings often are
applied by roller-coating large coils of sheet prior to stamping or other
forming of the final product. Consequently, these coatings must have
good adhesion and elastomeric properties.
F
Inhibitors
Inhibitors usually only reduce corrosion; but in some cases, like
closed water systems, proper inhibition can virtually eliminate
corrosion.
Most inhibitors prevent or reduce corrosion of aluminum by altering
either the anodic or the cathodic reactions. Chromates suppress the
anodic reactions and chromate conversion coatings have been a
widely used inhibitor for aluminum sheet. In recent years, use of
chromates has decreased because of concerns for personnel and
environmental toxicity. If anodic inhibitors are used in insufficient
amounts, they may only restrict corrosion to fewer sites and actually
increase corrosion at those localized sites.
Most other inhibitors (phosphates, silicates, nitrates, nitrites,
benzoates, and others), used either alone or in combination, affect the
cathode reactions. Inhibitors of this type have been used to effectively
treat water systems, particularly closed systems. Development and
maintenance of an effective water protection
Page 147
system is complex, and advice should be sought of experts in the
water treatment field. If a multimetal system is to be treated (e.g., an
automobile radiator) the inhibitor must be compatible with all metals
in the system.
An inhibitor used extensively by the airline industry and now finding
applications elsewhere is a water-displacing penetrant-type oil
sprayed into crevices and joints. Flexing fatigue action during flight
helps the oil penetrate into the joint and its hydrophobic nature
displaces water, stopping corrosion.
VII
Relative Resistance of Aluminum and Alloys
As explained earlier, regardless of environment, (1) pure aluminum is
more corrosion-resistant than alloys, (2) non-heat-treatable alloys tend
to be more resistant than heat-treatable alloys, and (3) the resistance to
corrosion tends to decrease as alloying content increases. However,
once the solid solubility limit of an element is exceeded, further
alloying has little effect. For example, alloy 2219 with a nominal
6.3% copper is used only in artificially aged tempers; but 2219-T851
or T87 has essentially the same resistance as 2024-T851 and T86,
which has a lower nominal copper of only 4.4%.
Aluminum products are used extensively in natural atmospheres.
Waters are probably the next most frequent natural environment,
followed by production equipment and containers for chemicals.
Hence it is worthwhile to conclude with a review the relative
resistance of alloys on that basis.
A
Atmospheric Weathering
The majority of aluminum products are used in what would be
considered an outdoor atmosphere. Outdoor atmospheres are
classified as seacoast, urban or industrial, and rural. Corrosion of
aluminum is self-limiting in all of these environments and the rate of
corrosion becomes very slow within 5 years, except for alloys and
tempers that incur exfoliation.
1
Seacoast Atmosphere
Seacoast atmosphere is by far the most damaging natural atmosphere
for aluminum alloys. The severe seacoast effect, however, is seen only
within 0.8 km (0.5 mile) of the shore line, and the seacoast effect will
be greatly diminished at distances of 38 km (25 miles) from the shore.
The corrosive effect is very dependent on prevailing wind direction
and roughness of surf action, which controls how far inland actual salt
mist will be carried. Differences have been reported between the two
beach test sites at Kure Beach, North Carolina, located 25 and 250 m
from the shoreline [31]. An even greater corrosion rate difference of
approximately 9:2 wa reported at two seashore sites at Fort Sherman,
Panama [31]. Corrosion of specimens exposed on the "breakwater"
site, incurring contin-
Page 148
ual wave spray, was much greater than at the "coastal" site located
about 50 m from the shoreline. The two sites are in close priminity
and were shown in a single aerial photograph.
If one wants to retain a shinny metallic surface in seacoast
atmosphere, even pure aluminum and all alloys will have to be
protected, usually by a clear lacquer or an anodized coating.
The 3XXX alloys and low-magnesium-content 5XXX alloys are too
low in strength to be considered as structural alloys. Hence they are
usually used as thin sheet or tubing. It is recommended that protection
be provided to prevent perforation of such thin products.
The higher strength 5XXX alloys and the 6XXX alloys have sufficient
inherent resistance to corrosion that they can be used without any
protective coating. These alloys will darken and show slight
roughening but can survive 20 or more years exposure without
appreciable degradation. This is particularly true if they are used in
relatively thick sections, about 12 mm or more, for which the amount
of corrosion should be slight enough to cause neglible loss on load
carrying strength [32].
The 2XXX and 7XXX alloys should never be used in seacoast
atmosphere in tempers that are susceptible to IGC, exfoliation, of
SCC. Likewise, thin products should be afforded protection against
perforation, even if they are susceptible only to pitting corrosion. For
exfoliation and SCC-resistant tempers, the need for protection
depends primarily on two things: (1) the desire to retain a pleasing
appearance and (2) whether the application is fatigue-critical, in which
case the depth of pitting corrosion might be a sufficient stress riser to
initiate a fatigue crack. For non-fatigue-critical, structural applications
using products of about 12 mm or more in thickness, weathering will
darken and stain the surface, and pitting corrosion will appreciably
roughen the surface to a maximum depth of about 0.15 mm; but the
loss in residual tensile strength and load carrying ability will be
neglible [32].
2
Urban or Industrial Atmospheres
Close proximity to a factory emitting specific chemical fumes can be
a problem if these condense on the aluminum surface as highly acidic
or basic electrolytes. However, urban/industrial atmospheres usually
are regarded as atmospheres that are high in CO and CO2, sulfates and
sulfites, and possibly various NXO gases. This type of atmosphere
definitely is less corrosive to aluminum and aluminum alloys than is
seacoast atmosphere. Cities using a large amount of sodium chloride
(rock salt) for snow removal may approach the corrosivity of the
seacoast during the winter months.
A few highly urban areas are capable of causing exfoliation, a notable
example being a Reynolds test site in the greater Chicago area [33].
Other cities with high NXO gases, like Los Angeles, have been found
capable of causing
Page 149
increased susceptibility to SCC [34]. However, for the most part,
urban/industrial environments do not cause exfoliation and the need
for protection is to (1) retain aesthetic appearance, (2) prevent pitting
perforation of thin parts, and (3) avoid failure by SCC. Even 2024T351 plate has not exfoliated after long periods in such environments
[32].
An aesthetic problem is the gradual darkening and eventual
blackening that occurs if there is no regulation on smoke emissions. A
noted example, is the performance in U.S. cities, such as Pittsburgh,
before and after smoke control regulations. This discoloration
problem is not limited to aluminum and occurs on other metals,
masonry, and stone.
Almost all of the aluminum alloys, at least with appropriate protective
measures, will provide usable corrosion resistance in urban/industrial
environments. Choice of alloy, therefore, generally is made on other
engineering requirements, plus cost and availability.
3
Rural Atmosphere
Rural sites, such as the ASTM test site at Pennsylvania State College,
were once considered as virtually noncorrosive for aluminum alloys.
However, in recent years, airborne pollution from coal-burning power
plants located upwind result in acid rain fall over a widespread area.
This has increased the corrosivity of many rural areas, particularly in
the ability to cause staining and darkening of the surface. It has not
resulted in the more severe forms of corrosion. The adverse effect of
acid rain increasing corrosion in rural areas has been considerably
more pronounced on steel than on aluminum alloys.
However, local problems do occur. A manufacturer of polished
highway trailers made of 5454-H34 sheet incurred dark brown
staining from rainfall on trailers stored outdoors at one plant location;
but not at a different plant location. Waxing the trailers or coating
them with a silicone based hydrophobic solution was necessary to
prevent this staining.
4
Indoor Atmosphere
Indoor air is relatively benign provided the temperature is relatively
constant (no marked, rapid cool-down) and particularly if the air is
dehumidified. Metallographers frequently store polished
metallographic mounts of aluminum specimens in sealed desiccators
for weeks without any staining occurring.
Staining, filiform corrosion, and other surface corrosion can be a
serious problem on products stored indoors in unheated buildings or
tractor trailers, etc. The problem is condensation on the metal during
cool nights after warm, humid days. Airborne pollutants, especially
SO2, dissolve in the condensed vapor resulting in a conducting
electrolyte. Serious staining problems can occur quickly and
surprisingly, and have actually occurred on the more resistant
materials, such as pure aluminum foil bonded to insulation board,
5182 ends on soft drink cans,
Page 150
and polished 6061-T6 truck wheels. The wheels were delivered
toward the end of a day and left overnight in an unheated truck trailer.
This problem is aggravated if the products are packaged in such a way
that moisture can condense on the aluminum surfaces, but then the
packaging helps retain moisture and prevent drying. The beverage
cans and truck wheels, for example, were held in place with plastic
"shrink" wrap which was not an air-tight envelope. The wrap acted to
prevent circulation and retard drying. Staining is usually a problem on
products with bright surfaces. Staining occurs to some extent on heattreated alloys stored in warehouses, but this has not been a serious
problem since these products already have a dark heat-treated film on
them. Also heat-treated plate and bar stored in distributor warehouses
usually are subsequently machined, often on all surfaces.
Another critical indoor atmosphere corrosion occurrence is that
normal humidity, in the typical 4055% relative humidity human
"comfort zone," can be a sufficient electrolyte to cause SCC in highly
susceptible low-copper or copperfree 7XXX alloys, such as 7079T651. Fortunately, this is now well known and these highly
susceptible alloys are no longer produced.
B
Waters
The heat-treatable 2XXX and 7XXX aluminum alloys are not suited
for use in waters. If they are required for their strength or other
engineering properties, then considerable protection will be provided.
Aluminum and the 3XXX, 5XXX, and 6XXX aluminum alloys have
good resistance to most natural waters. Any corrosion that occurs
usually is in the form of pitting, which follows a decreasing rate
curve. Appreciable rate test investigations were made in the 19501970
period.
1
Freshwaters
Soft waters have the least effect on aluminum alloys, in contrast to
their effect on iron and copper. Distilled, deionized, and steam
condensate are readily handled by aluminum systems.
Some special planning is necessary if multiple metals will be present
in a water-handling system. For example, soft water, especially if
acidic, will dissolve copper. If this water subsequently comes into
contact with bare aluminum the copper will plate out to form local
galvanic cells. Problems usually occur because of cathodic ions
entering the solution upstream of the aluminum, e.g., the corrosion of
a brass pump ahead of aluminum piping. One method of handling this
is to use a section of Alclad pipe as a heavy-metal trap at the start of
the aluminum piping, situated so that it can be readily replaced. The
Alclad "waster" pipe section scavenges the copper ions from the
flowing solution and is replaced periodically as it becomes
appreciably corroded. Another example of the
Page 151
same ion effect of this sort have been encountered with the use of
copper roof flashing and bare aluminum rain gutters, particularly since
rainfall has become more acidic.
Alloy 6061-T6 is well suited to water situations and is the preferred
metal of construction in most sanitary water treatment plants.
Aluminum canoes were original made of 6061-T9 (cold-rolled after
aging to increase strength). Subsequently the industry switched to
alloy 6010 and then to alloy 6013 for higher strength and better dent
resistance.
Care must be taken in treatments to prevent algae growth and the like
in small ponds and lagoons. The fungicides used often contain copper
or other heavy metals that will adversely affect aluminum and cause
severe, localized pitting.
The biggest corrosion concern in freshwaters is perforation by pitting.
Therefore, experimentally one wants to establish the maximum depth
of pitting, usually as a function of time, so that a corrosion rate can be
determined. Depth of pitting cannot be measured visually and must be
determined microscopically by a skilled microscopist. This requires
the use of multiple samples and several metallographic sections from
each sample to ensure a good assessment is obtained of the maximum
depth. Analysis of the measured pit depth data by extreme value
analysis can provide further confidence in the predicted maximum
depth of penetration.
Since the scanning electron microscope (SEM) became available,
investigators have sometimes exposed a tensile or fatigue specimen
which is precorroded and then tested to failure by tension or by
fatigue. Either method inherently causes failure at the site of the
deepest surface flaw. SEM examination of the fractured faces can then
accurately measure the maximum depth of pitting.
Early on Goddard [35] showed in several hundred tests that pitting of
aluminum in freshwater followed a cube root curve: di = Kti1/3 ,
where di is the maximum pit depth at time ti. The time to penetration
can then be calculated by the formula: t2 = t1(d2/d1)3. The significant
conclusion from the cube root curve is that doubling the thickness
increases the time to perforation by a factor of 8.
2
Seawater
Aluminum has been used appreciably in seawater as ships, off-shore
rigs, harbor piers, and retainment walls. For strength reasons, the
alloys of choice have been the high-magnesium 5XXX alloys and the
6XXX alloys, both of which are readily welded and have good weld
strength. Thick 5456-H116 plate has been used for cryogenic tanks on
compressed natural gas (CNG) tankers. Many of the newer naval
vessels now use aluminum superstructures for lighter weight and
increased speed.
Seawater tends to cause somewhat deeper pitting than does
freshwater. However, early tests [36, 37 and others] have established
that the rate of pitting again follows a cube root curve and becomes
self-limiting with time.
Page 152
Tests in the St. James River Estuary in the Chesapeake Bay and in
Lake Maracaibo, Venezuela indicate that brackish water with lower
salinity has about the same corrosiveness as seawater.
Tests at Miami, Florida have exposed identical specimens by
continuous total immersion and by intermittent immersion during high
tide. The continuously immersed specimens tended to develop fewer
but deeper sites of corrosion. Tests on pilings and pipe exposed above
and below water have shown that increased pitting sometimes occurs
at the waterline and in the splash zone. This most likely is the result of
increased oxygen content near the surface, plus concentration effects
due to partial drying in the splash zone.
3
Piping Applications
Another important use of aluminum has been the use of 3003 and
3004 piping in handling of seawater for desalinization plants and
high-purity water for cooling ion atomic energy plants. The predicted
life of the tubes can be extended through the use of Alclad tubing.
Pure aluminum, 3XXX, and lower magnesium 5XXX alloy tubing
have also been used extensively for various cooling applications on
both the commercial and residential scale. Examples are heat
exchangers, air conditioners, and automotive radiators.
C
Chemicals
Pure aluminum, the non-heat-treatable 3XXX and 5XXX alloys, and
in some cases 6XXX series alloys are suited to the production and
handling of many chemicals, so long as they are not excessively acidic
or alkaline. The 2XXX and 7XXX are not suited to chemical
applications. Applications are discussed in the chapters on the
chemical process industries in Ref. 38.
The subject is complex and performance depends on the physical
stage of the chemical (solid, liquid, or gas), the concentration,
temperature, and the presence of trace amounts of water or other
impurities. Consider the role of trace amounts of water vs. the
anhydrous chemical. In some chemicals, such as phenol, a trace
amount of water (0.1%) will decrease corrosion, while trace amounts
of water in liquid SO2 will form sulfuric acid and promote corrosion.
Temperature is also important. Laboratory tests showed that 3003 was
compatible with phenol up to a temperature of 50°C (122°F) but
became highly corrosive at higher temperatures [19].
The reader should consult handbooks on the compatibility of
aluminum with the chemical(s) of interest. Pertinent references are
(19, 39, and 40). Most testing has been with pure aluminum and alloys
3003 and 5052, which represent the lower strength, more resistant
aluminum alloys. If these alloys were incompatible, higher strength
alloys will also be incompatible. However, the reverse is not
necessarily true. Handbooks offer an initial guide and testing may be
Page 153
necessary to verify the suitability of the selected alloy and temper for
the user's particular application.
Acknowledgment
The author acknowledges Peter R. Bridenbaugh, vice president, and
director of research at Alcoa, for his gracious permission to use the
facilities of the Graphic Arts and Information Departments at Alcoa
Technical Center.
References
1. ASTM G1, Standard Practice for Preparing, Cleaning, and
Evaluating Corrosion Test Specimens, Annual Book of ASTM
Standards, Vol. 03.02, Annex A1, 1994, p. 28.
2. Aluminum Standards and Data, 10th ed., Aluminum Association,
Washington, DC, July 1990.
3. Registration Record of Aluminum Association Designations and
Chemical Composition Limits for Wrought Aluminum and Wrought
Aluminum Alloys, 1991.
4. Registration Record of Aluminum Association Alloy Designations
and Chemical Composition Limit for Aluminum Alloys in the Form
of Castings and Ingot, 1984.
5. K. R. Van Horn, (ed.), Aluminum, Vol. 1, Properties, Physical
Metallurgy and Phase Diagrams, American Society for Metals,
Metals Park, OH, 1967, p. 212, table 2.
6. ASTM B275, Standard Practice for Codification of Certain
Nonferrous Metals and Alloys, Cast and Wrought, Annual Book of
ASTM Standards, Vol. 02.02, 1994, pp. 282287.
7. SAE Handbook, General InformationChemical Composition,
Mechanical and Physical Properties of SAE Aluminum Casting
Alloys, Vol. 1, pp 10.0610.14, 1994, Society of Automotive
Engineers, Warrendale, PA.
8. D. O. Sprowls and R. H. Brown, What every good engineer should
know about stress corrosion of aluminum, Metals Progress,
April/May 1962.
9. John E. Hatch (ed.), Aluminum, Properties and Physical
Metallurgy, American Society for Metals, Metals Park, OH, 1984,
Chap. 7.
10. F. C. Frary, The electrolytic refining of aluminum, Trans. Am.
Electrochem. Soc., 4 (1925).
11. ASTM G34, Standard Test Method for Exfoliation Corrosion
Susceptibility in 2XXX and 7XXX Series Aluminum Alloys (EXCO
Test), Annual Book of ASTM Standards, Vol. 03.02, 1994, pp.
129137.
12. ASTM G44, Standard Test Method for Determining Susceptibility
to Stress-Corrosion Cracking of High-Strength Aluminum Alloy
Products, Annual Book of ASTM Standards, Vol. 03.02, 1994, pp.
184188.
13. M. Pourbaix, Atlas of Electrochemical Equilibria in Aqueous
Solutions, Pergamon Press, Oxford, 1966.
14. ASTM G110, Standard Practice for Evaluating Intergranular
Corrosion Resistance of Heat Treatable Aluminum Alloys by
Immersion in Sodium Chloride + Hydrogen Peroxide Solution,
Annual Book of ASTM Standards, Vol. 03.02, 1994, pp. 470472.
Page 154
15. B. W. Lifka and D. O. Sprowls, Significance of Intergranular
Corrosion of High Strength Aluminum Alloy Products, ASTM Special
Technical Publication 516, Localized Corrosion, Cause of Metal
Failure, 1972, pp. 120144.
16. ASTM G69, Standard Practice for Measurement of Corrosion
Potentials of Aluminum Alloys, Annual Book of ASTM Standards,
Vol. 03.02, 1994, pp. 265267.
17. M. S. Hunter, G. R. Frank, and D. L. Robinson, Mechanism of
corrosion of 2024 alloy as revealed by electron microscopy.
Proceedings of the Second International Congress on Metallic
Corrosion, National Association of Corrosion Engineers, 1963, p.
102.
18. D. O. Sprowls and R. H. Brown, Stress corrosion mechanisms for
aluminum alloys, International conference of Fundamental Aspects of
Stress-Corrosion Cracking, NACE, Houston, 1969, pp. 469512.
19. Guidelines for the Use of Aluminum with food and Chemicals,
Aluminum Assoc., 5th ed., 1984.
20. ASTM G3, Standard Practice for Conventions Applicable to
Electrochemical Measurements in Corrosion Testing, Annual Book of
ASTM Standards, Vol. 03.02, 1994, pp. 5461.
21. ASTM G82, Standard Guide for Development and Use of a
Galvanic Series for Predicting Galvanic Corrosion Performance,
Annual Book of ASTM Standards, Vol. 03.02, 1994, pp. 346351.
22. J. P. Moran and S. C. Byrne, Galvanic Corrosion, Its Principles
and Guidelines for Prevention (with Emphasis on Automotive
Applications), July 15, 1993. Request from Alcoa Technical Center,
Information Department, 100 Technical Drive, Alcoa Center, PA
15069.
23. R. B. Mears and R. H. Brown, Causes of corrosion currents, Ind.
Eng. Chem., 33:10011010 (1941).
24. ASTM G64, Standard Classification of Resistance to StressCorrosion Cracking of Heat-Treatable Aluminum Alloys, Annual
Book of ASTM Standards, Vol. 03.02, 1994, p. 245.
25. A. W. Peabody, Principles of cathodic protections, NACE Basic
Corrosion Course, (A. deS. Brasunas, ed.), National Association of
Corrosion Engineers, Houston, TX, 1970, pp. 1-5 to 37-5.
26. B. W. Lifka, Alcoa Center, Corrosion Test Number RD-33, Re:
7008 Alclad 7075-T76 Sheet, Nov. 17, 1970, unpublished data.
27. ASTM D3170, Standard Test Method for Chipping Resistance of
Coatings, Annual Book of ASTM Standards, Vol. 06.02, 1994, pp.
139141.
28. K. Nisacioglu, O. Lunden, and H. Holtan, Improving the corrosion
resistance of aluminum alloys by cathodic polarization in aqueous
media. Corrosion, 41 (5):247255 (1985).
29. F. Mansfeld and M. W. Kendig, Evaluation of anodized aluminum
surfaces with electrochemical impedance spectroscopy, Electrochem.
Soc., 135 (4):828833 (1988).
30. ASTM D1730, Standard Practices for Preparation of Aluminum
and Aluminum-Alloy Surfaces for Painting, Annual Book of ASTM
Standards, Vol. 06.01, 1992, pp. 216218. (Not in 1994 book).
31. G. F. Downs III and E. A. Baker, Comparative corrosion
evaluation: Fort Sherman,
Page 155
Panama and Kure Beach, North Carolina, TECOM Project Nol. 7CO-R87-TTO-003, TTC Report No. 891001, 1989.
32. B. W. Lifka, Corrosion resistance of aluminum alloy plate in rural,
industrial and seacoast atmospheres, Aluminum, 12:12561261 (1987).
33. S. J. Ketcham and E. J. Jankowsky, Developing an Accelerated
Test; Problems and Pitfalls ASTM STP 866, Laboratory Corrosion
Tests and Standards, ASTM, 1985, pp. 1423.
34. B. W. Lifka, SCC resistant aluminum alloy 7075-T73
performance in various environments, Aluminum, 12:750752 (1977).
35. H. P. Goddard, Can. J. Eng., 38:167 (1960).
36. R. B. Mears and R. H. Brown, Trans. Soc. Mar. Eng., 52:91
(1944).
37. T. E. Wright, H. P. Goddard, and I. H. Jenks, Corrosion, 13:481
(1957).
38. K. Van Horn (ed.), Aluminum, Vol, 2, Design and Application,
American Society for Metals, Metals Park, Ohio, 1967, Chaps. 12, 25.
39. H. H. Ulig (ed.), The Corrosion Handbook, John Wiley and Sons,
New York, 1948.
40. P. Juniere and M. Sigwaldt, Aluminum: Its Application in the
Chemical and Food Industries, translated from the French by W. C.
Barnes, First American Edition, Chemical Publishing Co., New York,
1964.
Page 157
8
Corrosion of Titanium
Philip A. Schweitzer
Fallston, Maryland
Titanium is the ninth most abundant element on earth and the fourth
most abundant metal. It is more plentiful than chromium, nickel, or
copper, which are commonly employed as alloys used to resist
corrosion. In spite of this, little use was made of titanium until the
commercialization of the Kroll process about 1950, which made
titanium sponge available. Although having the advantages of being
highly corrosion-resistant in oxidizing environments, a low density
(sp gr 4.5 approximately 60% that of steel), and high tensile strength
(60,000 psi), its widespread use has been somewhat limited by its
cost. However, as consumption has increased and new technologies
have been developed to reduce the high cost, usage has increased and
will probably continue to increase further. At the present time it is
competitive with nickel base alloys. Thinner sections, coupled with
decreased maintenance requirements and longer life expectancy in
many applications, permit titanium equipment installations to be costeffective despite a higher initial cost.
I
Alloys
The primary initial applications of titanium were in the aerospace
industry where mechanical properties were the primary consideration.
During this period many
Page 158
titanium alloys were developed for these applications. In industrial
applications, however, corrosion resistance is the most important
property. Consequently, this discussion of the titanium alloys will be
limited to those used for their corrosion resistance properties.
The four primary titanium alloys used industrially for corrosion
resistance are shown in Table 1. ASTM grade 2, unalloyed titanium is
the single alloy most often used for corrosion resistance. If better
formability or higher strength is required, ASTM grades 1, 3, and 4
are available containing more or less iron and oxygen.
ASTM grade 5 alloy is a general-purpose alloy used in the aerospace
industry for applications requiring higher strength or fatigue
resistance. Its corrosion resistance is somewhat inferior to that of the
unalloyed titanium.
Compared with unalloyed titanium grade 2, the titanium-palladium
alloy grade 7 offers improved corrosion resistance. This alloy, as
grade 11, is available with low oxygen and iron content for improved
formability.
TiCode 12 is a low-cost alternative for grade 7 in some applications.
II
Physical and Mechanical Properties
Titanium is a light metal with a density slightly over half that of ironbased or copper-based alloys. The modulus of elasticity is also
approximately half that of steel while its specific heat and thermal
conductivity are similar to those of stainless steel. Titanium has a low
expansion coefficient and a relatively high electrical resistivity. These
physical properties must be considered when designing and
fabricating process equipment.
Titanium and its alloys are available with strength and ductility
comparable to other corrosion resistant alloys. The TiCode 12 alloy
offers improved strength compared to grade 2 or grade 7 titanium. The
latter two grades have only moderate strength, which limits their
applications in many industrial areas. The physical and mechanical
properties of titanium are shown in Table 2.
Titanium has excellent fatigue properties. The fatigue/strength/tensile
strength ratio for titanium and its alloys at ambient temperature is
high, in the range of
Table 1 Titanium Alloys Used for Corrosion Resistance
Alloy
ASTM
UNS
Composition
grade designation
Ti-50A
2
R-50400 Unalloyed titanium
Ti-6A15
R -56400 Titanium-aluminum4V
vanadium alloy
Ti-Pd
7
R-52400 Titanium-palladium alloy
Ti-Code
12
R-53400 Titanium-nickel12
molybdenum alloy
Page 159
Table 2 Physical and Mechanical Properties of Titanium
Alloy grade
Property
2
5
7
12
Modulus of elasticity, in tension 106 psi 14.9 16.5 14.9 15.0
Modulus of elasticity, in torsion 106 psi 6.5 6.1 6.5 6.2
Density, lb/in.3
0.1630.1600.1630.164
Specific heat at 75°F, Btu/lb°F
0.1250.1350.1250.130
Thermal conductivity at 75°F, Btu/ft2h 114 50 114 132
°F in.
Coefficient of expansion 32600°F, 10-6 5.1 5.1 5.1 5.4
in./in. °F
Electrical resistivity at 75°F mW -cm 56 171 56.7 52
Tensile strength, ksi
50 130 50 70
0.2% yield strength, ksi
40 120 40 50
Elongation, 2 in. %
20 10 20 18
0.50.6. Unlike many metals, the fatigue properties of titanium are
relatively unaffected by many corrosive media. It has been shown that
the fatigue limit of unalloyed titanium in distilled water or simulated
seawater is actually higher than that for air.
Titanium has been approved for use under the ASME Pressure Vessel
Code for the construction of pressure vessels. The allowable stress
values for titanium alloys as a function of temperature are shown in
Table 3.
Unalloyed titanium grade 2 and the Ti-Pd grade 7 alloy allowable
stresses fade off more rapidly than those of Ti-Code 12 alloy. At
500°F (200°C) the design
Table 3 Allowable Design Stresses for Titanium
Plate
Allowable stress
values, ksi
For metal temperatures not
GradeGrade Gradea
exceeding (°F)
100
200
300
400
500
600
aCase BC78-326.
2
12.5
10.9
9.0
7.7
6.6
5.7
7
12.5
10.9
9.0
7.7
6.6
5.7
12
17.5
16.4
14.2
12.3
11.4
Page 160
stress for Ti-Code 12 is 70% greater than that of grade 2 or 7.
Titanium and its alloys maintain excellent properties to low
temperatures.
III
Types of Corrosion
Titanium, like any other metal, is subject to corrosion in certain
environments. The corrosion resistance of titanium is the result of a
stable, protective, strongly adherent oxide film. This film forms
instantly when a fresh surface is exposed to air or moisture. Additions
of alloying elements to titanium affect the corrosion resistance
because these elements alter the composition of the oxide film.
The oxide film of titanium is very stable though relatively thin, and is
attacked by only a few substances, most notable of which is
hydrofluoric acid. Because of its strong affinity for oxygen, titanium
is capable of healing ruptures in this film almost instantly in any
environment where a trace of moisture or oxygen is present.
Anhydrous conditions, in the absence of a source of oxygen, should
be avoided because the protective film may not be regenerated if
damaged.
The protective oxide film of most metals is subject to being swept
away above a critical water velocity. Once this takes place accelerated
corrosion attack occurs. This is known as erosion-corrosion. For some
metals this can occur at velocities as low as 23 feet/s. The critical
velocity for titanium in seawater is in excess of 90 feet/s. Numerous
corrosion-erosion tests have been conducted and all have shown that
titanium has outstanding resistance to this form of corrosion.
A
General Corrosion
General corrosion is characterized by a uniform attack over the entire
exposed surface of the metal. The severity of this kind of attack can be
expressed by a corrosion rate. With titanium this type of corrosion is
most frequently encountered in hot reducing acid solutions. In
environments where titanium would be subject to this type of
corrosion, oxidizing agents and certain multivalent metal ions have
the ability to passivate the titanium. Many process streams,
particularly sulfuric and hydrochloric acid solutions, contain enough
impurities in the form of ferric ions, cupric ions, and so forth to
passivate titanium and give trouble-free service.
B
Galvanic Corrosion
The coupling of titanium with dissimilar metals usually does not
accelerate the corrosion of titanium. The exception is in reducing
environments where titanium does not passivate. Under these
conditions titanium has a potential similar to aluminum and will
undergo accelerated corrosion when coupled to more noble metals.
For most environments titanium will be the cathodic member of any
Page 161
galvanic couple. It may accelerate the corrosion of the other member
of the couple but in most cases the titanium will be unaffected. As a
result of this, hydrogen will be evolved on the surface of the titanium
proportional to the galvanic current flow. This may result in the
formation of surface hydride films that are generally stable and cause
no problems. However, if the temperature exceeds 170°F (77°C)
hydriding can cause embrittlement.
The surest way to avoid problems with galvanic corrosion is to
construct equipment of a single metal. If this is not practical, select
two metals that are close in the galvanic series. If contact of dissimilar
metals with titanium is necessary, the critical parts should be
constructed of titanium because this is not usually attacked.
C
Hydrogen Embrittlement
The oxide film on titanium in most cases acts as an effective barrier to
penetration by hydrogen. However, embrittlement can occur under
conditions that allow hydrogen to enter titanium and exceed the
concentration needed to form a hydride phase (about 100150 ppm).
Hydrogen absorption has been observed in alkaline solutions at
temperatures above the boiling point. Acidic conditions that cause the
oxide films to be unstable may also result in embrittlement under
conditions in which hydrogen is generated on the titanium surface. In
any event, it appears that embrittlement occurs only if the temperature
is sufficiently high, i.e., above 170°F (75°C), to allow hydrogen to
diffuse into the titanium. Otherwise if surface hydride films do form
they are not detrimental.
Gaseous hydrogen has had no embrittlement effects on titanium. The
presence of as little as 2% moisture effectively prevents the absorption
of molecular hydrogen up to a temperature as high as 600°F (315°C).
This may reduce the ability of the titanium to resist erosion, resulting
in a higher corrosion rate.
D
Crevice Corrosion
Crevice corrosion of titanium is most often observed in hot chloride
solutions. However, it has also been observed in bromide, iodide, and
sulfate solutions.
Dissolved oxygen or other oxidizing species present in the solution
are depleted in the restricted volume of solution in the crevice. These
species are consumed faster than they can be replenished by diffusion
from the bulk solution. As a result the potential of the metal in the
crevice becomes more negative than the metal exposed to the bulk
solution. This establishes an electrolytic cell with the metal in the
crevice acting as the anode and the metal outside the crevice acting as
the cathode. Metal dissolves at the anode under the influence of the
resulting current. Titanium chlorides formed in the crevice are
unstable and tend to hydrolyze, forming small amounts of
hydrochloric acid. This reaction is very slow at first, but in the very
restricted volume of the crevice it can reduce the pH
Page 162
of the solution to a value as low as 1. This reduces the potential still
further until corrosion becomes quite severe.
Alloying with elements such as nickel, molybdenum, or palladium
improves the crevice corrosion resistance of titanium. Consequently
TiCode 12 and the titanium-palladium alloys are much more resistant
to crevice corrosion than unalloyed titanium.
E
Stress Corrosion Cracking (SSC)
Unalloyed titanium with an oxygen content of less than 0.2% (ASTM
grades 1 and 2) is susceptible to cracking only in absolute methanol
and higher alcohols, certain liquid metals such as cadmium and
possibly mercury, red-fuming nitric acid, and nitrogen tetraoxide. The
presence of halides in the alcohols accelerates cracking tendencies.
The presence of water (> 2%) tends to inhibit stress cracking in
alcohols and red-fuming nitric acid. Titanium is not recommended for
use in these environments under anhydrous conditions.
IV
Corrosion Resistance
In general, titanium offers excellent resistance in oxidizing
environments and poor resistance in reducing environments.
It has excellent resistance to moist chlorine gas, chlorinated brines,
and hypochlorites. Some corrosion rates for titanium in hypochlorite
solutions are given in Table 4. Titanium is not resistant to dry chlorine
gas. It is attacked rapidly and can ignite and burn if the moisture
content is sufficiently low. Approximately 1% water is required under
static conditions at room temperature. Somewhat less is required if the
chlorine is flowing. Approximately 1.5% water is required at 392°F
(200°C).
Table 4 Corrosion of Titanium in Hypochlorite Solutions
Test
Temperature,duration,Corrosion
Environment
°F/°C
days
rate Pitting
17% hypochlorous
acid with free chlorine
and chlorine
monoxide
50/10
203 <0.1
16% sodium
70/21
170 <0.1
None
hypochlorite
1820% calcium
7075/2124
204 Nil
None
hypochlorite
1.54% sodium
hypochlorite, 1215%
sodium chloride, 1%
sodium hydroxide
150200/6693 72 0.1
None
Page 163
Titanium is immune to all forms of corrosive attack in seawater and
chloride salt solutions at ambient temperatures. It is also very resistant
to attack in most chloride solutions at elevated temperatures.
Titanium offers excellent resistance to oxidizing acids such as nitric
and chromic acids. However, it is not recommended for use in redfuming nitric acid, particularly if the water content is below 1.5% and
the nitrogen dioxide content above 2.5%. Pyrophoric reactions have
occurred in this environment.
Titanium will be attacked by reducing acids such as hydrochloric,
sulfuric, and phosphoric.
It is also quite resistant to organic acids, which are oxidizing. Only a
few organic acids are known to attack titanium; these are hot
nonaerated formic acid, hot oxalic acid, concentrated trichloracetic
acid, and solutions of sulfamic acid.
Titanium is resistant to acetic acid, terephthalic acid, and adipic acids.
It also exhibits good resistance to citric, tartaric, carbolic, stearic,
lactic, and tannic acids. Good corrosion resistance is also shown to
organic compounds. In anhydrous environments when the temperature
is high enough to cause dissociation of the organic compound,
hydrogen embrittlement of titanium is a consideration.
Suggested Reading
1. Philip A. Schweitzer, Corrosion and Corrosion Protection
Handbook, 2nd ed., Marcel Dekker, New York, 1989.
2. Philip A. Schweitzer, Corrosion Resistance Tables, Parts AC, 4th
ed., Marcel Dekker, New York, 1995.
Page 165
Corrosion of Tantalum*
John B. Lambert
Lake Forest, Illinois
I
Overview
Tantalum has a unique place as a material of construction in the
chemical industry since its first application in the 1940s. Its chemical
inertness in highly corrosive chemical environments is remarkably
similar to that of glass, yet it has a number of advantages in
comparison to glass, as well as other nonmetallics such as graphite or
fluorocarbons.
1. The metal has excellent ductility and strength properties, which
approach those of mild steel. It can be readily fabricated using normal
metalworking techniques.
2. Tantalum has a high thermal conductivity, i.e., about 50 times that
of glass. Consequently, overall heat transfer coefficients for heat
exchanger designs are correspondingly higher than for glass or glasslined equipment.
3. Tantalum is resistant to impact damage and thermal shock. Whereas
nonmetallics are often susceptible to mechanical damage or breakage,
*Adapted with permission from M. Schussler and C. Pokross, Corrosion
Data Survey of Tantalum, 2nd ed., Fansteel, North Chicago, 1985.
Page 166
which may cause leakage and contaminate the process stream with
debris, tantalum is almost free of such problems.
In comparison to other corrosion-resistant metals such as Duriron,
titanium, or stainless steels, the better service life obtained with
tantalum because of lower corrosion rates frequently makes its use
cost-effective despite higher initial costs. In recent years considerable
progress has been made using fabrication techniques which reduce the
requirement for the high-cost tantalum material. For example, use of
thin-walled tubing, typically 0.0150.020 in. thick, is standard
engineering practice. Alloys, particularly tantalum-tungsten, are
significantly stronger than the base metal, and their use has permitted
optimization of mechanical designs without loss of corrosion
resistance. Also, cladding, by both resistance welding and explosive
bonding, or use of loose linings backed by lower cost steel, has given
the same reliability as unsupported solid tantalum designs at much
lower cost. Even if the installed cost of a tantalum unit is higher than
that of a competitive material of construction, this disadvantage is
often offset by longer service life, less frequent repair and
maintenance, and fewer process interruptions. Product quality
improvements may also result from better process reliability and by
reduced contamination from corrosion byproducts.
Tantalum has been used in the chemical and allied industries for a
variety of equipment, including both bayonet and tube-and-shell heat
exchangers, condensers, absorbers, spargers, valves and piping,
nozzles, spinnerettes, rupture diaphragms, thermowells, flow control
regulators, orifices, and repair kits for glass or glass-lined vessels.
Table 1 is a listing of a number of typical process applications for
tantalum equipment.
II
The Oxide FilmA Protective Barrier
On exposure to oxidizing or slightly anodic conditions, even at a
temperature as low as 25°C, tantalum forms a thin, impervious,
passive layer of tantalum oxide [2,3]. Tantalum's extraordinary
corrosion immunity indicates that this passivating oxide has the
broadest range of stability with regard to chemical attack or thermal
breakdown. Chemicals or conditions which attack tantalum, such as
hydrofluoric acid, are those which penetrate or dissolve this oxide
film, in the case of fluoride ion, by forming a complex
ion. Under
conditions where the oxide layer is lost, the metal loses its corrosion
resistance dramatically. Even this effect has a positive side in that the
limiting conditions for applicability can be readily defined. Since
tantalum oxide has a high dielectric constant, the formation of this
thin oxide film is also used in electronics for the manufacture of
tantalum rectifiers and capacitors. In these applications, when the
tantalum conductor is electrolytically anodized and the part is
processed to add a counterelectrode, a
Page 167
Table 1 Uses for Tantalum Chemical Equipment
Process or
Product
Equipment
operation
Amino acids Digesting proteins Bayonet heaters, condensers
in hydrochloric
acid
Ammonium Concentration
Heat exchangers
chloride, pure before
crystallizing
Aqua regia
Ore dissolving,
Bayonet heaters, pickling tank coils
pickling
Bromine, pure Purification from Boilers, condensers
chlorine and
organics
Chloral
HCl absorption
Absorbers
Chlorine
Brine cooling
Heat exchangers
ChlorobenzeneChlorinator, HCl Condensers, absorbers
absorption
Chromic acid Electroplating
Heat exchangers
Ethyl ether
Heating alcohol Bayonet heaters
reactor
Ethylene
Chlorination
Condensers
dichloride
Ethylene
Reactor, sulfuric Bayonet heaters
glycol
acid concentrator
Formic acid Distillation
Condensers
Fuming nitric Distillation
Multiple bayonet heaters, condensers
acid
Halogens,
Chlorine, bromine, Bayonet heaters, condensers,
except fluorine iodine genregulators, thermowells
erators and
recovery system
Hydrochloric Production,
Heat exchangers, gas coolers,
Remarks
High
pressure
Recovery of
byproduct
acid
acid
Hydroiodic
acid
purification,
recovery
Generation and
recovery
absorbers, chlorine burners, strippers,
thermowells
Heat exchangers, condensers
(table continued on next page)
Page 168
Table 1 Continued
Product
Process or
operation
Hydrolysis
Hydrogen
peroxide
Isopropyl
Concentration of
alcohol
sulfuric acid
Magnesium Concentration
chloride
Monosodium Glutamic
glutamate hydrochloride
production
Nitric acid Distillation,
recovery
NitroglycerinNitration, nitric
acid recovery
Perchloric Generation,
acid
concentration
Persulfuric Electrolysis,
acid
recovery
Phenol
Chlorination,
hydrolysis
Phosphoric Concentration
acid
Phosgene
Generation
Equipment
Remarks
Heat exchangers
Bayonet heaters
Heat exchangers
Complete plants, heat
exchangers
Bayonet heaters,
condensers
Condensers,
thermowells
Coils, condensers
Bayonet heaters,
electrode supports
Bayonet heaters, HCl
absorbers
Bayonet heaters
Bayonet heaters,
condensers
Rayon
Viscose process Bayonet heaters,
spinnerette cups,
thermowells
Sulfuric acid Concentration, Bayonet heaters
recovery
Trisodium Cleaning and
Bayonet heaters, coils
phosphate degreasing
Tantalum equipment used in
Raschig process
Tantalum cannot be used if
fluorine content exceeds 10
ppm
Vinyl
Chlorination
Anhydrous HCl plants
chloride
Source: Table condensed from table 7.1 from Ref. 1.
Page 169
reliable capacitor or rectifier device is produced capable of
withstanding high voltage potential without breakdown.
Tantalum occupies a position toward the electropositive end of the
electromotive force table. Therefore, when in contact with most other
metals, tantalum becomes cathodic. In galvanic couples in which
tantalum is the cathode, nascent (or atomic) hydrogen forms and is
absorbed by the tantalum causing hydrogen embrittlement. In
chemical equipment caution must be taken to electrically isolate
tantalum from other metals or otherwise protect it from becoming
cathodic. Cathodic protection is discussed later in this chapter.
III
Effect of Specific Corrosive Agents
A
Water
Tantalum is not attacked by hot or cold deionized, fresh-, or seawater.
It is also resistant to mine waters, which are often acidic. For
equipment exposed to boiler waters or condensates, the alkalinity
should be controlled. The pH should be less than 9. No failures have
been reported for exposure to steam condensate, and the metal is
considered resistant to saturated steam below 250°C (3.98 MPa or 577
psia). It has been reported that at 1127°C, water vapor is decomposed
by tantalum with absorption of oxygen and evolution of hydrogen. At
927°C and below, the reaction is negligibly slow.
Tantalum is resistant to hydrogen peroxide in all concentrations.
B
Acids
Like glass, tantalum is immune to attack by almost all acids under
normal conditions, except hydrofluoric. Tantalum is not attacked by
such acids as sulfuric, nitric, hydrochloric, aqua regia, perchloric,
hypochlorous, hydrobromic, or phosphoric when free of fluoride ion.
It is not attacked by organic acids, such as formic, acetic, oxalic,
lactic, monochloracetic, and phenol. It is attacked, even at room
temperature, by hydrofluoric acid and free sulfur trioxide. The charts
displayed in Figs. 14 outline the corrosion resistance of tantalum to
sulfuric, phosphoric, hydrochloric, and nitric acids, respectively, as a
function of temperature and acid concentration. Table 2 compares the
corrosion rates of tantalum, niobium, titanium, and zirconium in
various acid environments.
1
Sulfuric Acid
Tantalum is highly resistant to corrosion by sulfuric acid in all
concentrations to about 98%. Dilute acid has no effect even at boiling
temperature. A slow, uniform attack by concentrated acid begins at
about 175°C, but tantalum can be used successfully at a temperature
as high as 200°C with 98% acid.
Fuming sulfuric acid (oleum) attacks the metal much more rapidly, as
Page 170
Figure 1
Corrosion resistance of tantalum in sulfuric acid at various concentrations and
temperatures
(From Ref. 1.)
Figure 2
Corrosion resistance of tantalum in phosphoric acid at various concentrations
and temperatures
(From Ref. 6.)
Page 171
Figure 3
Corrosion resistance of tantalum in hydrochloric acid at various concentrations
and temperatures.
(From: Ref. 6.)
Figure 4
Corrosion resistance of tantalum and nitric acid at various concentrations and
temperatures.
(From: Ref. 6.)
Page 172
Table 2 Corrosion Resistance of Tantalum and Other Metals to Acids
Corrosion rate, mpy
Temp.,
Test period,
Solution
TantalumNiobium Zirconium Titanium
°C
days
HCl, 18% 1926
36
0.00000 0.00000 0.09a
4.5
HCl, conc. 1926
36
0.00000 0.12
0.08
698
HCl, conc. 110
7
0.00000 4b
18.75
Not tested
36
0.00000 0.00000 0.00000 0.05
HNO3, conc. 1926
1
1926
35
0.00000 0.02
Very
0.21
soluble
HNO3.2HCl
1
5060
1
0.00000 1.0
Very
Not tested
soluble
HNO3.2HCl
4
0.00000 0.02
0.18
Not tested
H2SO4, 20% 95100
35
0.00000 Not
0.00000 2.1
H2SO4, 50% 1926
tested
36
0.00000 0.02
Very
46.8
H2SO4, 98% 1926
soluble
30
0.00000 180b
Very
Very
H2SO4, 98% 145
soluble
soluble
30
0.01
Not
Not tested Not tested
H2SO4, 98% 175
tested
30
1.5
Not
Not tested Not tested
H2SO4, 98% 200
tested
6
29.
Not
Not tested Not tested
H2SO4, 98% 250
tested
36
0.00000 0.02
0.02b
6.75
H3PO4, 85% 1926
36
0.00000 0.00000 0.42c
0.03
FeCl3, 10% 1926
aBecame brittle.
bTarnished
cUneven corrosion
Source: Appears as table 2.5 in Ref. 1.
Page 173
shown in Fig. 5 [4]. However, the attack is uniform, even in the weld
areas. Grain boundary pitting does not occur, so that the life of
tantalum exposed to strong acid can be accurately predicted. The
presence of chloride, chromates, nitric acid, or ethyl alcohol does not
increase the corrosion rate in sulfuric acid. Although no instances of
hydrogen embrittlement in commercial practice using tantalum
equipment have been reported, at temperatures above 200°C, the
corrosion rate of tantalum and the amount of hydrogen absorption
increase with temperature and concentration. Embrittlement occurs
when the hydrogen content of the metal exceeds 100 ppm. The
corrosion rate and the amount of hydrogen absorbed decrease when an
oxidizer such as nitric acid or hydrogen peroxide is added to the
sulfuric acid.
2
Phosphoric Acid
As shown in Figs. 2 and 5, tantalum shows excellent resistance to
corrosion to reagent grade phosphoric acid at all concentrations below
85% and temperatures under 190°C. The boiling point of the acid as a
function of concentration is also
Figure 5
Corrosion rates of tantalum in fuming sulfuric acid, concentrated sulfuric acid,
and 85% phosphoric acid.
(From: Ref. 4.)
Page 174
shown in Fig. 2. The superiority of tantal0um vs. alternate materials of
construction becomes more evident as temperature and concentration
of the acid increase. However, if the acid contains more than a few
ppm of fluoride, as is frequently the case with commerical acid,
corrosion of the tantalum may occur, and corrosion tests should be run
to verify suitability.
In one study the corrosion resistance of tantalum in a vaporliquid
mixture from the system H3PO4-KCl-H2O containing 60250 ppm
fluoride has been investigated at 120°C and atmospheric pressure [5].
Corrosion rates calculated from the tests were on the order of 620 ×
10-3 mils per year (mpy) or 0.152 to 0.508 × 10-3 mm per year
(mm/y), indicating good corrosion resistance.
3
Hydrochloric Acid
Figure 3 gives data on the corrosion resistance of tantalum to aqueous
acid over the concentration range 037% and temperatures to 190°C
[6]. As previously, the curve shows the boiling point, and the metal is
resistant at all conditions at or below the boiling point. In sealed
capsule tests [7] corrosion rates were less than 10 mpy at 190°C and
concentrations below 30% and less than 50 mpy for 37% acid. At
concentrations above 30% at 190°C some hydrogen embrittlement
was detected, although this tendency was not noted at or below the
boiling point.
Tantalum is resistant to anhydrous hydrogen chloride gas to at least
250°C.
4
Nitric Acid
Tantalum is inert to nitric acid solutions in all concentrations and at all
temperatures to boiling (Fig. 4), and the presence of chlorides in the
acid does not reduce its resistance to corrosion. For acid at
temperatures below the boiling point, the corrosion rate is less than
0.015 mpy. Use of the tantalum under these conditions would
normally not be economical, however, as stainless steels will perform
adequately. Again, as the acid concentration and temperature
increases, the superiority of tantalum becomes increasingly evident.
For example, tantalum has been advantageously and successfully used
for years in handling fuming nitric acid at conditions up to 800 psig
and 315°C in chemical process equipment.
5
Hydrofluoric Acid
Hydrofluoric acid is the only good solvent for tantalum, with the rate
of attack ranging from slow for dilute acid to rapid for concentrated
solutions. The rate of dissolution can be accelerated by the addition of
nitric acid, hydrogen peroxide, and/or other oxidizing agents.
Embrittlement of the metal by absorption of nascent hydrogen can
occur, for instance, when the metal is undergoing pickling. When
sufficient nitric acid is present, embrittlement does not occur. The rate
of hydrogen absorption in dilute hydrofluoric acid may be greatly
reduced if the tantalum is made the anode in an electrolytic cell by
impressing 210 V on the material in the cell.
Page 175
6
Acid Mixtures and Other Acids
Over the temperature ranges commonly used for dissolution
processes, tantalum is inert to acid solvents such as aqua regia and
chromic acid ''cleaning solution" (H2SO4 + K2Cr2O7). Attack on
tantalum does not occur even in chromium plating baths containing
fluorides. For example, in one test [8] the solution contained 40%
CrO3 and 0.5% fluoride ion at temperatures of about 60°C for 2 1/2
months and the sample showed a corrosion rate of 0.02 mpy. Complex
ion formation between chromium and fluoride is thought to explain
the reduced fluoride activity.
Corrosion rates were measured on tantalum exposed to concentrated
sulfuric acid in the temperature range of 200270°C with additions of
90% nitric acid and 10% hydrochloric, which simulates the conditions
for a wet incineration process [9]. The corrosion rate was reduced by a
factor of 3 compared to that for sulfuric acid alone. The addition of
hydrochloric acid alone reduced the rate only slightly. When nitric
acid was added alone to the sulfuric, the corrosion rate was the same
as when both acids were added. Thus the oxidizing effect of nitric
acid apparently helps to stabilize the protective tantalum oxide film on
the metal.
Other inorganic acids, such as sulfamic, methylsulfuric, or
hydrobromic, do not corrode tantalum nor do the anhydrous acid
gases, hydrogen sulfide, phosphorous chlorides, SO2, SOCl2, and
chlorine oxides [8].
C
Alkali Salts, Organics, and Other Media
Although solutions of sodium and potassium hydroxide do not
dissolve tantalum, they tend to destroy the metal by forming
successive layers of surface scale. The rate of attack is accelerated by
both increasing temperature and concentration. It should be noted that
damage to tantalum chemical equipment has occurred unexpectedly
when strong alkaline solutions were used for cleaning. Tantalum is
attacked, even at room temperature, by strong alkali and is dissolved
by molten or fused caustic. It is fairly resistant to dilute alkaline
solutions, however. In one long-term exposure in a paper mill,
tantalum exhibited no attack in a solution having a pH of 10.
Tantalum has been used as anode baskets in a number of silver
cyanide barrel platers for several years of service life even though the
solutions were quite alkaline and contained free potassium hydroxide.
The tantalum is protected by the positive voltage of the cell itself and
remains bright and ductile.
Tantalum is not attacked by dry salts or salt solutions at any
temperature or concentration unless hydrofluoric acid is liberated
when the salt is dissolved or strong alkali is present. Salts which form
acidic solutions, such as ferric chloride, have no effect on tantalum.
However, fused sodium or potassium hydrosulfate dissolve tantalum
[8].
Most organic compounds, including acids, alcohols, ketones,
alkaloids, salts, and esters, have no effect on tantalum. Specific
exceptions should be made for chemicals which may hydrolyze to free
fluoride ion or contain (or liberate)
Page 176
sulfur trioxide or strong alkali. One other exception is worthy of note.
Mixtures of anhydrous methanol with chlorine, bromine, or iodine
cause a pit-type corrsion on tantalum at 65°C [10]. This observation is
unusual because tantalum is unattacked by either methanol, the
halogens alone, or the reaction product, methyl halide, even at some
what higher temperature. Furthermore, pit-type corrosion is rarely
observed with tantalum. It was concluded that the strong corrosive
attack of mixtures of methanol and halogens on tantalum depended on
the formation of a haloformic acid intermediate.
Tantalum is completely inert to body fluids and tissues. Bone and
tissue do not recede from tantalum, and this biocompatability makes it
an attractive material for body and dental implants. However, the
superior strength and rigidity of stainless steel and titanium and the
castability of high-cobalt alloys have led to their greater use for
prosthetic devices. Tantalum has nevertheless been used for bone
replacement and repair, for cranial repair plates, suture wire, and wire
gauze for abdominal muscle support in hernia surgery [11].
At red heat tantalum reacts with sulfur or hydrogen sulfide to form
tantalum sulfide (Ta2S4). At lower temperatures the metal is
completely inert. In like manner, at elevated temperatures of the order
of at least 800°C, tantalum powder or shavings react with elemental
carbon, boron, and silicon to form the corresponding binary
compounds. Tantalum also reacts with vapors of phosphorus,
selenium, and tellurium at comparable temperatures. In contrast, there
is only slight attack on the metal by liquid selenides and tellurides of
the rare earths and uranium in the range 13002100°C, and tantalum is
considered a satisfactory material for handling these intermetallic
compounds.
D
Gases
1
Oxygen and Air [8,12]
The kinetics of the reaction of tantalum with air may be considered an
extension of the reaction with oxygen, since tantalum forms oxides
preferentially over nitrides, although the rate of oxidation is generally
somewhat lower in air than in oxygen. Tantalum is quite stable in air
at 250°C and below. At 300°C it tarnishes after 24 h exposure. The
rate, as measured by weight gain, increases rapidly at higher
temperatures. At 500°C the white oxide, Ta2O5, begins to form.
Figure 6 is a plot of weight gain vs. temperature in air.
The presence of a few atomic percent of oxygen in tantalum increases
electrical resistivity, hardness, tensile strength, and modulus of
elasticity but decreases elongation, reduction in area, magnetic
susceptibility, and resistance to corrosion in hydrofluoric acid [8].
Since 1 at. % corresponds to 892 ppm, the effect of very small
contents of oxygen is evident. Figure 7 demonstrates the solubility of
oxygen in tantalum, as determined by an X-ray technique. The
Page 177
Figure 6
Corrosion rate of tantalum in air as a function of temperature.
(From: Ref. 4.)
conversion of tantalum into oxide has been shown to occur by
nucleation and growth of platelets along the {100} planes of the bodycentered metal.
The kinetics of oxidation of tantalum in pure oxygen have been
studied at temperatures up to 1400°C and at pressures ranging from
less than 1 to over 40 atmospheres (0.104.05 M Pa) [12]. The reaction
is initially parabolic, with a transformation to linear rate after a period
of time. Increasing the temperature not only increases the rate of
oxidation but also decreases the time before the reaction changes from
parabolic to linear behavior. Above about 500°C and pressures from
10 mm Hg to 600 psi (1333 Pa to 4.13 MPa) the transition occurs
almost at once. From 600°C to 800°C the oxidation shows a
pronounced increase in rate with pressure above 0.5 atm (0.05 MPa).
At 1300°C and 1 atm oxygen pressure,
Page 178
Figure 7
Solubility of oxygen in tantalum.
(From Ref. 12.)
tantalum oxidizes rapidly and catastrophically, but at 1250°C the
metal oxidizes linearly for a short time, then catastrophically. Unlike
tantalumoxygen reactions, however, tantalumair reactions do not
exhibit catastrophic oxidation at temperatures as high as 1400°C.
Oxygen attack is usually viewed as the primary mechanism for failure
of tantalum at low loads and elevated temperature. Consequently,
most attempts to protect tantalum against gas corrosion at high
temperature have aimed at imparting resistance to the base metal.
Although Ta2O5 forms thin, adherent, protective films below 500°C,
at higher temperatures the film becomes flaky and tends to spall. Two
approaches have been used to improve the oxidation resistance of
tantalum:
1. Form a denser, more adherent oxide film by alloy additions to the
tantalum to alter and modify the oxide phase.
2. Provide a protective coating to inhibit oxygen attack. Coatings
include silicides, aluminides, noble metals, and others [13].
Page 179
2
Nitrogen
It has been reported [12] that tantalum dissolves 4 at. % (3120 ppm)
nitrogen at 1000°C, and solubility decreases rapidly with decreasing
temperature. Another investigator [14] found using resistance
measurements that in the temperature range 16002000°C more than 7
at. % nitrogen dissolves in tantalum and forms a homogeneous solid
solution. However, when such a saturated, high-nitrogen solid
solution is cooled, fine particles of nitride precipitate as elongated,
platelet-shaped particles.
The effect of temperature on the rate of nitridation between 500°C and
850°C is shown in Fig. 8. The kinetics are "parabolic" above 600°C
[12], but at lower temperatures the data suggest that the reaction does
not obey the parabolic rate law. One investigator suggests that a plot
of cubic rate constants, although showing considerable scatter, is
linear with reciprocal temperature [15]. At pressures between 87 mm
Hg and atmospheric, the reaction with nitrogen appears to be
pressure-independent.
3
Hydrogen
Tantalum dissolves a considerable amount of hydrogen at
comparatively low temperatures [8,12]. The maximum solubility is 50
at. %, with the solubility decreasing rapidly with temperature.
Although tantalum does not react rapidly with hydrogen below 250°C,
it can absorb 740 times its own volume at red heat. Tantalum
containing more than 150 volumes of hydrogen loses its ductility [11].
Atomic or nascent hydrogen can be absorbed by tantalum even at
room temperature. Absorption of hydrogen is accompanied by an
expansion of the body-centered crystal lattice. When metal containing
absorbed hydrogen is heated to about 800°C or more in high
Figure 8
Effect of temperature between 500°C and 850°C on reaction of tantalum with
nitrogen.
(From: Ref. 16.)
Page 180
vaccum, it loses all of its hydrogen. If permanent damage to the metal
has not occurred during the lattice expansion, annealing or degassing
at 800°C or higher restores the metal to its original condition. In
addition to decreasing the ductility, strength, and density of tantalum,
the presence of hydrogen increases the hardness and electrical
resistivity. Figure 9 exhibits isothermal weight gain curves for the
reaction of tantalum and hydrogen between 350°C and 500°C [16]. It
should be noted that between 450°C and 540°C, as is evident in the
figure, the reaction exhibits a negative temperature dependence,
probably corresponding to a hydride phase transformation [12].
Above 540°C, the positive temperature dependence is resumed.
Failures due to hydrogen embrittlement have occurred in some severe
aqueous acid media in applications where tantalum was or became
electrically coupled to a less noble metal, such as mild steel. Under
these conditions, tantalum became the cathode in the galvanic cell so
created.
Because of the presence of stray currents, tantalum may become a
cathode in the system, and, consequently, absorb and become
embrittled by atomic (nascent) hydrogen in the electrolytic cell. The
presence of stray currents can result from induction from adjacent
lines, leakages, variable ground voltages, and others. Although stray
voltages may be transient, the effect of absorbed hydrogen is
cumulative in its effect on producing embrittlement. For applications
of pure tantalum in aggressive acids at high temperature, such
hydrogen embrittlement rather than the uniform corrosion is the main
concern [7,17]. Several methods have been proposed to reduce
hydrogen embrittlement of tantalum:
Figure 9
Effect of temperature between 350°C and 500°C on reaction of tantalum with
hydrogen.
(From: Ref. 16.)
Page 181
1. Complete electrical insulation of tantalum from all metals in the
system. For additional protection, the insulated tantalum may be
connected to the positive pole of a DC source (about 15 V) while the
other pole is connected to some other metallic part, which is exposed
to ground.
2. Addition of a selected oxidizing agent, e.g., nitric acid or nitrate, to
a mineral acid solution, such as sulfuric. Apparently, the oxidizer
prevents attack on, or immediately heals, the passivating oxide film.
3. Cathodic protection of the tantalum surface by contacting with a
noble metal such as platinum, which has a low hydrogen overvoltage
and is electrochemically cathodic to tantalum in the same environment
(other candidates are palladium, gold, rhodium, rhenium, and
ruthenium). Contact of the metal with tantalum is made by riveting or
welding a small spot of the noble metal to the surface. For example, in
concentrated HCl at 190°C, the corrosion rate of both the tantalum
and platinum have been found to be negligible, even though the
corrosion resistance of platinum is not good in concentrated
hydrochloric acid at high temperatures. Thus, both tantalum and
platinum are mutually benefited by the galvanic contact.
4. Anodizing the tantalum. When tantalum is anodized, the oxide film
increases thickness about 1517 Å/V as the formation voltage is slowly
increased [18]. Thus a 20-V film will be about 300340 Å thick. It has
been suggested that thicker anodic films, corresponding to such a
thickness or greater, may also be helpful in eliminating hydrogen
embrittlement.
5. Alloying. Substitutional alloying is another method of improving
the resistance of tantalum to hydrogen embrittlement. Small alloy
additions of about 13% of molybdenum or rhenium substantially
decrease the corrosion rate and hydrogen embrittlement of tantalum in
concentrated sulfuric acid at 250°C [7].
4
Halogens
Fluorine attacks tantalum at room temperature. The metal is inert to
wet or dry chlorine, bromine, and iodine up to 150°C. Chlorine begins
to attack tantalum at about 250°C. The reaction is rapid at 450°C and
occurs instantly at 500°C. The presence of water vapor sharply
decreases the corrosion by chlorine, so that with 3% water vapor
tantalum is useful to temperatures up to 400°C. Bromine and iodine
attack tantalum at about 300°C forming the respective tantalum
bromide or iodide [4,8].
5
Carbon Monoxide and Carbon Dioxide
Tantalum reacts with dry carbon dioxide at 8 atm (0.81 MPa) pressure
and 500°C [12]. In 10 days the weight gain is 6.7 mg/cm2 and in 60
days about 50 mg/cm2.
Page 182
At 1100°C tantalum reacts instantaneously with carbon dioxide to
form Ta2O5 and with carbon monoxide at about the same temperature
to form TaO, which converts to Ta2O5 when exposed to oxygen [8].
6
Nitrogen Monoxide and Nitrous Oxide
Below about 1125°C, the reaction rate of NO (as a 5% mixture in
argon) with tantalum cannot be detected, but thereafter the rate
increases rapidly with temperature [8]. The oxidation by nitrous oxide
(N2O) on an evaporated film of tantalum has been studied [19] over
the temperature range -56°C to 200°C. Fast dissolution and absorption
of N2O occurred at -56°C, accompanied by N2 evolution. Some
incorporation of N2O also occurred. The rate of N2O absorption was
independent of the pressure of the nitrous oxide.
7
Other Gases
Although there are few data, it is expected that oxygen-containing
gases, such as SO2 and NO2, react with tantalum at some elevated
temperature [19]. With hydrocarbons such as benzene or naphthalene,
tantalum reacts at temperatures between 1700°C and 2500°C to form
tantalum carbide. Tantalum has been used as a getter in vacuum tubes
to absorb residual gases at temperatures of 6501000°C.
Pure helium and argon do not react with tantalum. These gases are
used as inert "cover" gases for arc-melting and welding the metal.
E
Liquid Metals
Tantalum and tantalum-base alloys exhibit good resistance to many
liquid metals, as shown in Table 3, even to high temperature (1100°C)
in the absence of oxygen or nitrogen. Because liquid metals are good
candidates as coolants and heat transfer fluids, especially in nuclear
reactors and power generation systems, tantalum is a promising
material of construction for liquid metal containment. The specific
effects of a number of liquid metal systems on tantalum are described
in the following.
1
Aluminum
Molten aluminum reacts rapidly with tantalum to form the stable
intermetallic compound Al3Ta [8].
2
Antimony
Antimony vapor is said to severely attack tantalum at temperatures of
1000°C and higher [8].
3
Bismuth
The liquid has little action on tantalum below 1000°C but causes some
intergranular attack above this temperature [7,20].
Page 183
Table 3 Effects of Molten Metals on Tantalum
Media
Remarks Temp., Code
°C
Aluminum
Forms
Molten NR
Al3Ta
Antimony
to 1000 NR
Bismuth
to 900 E
Calcium
Molten E
Gallium
to 450 E
Lead
to 1000 E
Lithium
to 1000 E
Magnesium
to 1150 E
Mercury
to 600 E
Potassium
to 900 E
Sodium
to 900 E
Sodium-potassium
to 900 E
alloys
Tin
V
Uranium
V
Zinc
to 500 E/V
Mg-37% Th
in He
to 800 S
Bi-(510%) U
in He
to 1100 S
Bi-5% U-0.3% Mn
in He
to 1050 S
Bi-10% U-0.5% Mn in He
to 1160 S
Al-18% Th-6% U
Failed
to 1000 NR
U-10% Fe
Failed
to 900 NR
U-Cr (eutectic)
Failed
to 900 NR
Y-Sb intermetallic
18002000S
Cpd.
Y-Bi intermetallic
18002000S
Cpd.
Er-Sb intermetallic
18002000S
Cpd.
La-Sb intermetallic
18002000S
Cpd.
Pu-Co-Ce alloys
to 650 V
E, no attack; S, satisfactory (no or little attack); V,
variable depending on temperature and
concentration; NR, not resistant.
Source: Appears as table 5.1 in Ref. 1
4
Calcium
Tantalum is only slightly attacked at 1200°C. A crucible with a wall
thickness of 5.8 mils was reduced to 5.3 mils after 12 days exposure at
1200°C [7].
5
Cesium
Refluxing capsule tests indicated surface dissolution and severe attack
on tantalum after 720 h at 982°C and 1371°C, respectively [21]. In
contrast, the same tests
Page 184
with Ta-10%W alloy exposed at 1150°C for 528 hours showed no
mass transfer or attack.
6
Gallium
Tantalum resists molten gallium to 450°C but is attacked at
temperatures above 600°C.
7
Lead
Tantalum is very resistant to molten lead at temperatures up to
1000°C, the rate of attack being less than 1 mpy. No decrease in
stress-rupture life was observed when tests were conducted in molten
lead at 816°C [6].
8
Lithium
Tantalum has good resistance to molten lithium up to 1000°C as long
as the oxygen content of the tantalum is below 200 ppm [7,21,22]. In
static capsule tests conducted at Oak Ridge National Laboratory at
600°C, when oxygen of the tantalum exceeded a threshold level,
lithium penetrated the metal [23]. Penetration was confined to grain
boundaries at low oxygen levels, with the depth of attack and the
number of affected boundaries increasing with oxygen concentration.
At higher concentrations transgranular attack also occurred. The
mechanism involved the formation of a ternary oxide on the grain
boundaries and preferred crystallographic planes and proceeded as the
corrosion product became wedged into the boundaries.
9
Magnesium and Magnesium Alloys
Tantalum is unattacked by molten magnesium at 1150°C [8].
10
Mercury
In static tests, tantalum showed good resistance to mercury to
temperatures of 600°C [24]. Refluxing capsule tests showed no attack
up to 760°C. The excellent corrosion resistance of tantalum to
mercury was further verified in a two-phase natural circulation loop
test which ran for 19,975 h with a boiling temperature of 649°C and a
superheat temperature of 704°C. Posttest evaluation of the loop
showed no corrosion.
11
Potassium
The compatability of tantalum and potassium was studied at 600, 800,
and 1000°C in static capsule tests [25]. As the oxygen concentration
in the potassium increased, the amount of tantalum found in the
potassium after the tests also increased. The results suggested the
formation of an unidentified ternary oxide phase that is either
nonadherent or dissolved when the recovered potassium was
analyzed. When the tantalum specimens themselves contained oxygen
above a
Page 185
certain threshold level, potassium penetrated the tantalum, and
intergranular as well as transgranular attack was observed. The
threshold concentrations for intergranular attack at 600, 800, and
1000°C were 500, 700, and 1000 ppm oxygen, respectively. The
mechanism was believed to be the formation of the ternary oxide
phase. Other studies have shown that tantalum alloys which contain
an oxygen-gettering element (e.g., hafnium or zirconium) exhibit no
corrosion [24]. Thus, if oxygen is unavailable, as is the situation when
the getter element reacts with oxygen to form a stable oxide, corrosion
cannot occur by the formation and dissolution of a complex oxide.
12
Silver
Tantalum is only slightly attacked by silver at 1200°C. A tantalum
crucible tested at this temperature for 35 days showed a loss in wall
thickness of 0.8 mil.
13
Sodium
When free of oxygen, neither sodium nor sodium-potassium alloy has
any appreciable effect on tantalum [7]. Sodium does not alloy with
tantalum [6]. Oxygen contamination of the sodium (or alloy) causes
an increase in corrosion with slight weight loss in flowing liquid
metal. In an instance where the oxygen content of the tantalum was
390 ppm before exposure, extensive intergranular and transgranular
attack of the tantalum by sodium was found. In another test, sealed
capsules of tantalum two thirds filled with reactor grade sodium
showed no corrosion with the tantalum remaining bright and shiny
after exposure at 850°C for 5 h [27].
14
Tellurium
Corrosion of candidate materials for stills to separate radioactive
polonium-210 from bismuth by distillation at temperatures of
450950°C has been investigated [20,28]. Tellurium, which is
chemically similar to polonium, was used as a nonradioactive
simulant. Of the materials tested, tantalum was the most satisfactory
from the standpoints of fabricability and long-term corrosion
resistance. Tantalum corroded at rates up to 2 × 10-2 mph during the
initial 100200 h of exposure, and the rate decreased to less than 2 ×
10-3 mph for 400 h for concentrations of tellurium of less than 30% in
bismuth.
15
Thorium-Magnesium
In static tests the 63% thorium-37% magnesium eutectic had no
appreciable effect on tantalum at 1000°C, and no corrosion was noted
in dynamic tests for 28 days with a thermal gradient between 700°C
and 840°C [7].
16
Uranium and Plutonium Alloys
Short-term tests indicated that the practical upper limit for tantalum as
a container for uranium is about 1450°C. However, attack below this
temperature is signifi-
Page 186
cant. A tantalum crucible with a wall thickness of 0.06 in. was
completely corroded after 50 h at 1275°C [6]. Other investigations
have shown that tantalum is not attacked by uranium-magnesium and
plutonium-magnesium alloys at 1150°C. Extensive tests on
components for molten metal fuel reactors have demonstrated that
tantalum is a satisfactory material for several thousand hours of
service in liquidmetal environments [8].
17
Zinc
It is reported that molten zinc attacks tantalum at significant rates at
temperatures above 450°C. However, an industrial zinc producer has
observed excellent corrosion resistance at 500°C [8]. The maintenance
of an oxide film on the tantalum surface may explain the latter result.
IV
Corrosion Resistance of Tantalum-Base Alloys
Most of the foregoing has concerned the corrosion resistance of
unalloyed tantalum. However, alloys have been developed for
application-specific property improvements. For example, a tantalum40% niobium-0.5% tungsten alloy has been used as a cost reduction in
less demanding environments. The alloy is thought to be somewhat
less corrosion-resistant than pure tantalum but suitable in some media.
The major alloying addition used commercially is tungsten, which
gives significant improvement in strength without detracting from the
base metal's corrosion resistance. For example, Ta-2.5% W alloy has
about 50% higher tensile strength than the pure metal yet retains
excellent ductility, fabricability, and weldability. At 200°C this alloy
has yield strength of about twice that of pure tantalum, indicating that
the strengthening effects are increased at higher temperature.
Furthermore, as indicated previously, Ta-Mo and Ta-Re alloys have
shown promise in reducing the tendency for hydrogen embrittlement.
Figure 10 compares corrosion data for a number of binary
substitutional alloys tested in 95% H2SO4 at 250°C for 3 days [17].
As noted, additions of tungsten reduced the corrosion rate and
hydrogen absorption, but molybdenum and rhenium were more
effective. Additions of niobium and vanadium had only a slight
influence whereas lower valence elements, such as hafnium, increased
the corrosion rate.
A
Tantalum-Tungsten Alloys
Samples of several tantalum and alloy strips were exposed for selected
times in concentrated (95.598%) sulfuric acid at temperatures ranging
from 175°C to 200°C. Specimens were run for recrystallized powder
metallurgical (P/M) tantalum and for both pure electron-beam-melted
metal and Ta-2.5% W alloy. Both of the latter compositions were run
in as-rolled, stress-relieved, and fully recrystallized conditions. The
average corrosion rates in mpy per side are summarized
Page 187
Figure 10
Influence of solutes on the corrosion rate of tantalum exposed
for 3 days to 95% H2SO4 at 250°C.
(From: Ref. 17.)
in Table 4. Data showing the effect of tungsten content on corrosion
rate for the electron-beam-melted compositions are depicted in Fig.
11. It is concluded that at temperatures of 175°C and lower, the
tantalum-tungsten alloys have equivalent corrosion resistance to the
unalloyed metal. In special cases, where higher temperatures may be
required, there is an advantage in using the alloys, with the optimum
corresponding to the Ta-2.5% W composition. Furthermore, there are
no significant differences depending on the condition of the metal
(i.e., rolled, annealed, or fully recrystallized). Tests also run on
weldments in concentrated sulfuric acid up to the boiling point
indicate no selective corrosion of the weld metal or in the heataffected zone.
The design of dynamic equipment, such as pump impellers, requires
knowledge of the fatigue properties in specific corrosion
environments. An experimental program was developed to provide
high-frequency fatigue data to guide the design of equipment to pump
sulfuric acid at 150°C [29]. The results of the tests for both unalloyed
and Ta-2.5% W alloy are compared in Fig. 12, and, under the test
conditions, at 1010 cycles, the fatigue stress to have no failure is about
12% higher for the alloy.
Other corrosion tests for the Ta-2.5% W alloy in both concentrated
(37
Page 188
Table 4 Corrosion Rates for Tantalum Materials Exposed to
Concentrated Sulfuric Acid at 175200°C
Metallurgical
Test Exposure, Corrosion
Material
condition
temp., °C days
rate, mpy
Ta, EB Recrystallized
175
60
0.189
melted
Ta, P/M Recrystallized
175
60
0.217
Ta-2.5% Recrystallized
175
60
0.229
W
Ta-2.5% As-rolled
181
7
0.104
W
Ta-2.5% Stress-relieved
181
7
0.087
W
Ta-2.5% Recrystallized
181
7
0.087
W
Ta, EB As-rolled
199
3
0.72
melted
Ta, EB Recrystallized
199
3
0.96
melted
Ta-2.5% As-rolled
199
3
0.19
W
Ta-2.5% Stress-relieved
199
3
0.17
W
Ta-2.5% Recrystallized
199
3
0.18
W
Ta, EB Recrystallized
200
32
2.24
melted
Ta, P/M Recrystallized
200
32
2.27
Ta-2.5% Recrystallized
200
32
1.15
W
Ta-2.5% Recrystallized
200
13
1.24
W
Ta-5% W Recrystallized
200
13
1.34
Ta-10% Recrystallized
200
13
1.98
W
Source: Appears as table 6.1 in Ref. 1.
38%) hydrochloric acid for 24 h at 100°C and in 70% nitric acid for 3
days at 198°C have given corrosion rates of 0.04 mpy or less.
Researchers in other tests have found that both tantalum and
substitutional alloys became hydrogenembrittled in concentrated
hydrochloric acid at 150°C [17]. The conclusion has been made that
the Ta-2.5%W alloy is at least as resistant to acid media as the
unalloyed metal or perhaps slightly more resistant in strong acids near
the boiling point.
Tantalum-10% W, because of its higher tungsten content, is even
harder and stronger than lower alloys but is appreciably harder to
fabricate. This alloy has been used in some applications, such as pump
and valve parts, for which its improved physical properties are
desired. For example, Ta-10W is often used in valves as a hard plug in
combination with a softer tantalum seat. Corrosion tests were
conducted on this alloy in several environments. In acid media the
corrosion rates for the alloy were comparable to pure tantalum except
in concentrated H2SO4 (7090% concentration) at 230°C, where the
corrosion rate was about 50% higher for the alloy than for unalloyed
tantalum. At a slightly lower temperature of 205°C, the corrosion
rates were comparable. In 5% NaOH solution
Page 189
Figure 11
Corrosion rates for tantalum-tungsten alloys exposed to concentrated sulfuric
acid at 181°C and 208°C as a function of tungsten content.
(From Ref. 1.)
at 100°C, although there was little difference in the corrosion rates
between the alloy and pure metal, the Ta-10W alloy failed by
premature embrittlement. The pure tantalum did not but showed a
marked increase in yield strength, attributable to pickup of interstitial
oxygen, nitrogen, and hydrogen during exposure.
The corrosion resistance of a series of tantalum-tungsten alloys was
also studied in 50% KOH at 30°C and 80°C, 20% HF at 20°C, and a
mixture with 1 part KOH and 3 parts K3Fe(CN)6. In the hydroxide, a
maximum in corrosion rate was obtained at about 60 at. % tantalum.
A maximum in electrical resistivity was found at the same
composition. In 20% HF, the tantalum-tungsten alloy system exhibits
the relatively low corrosion rates associated with tungsten as long as
the tungsten content is at least 20%, below which corrosion rates
increase markedly. In the hydroxide-ferrocyanide mixture, alloys
exhibit little improvement over tantalum.
B
Tantalum-Molybdenum Alloys
Corrosion resistance was studied on a series of tantalum-molybdenum
solid solution alloys. The results of corrosion tests run for 500 h in
both concentrated sulfuric and hydrochloric acids are shown in Table
5. Although the corrosive attack was small for all cases, as long as the
alloy contained at least 50 at. % tantalum, the corrosion resistance of
tantalum was retained.
Page 190
Figure 12
Comparison of fatigue response of tantalum and Ta-2.5% alloy in 80% sulfuric
acid at 150°C.
(From Ref. 29.)
C
Tantalum-Niobium Alloys
Corrosion rates for Ta-Nb alloys run in hot and cold hydrochloric and
sulfuric acids increased roughly in proportion to the niobium content
of the alloy [30]. For example, even though a Ta-5% Nb alloy showed
excellent resistance under all test conditions, the rate of attack was
three times that for unalloyed tantalum. Other data have been reported
for corrosion tests of binary Ta-Nb alloys and ternary alloys based on
the Ta-Nb system [31]. Tests were carried out in 75% sulfuric acid at
room temperature and 185°C, in 70% sulfuric at 165°C, and in 20%
hydrochloric acid. The alloys containing 60% or more tantalum
appeared promising for boiling 70% H2SO4. Ternary alloys with
additions of zirconium, hafnium, chromium, and vanadium did not
offer advantages either in alloy fabricability or lowered corrosion
resistance.
D
Tantalum-Titanium Alloys
Dilution of tantalum with titanium is promising for the possibility of
providing an alloy with corrosion resistance almost comparable to
tantalum in some selected environments at a lower cost. Not only is
the cost of titanium less but also the density of alloy compositions is
lowered corresponding to the titanium content, and, on a volume
basis, less alloy weight is required. Corrosion tests in 1070% nitric
acid at the boiling point and at 190°C (in sealed glass tubes) were
conducted
Page 191
Table 5 Corrosion Rates of Tantalum-Molybdenum
Alloys in Concentrated Sulfuric and Hydrochloric
Acids at 150°Ca
Average corrosion rate,
mg/cm2-day
Conc.
Tantalum in
Conc. HC1
molybdenum (at. %)
H2SO4(98%)
(37%)
0
0.008
0.018
10.1
0.009
0.017
20.1
0.008
0.018
30.0
0.010
0.009
40.0
0.009
0.010
50.0
0.000
0.010
61.2
0.000
0.000
71.5
0.000
82.8
0.000
0.000
91.4
0.000
0.000
100.0
0.000
0.000
aSolutions saturated with oxygen.
Source: Ref. 6
on alloys ranging from pure tantalum to Ta-90Ti [7]. All of these
materials showed excellent behavior with corrosion rates less than 1
mpy and no hydrogen embrittlement. In hydrochloric acid solutions at
190°C, a media to which titanium is not resistant, the alloys rich in
titanium corroded at high rate while the tantalum-rich materials tend
to approach the resistance of pure tantalum. The tendency to hydrogen
embrittlement increased with acid concentration and titanium content.
Tests in hot, concentrated sulfuric acid led to similar conclusions.
E
Other Alloys
It has been observed that the presence of a small amount of iron or
nickel, e.g., in a tantalum weld, makes that site subject to the same
acid attack as would be experienced for iron or nickel alone [8].
Galvanic action as well as chemical attack is undoubtedly involved.
References
1. M. Schussler and C. Pokross, Corrosion Data Survey of Tantalum,
2nd Ed., Fansteel, North Chicago (1985).
2. J. Chelius, Use of refractory metals in corrosive environment
service, Mater. Eng. Quart., American Society for Metals, August
1957, pp. 5759.
Page 192
3. H. Diekmann and U. Gramberg, Tantalum as a construction
material in the chemical industry, Tantalum-Niobium International
Study Center Bull. No. 78, June 1994.
4. D. F. Taylor, Tantalum: its resistance to corrosion, presented at the
Chicago Section, Electrochemical Society, May 4, 1956.
5. A. Alon, M. Schor, and A. Vromen, Corrosion resistance of Ta to
mixtures of phosphoric acid and potassium chloride at 120°C,
Corrosion, 22(1): January 1966.
6. Tantalum, corrosion data, comparative charts and coating
characteristics, General Technologies Corporation, A Subsidiary of
Cities Service Corporation.
7. M. Stern and C. R. Bishop, Corrosion and Electrochemical
Behavior, Tantalum and Columbium (F. T. Sisco and E. Epremian,
eds.), John Wiley and Sons, New York, 1963.
8. C. A. Hampel, ''Tantalum," Rare Metals Handbook, Ch. 25, 2nd
Ed., Reinhold, New York, 1961.
9. J. Vehlow and H. Geisert, Tantalum corrosion under wet
incineration conditions: influence of the dosing components and study
of welded specimens, presented at International Corrosion Congress,
September 6, 1981.
10. E. Rabald, Werkstoffe Korrosion, 12: 695698 (1961).
11. D. F. Taylor et al., Tantalum and Tantalum Compounds,
Encyclopedia of Chemical Technology, Vol. 22, 3rd ed., John Wiley
and Sons, New York, 1983, pp. 541564.
12. F. E. Bacon and P. M. Moanfeldt, Reaction with common gases,
Columbium and Tantalum, (F. T. Sisco and E. Epremian, eds.), John
Wiley and Sons, New York, 1963.
13. C. T. Wang and R. T. Webster, Oxidation-resistant coatings for
tantalum, Metals Handbook: Surface Cleaning, Finishing, and
Coating, 9th ed., Vol. 5, American Society for Metals, Metals Park,
OH, 1982 pp. 665666.
14. H. D. Seghezzi, New investigations of the tantalumnitrogen
system. Proceedings of the 3rd Plansee Seminar, Reutte, Austria
(1958) pp. 593595.
15. F. F. Schmidt. W. D. Klopp, W. M. Albrecht, F. C. Holden, H. R.
Ogden, and R. I. Jaffee, U.S. Air Force Technical Report WADD-TR59-13 (1959).
16. E. A. Gulbransen and K. F. Andrews, Trans. AIME, 188:586599
(1950).
17. L. A. Gypen, M. Brabers, and A. Deruyttere, Corrosion resistance
of tantalum base alloys. Elimination of hydrogen embrittlement in
tantalum by substitutional alloying, Werkstoffe and Korrosion,
35:3746 (1984).
18. D. A. Vermilyea, The kinetics of formation and structure of
anodic oxide films on tantalum, Acta Metallurgica, 1(3):285 (1953).
19. J. M. Saleh and H. M. Matloob, Oxidation of titanium, tantalum,
and niobium films by oxygen and nitrous oxide, J. Phys. Chemi.,
76(24):24862489 (1974).
20. E. C. Miller, Liquid Metals Handbook, Atomic Energy
Commission, Dept. of Navy, Washington, DC 1952, pp. 144183.
21. E. E. Hoffman and R. W. Harrison, The compatability of
refractory metals with liquid metals, Refractory Metal Alloys
Metallurgy and Technology, Plenum Press, New York, 1968.
22. P. Cybulskis, Review of metals technology, liquid metals, Metals
and Ceramics Information Center, Batelle, Columbus Laboratories,
December 21, 1973.
23. R. L. Klueh, Effect of oxygen on the corrosion of niobium and
tantalum by liquid lithium, Report ORNL-TM-4069, Oak Ridge
National Laboratory, Oak Ridge, TN, contract W-7405-eng-26,
March 1973.
Page 193
24. J. R. Weeks, Liquidus curves and corrosion of Fe, Cr, Ni, Co, V,
Cb, Ta, Ti, Zr in 500750°C mercury, revised version of paper
presented at the 20th Conference, National Association of Corrosion
Engineers, Chicago, March 913, 1974.
25. R. L. Klueh, Effect of Oxygen on the compatability of tantalum
and potassium, Corrosion (NACE), 28(10):360367 (1972).
26. L. Rosenblum, C. M. Scheurmann, and T. A. Moss, Space-powersystem material compatability tests of selected refractory metal alloys
with boiling potassium," Symposium on Alkali Metal CoolantsCorrosion Studies and Systems Operating Experience, IAEA, Vienna,
Austria 1967.
27. J. K. Fink, J.J. Heilberger, R. Kumar, and R. A. Blomquist,
Interaction of refractories and reactor materials with sodium, Nucl.
Technol., 35:656662 (October 1977).
28. W. R. Kanne, Jr., Corrosion of metals by liquid bismuth-tellurium
solutions, Corrosion(NACE), 29(2):7582 (1973).
29. A. F. Conn and S. L. Rudy, High frequency fatigue tests of
tantalum in sulfuric acid at 150°C, Technical Report 7242-1,
Hydronautics, Inc., February 1973.
30. G. L. Miller, Tantalum and Columbium, Academic Press, New
York (1959).
31. D. Lupton and F. Aldinger, Possible substitutes for tantalum in
chemical plant handling mineral acids, Proceedings of the 10th
Plansee Seminar1981, Verlagsanstalt Tyrolia, Innsbruck (1981), pp.
101130.
Page 195
10
Corrosion of Zirconium
Te-Lin Yau
Teledyne Wah Chang
Albany, Oregon
I
Introduction
Zirconium, atomic number 40 and atomic weight 91.22, was identified
by the German chemist, Klaproth, in 1789. However, the metal itself
was not isolated until 1824, when Berzelius produced a brittle, impure
metal powder by the reduction of potassium fluorozirconate with
potassium. One hundred years later, van Arkel and de Boer developed
the iodide decomposition process to make a pure, ductile metal in
Einhoven, Holland. The "iodide crystal bar" process continues to be
used today as a method of purifying titanium, zirconium, and
hafnium, even though it is slow and expensive.
In the 1940s, several groups of scientists and engineers investigated
zirconium and other metals for use in nuclear reactors. For this
application, a suitable structural metal should have good hightemperature corrosion resistance, resistance to irradiation damage, and
transparency to thermal neutrons needed for the nuclear reaction.
There was a renewed interest in developing a process that could
produce a large quantity of zirconium at a much lower cost. In 1945,
only a few hundred pounds of zirconium was produced in the United
States. The cost was more than $300 per pound. Zirconium was
regarded as an exotic metal.
In 1945, development work on zirconium was initiated at the U.S.
Bureau of Mines in Albany, Oregon under Dr. Kroll's technical
direction. Dr. Kroll had
Page 196
already developed a production process for titanium by the reduction
of titanium tetrachloride with magnesium in an inert atmosphere
before 1940. A similar process for zirconium was developed in 1947
as a pilot plant with a weekly capacity of 60 pounds of zirconium
sponge.
About the time of Kroll's work, Dr. Kaufman of the Massachusetts
Institute of Technology and Dr. Pomerance of Oak Ridge found that
zirconium, as occurring in nature, was combined with hafnium. It was
the hafnium which gave the zirconium the high level of neutron
absorption. When the hafnium was removed, zirconium was found to
have a very low thermal neutron absorption cross-section. This was a
finding of great importance. At once Admiral Rickover, who directed
the U.S. Navy Nuclear Propulsion Program, decided to choose
zirconium for the naval reactor. This decision had stimulated an
avenue of R&D programs to advance zirconium technology in
production, Zr/Hf separation, property information, fabrication, and
applications.
It was found that highly or commercially pure zirconium was not ideal
because of its inconsistent corrosion and oxidation resistance in hightemperature water and steam. This abnormal behavior was attributed
to the presence of minor impurities. In particular, the effect of
nitrogen on the corrosion characteristics was very pronounced.
Various alloy development programs were established in the early
1950s to examine the effects of adding various elements to zirconium.
Independent discoveries by Battelle Memorial Institute and Iowa State
College revealed that tin proved most beneficial. The Zr-2.5% Sn
alloy was named zircaloy-1, which was recommended for the nautilus
reactor. By 1952, data showed that zircaloy-1 had an increasing rate
of corrosion over time. Activities on zircaloy-1 were stopped. An
urgent search for a new alloy was begun.
Fortunately, Bettis Atomic Power Lab already had an active program
of corrosion tests for a number of zirconium-based alloys. Included
was one ingot in which a small amount of stainless steel had
accidentally been added. Test results revealed the beneficial effects of
iron, nickel, and chromium. Quickly, zircaloy-2, the Zr-1.5% Sn0.12% Fe-0.1% Cr-0.05% Ni, was developed and specified for the
Nautilus reactor in August 1952. That reactor generated power on
December 30, 1954. The Nautilus got underway on January 17, 1955.
This marked the beginning of a new era.
Developmental work continued, since the limiting factor for zircaloy2 in a reactor was determined to be its absorption of hydrogen during
corrosion in high-temperature water. Bettis eventually discovered that
replacing nickel with iron produced an alloy which cut hydrogen
absorption in half. This alloy was named zircaloy-4. There is a
controversy on the effect of nickel. Some believe that the nickel
addition improves zirconium's corrosion resistance, but some don't.
Nevertheless, both zircaloys are important materials for nuclear
technology.
Demand for zirconium was on the rise, as the U.S. Congress had
authorized several nuclear submarines by the mid-1950s, and nuclear
power plants were on
Page 197
the horizon. The production cost for zirconium needed to be lowered.
This could be achieved by developing commercial sources, which
included Carborundum Metals, National Distillers Products, NRC
Metals, and Wah Chang. Wah Chang was contracted to provide
zirconium at a price just less than $10 per pound in April 1956. Only
Wah Chang (now Teledyne Wah Chang) remains as the most
experienced zirconium producer.
In 1958, zirconium became available outside the U.S. Navy programs.
Activities in developing applications for zirconium were booming.
The chemical process industry began to use zirconium by taking
advantage of its excellent resistance to a broad range of corrosives.
Thanks to its remarkable corrosion resistance and biocompatibility,
zirconium has found some medical applications, i.e., in surgical tools
and instruments, and in stitches for brain operations. Zirconium is
highly beneficial as an alloying element for iron-, copper-,
magnesium-, aluminum-, molybdenum-, and titanium-based alloys.
Zirconium is useful as a getter because of its ability to combine with
gases at elevated temperatures. Along with niobium, zirconium is
superconductive at low temperatures and is used to make
superconductive magnets. Zirconium is an engineering material and
should no longer bear an exotic image. In fact, certain chemical
process plants use more than 200 tons of zirconium in chemical
equipment.
In the nuclear industry, stainless steel was used to clad the uranium
dioxide fuel for the first-generation reactors. But by 1965, the force of
neutron economy had made zirconium alloys the predominant
cladding material for water-cooled reactors. There was a widespread
effort to develop strong, corrosion-resistant zirconium alloys.
Noticeably, the Ozhennite alloys were developed in the Soviet Union
for use in pressurized water and steam. These alloys contain tin, iron,
nickel, and niobium, with a total alloy content of 0.51.5%. The Zr-1%
Nb alloy also is used in the Soviet Union for pressurized water and
steam service. Researchers at Atomic Energy of Canada Ltd. took a
lead from the Russians' zirconium-niobium alloys and developed the
Zr-2.5% Nb alloy. This alloy is strong and heat-treatable. It is used
either in a cold-worked condition or a quenched-and-aged condition.
Also, zirconium is often stated as a rare metal. To the contrary,
zirconium is plentiful and is ranked 19th in abundance of the chemical
elements occurring in the earth's crust. Zirconium is more abundant
than many common metals, such as nickel, copper, chromium, zinc,
lead, and cobalt. The most important source for zirconium is zircon
(ZrO2 · SiO2), which appears in several regions throughout the world
in the form of beach sand. The supply of zirconium won't be a
problem in developing any application for this metal and its alloys.
Moreover, the cost of zirconium has been stable for many years and is
competitive with other high-performance materials.
Additional information is available in Refs. 13.
Page 198
II
General Characteristics
Zirconium and its alloys can be classified into two major categories:
nuclear and nonnuclear. They all have low alloy contents. They are
based on the a structure with dilute additions of solid solution
strengthening and a stabilizing elements like oxygen and tin.
However, in niobium-containing alloys, there is the presence of some
niobium-rich b particles.
One of the major differences between nuclear and nonnuclear
zirconium alloys is in their hafnium content. Nuclear grades of
zirconium alloys are virtually free of hafnium (not greater than 100
ppm). Nonnuclear grades of zirconium alloys may contain up to 4.5%
hafnium. Hafnium has an enormous effect on zirconium's nuclear
properties but has little effect on its mechanical and chemical
properties.
The majority of nuclear grade material are tubing which is used for
nuclear fuel rod claddings, guide tubes, pressure tubes, and ferrule
spacer grids. Flat materials, such as sheets and plates, are used for
spacer grids, water channels, and channel boxes for nuclear fuel
bundles. Bars are used for nuclear fuel rod end plugs.
Zirconium products of various types are available for nonnuclear
applications. They include ingots, forgings, pipes, tubes, plates,
sheets, foils, bars, wires, and castings. They are used to construct
highly corrosion-resistant equipment, such as the exchangers,
condensors, reactors, columns, piping systems, agitators, evaporators,
tanks, pumps, valves, and packings for use in the chemical process
industries.
A
Physical Properties
Zirconium is a lustrous, grayish white, ductile metal. A listing of the
physical properties of zirconium is given in Table 1. However, a few
comments can be made. First, zirconium's density is considerably
lower than those of iron-and nickel-based stainless alloys. Second,
zirconium has a low coefficient of thermal expansion favoring
equipment that requires a close tolerance. The coefficient of thermal
expansion of zirconium is about two-thirds that of titanium, about
one-third that of type 316 stainless steel (S.S.), and about one-half that
of Monel. Third, zirconium has high thermal conductivity, which is
more than 30% better than those of stainless alloys. These properties
make zirconium very fabricable for constructing compact, efficient
equipment.
B
Mechanical Properties
Zirconium ores contain a few percent of its sister element, hafnium.
Hafnium has chemical and metallurgical properties similar to those of
zirconium, although their nuclear properties are markedly different.
Hafnium is a neutron absorber but zirconium is not. As a result, there
are nuclear and nonnuclear grades of zirconium and zirconium alloys.
Some commercially available grades of zirconium and its alloys are
listed in Table 2.
Page 199
Table 1 Typical Physical and Mechanical Properties of Zirconium
Grades 702 and 705
Properties
Zr 702
Zr 705
Physical
Atomic number
40
Atomic weight
91.22
Atomic radius
1.601.62
Å (zero charge)
0.800.90
Å (+4 charge)
Density
6.510
6.640
(g/cm3 at 20°C)
0.235
0.240
(lb/in.3)
Crystal structure
a phase
Hexagonal
close-packed
(below 865°C)
b phase
Body-centered Body-centered cubic
cubic (above
(above 854°C)
865°C)
a + b phase
Hexagonal-close-packed +
body-centered cubic
(below 854°C)
Melting point
1852°C (3365°F)1840°C (3344°F)
Boiling point
4377°C (7910°F)4380°C (7916°F)
Coefficient of thermal
expansion per °C 25°C
(73°F)
5.89 × 10-6
6.3 × 10-6
Thermal conductivity
(300800K)
13
10
Btu-ft./h-ft2-°F
22
17.1
W/m-K
Specific heat [Btu/lb/ 0.068
0.067
°F(32212°F)]
Vapor pressure (mm
Hg)
2000°C (3632°F)
0.01
900.0
3600°C (6512°F)
Electrical resistivity [µ- 39.7
cm at 20°C (68°F)]
Temperature coefficient
of resistivity per °C
20°C(68°F)
0.0044
Latent heat of fusion
60.4
(cal/g)
Latent heat of
1550
vaporization (cal/g)
Mechanical
Modulus of elasticity
14.4
106 psi
99
GPa
Shear modulus
5.25
106 psi
36
GPa
Poisson's ratio (ambient 0.35
temperature)
55.0
14.0
97
5.0
34
0.33
Table 2 Commercially Available Grades of Zirconium Alloys
Composition (%)
Zr +
Alloy design Hf
Hf
Fe + Cr
(UNS no.) (min) (max) Sn
Nb
Fe
Cr
Ni Fe + Cr
Nuclear grades
Zircaloy-2
0.010 1.201.70
0.070.200.050.150.030.08
0.180.38
(R60802)
Zircaloy-4
0.010 1.201.70
0.180.240.070.13
0.280.37
(R60804)
Zr-2.5Nb
0.010
2.402.80
(R60901)
Chemical grades
Zr702
99.2 4.5
0.2
(R60702)
max
Zt704
97.5 4.5 1.02.0
0.20.4
(R60704)
Zr705
95.5 4.5
2.03.0
0.2 max
(R60705)
Zr706
95.5 4.5
2.03.0
0.2 max
(R60706)
Page 201
The presence of hafnium in zirconium does not significantly influence
mechanical properties other than the thermal neutron cross-section.
Moreover, hafnium is a valuable metal for many applications. The
source of hafnium comes as the byproduct in the production of
zirconium. The nonnuclear grades of zirconium alloys are also low in
hafnium content. Consequently, the counterparts of nuclear and
nonnuclear grades of zirconium alloys are interchangeable in
mechanical properties. However, specification requirements for
nuclear materials are more extensive than those for nonnuclear
materials. Only requirements for nonnuclear materials are given in
Tables 3 and 4. It can be seen that Zr 705 is the strongest one with an
excellent fabricability. Furthermore, Zr 706 has been developed for
severe-forming applications by lowering the oxygen content of Zr
705.
Additional typical property data are given in Table 5 and in Figs. 1
and 2.
C
Chemical and Corrosion Properties
Zirconium is highly reactive, as evidenced by its redox potential of
-1.53 V vs. the normal hydrogen electrode at 25°C. It has a strong
affinity for oxygen. In an oxygen-containing medium, zirconium
reacts with oxygen at ambient temperature and below to form an
adherent, protective oxide film on its surface. This film is self-healing
and protects the base metal from chemical and mechanical attack at
temperatures to 350°C. As a result, zirconium is a highly corrosionresistant metal.
Many engineering metals, such as iron, nickel, chromium, and
titanium, produce metal ions of a variable valency. Uniquely,
zirconium is predominantly quadrivalent in its oxides and many other
compounds. It forms very few compounds in which its valence is
other than 4. The chemistry of zirconium is characterized by the
difficulty of achieving an oxidation state less than 4. This character,
along with high oxygen affinity, allows zirconium to form protective
oxide films even in highly reducing media, such as hydrochloric acid
and dilute
Table 3 Nonnuclear Minimum ASTM Requirements
for the Room Temperature Mechanical Properties of
Zirconium Alloys
Minimum
tensile
Minimum yield
Bend
strength
strength (0.2% Minimum test
Alloy (MPa)
offset), % (MPa) elongationradiusa
Zr
380
207
16
5T
702
Zr
414
240
14
5T
704
Zr
552
380
16
3T
705
Zr
510
345
20
2.5 T
706
aBend tests are not applicable to material more than
4.75 mm thick. T is the thickness of the bend test
specimen.
Table 4 ASME Mechanical Requirements for Zr 702 and ZR 705 Used for Unfired Pressure
Vesselsa
Maximum allowable
stress in tension for met
temperature not exceedin
°C (MPa)
ASME
Tensile
Material form specification Alloy strength Minimum yield
and condition
number
grade
(MPa)
strength (MPa) 40 95 150205
Flat-rolled
SB 551
702
359
207
90 76 64
products
705
552
379
138115 98
Seamless
SB 523
702
359
207
90 76 64
tubing
705
552
379
138115 98
Welded
SB 523
702
359
207
77 65 55
tubingb
705
552
379
117 97 83
Forgings
SB 493
702
359
207
90 76 64
705
552
379
138115 98
Bar
SB 550
702
359
207
90 76 64
705
552
379
138115 98
aReaders should get the current version for possible revisions.
b85% joint efficiency was used to determine the allowable stress value for welded tube. Filler
material shall not be used in the manufacture of welded tube.
Page 203
Table 5 The 107 Fatigue Limits for Zirconium
Alloys
Fatigue limit
(MPa)
Alloy
SmoothNotched
Iodide Zr
145
55
Zircaloys or Zr 705 (annealed 2 283
55
h at 732°C)
Zr-2.5% Nb (Aged 4 h at
290
55
566°C)
sulfuric acid. Under these conditions, common metals and alloys may
form subordinate oxides or other compounds of low or no protective
capability.
Moreover, metal ions of a constant valency imply their stability. This
is an important requirement in many applications. For example, ZrCl4
is used as a catalyst in such reaction as the cracking of petroleum and
polymerization of ethylene. An ideal catalyst should not have any
reduction in its own mass. Also, zirconium can maintain the stability
of certain chemicals, such as hydrogen peroxide. Ions of a variable
valency are the common decomposition catalysts for hydrogen
peroxide.
Another advantage is that zirconium ions are colorless. This is
important when the color stability of products is a major concern.
Most transition metals produce ions of different colors depending on
their valence state.
Protective oxide films are difficult to form on zirconium's surface in a
few media, such as hydrofluoric acid, concentrated sulfuric acid, and
oxidizing chloride solutions. Consequently, zirconium is not suitable
or needs protective measures for handling these media.
1
Water and Steam
Zirconium has excellent corrosion and oxidation resistance in water
and steam at temperatures exceeding 300°C. Zirconium has a great
capability for taking oxygen from water for the formation of
protective oxide film. This capability is not weakened even when
zirconium is in a highly reducing medium. Most passive metals form
protective oxide films in aqueous solutions only when the solutions
are somewhat oxidizing. Consequently, zirconium is uniquely suitable
for nuclear applications since water-cooled reactors operate with
oxygen-or hydrogen-charged coolant at temperatures from 280°C to
300°C.
However, corrosion and oxidation of unalloyed zirconium in hightemperature water and steam were found to be irregular [4,5]. This
behavior is related to variations in the impurity content in the metal.
Nitrogen and carbon impurities are particularly harmful. The
oxidation rate of unalloyed zirconium increases markedly when
nitrogen and carbon concentrations exceed 40 and 300
Page 204
Figure 1
Tensile properties vs. temperature curves for zirconium
alloys: (a) Zr 702
ppm, respectively [4,5]. The irregular behavior of unalloyed
zirconium stimulated alloy development programs. Zircaloy-2,
zircaloy-4, Zr-2.5Nb, and Zr-1Nb are the most important ones
developed for nuclear applications because they are more reliable and
predictable for use in hot water and steam in addition to being
stronger.
As compared to unalloyed zirconium, zircaloy-2 has an improved
character in oxide formation at elevated temperatures. A tightly
adherent oxide film forms on this alloy at a rate that is at first quasicubic but after an initial period undergoes a transition to linear
behavior. Unlike the white, porous oxide films, on unalloyed
zirconium, the oxide film on zircaloy-2 remains dark and adherent
throughout transition and in the posttransition region.
Zircaloy-4 differs in composition from zircaloy-2 in having a slightly
higher iron content but no nickel. Both variations are intended for
reducing hydrogen pickup with little effect on corrosion resistance in
reactor operation. For example, in water at 360°C, hydrogen pickup
for zircaloy-4 is about 25% of theoretical, or less than half that of
zircaloy-2. In addition, hydrogen pickup for zircaloy-4 is less
sensitive to hydrogen overpressure than that for zircaloy-2. For both
alloys, hydrogen pickup is greatly reduced when dissolved oxygen is
present in the medium [4].
Zr-2.5Nb is considered to be somewhat less resistant to corrosion than
the
Page 205
(b) Ze 704
(c) Ze 705.
Page 206
Figure 2
Minimum creep rate vs. stress curves for zirconium alloys: (a) Zr 702
(b) Zr 705.
zircaloys with exception. Nevertheless, Zr-2.5Nb is suitable for many
applications, such as pressure tubes in the primary loops of some
reactors. Furthermore, the corrosion resistance of Zr-2.5Nb can be
substantially improved by heat treatments [4,5]. Also, Zr-2.5Nb is
superior to zircaloys in steam at temperatures above 400°C [6].
Page 207
2
Salt Water
Zirconium has excellent corrosion resistance to seawater, brackish
water, and polluted water. The corrosion properties of zirconium
grades 702 and 704 in natural seawater can be found in Ref. 9.
Zirconium's advantages include its insensitivity to variation in factors
like chloride concentration, pH, temperature, velocity, crevice, and
sulfur-containing organisms. Some of the results are summarized as
follows.
Zr 702 specimens with or without a crevice attachment were placed in
the Pacific Ocean at Newport, Oregon for up to 129 days. All welded
and nonwelded specimens exhibited nil corrosion rates. Marine
biofouling was observed; however, no attack was found beneath the
marine organisms or within the crevices.
Laboratory tests were performed on Zr 702 and Zr 704 in boiling
seawater for 275 days and in 200°C seawater for 29 days. Both alloys
were resistant to general corrosion, pitting, and crevice corrosion.
Tests of U-bend specimens, with or without steel coupling, of Zr 702,
nickel-containing Zr 704, and nickel-free Zr 704 were conducted in
boiling seawater for 365 days. No cracking was observed during the
test period. Overstressing of the tested U bends indicated that all
specimens remained ductile except for the welded nickel-containing
Zr 704 with steel coupling. Steel-coupled nickel-containing Zr 704
showed much higher hydrogen and oxygen absorption and formed
hydrides. Chemical analyses and metallographic examinations on
other U bends did not show evidence of hydrogen absorption and
hydride formation. Results of chemical analyses are given in Table 6.
3
Halogen Acids
Zirconium resists attack by all halogen acids, except hydrofluoric acid
(HF). It is vigorously attacked by HF at all concentrations and even
acid fluoride solutions [10]. It should be noted that zirconium's
corrosion resistance is not as poor in
Table 6 Chemical Analyses for Hydrogen and Oxygen (ppm)
of Tested U Bends in Boiling Seawater for 365 Days
Metal
HydrogenOxygen
Nonwelded Zr 702 U bend with steel
6
1350
coupling
Nonwelded Zr 704 (Ni-containing) U bend
8
1480
with steel coupling
Nonwelded Zr 704 (Ni-free) U bend with
9
1440
steel coupling
Welded Zr 702 U bend with steel coupling
8
1250
Welded Zr 704 U bend (Ni-containing) with 450
5000
steel
Welded Zr 704 (Ni-free) U bend with steel
5
1480
coupling
Page 208
fluoride salt solutions until fluorides become HF in pH < 3 solutions
[11]. The effect of pH on the corrosion of zirconium in fluoridecontaining solutions is shown in Table 7 [12]. This fact is taken
advantage of in the preparation of zirconium surfaces using mixtures
of hydrofluoric and nitric acids for various fabrication steps and for
improved corrosion resistance in certain nuclear and chemical
applications. In recent years, it appears that the change to have
fluorides in the process media has increased somewhat. One of the
possibilities is the increased usage of recycled chemicals. For
example, recycled sulfuric acid may contain more than 100 ppm
fluorides [13]. When zirconium equipment faces fluoride-containing
acids, inhibitors that form strong fluoride complexes should be added
for protecting zirconium equipment [12]. Effective inhibitors include
zirconium sponge, zirconium nitrate, zirconium sulfate, and
phosphorous pentoxide. Table 8 gives some results about the effect of
fluoride complexes on zirconium's corrosion in fluoride-containing
solutions.
The other halogen acids, i.e., hydrochloric (HCl), hydrobromic (HBr),
and hydriodic acids (HI), do not attack zirconium [1416]. According
to results generated at Teledyne Wah Chang, corrosion rates for
zirconium in boiling 20, 45, and 48% HBr are less than 5 mil per year
(mpy), and in up to 57% HI at temperatures up to 250°C are less than
1 mpy.
Table 7 Corrosion of Zirconium in Fluoride-Containing
Chloride Solutions at 80°C After Four 1-Day Cycles
Solution
F (as
F (as
Corrosion
NaF)
CaF2) P2O5
rate
2
2
CaCl (%)MgCl (%) (ppm)
(ppm) (ppm) pH (mm/yr)
0.2
0.1
200
100
1
8.79
0.2
0.1
200
100
3
0.17
0.2
0.2
0.2
2.0
2.0
2.0
2.0
2.0
6.6
6.6
6.6
6.6
6.6
0.1
0.1
0.1
1.0
1.0
1.0
1.0
1.0
3.3
3.3
3.3
3.3
3.3
200
200
200
200
200
200
200
100
300
300
2800
2800
2800
300
300
9800
9800
9800
300
300
1200 1
1
1200 1
1
3
800 1
1
1200 1
1
3
800 1
1
1200 1
3.54
8.79
0.39
2.87
0.01
0.13
3.71
0.01
1.92
0.01
0.00
1.02
0.02
Page 209
Table 8 Effect of Zr Sponge or P2O5 on the Corrosion
of Zirconium in Fluoride-Containing Solutions
Temp. Corrosion
Medium
Inhibitor (°C) rate (mpy)
None
90
7.2% AlF3 + 0.5% HF
16% Zr
90 >1000
sponge
<1
80 350
0.2% CaCl2 + 0.1% MgCl2 None
1200 ppm 80 15
+ 620 ppm CaF2; pH 1
P2O5
80 150
2% CaCl2 + 1% MgCl2 + None
1200 ppm 80 <1
620 ppm CaF2; pH 1
P2O5
80 40
6.6% CaCl2 + 3.3% MgCl2 None
1200 ppm 80 <1
+ 620 ppm CaF2; pH 1
P2O5
25 >1000
90% HNO3 + 200 ppm HF None
800 ppm 25 1
Zr sponge
Yet one of the most impressive corrosion properties for zirconium is
its excellent resistance in HCl at all concentrations and temperatures
even above boiling [1518]. Because of its strong reducing power, it is
very difficult for most metallic metals to form protective oxide films
in HCl. The presence of even a small amount of HCl in a medium
may cause common metals and alloys to suffer general corrosion,
pitting, and/or stress corrosion cracking (SCC).
The isocorrosion diagram for zirconium in HCl is shown in Fig. 3.
Zirconium is suitable for handling HCl at all concentrations.
Moreover, zirconium is not as susceptible to hydrogen embrittlement
in HCl as tantalum is [19,20]. As indicated in Ref. 20, tantalum lost
33% and 18% of its ductility after 1000 h in 11 M HCl and 11 M HCl
+ 7% GaCl3, respectively, at 70°C. Under the same testing conditions,
zirconium remained unattacked and retained 100% of its ductility.
Unlike most passive metals, the anodic polarization curves of
zirconium do not have the active region in this highly reducing acid
(Fig. 4). This corrosion property explains the resistance of zirconium
to crevice corrosion in chloride solutions. However, Fig. 4 shows that
zirconium is susceptible to localized corrosion, such as pitting,
intergranular corrosion, and SCC, when it is anodically polarized to
breakdown potentials. Zirconium is susceptible to pitting in £ 20%
HCl but to intergranular corrosion in > 20% HCl [21]. The same types
of corrosion problems may be developed in HCl when highly
oxidizing ions, such as ferric and cupric ions, are present. Figure 5
demonstrates the detrimental effect of ferric ions in 20% HCl at
100°C. It can be seen that the presence of ferric ions polarizes the
zirconium surface to a potential exceeding the breakdown potential.
Thus a local breakdown of the passive surface at preferred sites
occurs, and a
Page 210
Figure 3
The isocorrosion diagram for zirconium in HCl.
condition favors the occurrence of localized corrosion. To eliminate
preferred sites by pickling zirconium in a mixture of hydrofluoric and
nitric acids can suppress the breakdown process of passive films
[22,23]. Alternatively, maintaining zirconium at a potential in its
passive region, which is arbitrarily set at 50100 mV below the
corrosion potential, can counteract the detrimental effects resulting
from the presence of ferric ions [21].
4
Nitric Acid
Nitric acid (HNO3), because of its passivating power, is not
considered to be a difficult acid for passive metals to handle.
Nevertheless, HNO3 becomes highly corrosive when its temperature
is high, its concentration is too high or not high enough, or its purity is
poor. The passivating power favors the formation of oxide films but
may also cause the passive films to break down.
Zirconium is considerably more suitable than most passive metals for
handling HNO3, particularly, when the acid is hot, impure, and/or
variable in concentration. Under certain conditions, zirconium is even
more resistant than the noble metals to the acid. Zirconium's
temperature limit is somewhat higher than
Page 211
Figure 4
The anodic polarization curves of zirconium in HCl at near boiling temperatures.
that of noble metals. Traces of chloride may lead to rapid attack of
noble metals, but not of zirconium.
The excellent corrosion resistance of zirconium in HNO3 has been
recognized for more than 30 years [4,15,18]. Below the boiling point
and at 98% HNO3 and up to 250°C and at 70% HNO3, the corrosion
rate of zirconium is conservatively put at less than 5 mpy (Fig. 6).
Results of autoclave tests showed that the corrosion rates of zirconium
were less than 1 mpy in 80% HNO3 and 90% HNO3 at 120°C and
150°C [24]. Moreover, the corrosion rates were less than 1 mpy when
zirconium was tested in boiling 30 to 70% HNO3 with up to 1%
FeCl3, 1% NaCl, 1% seawater, 1% iron, or 1.45% type 304 stainless
steel at 205°C [24]. These results indicate that the presence of heavymetal ions and chlorides has little effect on the corrosion resistance of
zirconium.
Zirconium is normally susceptible to pitting in acidic oxidizing
chloride solutions. However,
ion is an effective inhibitor for the
pitting of zirconium [18,25,26]. The minimum
molar ratio
required to inhibit pitting of zirconium was determined to be 1 [18,25]
or 5 [26]. Still, the presence of an appreciable amount of HCl should
be avoided since zirconium is not resistant to aqua regia.
The anodic polarization curves of zirconium in HNO3 are shown in
Fig. 7.
Page 212
Figure 5
Effect of 500 ppm Fe3+ on the anodic polarization curve of zirconium in 20% HCl at
100°C.
Zirconium exhibits the passive-to-active behavior in HNO3. It has very
noble corrosion potentials because of the oxidizing nature of HNO3. The
transpassive potential decreases with increasing acid concentration.
However, common oxidizing agents, such as oxygen and ferric ion, will
not affect the corrosion resistance of zirconium. The polarization curves
do imply that zirconium may be sensitive to stress in concentrated HNO
This is consistent with the observation of SCC in U-bend specimens in
greater than 70% HNO3 [27]. The slow strain-rate technique can reveal
zirconium's SCC susceptibility in less than 70% HNO3 [28]. The primary
concern for using zirconium in HNO3 service would be SCC in
concentrated acid. Results of C-ring tests indicate that zirconium will
have a long life in concentrated acids when they are stressed below the
yield point [27]. Cracking can be prevented by avoiding high, sustained
tensile stresses or by applying other preventive measures [29].
Other concerns include the accumulation of chlorine gas in the vapor
phase and the presence of fluoride ions [34]. Chlorine gas can be
generated by the oxidation of chlorides in HNO3. Areas that can trap
gases should be avoided when Cl- is present in HNO3. Or, zirconium
equipment should be pickled for having
Page 213
Figure 6
The isocorrosion diagram for zirconium in HNO3.
much improved resistance to pitting in wet-chlorine-containing
vapors. As indicated previously, the corrosion of zirconium in
fluoride-containing acids can be controlled by adding an inhibitor,
such as zirconium sponge and its compounds, to convert fluoride ions
to noncorrosive complex ions.
5
Sulfuric Acid
Sulfuric acid (H2SO4) is the most important acid for use in the
manufacture of many chemicals. For example, it is used as a
dehydrating agent, an oxidizing agent, an absorbant, a catalyst, a
reagent in chemical syntheses, and more. The consumption of sulfuric
acid indicates a nation's industrial activity. These highly versatile
capabilities can be attributed to the complicated nature of this acid.
Dilute solutions are reducing, which makes passive metals vulnerable
to corrosion. In fact, hot, dilute solutions can be used to pickle steel
and stainless steel. Solutions become increasingly oxidizing at or
above 65%. The usefulness of common metals depends strongly on
acid concentration, temperature, and the presence of other chemicals.
The corrosion of zirconium is H2SO4 is rather straightforward (Fig.
8). It can be seen that zirconium resists attack by H2SO4 at all
concentrations up to
Page 214
Figure 7
The anodic polarization curves of zirconium in HNO3 at near boiling temperatures.
70% and at temperatures to boiling and above. In 7080% H2SO4, the
corrosion resistance of zirconium depends strongly on temperature. In
higher concentrations, the corrosion rate of zirconium increases rapidly
with concentration.
In the range in which zirconium shows corrosion resistance in H2SO4, a
passive film is formed on zirconium that is predominantly cubic
zirconium oxide (ZrO2) with only traces of the monoclinic phase [30].
Zirconium corrodes in highly concentrated H2SO4 (e.g., 80%) because
loose films are formed that prove to be zirconium disulfate tetrahydrate
[Zr(SO4)2 · 4H2O] [31]. Also, at the higher acid concentrations, films
that flake off are formed and are probably partly zirconium hydrides
[31].
The anodic polarization curves of zirconium in 4.972.5% H2SO4 at
near-boiling temperatures are shown in Fig. 9. As indicated in Fig. 9,
zirconium experiences a passive-to-transpassive transition in the acid
with increasing potential. Again, zirconium does not have the active
region in H2SO4, as do common metals and alloys. Figure 9 shows that
the transpassive potential of zirconium in the acid decreases with
increasing concentration. Figure 10 indicates that zirconium can tolerate
some amounts of strong oxidizing agents in £ 65% H2SO4
Page 215
Figure 8
The isocorrosion diagram for zirconium in H2SO4.
without a reduction in corrosion [32]. Moreover, in £ 40% H2SO4,
zirconium can tolerate large amounts of strong oxidizing agents.
Consequently, zirconium equipment is often used in steel pickling
[3336]. In > 65% H2SO4, zirconium becomes sensitive to the
presence of oxidizing agents [37].
The presence of chlorides in H2SO4 has little effect on the corrosion
resistance of zirconium unless oxidizing agents are also present.
Therefore, in the presence of oxidizing agents, chloride ions should be
controlled within a limit to avoid detrimental attack. Figure 11 can be
used as a guide when chloride ions and oxidizing ions coexist. Table 9
indicates that 6070% H2SO4 would be the optimum range for
zirconium to face the coexistence of chloride ions and oxidizing ions.
Zirconium weld metal may corrode preferentially when H2SO4
concentra-
Page 216
Figure 9
The anodic polarization curves of zirconium
in H2SO4 at near boiling temperatures.
tion is approximately 55% and higher [38]. Heat treatment at 775 ±
15°C for 1 h per 25.4 mm of thickness can be applied to restore the
corrosion resistance approaching that of the parent metal [3941].
However, this high-temperature heat treatment is not suitable for
equipment made of zirconium/steel-clad materials because of the big
difference in thermal expansion coefficients between these two alloys.
Heat treatment at a much lower temperature, such as 510 ± 20°C,
should be applied when there is concern about SCC. Zirconium is
susceptible to SCC in a narrow range of H2SO4, i.e., 6469% [42,43].
This susceptibility was suppressed in the past because it was typical to
heat-treat zirconium equipment for service in this narrow range. It was
detected recently when a zirconium-clad vessel, which was not heattreated, failed by SCC in this range. Consequently, heat treatment at a
temperature as low as 450 ± 25°C has been developed for dealing with
clad materials.
For zirconium equipment, it is very important to maintain acid
concentration within the limits indicated in Fig. 8. When the limit is
exceeded, zirconium may corrode rapidly. In £ 65% H2SO4, the vapor
phase is almost entirely water [44]. However, the concentration
change is neglible when a system is under a
Page 217
Figure 10
Effect of oxidizing ions on the isocorrosion
diagram for zirconium in H2SO4.
pressurized condition [45]. Acid concentration may change
significantly when, for example, the system is imperfectly sealed. In a
leaking system, the acid concentration can exceed the concentration
limit. Acid concentration can easily increase when the system is under
vacuum operation because water vapor is continuously taken away.
Factors that affect the corrosion properties of zirconium in H2SO4 are
discussed in Ref. 46. It is essential to operate zirconium equipment
within the limits. When the corrosion resistance limits of zirconium in
H2SO4 are exceeded, a pyrophoric surface layer may be formed on
zirconium under certain conditions [47,48]. The pyrophoric surface
layer on zirconium formed in 77.5% H2SO4 + 200 ppm Fe3+ at 80°C
consisted of g-hydride, ZrO2, Zr(SO4)2, and fine metallic particles
[48]. The combination of hydride and fine metallic particles is
suggested to be responsible for the pyrophoricity. Treating in hot air
or steam can eliminate this tendency [48].
Page 218
Figure 11
Chloride allowable for zirconium in H2SO4 under oxidizing condition.
6
Phosphoric Acid
Phosphoric acid (H3PO4) is less corrosive than other mineral acids.
Many stainless alloys demonstrate useful resistance in the acid at low
temperatures. As often occurs, corrosion rates of common alloys in
the acid increase with increasing temperature, concentration, and
impurities. Such areas as the liquid level line or the condensing zones
are particularly vulnerable to attack.
Table 9 Inhibiting Effect of Sulfate Ion on the
General and Local Corrosion of Zirconium in
Boiling Sulfuric Acid and Containing FeCl3
Temp. Corrosion rate
(°C)
(mpy)a
115
199b
40% H2SO4 + 2%
FeCl3
124
80
50% H2SO4 + 2%
FeCl3
131
55% H2SO4 + 2%
FeCl3
141
60% H2SO4 + 2%
FeCl3
166
70% H2SO4 + 2%
FeCl3
aFour 1-day cycles.
bLocalized corrosion.
83
26
33
Page 219
Figure 12
The isocorrosion diagram for zirconium in H3PO4.
Zirconium resists attack in H3PO4 at concentrations up to 55% and
temperatures exceeding the boiling point. Above 55% H3PO4, the
corrosion rate could increase greatly with increasing temperature (Fig.
12). The most useful area for zirconium would be dilute acid at
elevated temperatures. Zirconium outperforms common stainless
alloys in this area [49].
Figure 13 shows the anodic polarization curves of zirconium in
H3PO4 at near-boiling temperatures. As the concentration increases,
the passive range diminishes gradually and the passive current
increases progressively. It appears that zirconium passivates more
slowly in H3PO4 than in other mineral acids.
If H3PO4 contains more than a trace of fluoride ions, attack on
zirconium may occur. Since fluoride compounds are often present in
the ores used for making H3PO4, the employment of zirconium has
always been questioned. However, because P2O5 is an effective
fluoride inhibitor and is usually present in large amounts in H3PO4
processes, tests should be conducted to determine zirconium's
suitability in the actual stream. Furthermore, zirconium sponge and its
compounds can be used to complex fluorides.
Page 220
Figure 13
The anodic polarization curves of zirconium in H3PO4 at near boiling temperatures.
7
Other Acids
Zirconium has excellent corrosion resistance in up to 30% chromic acid
at temperatures to 100°C [50]. It is not suitable for handling chromeplating solutions that contain fluoride catalysts.
Zirconium is also resistant to some mixed acid systems. It can be in acid
mixtures of sulfuric-nitric, sulfuric-hydrochloric, and phosphoric-nitric.
The sulfuric acid concentration must be below 70% [18,51]. Zirconium
is aggressively attacked in 1:3 volume mixtures of nitric and
hydrochloric acids (aqua regia). In the 1:1 volume mixture, zirconium is
attacked by much more slowly than in the 1:3 mixture [18]. In mixtures
greater than the 3:1 ratio, zirconium is resistant.
Some data for the mixed acid systems are given in Table 10.
8
Alkalies
Zirconium resists attack in most alkalies, which include sodium
hydroxide, potassium hydroxide, calcium hydroxide, and ammonium
hydroxide [50,52,53]. This makes zirconium distinctly different from
other highly corrosion-resistant
Page 221
Table 10 Corrosion Rates of Zirconium in Some Mixed
Acids
Test solutiona (wt %) Temperature Corrosion rates
(°C)
(mpy)
RT,b 100
0.06
1% H2SO4 99%
HNO3
10%
90%
RT, 100
WGc
H2SO4
HNO3
14%
14%
Boiling
0.1
H2SO4
HNO3
25%
75%
100
150
H2SO4
HNO3
50%
50%
RT
0.63
H2SO4
HNO3
68%
Boiling
2000
5% HNO3
H2SO4
68%
Boiling
11
1% HNO3
H2SO4
75%
25%
RT
260
H2SO4
HNO3
88%
0.5%
RT
0.0
H3PO4
HNO3
88%
RT
WG
5% HNO3
H3PO4
Aqua regia
RT
Dissolved
20% HCl 20%
RT
Dissolved
HNO3
10% HCl 10%
RT
Dissolved
HNO3
7.5%
19% HCl
Boiling
0.5
H2SO4
34%
17% HCl
Boiling
0.3
H2SO4
40%
14% HCl
Boiling
0.2
H2SO4
56%
10% HCl
Boiling
2.0
H2SO4
60%
1.5% HCl
Boiling
1.0
H2SO4
69%
1.5% HCl
Boiling
5.0
H2SO4
69%
4% HCl
Boiling
15.0
H2SO4
72%
1.5% HCl
Boiling
20.0
H2SO4
aCorrosion test with HCl/H2SO4 and HCl/HNO3
solutions were conducted at TWC.
bRT. room temperature.
cWG. weight gain.
Source: Refs. 18 and 51.
materials, such as titanium, tantalum, graphite, glass, and
polytetrafluoroethylene (PTFE).
Zirconium U-bend specimens were tested in boiling, concentrated
NaOH at TWC. During the test period, the concentration changed
from 50% to about 85%, and temperature increased from 150°C to
300°C. The PTFE washers and tubes used to make the U bends
dissolved. However, the zirconium U bends remained ductile and did
not show any cracks after 20 days. It should be noted that stainless
steel is susceptible to SCC in alkaline solutions including NaOH
solutions.
Zirconium coupons were tested in a white liquor, paper-pulping
solution, which contained NaOH and sodium sulfide, at 120, 175, and
225°C. All coupons had corrosion rates of less than 1 mpy [54]. In the
same solution, graphite and
Page 222
glass both corroded badly at 100°C. Zirconium also exhibits excellent
resistance to SCC in simulated white liquors [54].
9
Salt Solutions
Zirconium is resistant to most salt solutions, which include halogen,
nitrate, carbonate, and sulfate [51]. Corrosion rates are usually very
low at temperatures at least up to the boiling point. Solutions of strong
oxidizing chloride salts, such as FeCl3 and CuCl2, are examples of the
few exceptions. In strong oxidizing chloride solutions, zirconium's
performance is very dependent on surface conditions [22,23].
Zirconium becomes quite resistant to pitting when it has a good
surface finish like the pickled surface [22,23].
Although zirconium has good corrosion resistance in sodium fluoride
and potassium fluoride at low temperatures, its resistance decreases
rapidly with increasing temperature or decreasing pH. Consequently,
zirconium is not ideal for handling most fluoride-containing solutions
unless fluoride ions are complexed.
Zirconium is considerably more resistant to chloride SCC than
stainless steels are. No failure was observed in U-bend tests conducted
in boiling 42% magnesium chloride (MgCl2) at TWC. Another
attractive property of zirconium is its high crevice corrosion
resistance. Zirconium is not subject to crevice corrosion even in acidic
chloride at elevated temperatures. No attack was observed on
zirconium in a salt spray environment [55].
Unlike many common metals, zirconium has very little affinity for
sulfur. Zirconium-sulfur compounds form only at temperatures above
500°C [56]. Furthermore, there is no instance of zirconium-sulfur
bonds forming in aqueous systems [57]. Hence, hydrogen sulfide
(H2S), which is highly corrosive to common reactions of zirconium in
sulfide-containing solutions. Zirconium coupons and U bends were
tested in numerous NaCl-H2S solutions at temperatures to 232°C
[9,5860]. No general corrosion, pitting, crevice corrosion, or SCC was
observed.
10
Organic Solutions
Zirconium has excellent corrosion resistance in most organic
solutions, except certain chlorinated compounds, as indicated in Table
11 [6163]. It has been extensively tested in organic-cooled reactors
where the coolant consisted of mixtures of high-boiling aromatic
hydrocarbons, e.g., terphenyls [61]. These coolants are noncorrosive
to zirconium. However, early experiments in the organic coolants
indicated that hydriding was a major concern. It was found that
chlorine impurity in the organic coolants was the major cause of gross
hydriding. Elimination of the chlorine and maintenance of a good
surface oxide film by ensuring the presence of adequate water (> 50
ppm) alleviates the hydriding problem.
Indeed, the combination of a lack of water and the presence of
halogens or halides is the major reason for zirconium to experience
corrosion problems in
Page 223
Table 11 Corrosion Rates for Zirconium in Organic Solutions
Environment
Concentration Temperature Corrosion rate
(wt %)
(°C)
(mpy)
Acetic acid
599.5
35-boiling <0.07
Acetic anhydride
0.03
99.5
Boiling
Aniline
<0.01
5, 20
35100
hydrochloride
Chloroacetic acid
<0.01
100
Boiling
Citric acid
<0.2
1050
35100
Dichloroacetic
<20
100
Boiling
acid
Formic acid
1090
35-boiling <0.2
Lactic acid
1085
35-boiling <0.1
Oxalic acid
<0.5
0.525
35100
Tartaric acid
<0.05
1050
35100
Tannic acid
<0.1
25
35100
Trichloroacetic
>50
100
Boiling
acid
Urea reactor
<0.1
58% urea
193
17% NH3
15% CO2
10% H2O
Source: Excerpted by special permission from Chemical
Engineering, June 2, 1980. Copyright 1980 by McGraw-Hill, New
York.
organic solutions. For example, addition of some water can suppress
zirconium's susceptibility to SCC in alcohol solutions with halide
impurities [6466]. On the other hand, zirconium shows excellent
corrosion resistance in certain chlorinated carbon compounds, e.g.,
carbon tetrachloride and dichlorobenzene at temperatures up to
200°C. A general discussion on the corrosion mechanisms of
zirconium in organic solutions is given in Ref. 63.
From a corrosion point of view, organic halides can be classified into
three groups, i.e., water-soluble, water-insoluble, and waterincompatible.
Water-soluble halides, such as aniline hydrochloride, chloroacetic
acid, and tetrachloroethane, are not corrosive to zirconium. They may
become more corrosive when water content is low and/or zirconium is
highly stressed. More active halides, such as dichloroacetic and
trichloroacetic acids, are more corrosive to zirconium. It is suspected
that these halides may attack zirconium or intermetallic compounds at
grain boundaries to form organometallic compounds. It should be
noted that certain organic compounds, such as alkyl and aryl halides,
are the common ones to react with most metals, including noble
metals, to form organometallic compounds. These reactions can be
suppressed when there is some water present in the media.
Consequently, addition of water and/or stress relief would be effective
in preventing the corrosion of zirconium in water-soluble
Page 224
halides. However, water addition may increase the corrosivity of
many organic solutions toward common metals and alloys, but it
seems always to be beneficial to zirconium.
Water-insoluble halides, such as trichloroethylene and
dichlorobenzene, are not corrosive to zirconium, probably because of
their stability. They won't dissolve in water, and they won't exclude
water either. They and water can be physically mixed together.
Water-incompatible halides, such as acetyl chloride, may be highly
corrosive to zirconium. They are not stable. They react violently with
water. There is no chance of water to be present in this type of halide,
which is the most undesirable organic for zirconium, and maybe other
metals, to handle.
11
Gases
Zirconium forms a visible oxide film in air at about 200°C. The
oxidation rate becomes high enough to produce a loose, white scale on
zirconium at temperatures above 540°C. At temperatures above
700°C, zirconium can absorb oxygen and become embrittled after
prolonged exposure.
Zirconium reacts more slowly with nitrogen than with oxygen since it
has a higher affinity for oxygen than for nitrogen and it is normally
protected by a layer of oxide film. Once nitrogen penetrates through
the oxide layer, it diffuses into the metal faster than oxygen because
of its smaller size. Clean zirconium starts the nitride reaction in
ultrapure nitrogen at about 900°C. Temperatures of 1300°C are
needed to fully nitride the metal. The nitriding rate can be enhanced
by the presence of oxygen in the nitrogen or on the metal surface.
The oxide film on zirconium provides an effective barrier to hydrogen
absorption up to 760°C, provided that small amounts of oxygen are
also present in hydrogen for healing damaged spots in the oxide film.
In an all-hydrogen atmosphere, hydrogen absorption will begin at
310°C. Zirconium will ultimately become embrittled by forming
zirconium hydrides when the limit of hydrogen solubility is exceeded.
Hydrogen can be effectively removed from zirconium by prolonged
vacuum annealing at temperatures above 760°C.
The corrosion resistance of zirconium and its alloys in steam is of
special interest to nuclear power applications. They can be exposed
for prolonged period without pronounced attack at temperatures up to
425°C. In the 360°C Steam, up to 350 ppm chloride and iodide ions,
100 ppm fluoride ions, and 10,000 ppm sulfate ions are acceptable for
zirconium in general applications but not in nuclear power
applications.
Zirconium is stable in NH3 up to about 1000°C, in most gases (CO,
CO2, and SO2) up to about 300 to 400°C, and in dry halogens up to
about 200°C. At elevated temperatures, zirconium forms volatile
halides. Depending on the surface condition, zirconium may or may
not be resistant in wet chlorine
Page 225
[22,23]. Zirconium is susceptible to pitting in wet chlorine unless it
has a properly cleaned surface.
12
Molten Salts and Metals
Zirconium resists attack in some molten salts. It is very resistant to
corrosion by molten sodium hydroxide to temperatures above 1000°C.
It is also fairly resistant to potassium hydroxide. The oxidation
properties of zirconium in nitrate salts are similar to those in air.
Zirconium resists some types of molten metals, but the corrosion
resistance is affected by trace impurities, such as oxygen, hydrogen,
or nitrogen. Zirconium has a corrosion rate of less than 1 mpy in
liquid lead to 600°C, lithium to 800°C, mercury to 100°C, and sodium
to 600°C. The molten metals known to attack zirconium are
aluminum, zinc, bismuth, and magnesium.
Table 12 illustrates zirconium's corrosion resistance in several molten
metal systems [67].
D
Selected Corrosion Topics
1
Pitting
Zirconium is susceptible to pitting in all halide solutions except
fluoride [68]. This susceptibility is greatest in chloride solutions and
decreases as the halide ion
Table 12 Corrosion of Zirconium in Some Liquid Metals
Temp. (°C)
Liquid
Melting temp.
300
600
800
metal
(°C)
Bi
271.3
UnknownPoor
Poor
Bi-In-Sn 60
Bi-Pb
125
Bi-Pb-In 70
Bi-Pb-Sn 97
Ga
298
Hg
-38.4
Li
180.5
Mg
650
Na, K, or 12.397.9
NaK
Pb
327.4
Sn
231.9
aAt its melting point.
Source: Ref. 67.
UnknownPoor
Good
Limited
UnknownPoor
Good
Limited
Limited Poor
Poor
Poor
Good
Limited
Poora
Good
Good
Gooda
Good
Unknown
Unknown
Unknown
Unknown
Poor
Unknown
Limited
Unknown
Unknown
Limited Limited
UnknownUnknown
Page 226
becomes heavier, i.e., the pitting potentials in 1 N solutions of Cl-, Br, and I- are 380, 660, and 910 mVNHE, respectively. However,
zirconium does not pit in most halide solutions, except acidic chloride
solutions, because its corrosion potential is often lower than the
pitting potential. The presence of oxidizing ions, such as ferric and
cupric ions, in acidic chloride solutions may increase the corrosion
potential to exceed the pitting potential. Therefore, pitting may occur.
Nitrate and sulfate ions can inhibit the pitting of zirconium under
certain conditions [25,26,69,70]. Other protective measures, which
will be discussed later, are also available in controlling pitting.
2
Stress Corrosion Cracking
Zirconium and its alloys resist SCC in many environments, such as
NaCl, MgCl2, NaOH, and H2S, which are strong SCC-inducing
agents on common metals and alloys. Zirconium service failures
resulting from SCC are few in chemical applications. The high SCC
resistance of zirconium can be attributed to its high repassivation rate.
In the presence of some water or oxygen, any breakdown in the
surface oxide film will be quickly healed.
The environments known to cause SCC of zirconium include FeCl3,
CuCl2, halogen or halide-containing methanol, concentrated HNO3,
6469% H2SO4, and liquid mercury or cesium [42,43,7,72].
Preventive measures for the SCC of zirconium include
1. Avoiding high sustained tensile stresses
2. Modifying the environment, e.g., changing pH, concentration, or
adding an inhibitor
3. Maintaining a high-quality surface film, i.e., one low in impurities,
defects, and mechanical damages
4. Applying electrochemical techniques
5. Shot peening
6. Achieving a crystallographic texture with the hexagonal basal
planes perpendicular to the cracking path
3
Fretting Corrosion
Fretting corrosion takes place when vibration contact is made at the
interface of tight-fitting, highly loaded surfaces, such as between the
leaves of a spring or the pats of ball and roller bearings. Fretting of
zirconium occurs when its protective oxide coating is damaged or
removed. It can be overcome. If it cannot be eliminated mechanically,
addition of a heavy oxide coating on the zirconium may eliminate the
problem. This coating reduces friction drastically and prevents the
removal of the passive protection oxide.
Page 227
4
Galvanic Corrosion
Because of the protective oxide film that forms on zirconium in air
and most oxygen compounds, zirconium assumes a noble potential
similar to that of silver (Table 13). Zirconium may become activated
and can therefore corrode at vulnerable areas when in contact with a
noble material, particularly in chloride solutions. Vulnerable areas
include areas with damaged oxide films and grain boundaries.
Graphite and carbides in the powder form can be very effective in
promoting galvanic corrosion since even a small amount of material
can produce a very large cathode area.
Other less noble metals will corrode when in contact with zirconium
when
Table 13 Galvanic Series in Seawater
Cathodic
Platinum
(most
noble)
Gold
Graphite
Titanium
Silver
Zirconium
Type 316, 317 stainless steel
(passive)
Type 304 stainless steel
(passive)
Type 410 stainless steel
(passive)
Nickel (passive)
Silver solder
Cupronickels (70:30)
Bronzes
Copper
Brasses
Nickel (active)
Naval brass
Tin
Lead
Type 316, 317 stainless steels
(active)
Type 304 stainless steel
(active)
Cast iron
Steel or iron
Aluminum 2024
Cadmium
Aluminum (commercially
pure)
Anodic
Zinc
(active) Magnesium and magnesium
alloys
Page 228
its oxide film is intact. For example, in seawater or acidic solutions,
corrosion of steel, aluminum, and zinc is accelerated in electrical
contact with zirconium. It is advisable to keep them electrically
insulated.
Since zirconium oxide is an excellent insulator, a thick coating of this
oxide on zirconium can be used to minimize galvanic corrosion.
Oxide formation methods will be discussed later.
5
Crevice Corrosion
Of all the corrosion-resistant metals and alloys, zirconium is among
the most resistant to crevice corrosion. In low-pH chloride solutions
or chlorine gas, for example, zirconium is not subject to crevice
corrosion. This resistance can be attributed to zirconium's excellent
corrosion resistance in HCl solutions.
Still, zirconium is not immune to crevice corrosion in a broad sense.
For example, when a dilute H2SO4 solution is allowed to concentrate
within a crevice, crevice corrosion of zirconium becomes possible.
E
Corrosion Protection
1
Oxide Film Formation
One of the unique properties of zirconium that makes it attractive for
chemical applications is the inert nature of its oxide film. Zirconium
oxide (ZrO2), which forms on zirconium's surface, is among the most
insoluble compounds in a broad range of chemicals. This film gives
excellent corrosion protection in most media, in spite of the reactive
nature of the metal. If it is mechanically destroyed, this impervious
oxide barrier will regenerate itself in many environments. For
corrosion resistance, there is no need to thicken the oxide film before
zirconium is placed in a corrosive medium.
However, for mechanical reasons, it is desirable to preoxidize
zirconium. Properly oxidized zirconium has a much improved
performance against sliding forces, although it can be damaged by
striking action. Oxidized zirconium pump shafts are an example of a
common application. Bolts and nuts are often oxidized for the purpose
of preventing galling.
Several methods of oxide formation are possible, depending on the
properties desired. They include anodizing, autoclaving in hot water
or steam, formation in air, and formation in molten salts.
Anodizing
Anodizing forms a very thin film (< 0.5 µm). The surface of
zirconium with anodized films appears in different colors, varying
through the entire spectrum. The thickness of the film is in the range
of the wavelengths of visible light and, because of interference of this
light, only certain wavelengths are selectively reflected through the
thin film from the zirconium metal underneath. Since the selected
wavelengths depend on the thickness of such a film, the
Page 229
change in color observed with increasing voltage indicates that the
film is growing in thickness. Nevertheless, the film is formed at
ambient temperatures. It does not have the adhesion to the underlying
metal of thermally produced films. Anodized films look great but
have very limited capability to protect the metal from mechanical
damage.
Autoclave Film Formation
Autoclave film formation is a practice common to the nuclear reactor
industry. In this method, the uniform film of high integrity is formed
in pressurized (19 MPa) deionized water at 360°C for 14 days or in
high-purity steam (10 MPa) at 400°C for 13 days. In addition to the
slower corrosion rate, the rate of hydrogen absorption is drastically
reduced.
Film Formation in Air or Oxygen
Film formation in air is the most common method used in the
chemical process industry. This film is formed during the final stress
relief of a component in air at 550°C for 0.54 h. This film ranges from
a straw yellow, to an iridescent blue or purple, to a powdery tan or
light grey. Such films need not be taken as a sign of metal
contamination. This treatment does not cause significant penetration
of oxygen into the metal but it does form an oxide layer that is
diffusion-bonded to the base metal.
The oxide film formed under a well-controlled condition serves as an
excellent bearing surface against a variety of materials in several
media and over a wide temperature range [73,74]. For example, a
layer of black oxide film can be formed on a cleaned zirconium
component in air at 550°C for 46 h [73] or in a fluidized bed using
oxygen during the oxide formation stage but using an inert gas during
the heating and cooling stages [74]. The resultant oxide layer,
approximately 5 µm [73] or 20 µm [74], is equivalent to sapphire in
hardness and is diffusion-bonded to the base metal. The oxide layer
can be damaged by a striking action, but it serves as an excellent
surface for sliding contact.
Film Formation in Molten Salts
A layer of thick, protective, strongly cohesive oxide film can be
formed on zirconium by a patented process developed at TWC [75].
In this process, a zirconium subject is treated in fused sodium cyanide
containing 13% sodium carbonate, or in a eutectic mixture of sodium
and potassium chlorides with 5% sodium carbonate. Treatment is
carried out at temperature ranging from 600 to 800°C for up to 50 h.
A treatment time of several hours typically is used. The thickness of
oxide film formed in the fused salt bath ranges from 20 to 30 µm. This
film has greatly improved resistance to abrasion and galling over thick
oxide films grown by many other means.
2
Electrochemical Protection
Zirconium exhibits a passive-to-transpassive transition with increasing
potential in all mineral acids except HF [76]. The commonly observed
active node in many metal-acid systems is not observed for zirconium.
Consequently, zirconium
Page 230
performs well in most reducing environments. This can be attributed
to zirconium's ability to take oxygen form water to form stable passive
films. Most passive metals and alloys would need the presence of an
oxidizing agent, such as oxygen, in order to form protective oxide
films. In fact, zirconium is one of the best metals for handling
reducing media. Zirconium's corrosion problems can be controlled by
converting the corrosive condition to a more reducing condition by
various means.
Electronically, by impressing a potential that is arbitrarily 50100 mV
below its corrosion potential, zirconium becomes corrosion-resistant
in oxidizing chloride solutions [21]. Tables 14 and 15 demonstrate the
benefits of electrochemical protection in controlling pitting and SCC.
The general corrosion rates of unprotected zirconium in oxidizing
chloride solution may be low. However, the penetration rates are
much higher than the general corrosion rates and increase with
exposure time. Electrochemical protection eliminates this local attack.
Similarly, unprotected U bends of welded zirconium cracked in all but
one case shortly after exposure. On the other hand, protected U bends
resisted cracking for the 32-day test interval in all but one acid
concentration. Thus electrochemical protection offers a very definite
improvement to the corrosion properties of zirconium in oxidizing
chloride solutions.
This technique can also be used to combat the SCC of zirconium in
concentrated HNO3 [29]. Because of strong oxidizing, passivating
power of the acid, zirconium exhibits a wide passive range and a very
noble corrosion potential. Also, there is a large difference between the
corrosion potential and the critical potential to cause SCC. It is
desirable to control the potential of zirconium a few hundred
millivolts below the corrosion potential or at 740 mVNHE.
3
Others
Zirconium is susceptible to localized corrosion in oxidizing chloride
solutions. Strong oxidizing agents like ferric and cupric ions are the
common ones. There
Table 14 Corrosion Rate of Zirconium in 500 ppm Fe3+
Solution After 32 Days
Penetration rate
(mpy)
Environment
Acidity Temp. UnprotectedProtected
(°C)
10% HCl
3N
60
7.1
< 0.1
120
51
< 0.1
Spent acid (15% 5 N
65
36
< 0.1
Cl)
80
36
< 0.1
20% HCl
6N
60
3.6
< 0.1
107
59
< 0.1
Page 231
Table 15 Time to Failure of Welded Zirconium U
Bends in 500 ppm Fe3+ Solution After 32 Days
Time to failure
(days)
Environment
Acidity Temp. UnprotectedProtected
(°C)
10% HCl
3N
60
<0.1
NFa
120
<0.1
NF
Spent acid (15% 5 N
65
<0.3
NF
Cl)
20% HCl
6N
60
NF
NF
107
<0.1
NF
28% HCl
9N
60
2
NF
94
<0.1
NF
32% HCl
10 N
53
1
32
77
<0.1
20
37% HCl
12 N
30
0.3
NF
53
1
NF
aNF, no failure.
are less occurring stronger oxidants, such as ions of noble metals and
certain ions of radioisotopes (Pu4+, Am4+, etc.), that may cause
zirconium to pit in chloride solutions too. More than just inducing
pitting, ions of noble metals may be reduced to metal plating on
zirconium's surface in order to produce the galvanic effect.
Of the common ones, cupric ion seems to be more detrimental than
ferric ion in promoting the general corrosion and pitting of zirconium
in acidic chloride solutions. The test results of zirconium in NaCl +
CuCl2 given in Tables 16 and 17 demonstrate the effects of pH and
heat treatment. As indicated in these tables, the corrosion problems of
zirconium in cupric ioncontaining solutions can be controlled by
adjusting the pH to 6 or higher (Table 16) or by high-temperature heat
treatment (Table 17). In ferric ioncontaining solutions, it is sufficient
to adjust the pH to 3 or higher.
Tensile stresses provide a driving force for not just SCC but corrosion
in general to occur. Lowering residual stresses by a stress-relieving
treatment can be effective in controlling pitting as well.
Nevertheless, one of the critical factors in pitting is surface condition.
A metal with a homogeneous surface is always less vulnerable to
pitting and other forms of localized corrosion. Pickling is a common
way to homogenize metals' surface. Indeed, recent results show that
pickled zirconium may perform well in boiling 10% FeCl3 and even
in ClO2 [22,23]. It is well known that zirconium with a normal surface
finish is unsuitable for handling these solutions.
Page 232
Table 16 Test Results of Zirconium (As-Received
Condition) in Boiling 500 ppm Cu2+ Containing
NaCl Solutions After Seven 1-Day Runs
Average corrosion rate (mpy)
No.% NaClpHNonwelded couponsWelded coupons
1.
3.5 1
2.1a
23.6a
2.
25 1
1.6
21.7a
3.
3.5 4.8
0.38a
23.8a
4.
25 4.0
1.0
21.9a
5.
3.5 5.0
0.7
25.3a
6.
25 5.0
Nil
Nil
7.
3.5 6.0
Nil
Nil
8.
25 6.0
Nil
Nil
9.
3.5 7.5
Nil
Nil
10. 25 7.5
Nil
Nil
aPitting.
Stress relieving and/or surface conditioning could be used to expand
the usefulness of zirconium in oxidizing chloride solutions.
III
Typical Applications
A
Nuclear Industry
The development of water-cooled nuclear power reactors brought
about the use of zirconium and its alloys for uranium fuel cladding
and for structural components. As a result of these developments in
the nuclear industry, the cost of zirconium and its alloys decreased
considerably and became competitive with that of other corrosionresistant materials. Zirconium and its alloys have emerged as
engineering materials instead of laboratory curiosities.
Materials for fuel-cladding and structural components in nuclear
reactors are restricted because of the following crucial requirements:
1. Low-absorption cross-section for thermal neutrons
2. Excellent corrosion and oxidation resistance
3. Adequate strength and creep resistance
4. High resistance to radiation damage
5. Lack of reactions with the fuel material and fission products
Zirconium alloys, such as the zircaloys and Zr-2.5Nb, have been
developed to better meet these requirements. In water-cooled reactors,
zirconium alloys have found extensive use for fuel cladding and as
pressure tubes. In systems in which
Page 233
Table 17 Effects of Heat Treatment on the Corrosion of
Zirconium (Sand-Blasted and Pickled) in Boiling NaCl +
CuCl2 Solutions After Seven 1-Day Runs
Average corrosion rate
(mpy)
35% NaCl
25% NaCl
+ 500 ppm
+ 500 ppm
No. Metal
Condition
Cu2+
Cu2+
1. NonweldedAs
0.27a
0.98a
conditioned
2. Welded
As
0.45a
1.30a
conditioned
3. Welded
760°C/AC
<0.1a
0.23a
4. Welded
760°C/WQ
<0.1a
0.17a
5. Welded
871°C/AC
0.13
0.23a
6. Welded
871°C/WQ
0 13
0.17a
7. Welded
982°C/AC
0.20
0.27
8. Welded
982°C/WQ
0.20
0.27
aPitting.
AC, air-cooled; WQ, water-quenched.
the first of the listed requirements is of overriding important for
reason of neutron physics, the choice is virtually restricted to
zirconium or one of its alloys.
Zirconium and its alloys also find applications in other nuclear reactor
systems, such as gas-cooled or organic coolantcooled reactors.
B
Chemical Processing and Other Industries
For more than 30 years, many diverse applications have been
developed for zirconium and its alloys. Because of its reactivity,
zirconium is used in military ordnance, including percussion-primer
compositions, delay fuses, tracers, and pyrophoric shrapnel, in getters
for vacuum tubes and inert gas glove boxes, and as shredded foil in
flashbulbs for photography and excitation of lasers. Alloying
application of zirconium include zirconium-niobium superconductors,
titanium or niobium alloys for the aerospace industry, and
strengthening of copper alloys. Currently, there are active programs in
developing zirconium alloys for rechargeable batteries and hydrogen
storage.
Increasingly, zirconium and its alloys are being used as structural
materials in fabricating, e.g., columns, reactors, heat exchangers,
pumps, piping systems, valves, and agitators for the chemical
processing industry (CPI). This will be discussed to a great extent.
Modern processes emphasize low costs (raw materials, operation,
maintenance, etc.), improved efficiency, high quality, safety, and
environmental friendliness. These emphases often require that
chemical reactions take place at elevated temperatures and pressures
in the presence of a catalyst. Therefore,
Page 234
as the CPI modernizes its technologies, process environments may
become more corrosive and require more corrosion-resistant
equipment to cope with corrosion problems.
Corrosion problems continue to mean that process equipment has
suffered visible, excessive corrosion. Hence process equipment
becomes unsafe and requires repairing or replacement. Today
corrosion problems have a much broader meaning because of
increasing concerns for efficiency, quality, and environment. A
corrosion problem may exist even if process equipment does not show
any visible corrosion. For example, type 316L stainless steel
equipment corrodes at 50 µm/year, which is considered low. Still,
each day for every 1000 m2, about 700 g of iron, 175 g or chromium,
110 g of nickel, and 20 g of molybdenum is released into the product,
the process medium, or the environment. These low levels of
discharge can be undesirable and unacceptable. Such corrosion
products may damage the quality of fine chemicals, drugs, foods, and
fertilizers. They can poison certain catalysts to reduce process
efficiency and/or can cause certain chemicals, such as hydrogen
peroxide, to catalytically decompose. They may be harmful to the
environment.
Zirconium is compatible with many corrosives. The CPI recognized
the advantage of zirconium for solving corrosion problems from
zirconium's inception. Dr. Kroll predicted that zirconium would be
useful in HCl applications, since HCl is one of the most difficult acids
for common metals to handle. Many R&D programs were established
to evaluate zirconium for applications that involved harsh conditions
[51,52,55,7784]. Zirconium has proved itself to be one of the most
corrosion-resistant materials in the CPI.
Zirconium is more than just a highly corrosion-resistant metal. It is
lighter than iron-and nickel-based alloy. It has good thermal
conductivity and adequate strength. It appears to be nontoxic and
biocompatible. It is not just used to solve corrosion problems but to
cope with current social/economic challenges.
1
Urea
One of the earliest applications of zirconium was in the production of
urea. Certain zirconium vessels and heat exchangers have been in
service for more than 30 years and show no signs of corrosion. In
modern processes, urea is produced by the combination of NH3 with
CO2 to form ammonium carbamate followed by the dehydration of
the ammonium carbamate. In order to have a high conversion rate,
reactions have to take place at elevated temperatures and pressures.
The reactants, particularly the carbamate solution, are too corrosive
for stainless alloys unless oxygen is injected carefully for passivation.
Indeed, oxygen injection has been popular for some years in urea
plants to control the corrosion of stainless steel equipment. This
measure has certain drawbacks: lower plant efficiency and greater
safety concerns. The safety concerns are real. The explosion of
stainless steel equipment as given in Ref. 85 has
Page 235
occurred in different countries. Moreover, there is an increasing
concern for the presence of heavy metal ions in fertilizers.
Interest in using zirconium in urea production has been renewed in
recent years. For example, stainless steel tubes with zirconium lining
have been developed for carbamate decomposers and/or condensers
[86,87].
2
Acetic Acid
Zirconium is an important material for the production of acetic acid
by the reaction of methanol and carbon monoxide. This technology
has been studied for more than 40 years. It could be commercialized
in the 1970s only when the corrosion problems of structural materials
were managed. In this technology, the reaction must proceed at a high
temperature (³ 150°C) and a high pressure (3.36.6 MPa) in the
presence of a halide as the catalyst. The crude acid produced first is
separated from the catalyst and then dehydrated and purified in an
azeotropic distillation column. The final product is highly pure acetic
acid, allowing it to be used in food and pharmaceutical applications.
Here all factors inducing corrosion problems on stainless alloys are
encountered. These factors include
1. An intermediate acid concentration
2. An elevated temperature
3. The presence of highly corrosive methanol and iodides
The process equipment must be made of the most corrosion-resistant
materials, such as zirconium and its alloys. Zr 702 and Zr 705 are
often used to construct process equipment, such as reactors, columns,
heat exchangers, pumps, valves, piping systems, trays, and packings.
In recent years, zirconium has been replacing stainless alloys after
their failures in acetic acid service [8890]. Zirconium could be the
most cost-effective structural material when all issues, such as process
efficiency, product yield and quality, safety, maintenance and
replacement costs, toxicity, and environmental protection are
considered.
Titanium often shows very low corrosion rates in many acetic acid
media as indicated in Table 18. However, most acetic acid media are
too reducing for titanium to form high-quality oxide films on its
surface. Titanium, like tantalum, is susceptible to hydrogen
embrittlement under a reducing condition. There were cases of
hydrogen embrittlement of titanium equipment in acetic acid service
after a few years. Results of recent autoclave tests, as shown in Table
19, confirm this [91]. In general, zirconium is more suitable than other
reactive metals for handling reducing media.
3
Formic Acid
Formic acid production by the hydrolysis of methyl formate, such as
the Leonard-Kemira process, is another modern process that requires
zirconium. Formic
Page 236
Table 18 Resuls of 48-h
Tests in 50% Acetic Acid
with Cobalt Acetate and
Potassium Iodide as
Catalysts with Pressurized
Carbon Monoxide at 260°C
Metal
Corrosion rate
(µm/y)
Type 316
22,350
S.S.
Alloy 825
5,080
Alloy C
1,780
Alloy B
360
Nickel
5,840
Duriron
2,670
Titanium
<25
Zirconium
<25
Source: Ref. 11.
acid is more highly ionized and therefore more corrosive than acetic
acid. Stainless steels can be seriously attacked by intermediate
strengths of hot formic acid. Nickel-based alloys may corrode at high
rates in the presence of certain impurities (see Table 20) and under
heat transfer conditions. Titanium's performance in formic acid may
be affected by minor factors, such as aeration.
In the LeonardKemira process, CO gas contacts methanol in the
presence of a catalyst to form methyl formate in a reactor. The methyl
formate is hydrolyzed in the presence of a catalyst to yield formic acid
and methanol, which are separated by distillation. The methanol is
recycled to the first stage of the process. Factors such as elevated
temperatures and pressure and presence of water
Table 19 Hydrogen
Absorption in 95% Acetic
Acid Plus 1000 ppm
Hydrobromic Acid at
210°C
Hydrogen
concentration
(ppm)
Test
Zr
Ti
duration
(days)
0
8
15
30
7
29
60
8
46
90
9
73
Page 237
Table 20 Corrosion of Metals in Boiling Formic Acid Solutions for 8 days
Corrosion rate (µm/yr)
Zr
316 S.S.
Alloy B-2
Alloy C276
Formic acid
(%)
Impurity NWa Wa NW W
NW W
NW W
50
None
<2.5 <2.5 600625
22.5 25
50 45
50
1% Fe pdr <2.5 <2.5 650700
1825 2150 160 120
50
1% Cu2+ <2.5 <2.5
3838
>5000b>5000b178 128
50
5%
<2.5 5.0 22252650 28
43
46 56
H2SO4
50
5% HCl <2.5 2.5 >5000>5000 46
50
27183200
50
5% HI
<2.5 <2.5 6101118 33
38
230 250
70
None
2.5 <2.5 483483
69
69
58 48
70
1% Fe pdr <2.5 <2.5 560530
710
710
205 205
70
1% Cu2+ <2.5 <2.5
1013
WG WG 155 100
b,c
b,c
70
5%
<2.5 <2.5 37103710 43
46
58 50
H2SO4
70
5% HCl 2.5 2.5 >5000>5000 38
43
940 965
70
5% HI
<2.5 <2.5 4370>5000d1500 1900 32753125d
98
None
<2.5 <2.5 125130
43
43
23 20
98
1% Fe pdr <2.5 <2.5
4550
38
38
30 28
96
1% Cu2+ <2.5 <2.5
55
3175b 3860 53 36
93
5%
<2.5 <2.5 45705000 15
15
23 20
H2SO4
85
5% HCl <2.5 <2.5>5000>5000 18
18
355 380
90
5% HI
<2.5 <2.5 635790
160
185
585 560d
aNW, nonwelded; W, welded.
bCoupons were plated with copper.
cW G, weight gain.
dWeld metals were preferentially attacked.
and catalyst make common materials, including glass lining, resin,
and plastic coatings, and stainless alloys inadequate as structural
materials for this process [92]. Zirconium proves to be the most
economical structural material for use in the main equipment for this
process.
4
Sulfuric AcidContaining Processes
Zirconium was used in H2O2 manufacturing by the electrolysis of
acid sulfates. This production process is very corrosive. At one time,
graphite equipment was standard for this process. The FMC plant in
Vancouver, Washington found that zirconium was superior to
graphite. FMC used zirconium equipment to produce up to 90%
H2O2. The average maintenance-free life of the heat exchanger was
10 years. Graphite exchangers had failed after 1218 months of service.
The
Page 238
graphite equipment failure was attributed to the leaching of the binder
from the graphite by the 35% H2SO4 feed, which created a porous
condition and caused the ultimate failure.
Hydrogen peroxide is becoming a preferred oxidant because of its
environmentally safe nature. It has been made inroads to many
industrial applications, and its consumption is increasing at a rapid
pace. It is, however, not a stable chemical. It can be catalytically
decomposed by many heavy metal ions. The decomposition reaction
is wasteful and may create a condition for fire or explosion. Certain
peroxide solutions are corrosive too. Zirconium is one of very few
metals that is highly compatible with a broad range of peroxide
solutions. It is corrosion-resistant. It does not produce active ions to
catalytically decompose peroxide. It is attractive for many peroxide
applications, such as production and handling of fine H2O2, pulp
bleaching, waste treatment, etc.
The experience of zirconium in peroxide production led FMC to
replace the graphite heat exchangers with zirconium shell and tube
exchangers used in the manufacturing of acrylic films and fibers. In
this application, the H2SO4 concentration was as high as 60% at
150°C.
Another major application in H2SO4 concerns the manufacture of
methyl methacrylate and methacrylic acid. The system at the Rohm
and Haas plant in Deer Park, Texas includes pressure vessels,
columns, heat exchangers, piping systems, pumps, and valves made
from zirconium. A zirconium unit built more than 20 years ago is still
in service.
Zirconium is also widely used for column internals and reboilers in
the manufacture of butyl alcohol. The operating conditions are 6065%
H2SO4 at temperatures to boiling and slightly above. Zirconium may
corrode under upset conditions of elevated concentrations and when
impurities such as Fe3+ are present. Zirconium has been used in
H2SO4 recovery and recycle systems in which fluorides are not
present and the acid concentration does not exceed 65%. A major
application for zirconium is in iron and steel pickling, using hot 540%
H2SO4.
Rayon is a man-made textile fiber. Most of today's rayon is made by
the viscose process. Equipment made of graphite was popular for use
in the H2SO4- affected areas of this process. It is vulnerable to
breakdowns. Avtex Fiber began experimenting with zirconium
equipment in 1970. Zirconium's excellent performance prompted
Avtex to convert more pieces of equipment to zirconium, which
included 10 acid evaporators, 14 shell and tube heat exchangers, and
12 bayonet heat exchangers [93,94]. In addition to dramatically
reducing maintenance costs and downtime, the zirconium equipment
improved operating efficiency and lowered overall energy costs.
Hydroxyacetic acid (HAA) is also known as glycolic acid. It can be
produced in a synthetic process other than extracted from natural
sources. Under high pressure (3090 MPa) and temperatures
(160200°C), formaldehyde reacts
Page 239
with carbon monoxide and water in the presence of an acidic catalyst,
such as sulfuric acid, to form HAA. DuPont could not even rely on a
silver lining for reliable service in this process [95]. Silver showed
poor erosion resistance in the piping system. There were cases of
blowouts in the piping due to failure of the lining. By the mid-1980s,
zirconium lining was evaluated when other materials, such as glass,
ceramic, stainless alloys, and titanium, were found unsuitable.
Zirconium is well known for its corrosion resistance in weak sulfuric
acid at temperatures up to and above 260°C. The excellent corrosion
resistance of zirconium in HAA at 205°C was confirmed at TWC. An
8-month field test at DuPont indicated that a zirconium tube would
not corrode in the most severe service section of the process.
Zirconium's excellent resistance to erosion is also apparent.
Consequently, DuPont replaced silver lining with zirconium lining in
piping sections more than 5 years ago. it was estimated that zirconium
lining would last at least three times longer than silver lining.
5
Halide-Containing Processes
Zirconium has many applications in HCl, such as the production of
concentrated HCl and polymers. Zirconium heat exchangers, pumps,
and agitators have been used for more than 15 years in an azo dye
coupling reaction. In addition to being very corrosion-resistant in this
medium, zirconium does not plate out undesirable salts that would
change the color and stability of the dyes.
Lactic acid is commercially produced either by fermentation or by
synthesis. The synthetic process is based on lactonitrile which is
prepared by reacting acetaldehyde with hydrogen cyanide at up to
200°C. Lactonitrile is then hydrolyzed in the presence of HCl to yield
lactic acid. In the HCl-affected areas, suitable materials are limited.
Glass-lined materials are prone to breakdowns. Stainless alloys
corrode and introduce toxic materials to the process stream. Titanium
and its alloys are susceptible to crevice corrosion in hot chloride
solutions. Zirconium is virtually ideal for this process. Since lactic
acid is produced as a fine chemical, contamination has to be prevented
in all areas. Oxidizing HCl conditions resulting from the presence of
ferric or cupric ions are avoided. Moreover, zirconium is highly
resistant to crevice corrosion in chloride solutions. Since the 1970s,
zirconium equipment has provided excellent service in lactic acid
production.
Other applications in HCl include the breaking down of cellulose in
the food industry and the polymerization of ethylene chloride, which
is carried out in HCl and chlorinated solvents.
Zirconium and its alloys have been identified to offer the best
prospects from a cost standpoint as materials for an HI decomposer in
hydrogen production. They resist attack by HI media (gas or liquid)
from room temperature to 300°C. Most stainless alloys have adequate
corrosion resistance only at low temperatures.
Page 240
6
Nitric AcidContaining Processes
There is an increasing interest in the use of zirconium for HNO3
service. For example, because of the high degree of concern over
safety, zirconium is chosen as the major structural material for critical
equipment used to reprocess spent nuclear fuels.
In most HNO3 service, stainless steel has been the workhorse for
decades. The excellent compatibility between zirconium and HNO3
was thought not to be needed. This situation changed when nitric acid
producers started to modernize their technology in the late 1970s.
Conventionally, HNO3 is manufactured by oxidation of ammonia
with air over platinum catalysts. The resulting nitric oxide is further
oxidized into nitrogen dioxide and then absorbed in water to form
HNO3. Acid of up to 65% concentration is produced by this process.
Higher concentration acid is produced by distilling the dilute acid with
a dehydrating agent.
Before the 1970s, dual-pressure processes were the dominant means
of HNO3 production. A typical dual-pressure process operates the
converter at about 500 MPa and the absorber at about 1100 MPa. In
the late 1970s. Weatherly Inc. introduced a high-monopressure
process which operates at 13001500 MPa. The advantages of this new
process are
1. Greater productivity due to higher operating pressure
2. Smaller equipment resulting in a lower capital cost
3. Higher energy recovery capabilities
The first try of this new process came in 1979 when Mississippi
Chemical in Yazoo City retrofit their existing plant with a new
compressor system to increase pressure for greater productivity and
energy efficiency. It was at this point that severe corrosion problems
were discovered.
Prior to the upgrade, the cooler condenser was constructed of type
304L stainless steel tubesheets and type 329 stainless steel tubes.
Under the previous operating conditions, the cooler condenser had
experienced some corrosion, which was managed by plugging tubes
and replacing the unit every 34 years. Shortly after the upgrade, with
an operating temperature and pressure of 200°C and 1035 MPa, 10%
of the type 329 stainless steel tubes were found to be leaking. This
condenser was replaced with a unit using type 310L stainless steel,
which had to be replaced after 13 months of operation. The original
condenser with new tubes of improved grade 329 stainless steel
replaced the 310L unit. Mississippi Chemical began looking for
alternatives.
In an attempt to find a solution to this problem, TWC conducted
autoclave tests on many newer types of stainless steels and zirconium
in solutions up to 204°C and at concentrations up to 65%. Clearly,
zirconium was the only suitable material for the monopressure
process. Corrosion rates of zirconium coupons
Page 241
were consistently below 1 mpy. The next step was to test zirconium
tubes in service. Several tubes were installed in a rebuilt stainless steel
condenser. They were destructively examined after 13 months. There
were no signs of corrosion. Zirconium tubes were then placed in
another condenser for a year. Once again, there were no signs of
corrosion.
Consequently, Mississippi Chemical replaced its stainless steel cooler
condenser with one constructed from zirconium tubes and
zirconium/304L stainless steel explosion-bonded tubesheets. This unit
contains more than 18 km of zirconium tubing. In service since 1984,
the zirconium unit has already outperformed the stainless steel
predecessors. Thereafter, several zirconium cooler condensers have
been built for other HNO3 producers.
Monopressure plants are not the only ones to use zirconium as a
solution to corrosion problems. Certain plants use a distillation
process to increase concentration with the acid passing through a
reboiler and entering a distillation column to drive off water and
concentrate the acid. In 1982, Union Chemicals replaced the bottom
portion of each of two distillation columns and the tube bundles of
each of two reboilers. The lower parts of the columns had been
constructed originally with type 304L stainless steel which
experienced corrosion problems. Titanium was tried but also failed.
While glass-lined steel did not have the corrosion problems
experienced by type 304L stainless steel and titanium, the
maintenance costs were found to be unacceptable. Zirconium provides
significantly improved corrosion resistance without adding
maintenance costs. Zirconium also solved the corrosion problem in
the reboilers. Prior to the installation of zirconium tube bundles, both
304L and titanium tube bundles had failed with less than 18 months of
operation.
With proper design and fabrication, zirconium's susceptibility to SCC
can be suppressed in highly concentrated HNO3. For example, an
Israeli chemical plant uses zirconium tubes in a U-tube cooler that
processes bleached HNO3 at concentrations between 98.5% and 99%.
The unit cools the acid from 7075°C to 3540°C. Previously, U-tube
coolers were made from aluminum, which failed in 212 weeks. The
zirconium unit has been in service for more than 2 years, operating 24
hours a day, 6 days a week.
Adipic acid is produced primarily for use in the manufacture of nylon6,6. Major commercial routes to adipic involve the oxidation of
cyclohexanol or oxidized cyclohexane with nitric acid at elevated
temperatures. Because of the excellent performance of zirconium in
the production of HNO3, it started to find its way into adipic acid
plants.
7
Others
A unique application for zirconium is in processes that cycle between
HCl or H2SO4 and alkaline solutions. One company replaced a leadand-bricklined carbon steel reactor vessel with zirconium because the
reaction alternated be-
Page 242
tween hot H2SO4 and caustic. The vessel has been in use for several
years with no corrosion problems.
A zirconium distillation column proved to be suitable and economical
in a chlorinated hydrocarbon environment at more than 150°C. The
column was built as a replacement for an old brick-lined column.
Corrosion of the old column necessitated frequent renovations, with
resultant high maintenance costs and plant downtime.
In the zirconium extraction process, several corrosive chemicals are
used, including methyl isobutyl ketone, HCl, ammonium thiocyanate,
H2SO4, and zirconyl chloride. Zirconium has been found to withstand
the rigors of this manufacturing process. Six examples are given, as
follows:
1. The pumps used to transport the process chemicals are primary
candidates for failure. Various materials, including stainless alloys,
plastics, and high-silicon cast iron, had been tried but each had some
limitation in this severe service. All wetted pump parts were
converted to zirconium. The oldest one has been operating for more
than 5 years and shows no sign of corrosion.
2. For heat exchanger service in the same process chemicals,
zirconium tubes replaced impervious graphite tubes that were prone to
failure. The oldest zirconium heat exchanger had been in service for
more than 7 years and shows no signs of corrosion.
3. Steam stripper columns were converted from a furan resin to
zirconium. At the 105°C operating temperature of the columns, the
plastic would embrittle and crack. Failure of these columns often
resulted in expensive spills. By contrast, zirconium columns require
no maintenance due to material failure.
4. Zirconium is used in an electrostatic precipitator installed to scrub
corrosive gases emitted from rotary kilns. The emitting electrodes and
all fixtures exposed to the ammonium sulfate and chloride gas process
stream were constructed from zirconium. The emitting electrodes are
charged with 45,000 V.
5. Several parts of a crude chlorination scrubber were converted from
plastic to zirconium.
The plastic parts were replaced approximately every 6 months. The
zirconium replacement is expected to last about 25 years.
6. Water-type vacuum pumps were used to scrub HCl-containing offgases.
Zirconium appears to be nontoxic and biocompatible. It has very low
corrosion rates in many media. It is ideal for making equipment used
in food processing, in the manufacture of fine chemicals, and in
pharmaceutical preparations. Zirconium and its alloys are suitable for
certain implant applications as well.
Other applications for zirconium include heat exchangers exposed to
hot
Page 243
seawater, cladding for high-strength or light-weight alloys, and
reclaiming valuable chemicals for wastes.
Finally, zirconium is an excellent material for laboratory equipment,
such as crucibles and autoclaves. Zirconium crucibles have replaced
platinum crucibles for handling certain molten salts. Properly oxidized
zirconium and zircaloys can be used as insulating washers and
measuring devices exposed to high-temperature corrosives.
IV
Fabrication Practices
Zirconium and its alloys are ductile and workable. They can be
fabricated using standard shop equipment with a few modifications
and special techniques. For example, zirconium has a low modulus of
elasticity. Therefore, zirconium will deform considerably under load
and then spring back. This characteristic must be figured in when
fabricating at low temperatures. Springback is not a concern in hot
forming since lower forming loads may be used.
Of primary concern is the tendency of zirconium to react with gases in
air at elevated temperatures and to gall and seize under sliding contact
with other metals. Fabricability depends on impurities and alloying
elements. For example, the lower the oxygen content the easier the
metal is to form. On the other hand, alloying elements, such as
niobium, may improve zirconium's fabricability. Proper fabrication
practices are described in Reference 96.
A
Forming
In forming operations, a thin oxide layer on zirconium acts as a
lubricant against galling. If additional lubricant is required, use any oil
or grease that does not contain halogen or sulfur, except that
molydisulfide may be used. It is important to remove all traces of oil
and dirt before heat treatment.
Zirconium can be easily bent and formed using standard shop
equipment. Zr 702 sheet and strip can be bent on conventional press
brake or roll-forming equipment to a 5T bend radius at room
temperature and to 3T at about 200°C. Zr 705 can be bent to 3T bend
radius at room temperature and to 1.5T at about 200°C.
B
Punching
Pure zirconium and zirconium alloys act differently in punching
operations. Pure zirconium requires high-pressure plate pressures or
excessive side flow may occur. Very close punch and die tolerances
(12% of the metal thickness) yields the best results. Larger punch and
die clearances would be acceptable when zirconium alloys are
punched.
Page 244
C
Tube Bending
The same techniques and equipment used to cold-form stainless steel
are also used on zirconium tube. Provisions for the springback
behavior should be made for any bending operation. A minimum bend
radius of approximately three times the outside diameter (OD)
dimension is recommended for cold bending. Hot bending at
200425°C or the use of special bending techniques are required for
small-radius bends. To prevent buckling and wall thinning, both the
inside and outside surfaces of the bend area must be in tension during
the bending operation.
D
Machining
Zirconium and its alloys can be machined by conventional methods
according to the following three rules:
1. Slow speeds
2. Heavy feeds
3. A flood coolant system using a water-soluble oil lubricant
Zirconium exhibits a marked tendency to gall and work-harden. This
indicates that higher than normal clearance angles on tools are needed
to penetrate the previously work-hardened surface and cut a clean
coarse chip. Satisfactory results can be obtained with both cemented
carbide and high-speed tools. The carbide tools usually give better
finishes and productivity. Polishing or honing the cutting edges will
prolong the tool life. Zirconium alloys machine to an excellent finish,
requiring relatively light horsepower compared with alloy steel. The
tool forces are relatively low.
E
Welding
Common to all reactive metals, welding zirconium by most methods
would require proper shielding of the weld puddle and hot bead from
air. The shielding gas must be highly pure argon, helium, or a mixture
of these two. Absorption of elements like oxygen and nitrogen by the
metal will results in the weld embrittlement. Cleanliness is another
important requirement in preventing weld embrittlement.
Zirconium is most commonly welded by the gas tungsten arc welding
(GTAW) technique. Other welding methods include metal arc gas
welding (MAGW), plasma arc welding, electron beam welding, spot
welding, friction welding, and resistance welding.
There are several convenient aspects in zirconium welding. Zirconium
has a low coefficient of thermal expansion, which contributes to a low
distortion during welding. The low modulus of elasticity will induce
low residual stresses in a finished weld. Because no fluxes are needed
in welding, flux entrapment is
Page 245
not possible. Preweld heat treatment is not required. Postweld heat
treatment, such as stress relieving, should be considered only when
there is a concern for SCC. However, stress relieving of Zr 705 welds
for preventing delayed hydride cracking is required. It should be noted
that as Zr 705 is a heat-treatable alloy, all welding variables and
postweld heat treatments need to be carefully reviewed to ensure weld
ductility.
1
Equipment
The most common techniques used to weld zirconium are the inert gas
welding methods of GTAW and MAGW. This equipment can be set
up and used in the manual or automatic welding of zirconium. Either
AC or DC current can be used in an arc welding. However, DC
current is preferred for welding with a consumable zirconium
electrode because it results in a more stable arc. A high-frequency arc
is desirable at the starting stage for reducing the possibility of
contamination by a nonconsumable electrode.
2
Weld Preparation
In preparation for welding zirconium, the edges to be joined should be
draw-filed or wire-brushed with a stainless steel brush prior to
welding. This should be followed by a thorough cleaning with alcohol
or acetone to ensure a clean area for welding. Chlorinated solvents
should be avoided in preparing the area for welding.
3
Shielding
Zirconium is subject to severe embrittelment by relatively minute
amounts of impurities, especially nitrogen, oxygen, carbon, and
hydrogen. Zirconium is highly reactive at welding temperatures.
Because of this high activity, zirconium must be welded under an inert
environment, such as argon or helium, or both.
Arc welding in an inert blanket with use of either a tungsten electrode
or a consumable electrode of zirconium gives the best results. The
weld puddle, the bead just behind the weld puddle, and the backside
of the weld must all be protected from the atmosphere in some
manner. The weld puddle and the bead just behind the weld puddle
can be protected by a secondary shield, such as a trailing shield. This
trailing shield is a device attached to the torch head that provides
protection (inert gas covering) to the weld area while it cools.
Other protection may include side shields that pass protective gas over
the area being welded or by covering the weld area with a plastic or
cardboard enclosure purged with argon. The size and configuration of
the work being welded determine the means of shielding the weld.
When the backside of the weld is inside an enclosed area, the entire
enclosure can be sealed and purged with an inert gas.
Circumferential and longitudinal welds are commonly made using a
fixture positioned which both chills and shields. Rapid cooling of the
weld and heat-
Page 246
affected zone is promoted by the use of copper backup plates and
hold-down clamp. Hold-down bars, backing blocks, and trailing
shields are necessary to protect the weld and weld zone from
contamination until the metal has cooled to below 250300°C.
Shielding lines should permit uniform flow of the inert gas, and care
should be taken to prevent any pockets in the system that might bleed
contamination into the shielding gas during welding. The shielding
system should be thoroughly cleaned and purged before use. Welding
grade shielding gases should be specified.
In consumable electrode welding, the torch should have provision for
purging at the point of the wire entry to prevent drag of air into the
torch with wire.
When desired, a flow purge or vacuum purge chamber may be used.
The same precautions against contamination apply in the use of these
chambers.
4
Welding Process
Welding techniques similar to those for stainless steel and titanium are
applied to weld zirconium. After purging the torch, the weld should be
made using a smooth, continuous motion along the weld joint. When
welding with zirconium wire as a filler, the wire should be fed
continuously into the puddle to prevent contamination from the air.
After the welding has stopped, the contaminated end of the weld wire
should either have the black oxide film removed or be cut off before
beginning the next weld. This measure prevents the contamination of
the next weld from the wire end. Sample welds should be made before
production welding to check the process. The sample welds can be
checked easily by bend testing. The welds should withstand a 90°
bend, without cracking, around a rod with a diameter of six times the
thickness of the weld sample.
5
Weld Inspection
A clean, bright weld is obtained through the use of a proper shielding
system. Light discoloration of a weld is not necessarily an indication
of its acceptability. White deposits or a black color in the weld area is
not acceptable. A bend test is probably the best test to determine weld
quality.
6
Welding of Dissimilar Metals
Zirconium cannot be welded directly to most other structural metals
by techniques such as GTAW. The exceptions include titanium,
vanadium, and niobium. Hence it is necessary to fabricate allzirconium equipment or line other materials with zirconium. In the
latter, the liner can be fastened to the shell by mechanical means.
Proper design of the vessel provides the combination of excellent
corrosion resistance and heat transfer in the use of lined systems.
If the equipment is going to be subject to vacuum, high temperature,
or high internal pressures that require 25.4-mm or more wall
thickness, clad materials should be considered. Thin zirconium plates
can be clad to thicker, lower cost
Page 247
materials, such as carbon steel and type 304 stainless steel. Applicable
cladding techniques include explosive bonding and resistance
cladding. Also, zirconium can be clad to titanium by roll-bond lining.
The metallurgically bonded cladding will utilize the corrosion
resistance and excellent heat transfer characteristics of zirconium,
with the substrate contributing the mechanical strength needed for the
equipment.
7
Welding Clad Plate
Butt welding of clad materials can be accomplished by two methods.
A single V or U joint is used for welding the substrate metal by any
conventional welding process. The joint is then back-chipped to clean
metal and back-welded with compatible filler wire. The zirconium
cladding should be machined back from the weld joint to prevent
dilution from the substrate metal. The zirconium clad can then be
joined with a zirconium cover strip.
The other technique is to overlay the steel weld with silver or
vanadium and then weld the zirconium cladding with zirconium filler
wire. A zirconium coverstrip should then be welded over the previous
weld.
V
Zirconium Products
The variety of mill products available is comparable with those of
stainless steels. This includes forging, sheet, plate, billet, wire, foil,
tubing, pipe, fastener, and the like. Also, zirconium can be cast by
investment or rammed graphite mold casting. These materials can be
fabricated into tanks, columns, reactors, heat exchangers, agitators,
exhausters, packings, valves, pumps, pipings, and nozzles, to name a
few.
VI
Health and Safety
Zirconium is nontoxic and consequently does not require serious
limitations on its use because of health hazards. In fact, certain
antiperspirant products contain zirconium compounds, such as
aluminum zirconium tetrachlorohydrez glycine.
The oxidation of zirconium is an exothermic reaction. Large pieces of
sheet, plate, bar, tube, and ingot can be heated to high temperatures
without an excessive oxidation/mass ratio. But small pieces, such as
machine chips and turnings, are easily ignited and burn at extremely
high temperatures. Like most materials, zirconium fines with a
dimension smaller than 0.05 mm are flammable. Large accumulations
of chips and other finely divided materials should be avoided. When
storing the chips and turnings, care should be taken to place the
material in a nonflammable container and remove it to an isolated
storage area.
One effective storage method is to keep the material covered with
water in the containers and in turn use oil on the water to keep it from
evaporating. If a fire accidentally starts in zirconium, do not attempt
to put it out with water or
Page 248
ordinary fire extinguishers. Use dry sand salt or commercially
available Metal-X power (Ansul manufacturing, Marinette,
Wisconsin). Large quantities of water can be used to control and
extinguish fires in other flammables in the vicinity of a zirconium fire.
When the corrosion resistance of zirconium is exceeded, a corrosion
product containing fine metal particles may be generated on its
surface in certain very aggressive, stagnant media, such as > 70%
H2SO4 with the presence of some ferric ions. This type of corrosion
product may be very pyrophoric. This film can be rendered
nonpyrophoric by simple oxidation treatments with hot steam after
water rinsing [48].
VII
Concluding Remarks
Zirconium possesses a set of unique properties, which include the
following:
1. Zirconium is highly transparent to thermal neutrons.
2. Zirconium is one of very few metals that resists attack by strong
acids and caustics, as well as many salt solutions and molten salts.
3. Zirconium appears to be nontoxic and biocompatible.
4. Zirconium has adequate strength.
5. Compared to stainless alloys, zirconium is lower in density, higher
in thermal conductivity, and lower in coefficient of thermal
expansion.
6. Zirconium can be fabricated into almost anything by conventional
methods.
7. Zirconium is reactive when its surface area-to-mast ratio is high.
As a result, many nuclear and industrial applications have been
developed for zirconium and its alloys. These applications include
fuel cladding and pressure tubes for nuclear reactors, process
equipment for the CPI, superconducting materials, battery alloys,
hydrogen storage alloys, ordnance applications, implant materials, and
consumer goods.
The manufacturing industry is emphasizing quality, efficiency, and
environmental compatibility. Zirconium is well positioned to meet
these needs. Interest in zirconium and its chemicals is on the rise.
However, there is still a persistent perception that zirconium is exotic
and costly. Actually, zirconium is plentiful. In the earth's crust,
zirconium is more abundant than many common elements, such as
nickel, copper, chromium, zinc, lead, and cobalt. The prices of
zirconium and its alloys have been relatively stable for many years.
They are very competitive with other high-performance materials.
Life cycle costs of zirconium equipment can be particularly attractive.
There is much room for zirconium to grow in the coming years.
Page 249
References
1. Zirconium: Its Production and Properties, USBM Bull. 561, 1956.
2. H. G. Rickover, L. D. Geiger, and B. Lustman, History of the
Development of Zirconium Alloys for Use in Nuclear Reactors, U.S.
Energy Research and Development Administration, Division of Naval
Reactors, TID-26740, U.S. GPO, Washington, DC, 1975.
3. J. H. Schemel, ASTM Manual on Zirconium and Hafnium, ASTM
STP 639, ASTM, Philadelphia, 1977.
4. S. Kass, The Development of the Zircaloys in Corrosion of
Zirconium Alloys, STP 3687, ASTM, Philadelphia, 1964.
5. D. E. Thomas, Corrosion in water and steam, Metallurgy of
Zirconium (B. Lustman and F. Kerze, Jr. eds.), McGraw-Hill, New
York, 1995.
6. S. B. Dalgaard, Corrosion and hydriding behavior of some Zr2.5Nb alloys in water, steam and various gases at high temperature,
Proceedings of the Conference on Corrosion Reactor Materials, Vol.
2, International Atomic Energy Association, 1962.
7. J. E. LeSurf, The corrosion behavior of 2.5Nb zirconium alloys,
Symposium on Applications, Related Phenomena in Zirconium and Its
Alloys, STP 458, ASTM, 1969.
8. H. H. Klepfer. J. Nucl. Mater, 9: 65 (1963).
9. T. L. Yau, Proceedings of Fourth Asian-Pacific Corrosion Control
Conference, Vol. 2, 1985, p. 136.
10. T. Smith and G. R. Hill, J. Electrochem. Soc., 105: 117, (1958).
11. M. Pourbaix, Atlas of Electrochemical Equilibria in Aqueous
Solution, NACE, Houston, 1974, p. 579.
12. T. L. Yau, Performance of zirconium and columbium in simulated
FGD scrubber solutions, Corrosion in Desulfurization Systems,
NACE, Houston, 1984.
13. T. L. Yau, Chemical impurities create new challenges in material
selection, Outlook, Teledyne Wah Chang, 9 (4): 4 (1988).
14. J. H. Schemel, Corrosion Resistance of Metals and Alloys (F. L.
LaQue and H. R. Copson, eds.), Reinhold, New York, 1963, p. 666.
15. C. R. Bishop, Corrosion, 19 : 308t (1963).
16. W. E. Kuhn, Corrosion, 15: 103t (1959); 16: 136t (1960).
17. D. R. Knittel and R. T. Webster, Corrosion Resistance of
Zirconium and Zirconium Alloys in Inorganic Acids and Alkalies, STP
728, ASTM, 1981, p. 191.
18. V. V. Andreeva and A. I. Glukhova, J. Appl. Chem., 11: 390,
(1961); 12: 457 (1962).
19. M. N. Fokin, R. L. Barn, and M. M. Kurtepox, Corrosion and
Metal Protection (J. L. Rosenfeld, ed.), Indian National Scientific
Documentation Letta, 1975, p. 113.
20. L. L. Migai, et al., Nauchyne Trudy Nauchno-Issledovatel' shii,
Proekt. IN-T Red-komet, Prom-Sti, 106: 123 (1981).
21. T. L. Yau and M. A. Maguire, Corrosion, 40: 289 (1984); 41: 397
(1985).
22. TR. L. Yau, J. Fahey, And D. R. Holmes, Performance of
zirconium in bleaching solutions, Proceedings of 1993 TAPPI
Engineering Conference, Orlando, September 2023, 1993, TAPPI,
Atlanta, Georgia (1993).
23. Fahey, Holmes, and T. L. Yau, Control and monitoring of the
localized corrosion of zirconium in acidic chloride solutions,
Corrosion 95, March 2631, 1995, Orlando, Paper No. 246, NACE,
Houston.
Page 250
24. T. L. Yau, Industrial Applications of Titanium and Zirconium:
Fourth Volume, STP 917, ASTM, 1986, p. 57.
25. G. Jangg, R. T. Webster, and M. Simon, Werkst, Korros, 29: 16
(1978).
26. M. Maraghini, G. G. Adams, and P. V. J. Russelberghe, J.
Electrochem. Soc., 101: 400 (1954).
27. T. L. Yau, Corrosion, 39: 167 (1983).
28. J. A. Beavers, J. C. Griess, and W. K. Boyd, Corrosion, 36: 292
(1981).
29. T. L. Yau, Corrosion, 44: 745 (1988).
30. B. Cox, J. Electrochem. Soc., 117 (5): 654 (1970).
31. T. Smith, J. Electrochem. Soc., 107 (2) : 82 (1960).
32. Zircadyne Corrosion Properties, Teledyne Wah Chang, 1986, p.
9.
33. Zirconium heat exchanger: a new direction for pickling
operations, Iron Steel Eng. 52: 55 (September 1975).
34. Zirconium coils make inroads into pickling tanks, Iron Age, 216:
39 (September 1, 1975).
35. B. Adams and D. Byrnes, Chem. Process. 26 (August 1981).
36. T. L. Yau, Zirconium in a sulfuric acid pickling application,
Corrosion 88, Paper No. 319, NACE, Houston, 1988.
37. T. L. Yau, Industrial Applications of Titanium and Zirconium:
Third Conference, STP 830, ASTM, 1984, p. 203.
38. R. E. Smallwood, Mater. Perform. 16 (2): 27 (1977).
39. T. Gunter, Werkst. Korros, 30: 308 (1979).
40. B. F. Frechem, J. G. Morrison, and R. T. Webster, in Industrial
Applications of Titanium and Zirconium, STP 728, ASTM, 1981, p.
85.
41. T. L. Yau and R. T. Webster, Corrosion, 39: 218 (1983).
42. B. J. Fitzgerald, T. L. Yau, and R. T. Webster, Stress corrosion
cracking of zirconium and its control in sulfuric acid, Corrosion 92,
Paper no. 154, NACE, Houston, 1992.
43. B. J. Fitzgerald and T. L. Yau, The mechanism and control of
stress corrosion cracking of zirconium in sulfuric acid, 12th
International Corrosion Congress, Houston, September 1924, 1993.
44. O. T. Fasullo, Sulfuric Acid, McGraw-Hill, New York, 1965, p.
290.
45. T. L. Yau and R. T. Webster, in Laboratory Corrosion Tests and
Standards, STP 866, G. S. Haynes and R. Baboian, Eds., ASTM,
1985, p. 36.
46. B. J. Fitzgerald, R. G. Webber, B. S. Frechem, K. F. Briegel, and
T. L. Yau, Factors to consider for using zirconium in sulfuric acid
services, Corrosion 95, March 2631, 1995, Orlando, Paper no. 247,
NACE, Houston.
47. J. H. Schemel, ASTM Manual on Zirconium and Hafnium, ASTM,
Philadelphia, 1977.
48. T. L. Yau, in Industrial Applications of Titanium and Zirconium:
Third Conference, STP 830, R. T. Webster and C. S. Young, Eds.,
ASTM, 1984, p. 124.
49. R. T. Webster and T. L. Yau, Zirconium in Process Industries
Corrosion (B. J. Moniz and W. I. Pollock, ed.), NACE, 1986, p. 540.
50. S. M. Shetlon, Zirconium: Its Production and Properties, U.S.
Bureau of Mines Bull. 561, 1956.
51. L. B. Golden, I. R. Lane, Jr., and W. L. Acherman, Ind. Eng.
Chem., 44: 1930 (1952); 45: 782 (1953).
Page 251
52. P. J. Gegner and W. L. Wilson, Corrosion, 15: 341t (1959).
53. C. M. Graighead, L. A. Smith, and R. I. Jaffee, Screening Tests on
Metals and Alloys in Contact with Sodium Hydroxide at 1000 and
1500 °F, U. S. Energy Commission Report BMI-706, Battelle
Memorial Institute, November 1951.
54. T. L. Yau, Zirconium for the changing pulp and paper industry,
Mater. Perform. 32: 65 (June 1993).
55. L. B. Golden, Zirconium and Zirconium Alloys, ASM, Cleveland,
1953.
56. H. Bloom, PhD thesis, University of London, 1947.
57. W. B. Blumenthal, Zirconium Compounds, National Lead Co.,
September 1969.
58. T. L. Yau, Corrosion, 38: 615 (1982).
59. T. L. Yau, Zirconium versus corrosive species in geothermal
fluids, Corrosion 84, Paper no. 140, NACE, 1984.
60. T. L. Yau, Corrosion behavior of zirconium in H2S-containing
environments, Proceedings of the 5th Asian-Pacific Corrosion
Conference, Melbourne, November 2327, 1987.
61. B. Cox, Advance in Corrosion Science and Technology, Vol. 5
(M. G. Fontana and R. W. Staehle, eds.), Plenum Press, New York,
1976, p. 173.
62. N. E. Hamner, Corrosion Data Survey, NACE, Houston, 1974.
63. T. L. Yau, Zircadyne improves organics production, Outlook,
Teledyne Wah Chang, 16 (1): 1 (Spring 1995).
64. K. Mori, A. Takamura, and T Shimose, Corrosion, 22: 29 (1966).
65. P. K. De, K. Elayaperumal, and J. Balachandra Corrosion Sci. 11:
579 (1971).
66. B. Cox, Reports AECL-3551, 3612 and 3799, Atomic Energy of
Canada Ltd.
67. R. F. Koenig, Corrosion of Zirconium and Its Alloys in Liquid
Metals, prepared for the U.S. Atomic Energy Commission by the
General Electric Co., Report No. KAPL-982, October 1, 1953.
68. Y. M. Kolotyrkin, Electrochemical behavior of metals during
anodic and chemical passivation in electrolytic solutions, First Int.
Cong. on Metallic Corrosion, London, 1961, p. 10.
69. T. L. Yau and M. A. Maguire, Control of localized corrosion of
zirconium in oxidizing chloride solutions, Advances in Localized
Corrosion, Vol. 8, NACE, Houston, 1990, p. 311.
70. T. L. Yau and K. Bird, Zirconium is the answer to sulfuric acid
material selection equation, Outlook, Teledyne Wah Chang, 14 (3): 3
(1993).
71. B. Cox, Oxidation of Metals, 3: 399 (1971).
72. B. Cox, Corrosion, 28: 207 (1972).
73. R. D. Watson, Oxidized Zirconium as a Bearing Material in
Water Lubricated Mechanisms, Report CRE-996, Atomic Energy of
Canada, 1960.
74. W. E. Kemp, Nobleizing: creating tough, wear resistance surfaces
on zirconium, Outlook, Teledyne Wah Chang, 11 (2): 4(1990).
75. J. Haygarth and L. Fenwick, Improved wear resistance of
zirconium by enhanced oxide films, Thin Solid Films, 351 (1984).
76. M. A. Maguire and T. L. Yau, Electrochemical properties of
zirconium in mineral acids, Corrosion 86, Paper no. 265, NACE,
Houston, 1986.
77. E. A. Gee, L. B. Golden, and W. E. Lusby, Jr., Ind. Eng. Chem.,
41: 1668 (1949).
78. W. E. Kuhn, Chem. Eng. 67: 154 (Feb. 1960).
Page 252
79. I. R. Lane, Jr., L. B. Golden, and W. L. Acherman, Ind. Eng.
Chem., 45: 1967 (1953).
80. T. Smith and G. R. Hill, J. Electrochem. Soc., 105: 117 (1958).
81. V. V. Andreeva and A. I. Glukhova, J. Appl. Chem., 12: 457
91962).
82. W. E. Kuhn, Corrosion, 16: 136t (1960); 18: 103t (1962).
83. R. S. Sheppard, D. R. Hise, P. J. Gegner, and W. L. Wilson,
Corrosion, 18: 211t (1962).
84. W. E. Lusby, Jr., Corrosion, 13: 654t (1957).
85. Explosion rips Louisiana nuclear plant, Chem. Eng. News, 14
(August 3, 1992).
86. S. Menicatti, C. Miola, S. D. Milanese, and F. Granelli, Process
and apparatus for the synthesis of urea and material used in it, U.S.
Patent 4, 899, 813 (1990).
87. C. Miola and H. Richter, Werst. Korros. 43: 396 (1992).
88. Zirconium heat exchangers resist corrosion in the production of
acetic acid and anhydride, Outlook, Teledyne Wah Chang, 4 (2): 1
(1983).
89. In acetic acidICI Australia increases equipment life by converting
to Zircadyne 702, Outlook, Teledyne Wah Chang, 11 (1): 6 (1990).
90. K. Bird, Outlook, Teledyne Wah Chang, 13 (4): 1 (1994).
91. T. L. Yau and K. Bird, A. comparison of zirconium and titanium
in acetic acid, First NACE Asian Conference, Singapore, September
711, 1992.
92. Kemira specifies Zircadyne 702 for use in a formic acid
application, Outlook, Teledyne Wah Chang, 11 (1): 1 (1990).
93. L. B. Bowen, ASTM STP 728, 1981, p. 119.
94. Rayon producer converts 26 heat exchangers to Zircadyne
Zirconium, Outlook, Teledyne Wah Chang, 10 (3): 1 (Fall 1989).
95. DuPont acid dissolves problems: Zircadyne-lined tube meets Belle
Plant production challenge, Outlook, Teledyne Wah Chang, 14 (3): 1
(1993).
96. Zircadyne Properties and Applications, Teledyne Wah Chang,
1986.
Page 253
11
Corrosion Resistance of Cast Alloys
James L. Gossett
Fisher Controls
Mashalltown, Iowa
I
Introduction
The corrosion resistance of a cast alloy is often different from that of
the wrought equivalentsometimes better, sometimes worse. Wrought
metal is metal that has been formed into a desired shape by working
(rolling, extruding, forging, etc.) This includes bar, plate, sheet,
tubing, pipe, and forgings. Most available corrosion data, supplied by
producers or in the literature, is for wrought material. Use of these
data for selection of critical castings may have undesirable results.
It is important for the end-user to be familiar with how various
components are fabricatedfrom castings, forgings, bar stock, or
whatever the case may be. Cast and wrought alloys often behave
differently in identical service conditions. In nearly all cases, the cast
compositions are altered from the wrought alloys to improve
castability. For example, silicon is added for improved fluidity when
pouring the molten metal into the mold. In addition, there may be a
number of cast alloys commonly used as equivalents for a single
wrought alloy. CW12MW, CW6M, CW2M, and CX2MW are all cast
equivalents for nickel base Hastelloy C (trademark of Haynes
International).
It can also be true that some cast alloys can be produced with superior
properties, but the composition cannot be produced in a wrought form.
Some cast alloys have high silicon and/or carbon contents for superior
corrosion or abrasion
Page 254
resistance; however, the low ductility and high strength may make
rolling and/or forging impossible.
Castings are produced by several different molding processes: green
sand, air-set sand, resin-bonded sand, rammed graphite, investment,
etc. The corrosion resistance of the as-cast surface is a function of the
molding process, pouring temperature; and mold surface treatments or
mold washes. Carbon pickup and mold reactions are just two of the
factors which influence corrosion resistance. The corrosion resistance
of most machined surfaces will be independent of the molding process
provided 1/161/8 in. of material is removed.
To make an accurate prediction of the corrosion resistance of a cast
material in a specific environment, laboratory or field corrosion
testing is needed. Cast coupons should be removed from larger cast
pieces similar to the equipment which they are to represent. The
minimum section thickness for sand-type processes should be 1 in.
Both as-cast and machined surfaces should be tested.
Before going further, the common conventions used to identify cast
materials must be explained.
A
Cast Material Designations
Over the years, materials have come to be known by many names,
including
Popular trade names, such as Inconel (trademark of Inco Alloys
International), Hastelloy, Nitronic (trademark of Armco), Meehanite
(trademark of Meehanite Metal Corp.), Illium (trademark of Stainless
Foundry and Engineering), Wiscalloy (trademark of Wisconsin
Centrifugal), and Escoloy (trademark of Esco Foundry and Machine
Co)
Industry group designations, such as the American Iron and Steel
Institute (AISI), the Aluminum Association (AA), and the Copper
Development Association (CDA)
Standards organization designations, such as American Society for
Testing and Materials (ASTM), American National Standards
Institute (ANSI), American Society of Mechanical Engineers
(ASME), and others
Unfortunately, the less knowledgeable people are regarding materials,
the more likely they are to use a trade name. It is only legal to utilize a
trade name if the material is actually purchased from the owner or
licensee of the trade name. Since many of these materials are now
available from multiple suppliers, the trade name owner may not
actually supply the material. In addition, the use of a trade name, such
as Hastelloy without the grade designation, can lead to disastrous
results. There are over 15 Hastelloy alloys (B, B2, B3, C, C4, C276,
C22, D, F, G, G2, G3, G30, etc.) which have been produced over the
years; some have radically different properties. One alloy may
perform well in a specific environment while others may fail
catastrophically.
Page 255
Trade names are widely used because they are easily remembered.
Another reason is that over the years there have been multiple
designation schemes for different families of materials. The lack of
standard designations has caused a great deal of confusion between
customers and suppliers throughout the industry. In the recent past, a
number of standards organizations and trade associations have made
an effort to solve this problem. Three designation systems are now
commonly used: Unified Numbering System (UNS), Alloy Casting
Institute (ACI), and ASTM.
B
Unified Numbering System
The most widely accepted designation system is the Unified
Numbering System (UNS) developed by the Society of Automotive
Engineers (SAE) and ASTM [1]. The UNS system divides metals and
alloys into 18 series. UNS designations start with a single a character,
followed by five numerical digits. Where possible, the a character is
suggestive of the family of metals it identifies (i.e., ''A" for aluminum
alloys, "C" for copper alloys, "N" for nickel alloys, "S" for stainless
steels, etc.). Table 1 gives a complete listing of the prefixes. The
alloys within a series are identified using the five numerical digits.
Where possible, common designations are used within the five
numerical digits for user convenience (e.g.,
Table 1 Alpha Prefixes Used in UNS
Prefix
Alloy Series
A
Aluminum and aluminum alloys
C
Copper and copper alloys
D
Steels with specified mechanical
properties
E
Rare earths and rare earthlike metals
and alloys
F
Cast irons
G
H
J
K
L
M
N
P
R
S
T
W
Z
Carbon and alloy steels
AISI H-steels (hardenability
controlled)
Cast steels (except tool steels)
Miscellaneous steels and ferrous
alloys
Low-melting metals and alloys
Miscellaneous nonferrous metals
and alloys
Nickel and nickel alloys
Precious metals and alloys
Reactive and refractory metals and
alloys
Heat and corrosion-resistant
(stainless) steels
Tool steels
Welding filler materials
Zinc and zinc alloys
Page 256
A92024 for 2024 aluminum, C36000 for copper alloy 360, S31600 for
type 316 stainless steel, N10276 for nickel alloy C276, G10180 for
1018 steel, etc.). The publication Metals and Alloys in the Unified
Numbering System is a complete listing of all UNS numbers assigned
to date [2].
C
ACI Designations
The Alloy Casting Institute was a former division of the Steel
Founder's Society. ACI developed a system for naming stainless and
heat resisting casting alloys. Most ACI designations begin with 2 a
characters followed by 2 or 3 numeric digits. Some also end with
additional a characters and/or numeric digits. They generally begin
with either a "C" (for corrosion resistant materials) or an "H" (for heat
resistant materials). The second letter in the designation ranges from
"A" to "Z" depending upon the nickel, and to a lesser degree, the
chromium content. For example, a corrosion resistant material with
12% chromium and no nickel begins with "CA'' (example, CA15). An
alloy with 100% nickel begins with "CZ" (example, CZ100). Alloys
in between begin with intermediate letters. The numerical digits
indicate the maximum carbon content (percent × 100). Additional
letters, following the numerical digits, indicate the presence of other
alloying elements. Table 2 gives the examples. There are two groups
of materials whose designations do not follow the scheme shown in
the table. Nickel-copper materials use "M" as the first letter (examples
are M35-1 and M25S). Nickel-molybdenum materials utilize "N" as
the beginning letter (such as N7M and N12MV). Although ACI no
longer exists, the system has been adopted by ASTM, and
designations for new cast alloys are assigned by the appropriate
ASTM committees.
Table 2 Examples of ACI Designations
Alloying elements (%)
Chromium, Nickel, Carbon, Other alloying
Designation nominal nominal maximum elements, nominal
CA15
12
0.15
CD4MCu
25
6
0.04
Mo 3, Cu 3
CF8M
19
10
0.08
Mo 2.5
CF3M
19
10
0.03
Mo 2.5
CN7M
21
29
0.07
Mo 2.5
CW2M
16
68
0.02
Mo 16
CZ100
0
100
1.00
HK40
25
20
0.40
Page 257
D
ASTM Designations
ASTM has retained their own naming convention for many special
carbon and alloy steel products and for cast iron. Some designations
indicate the material type; WCA, WCB, and WCC are weldable
carbon steel, grades A, B, and C. Some grades are numbered in
sequence as they are added to a specification and others indicate a
property, such as strength. The UNS numbers have not been adopted
for these materials because they have no relation with any common
designation. Several examples are given in Table 3.
E
Materials Covered
Most cast alloys are derived from their wrought counterparts. Others
are proprietary alloys developed as casting alloys. The only alloys
discussed here are those which are covered by ASTM specifications.
Use of industry specifications is not a guarantee that the required
casting quality will be obtained. The more critical the applications, the
more the end-user should know about the material, the foundry, and
any intermediate fabricators. When there are no industry
specifications, a private specification should be developed, either by
the end-user, or by the foundry, and carefully reviewed by the enduser. Extensive testing is required to develop a melting practice,
compositional modifications for castability, weld procedures, filler
material, optimum heat treatment, etc. All of these affect corrosion
resistance and casting quality.
II
Cast Irons
Cast iron is the generic term used for the family of high carbonsilicon-iron casting alloys [3]. They contain over 2% carbon, which
can take several different forms: graphite flakes, irregular graphite
nodules, graphite spheres, iron carbides, or cementite, and
combinations of these. Table 4 gives worldwide production figTable 3 Examples of ASTM and UNS Designations
ASTM designation
UNS
designation
Grade WCC Weldable Carbon Steel Casting,
J02503
Grade C
Grade LCB Low Temperature, Weldable
J03003
Carbon Steel Casting, Grade B
Class C Cast Iron
F12802
Grade 135-125 135 Min Tensile, 125 Min
None
Yield Strength Steel Casting
Page 258
Table 4 Worldwide Casting Production
Material
Metric tons
Gray iron
39,462,646
Compacted graphite
Not reported
iron
separately
Ductile iron
10,578,569
Malleable iron
2,151,665
Steel
7,083,837
Copper-based
2,072,642
Aluminum
3,617,034
Magnesium
93,229
Zinc
1,611,360
Other
159,828
Total
66,830,810
Percent of total
59.05
Not reported
separately
15.83
3.22
10.60
3.10
5.41
0.14
2.41
0.24
100
ures for iron and other materials. Use of the term "cast iron" by itself
is discouraged, as cast iron includes gray, malleable, ductile, white,
and high-alloy irons.
A
Gray Iron
Gray iron is the most common cast iron, representing 59% of total
worldwide casting production in 1993 [4]. Gray cast iron has a
relatively large percentage of the carbon present as graphite flakes.
The gray irons have good fluidity at pouring temperatures, which
makes them ideal for casting intricate shapes in all sizes. Most show
little or no shrinkage during solidification, so that pattern making is
simplified compared to other alloys. Gray iron has relatively poor
toughness because of the stress concentration effect of the graphite
flake tips. Gray irons are generally purchased to ASTM specifications.
ASTM A48 and A126, as well as other gray iron ASTM
specifications, use tensile strength as the main acceptance criteria.
Graphite is essentially an inert material and is cathodic to iron. This
results in rapid attack of the iron in even mildly corrosive
environments. As the iron is removed, the remaining graphite flakes
and corrosion products may form a barrier to further attack. This
process is called graphitization because the remaining film is often
black. In the extreme case, the part may appear unaffected, but the
loss of iron may be so severe that significant structural integrity is
lost.
The corrosion resistance of gray iron is slightly better than carbon
steel in water, seawater, and various atmostpheric environments. In
general, however, the corrosion properties of gray iron are similar to
those for carbon steel. Corrosion rates in rural, industrial, and seacoast
environments are generally acceptable. Gray iron is also commonly
used for flue gas applications, such as wood and coal-fired furnaces
and heat exhangers. The life of buried gray iron pipe is
Page 259
generally longer than that of steel, but it is highly dependent on soil
type, drainage, and other factors. Gray iron is not resistant to
corrosion in acid except for certain concentrated acids where a
protective film is formed [3].
Gray iron has good resistance to alkaline solutions, such as sodium
hydroxide and molten caustic soda. Resistance is good in alkaline salt
solutions, such as cyanides, silicates, carbonates, and sulfides. Acid
and oxidizing salts rapidly attack gray iron. Gray iron is used to
contain sulfur at temperatures of 350400°F (149205°C) [5]. Molten
sulfur must be air-free and solid sulfur must be water-free. Gray iron
melting pots are commonly used for aluminum, cadmium, lead,
magnesium, and zinc.
B
Compacted Graphite Iron
Compacted graphite iron is a relatively new type of cast iron [3]. Its
structure is between that of gray and ductile iron. The graphite is
present as blunt flakes which are interconnected. Production is similar
to ductile iron with an additional alloying element like titanium.
Compacted graphite iron retains many of the attractive casting
properties of gray iron but has improved strength and ductility. There
is little difference in the corrosion resistance of compacted graphite
iron and gray iron.
C
Ductile (Nodular) Iron
Ductile iron contains graphite nodules which are formed during
solidification by adding an appropriate inoculant, like magnesium, to
the molten iron just prior to pouring. Ductile iron is also called
nodular iron. This material has many of the advantages (cost, ease of
manufacturing, wear resistance, etc.) as gray iron, with the added
benefit of good ductility. ASTM A395 covers ductile iron castings for
pressure retaining and/or high-temperature applications, and A536
covers ductile irons for structural applications. There is little
difference in the corrosion resistance of ductile iron and malleable
iron.
D
White Iron
White iron solidifies with a "chilled" structure, meaning that instead
of forming free graphite, the carbon forms hard, abrasion-resistant,
iron-chromium carbides. Some white iron contains as much as 25%
chromium to permit casting thicker sections. White irons are primarily
used for abrasive applications. After final machining, the material is
generally heat-treated to form a martensitic matrix for maximum
hardness and wear resistance. White irons are very brittle. Elongation
in the hardened condition is typically about 2%. The only property
required by the ASTM specificiation is the Brinell hardness. In
general, there is little
Page 260
difference in the corrosion resistance of gray iron and white iron. The
high chromium iron has only slightly better corrosion resistance.
E
Malleable Iron
The properties of malleable iron and ductile iron are very similar, but
malleable iron is declining in popularity for economic reasons.
Malleable iron is only available in thin sections because rapid cooling
in the mold is required to form white iron. Irregularly shaped graphite
nodules are formed from the white iron by a heat treatment of over 8
h. The costs involved with such a long heat treatment in a controlled
atmosphere furnace have become prohibitive. The graphite nodules
provide much better ductility than is found in the gray irons. ASTM
A47 and A197 are the two most widely used industry specifications.
In general, there is little difference in the corrosion resistance of gray
iron and malleable iron. Under flowing conditions, malleable iron
may be inferior to gray iron. Without the graphite flakes to hold the
corrosion products in place, attack continues at a constant rate rather
than declining with time.
F
High-Alloy Cast Irons
The high-alloy cast irons are generally divided into three groups:
austenitic, austenitic ductile, and high-silicon cast irons, described in
the following subsections. Table 5 lists attributes for several
specifications.
1
Austenitic Gray Cast Irons
The austenitic gray irons are gray irons with additions of nickel (and,
in some instances, copper) to produce an austenitic matrix structure
like the 300 series stainless steels (SST). They have a flake graphite
structure and mechanical properties similar to the gray irons. The
austenitic gray cast irons (often called Ni-Resist, which is a trademark
of Inco Alloys International) offer better corrosion and wear
resistance, toughness, and high-temperature properties than the
standard gray irons. Austenitic gray irons are purchased per ASTM
A436. The specification governs the chemistry, minimum tensile
strength, and Brinell hardness. Types 1 and 2 are the most commonly
used grades.
The corrosion resistance of austenitic gray iron falls between that of
gray iron and the 300 series SSTs. The largest use is in hydrogen
sulfide-containing oil field applications. A protective sulfide film is
formed preventing excessive attack. Austenitic gray iron also resists
erosion from sand often entrained in crude oil. It is superior to gray
iron in atmospheric exposure, seawater, caustic soda, or sodium
hydroxide and dilute and concentrated (unaerated) sulfuric acid. The
copper in type 1 provides the best resistance to sulfuric acid [6].
Page 261
2
Austenitic Ductile Cast Irons
The austenitic ductile irons (commonly called Ductile Ni-Resist,
trademark of Inco Alloys International) are similar to the austenitic
gray irons except that they are treated with magnesium to produce a
nodular graphite structure. This structure produces higher strength and
ductility than the flake graphite structure. Several different austenitic
ductile irons are produced to obtain desired properties. Type D-2 is
the most commonly used grade. Austenitic ductile iron and austenitic
gray iron have similar corrosion resistance. Grades containing 2% or
more chromium are superior [6].
3
High-Silicon Cast Irons
High-silicon cast irons contain 1218% silicon. Two common
tradenames are Duriron and Durichlor 51 (Duriron Company, Inc.). A
minimum of 14.2% silicon is required for good corrosion resistance
[7]. Chromium and molybdenum may also be added to enhance
corrosion resistance. When high-silicon cast irons are first exposed to
a corrosive environment, surface iron is removed leaving behind a
silicon oxide layer. This layer is very adherent and corrosion-resistant
[3]. High-silicon iron is extremely brittle and difficult to machine.
Great care must be taken to limit stresses and prevent brittle fractures.
This lack of toughness prevents much wider application of highsilicon cast iron.
High-silicon cast iron has the best across-the-board corrosion
resistance of all the commercial metals [8]. It has good resistance to
corrosion in sulfuric, nitric, and organic acids to temperatures at least
as high as the boiling point. Corrosion resistance in mixtures of
oxidizing acids is also good. It has good resistance to neutral salt
solutions. Hydrofluoric acid causes rapid attack.
III
Carbon and Low-Alloy Steels
Cast carbon and low-alloy steels are widely used because of their low
cost, versatile properties, and wide range of grades (Table 6). They are
seldom selected for their corrosion resistance [9]. Carbon steels are
basically alloys of iron and carbon (< 2%). They also contain
manganese (< 1.65%), silicon, sulfur, phosphorus, and other elements
in small quantities either for their desirable effects or because of the
difficulty and expense of removing them. Low-alloy steels contain
less than 5% total alloying elements but more than carbon steel.
There is no significant difference in the corrosion resistance of cast
and wrought carbon and low-alloy steels. Wrought "weathering" lowalloy steels gradually form a protective rust layer after a few years of
exposure to rural and urban atmospheres. These steels contain both
chromium and copper and may also contain silicon, nickel,
phosphorus, or other alloying elements [10]. ASTM G101 is a guide
for estimating the atmospheric corrosion rates of wrought weathering
Page 262
Table 5 Typical Cast Iron Attributes
Minimum strength, >
ksi (MPa)
ASTM specification, grade and
form
A48 Class 30
Gray Iron Castings
A48 Class 40
Gray Iron Castings
A126 Class B
Valves, Flanges and Pipe Fittings
A126 Class C
Valves, Flanges and Pipe Fittings
A842 Grade 250
Compacted Graphite Iron
Castings
A842 Grade 400
Compacted Graphite Iron
Castings
A4732510
Ferritic Malleable Iron Castings
A197
Cupola Malleable Iron
SA395 (60-40-18)
Ferritic Ductile Iron Pressure
Retaining Castings
Chemical
requirements
Tensile
Yield
None
30 (207) NA
Typical
hardness (HB)
195230
None
40 (276) NA
210240
S 0.15a, P 0.75a 31 (214) NA
160220
S 0.15a, P 075a 41 (283) NA
230
None
36.3 (250)NA
179 max
None
58.1 (400)NA
197255
None
50 (340) 32.5 (224)110145
None
40 (275) 30 (207) 110145
C 3 min, Si 2.5a, 60 (415) 40 (275) 143187
P 0.08a
(table continued on next page)
Page 263
(table continued from previous page)
Minimum strength, ksi
(MPa)
Typical
hardness
ASTM specificiation,
grade and form
Chemical requirements
Tensile
Yield
A536 (80-55-06)
None
80 (550)
55 187250
Ductile Iron Castings
(380)
A532 Class I, Type A
C 2.83.6, Cr 1.44, Ni 3.35 NA
NA 550 min
Abrasion Resistant Cast
Irons
A436 Type 1
NA 131183
C 3a, Si 12.8, Ni 13.517.5, 25 (172)
Austenitic Gray Iron
Cu 5.57.5, Cr 1.52.5
Castings
A436 Type 1b
NA 149212
C 3a, Si 12.8, Ni 13.517.5, 30 (207)
Austenitic Gray Iron
Cu 5.57.5, Cr 2.53.5
Castings
A436 Type 2
NA 118174
C 3a, Si 12.8, Ni 1822, Cr 25 (172)
Austenitic Gray Iron
1.52.5
Castings
A439 Type D-2
30 139202
C 3a, Si 1.53, Ni 1822, Cr 58 (400)
Austenitic Ductile Iron
(207)
1.752.75
Castings
A439 Type D-2B
30 148211
C 3a, Si 1.53, Ni 1822, Cr 58 (400)
Austenitic Ductile Iron
(207)
2.754
Castings
A439 Type D-3
30 139202
C 2.6a, Si 12.8, Ni 2832, 55 (380)
Austenitic Ductile Iron
(207)
Cr 2.53.5
Castings
A439 Type D-4
NA 202273
C 2.6a, Si 56, Ni 2832, Cr 60 (413)
Austenitic Ductile Iron
4.55.5
Castings
A518 Grade 1
C 0.651.1, Si 14.214.75 930 lbf. (4090 N) NA 520
Corrosion-Resistant Highmin bend test
Silicon Iron Castings
A861
High-Silicon Iron Pipe
and Fittings
aMaximum content
C 0.751.15, Si 14.214.75, 930 lbf. (4090 N) NA 520
Cr 3.255, Mo 0.40.6
min bend test
Table 6 Cast Carbon and Low-Alloy Steels
Specification, grade and form
ASTM A148 Grade 8040
High-Strength Structural Steel Castings
ASTM A148 Grade 9060
High-Strength Structural Steel Castings
ASTM A148 Grade 11595
High-Strength Structural Steel Castings
ASTM A148 Grade 135125
High-Strength Structural Steel Castings
ASTM A148 Grade 165150
High-Strength Structural Steel Castings
ASTM A216 Grade WCB
Weldable, Pressure-Containing Carbon
Steel Castings
(table continued on next page)
Minimum
strength,
ksi (MPa)
C Mn Si
S
P Other Tensile
max max max
b
b b
0.06 0.05b
80 (550)
b
b
b
0.06 0.05b
90 (620)
b
b
b
0.06 0.05b
115 (795)
b
b
b
0.06 0.05b
135 (930)
b
b
b
0.02 0.02b
165
(1140)
7095
(485655)
0.30a1.0a0.6 0.0450.04
(table continued from previous page)
Specification, grade and form
C
Mn Si
S P Other Tensile
max max max
1.2a 0.6 0.0450.04
7095
(485655)
ASTM A216 Grade WCC
0.25a
Weldable, Pressure-Containing Carbon
Steel Castings
ASTM A352 Grade LCB
0.30a 1.0a 0.6 0.0450.04
6590
(450620)
Weldable, Pressure-Containing Steel
Castings for LowTemperature Service
ASTM A352 Grade LCC
0.25a 1.2a 0.6 0.0450.04
7095
(485655)
Weldable, Pressure-Containing
Steel Castings for LowTemperature Service
ASTM A217 Grade WC6
0.050.2 0.50.80.6 0.0450.04 Cr 11.5, 7095
(485655)
Pressure-Containing Alloy Steel
Mo
Castings for High-Temperature Serfice
0.450.65
ASTM A217 Grade WC9
0.050.180.40.70.6 0.0450.04 Cr 22.75, 7095
(485655)
Pressure-Containing Alloy Steel
Mo
Castings for High-Temperature Service
0.91.2
aMaximum content.
bSelected by foundry to obtain required mechanical properties.
Page 266
steels. Care should be used when utilizing these estimating methods
for cast steels. The natural segregation in cast steels may produce
different results. Weathering steels are of little benefit in submersed
service.
Cast carbon and low-alloy steels are usually protected from
atmospheric corrosion by painting and/or coating systems [11].
Coatings may also be used to prevent rust contamination where
product purity is a requirement. Carbon and low-alloy steels are used
for water, steam, air, and many other mild services. They are also
resistant to many gases provided the moisture content is below the
saturation point. These applications include carbon dioxide, carbon
monoxide, hydrogen sulfide, hydrogen cyanide, sulfur dioxide,
chlorine, hydrogen chloride, fluorine, hydrogen fluoride, and nitrogen.
It must be emphasized that these gases must be dry. Contamination
from air and humidity will cause excessive attack and/or stress
corrosion cracking (SCC).
In certain corrosive environments, a protective surface layer may be
formed which will prevent excessive corrosion. Examples include
concentrated sulfuric acid where a ferrous sulfate film protects the
steel and concentrated hydrofluoric acid which forms a fluoride film.
Extreme care must be taken, however, to prevent conditions which
may damage the film and lead to extremely high corrosion rates.
These conditions include high velocities, condensing water (humidity
from the air), and hydrogen bubbles floating across a surface.
Steel is used for alkaline compounds such as sodium hydroxide and
potassium hydroxide. At temperatures above 150°F (66°C), however,
SCC and excessive corrosion may develop [12]. Neutral salts, brines,
and organics tend to be noncorrosive to steel. Acidic and alkaline salts
are more corrosive. The ''NACE Corrosion Data Survey" is a good
reference for these applications [13].
There is generally little difference in the corrosion resistance of
carbon steels and low-alloy steels. In boiler feedwater, however, the
Cr-Mo grades such as WC6 and WC9 offer definite advantages over
WCB and WCC. The Cr-Mo grades form a more adherent iron oxide
film making them more resistant to erosion/corrosion. Very high
velocities may erode the protective film from WCB and WCC while
WC6 and WC9 are unaffected. When the film is removed, corrosion
will proceed at a high rate until the film is reformed. Under conditions
of high-velocity impingement, carbon steel may be perforated in a few
months whereas a Cr-Mo replacement will last years.
Cast carbon and low alloy steels are routinely used for hydrogen
service. As temperature and hydrogen partial pressure increase,
however, a phenomenon called hydrogen attack can occur. In
hydrogen attack, atomic hydrogen diffuses into the steel and combines
with carbon to form methane. Over time, high-pressure methane
pockets are formed and structural integrity is lost [14]. Following the
guidelines of the American Petroleum Institute (API) will prevent this
attack [15].
Page 267
Anhydrous ammonia is routinely handled by cast carbon and lowalloy steels. To prevent SCC, however, small amounts of water are
added. Oxygen contamination should also be avoided. Other media
known to cause SCC are high-temperature hydroxides, nitrates,
carbonates, moist gas mixtures of carbon dioxide and carbon
monoxide, hydrogen cyanide solutions, amine solutions, and
hydrogen sulfide.
Postweld heat treatment (PWHT) or stress relieving is a relatively
inexpensive process to minimize the occurrence of SCC. NACE
International committees have developed several standards and
recommended practices along these lines, including MR0175,
RP0472, RP0590, and 8X194 [1619].
IV
Stainless Steels
Stainless steels are ferrous alloys with a minimum of 12% chromium.
The chromium forms a uniform, adherent chromium oxide film,
providing greatly improved corrosion resistance compared to carbon
and low-alloy steel. SSTs also contain varying amounts of nickel,
molybdenum, nitrogen, copper, and/or other elements. The widely
different compositions result in a range of properties. To review their
corrosion properties, they will be grouped as follows: martensitic,
ferritic, austenitic, superaustenitic, precipitation hardenable, and
duplex stainless steels. The Avesta Sheffield Corrosion Handbook is a
good general reference for SSTs [20].
A
Martensitic Stainless Steels
Martensitic stainless steels were the original SST's developed in the
early 1900's. Since then a range of martensitic grades have been
developed. The main advantages they offer are low cost and the
ability to be hardened for wear resistance. The martensitic grades can
be heat-treated similar to the low-alloy steels to produce hardnesses,
varying by grade, as high as 60 HRC.
Cast CA15 is the modern version of the original 12% chromium
stainless steel. CA15 is often replaced by a newer grade called
CA6NM. CA6NM is modified (with additions of nickel and
molybdenum) for improved castability, mechanical properties, lowtemperature toughness, and resistance to sulfide stress cracking (SSC).
CA28MWV is also a modified 410 with improved high temperature
strength. CA40F is the free machining version of 420 SST.
The martensitic grades are resistant to corrosion in mild atmospheres,
water, steam, and other nonsevere environments (Table 7). They will
quickly rust in marine and humid industrial atmospheres, and are
attacked by most inorganic acids. They are susceptible to several
forms of SCC when used at high hardness levels. Hardened
martensitic SSTs have poor resistance to sour environments and
Table 7 Cast Martensitic Stainless Steels
Minimum
strength, ksi
(MPa)
ASTM specification, grade and Wrought
form
equivalent
A217 Grade CA15
410
Steel Castings for Pressure
Retaining Service
A743 Grade CA40F
420F
Corrosion Resistant Castings for
General Application
A743 Grade CA28MWV
422
Corrosion Resistant Castings for
General Application
A487 Grade CA6NM, Class B 410
Corrosion Resistant Castings for modified
General Application
C
0.15
max
Other
Cr elements Tensile Yield
11.514
90115 65
(620795)(450)
0.20.4 11.514S 0.20.4 100
(690)
70
(485)
0.20.281112.5 Mo
140
0.91.25, (965)
W
0.91.25,
V 0.20.3
0.06 11.514 Ni
100
max
3.54.5, (690)
Mo 0.41
110
(760)
75
(520)
Page 269
may crack in humid industrial atmospheres. In the quenched and fully
tempered condition (usually below 25 HRC), SCC resistance is
greatly improved, especially for CA6NM. The martensitic grades are
generally less corrosion-resistant than the austenitic grades.
B
Ferritic Stainless Steels
When the chemistry of stainless steel is properly balanced, the
structure will be ferritic at room temperature just like a plain carbon
steel. The ferritic SSTs have properties much different from those of
the austenitic SSTs (see Table 8), some of which can be very
advantageous in certain applications. The two most common cast
ferritic SSTs are CB30 and CC50. These alloys have very poor impact
resistance compared to the cast austenitic grades. Due to the formation
of a brittle s phase at elevated temperatures, most ferritic SSTs are
limited to use below about 650°F (343°C). In general, ferritic SSTs
have poor weldability. There are few instances where these materials
would be preferred over an austenitic SST.
CB30 is resistant to nitric acid, alkaline solutions, and many inorganic
chemicals. CC50 is used for dilute sulfuric acid, mixed nitric and
sulfuric acids, and various oxidizing acids. Their resistance to chloride
SCC is better than austenitic SSTs due to their low nickel contents.
C
Austenitic Stainless Steels
The early austenitic stainless steels had compositions of
approximately 18% chromium and 8% nickel, and were commonly
called "18-8" SST. Austenitic SSTs have much better general
corrosion resistance than the 12% chromium SST's (Table 9). While
the wrought austenitic SSTs have completely austenitic structures in
the annealed condition, the castings are chemically balanced to form
some ferrite as they solidify. The ferrite is necessary to prevent hot
cracking of the castings. While most grades contain at least 5% ferrite
and attract a magnet weakly, it is not unusual for CG8M to contain as
much as 30% ferrite and attract a magnet strongly. Another benefit is
that the ferrite phase is resistant to SCC in some environments, and its
presence can retard cracking.
CF8M is the most widely used cast stainless steel. CF8M is the cast
equivalent of 316. CF8M and 316 have a good balance of corrosion
resistance, availability, strength, and cost. Although 304 is considered
the standard wrought SST, CF8M is the standard cast SST. CF8
castings are more expensive than CF8M and should only be specified
when CF8M cannot be used. Most other cast SSTs are used for
specific niches where a small compositional difference gives better
performance in that application.
Table 8 Cast Ferritic Stainless Steels
Minimum
strength, ksi
(MPa)
ASTM specification, grade and
form
A743 Grade CB30
Corrosion-Resistant Castings for
General Application
A743 Grade CC50
Corrosion-Resistant Castings for
General Application
Wrought C
Ni
equivalent max Cr max Tensile Yield
431
0.3 1821 2 65 (450) 30 (205)
446
0.5 2630 4 55 (380) NA
Table 9 Cast Austenitic Stainless Steels
ASTM specifications
and grade
A743 Grade CF20
A743 Grade CF16F
A351, A743, A744
Grade CF3
A351, A743, A744
Grade CF8
A743 Grade CF8C
Wrought
equivalent
302
303
304L
C
max Cr
0.151719
0.151719
0.031721
Ni Mo Other elements
810
810 0.6aS 0.15a
812 0.5a
Tensile
70 (485)
70 (485)
70 (485)
304
0.081821
811
0.5a
70 (485)
347
0.081821
912
A351, A743, A744
Grade CF3M
A351, A743, A744
Grade CF8M
A351, A743, A744
Grade CG8M
A351, A743 Grade
CG6MMN
316L
0.031721
913
0.5a Cb = 8 times C, 70 (485)
1% minimum
23
70 (485)
316
0.081821
912
23
70 (485)
317
0.081821
913
34
75 (520)
A351, A743 Grade
CF10SMnN
aMaximum content.
Nitronic 50 0.0620.523.511.5131.53 Mn 46, Cb 0.1 85 (585)
0.3, V 0.10.3,
N 0.20.4
Nitronic 60 0.1 1618 89
85 (585)
Mn 79, Si 3.5
4.5, N 0.080.18
Page 272
CF8M has excellent corrosion resistance in normal atmospheric
conditions including seacoast exposure. At worst, some slight staining
may develop. It resists most water and brines at ambient temperature.
Seawater may cause pitting corrosion particularly under low-flow or
stagnant conditions or at elevated temperatures. CF8M is used for
80100% sulfuric acid at ambient temperature.
Corrosion is reduced further under oxidizing conditions, such as small
additions of nitric acid, air, or copper salts. CF8M has good resistance
to phosphoric acid at all concentrations up to 170°F (77°C). It is used
for nitric acid up to boiling at all concentrations to 65%. CF8M resists
attack by most organic acids including acetic, formic, and oxalic acids
at all concentrations at ambient temperature. It is used for citric acid at
all concentrations. It is not attacked by organic solvents; however,
chlorinated organics may attack CF8M especially under condensing
conditions or when water is present. CF8M resists many alkaline
solutions and alkaline salts; ammonium hydroxide at all
concentrations to boiling and sodium hydroxide at all concentrations
up to 150°F (65°C) above which SCC may occur [21].
Metallic chloride salts, such as ferric chloride and cupric chloride, can
be very corrosive to CF8M. Above 160°F (71°C) chlorides can also
cause SCC. The combination of chlorides, water, oxygen, and surface
tensile stress can result in cracking at stresses far below the tensile
strength of all austenitic SSTs. Although a threshold chloride level
may exist, one is difficult to set because chlorides concentrate in pits,
crevices, and under deposits until the minimum concentration is
reached. One must be concerned about SCC any time a few hundred
ppm chlorides is present and the temperature exceeds about 160°F
(71°C). SCC may develop at lower temperatures if the pH is low.
Sensitization of austenitic SSTs develops from exposure to
temperatures between 950 and 1450°F (510 and 788°C). Chromium
carbides form at the grain boundaries leaving a zone which is
chromium-depleted. In aggressive environments, the grain boundaries
are corroded. This is called intergranular attack (IGA). When attack
surrounds an entire grain, grain dropping occurs resulting in extremely
high rates of attack. Welding can also produce sensitization in the
weld and in the heat-affected zone (HAZ). In most applications, this
attack can be prevented by welding with low-carbon filler material
and using minimum heat input. Only the most aggressive
environments will produce IGA.
CF3M is the cast equivalent of 316L. It has a maximum carbon
content of 0.03% vs. 0.08% for CF8M. With <0.03% carbon,
sensitization is largely eliminated. CF3M can be specified for
applications where IGA has been a problem. With today's improved
foundry technology, many heats of CF8M are at or near the 0.03%
carbon limit.
Page 273
Molybdenum is added to SSTs to increase pitting resistance.
Molybdenum makes the surface oxide layer tougher, so that chlorides
and other pitting agents are less likely to break it down. CF3 and CF8
are the cast equivalents of 304L and 304, respectively. CF3 and CF8
contain a maximum molybdenum content of 0.5% vs. the 23%
molybdenum of CF3M and CF8M, which does sacrifice pitting
resistance and general corrosion resistance in some environments. In
strongly oxidizing environments, the lower molybdenum of CF8
provides superior corrosion resistance. CF3 and 304L are the standard
materials in hot, concentrated nitric acid. CF8 and 304 are not
generally used because they are more susceptible to IGA in nitric acid.
Cast CF20 (cast equivalent of 302) is the modern version of the
original 18-8 composition. CF16F (cast equivalent of 303) is a freemachining version of CF20. The lower alloy content in these grades
sacrifices some corrosion resistance. The added sulfur reduces
resistance further due to the galvanic effects between the matrix and
the manganese sulfide inclusions.
CF8C (cast equivalent of 347) contains columbium to stabilize the
material against chromium carbide formation. A narrow line of attack
adjacent to a weld can occur if the casting is not properly heat-treated.
CF8C must be solution heat-treated at 19502048°F (10661120°C) and
stabilized at 15981652°F (870900°C) [22]. The corrosion resistance of
CF8C is about the same as that for CF3 and CF8.
CG8M (cast equivalent of type 317) is essentially a modified CF8M.
The chromium, nickel, and molybdenum contents are all increased
slightly, imparting better overall corrosion and pitting resistance.
CG8M is widely used in the pulp and paper industry where it better
resists the attack from pulping liquors and bleach-containing water.
These applications are becoming increasingly corrosive and even
higher grades of SST are often needed.
CG6MMN is the cast equivalent of Nitronic 50 (trademark of Armco,
Inc.). It is a nitrogen-strengthened alloy with 22% chromium, 13%
nickel, 5% manganese, and 2.2% molybdenum. The material is used
in place of CF8M when higher strength and/or better corrosion
resistance is needed.
CF10SMnN is the cast equivalent of Nitronic 60 (trademark of
Armco, Inc.). It has better galling resistance than the other CF grades.
The corrosion resistance is similar to CF8 but not as good in hot, nitric
acid.
Austenitic SST castings are purchased to three specifications. ASTM
A743 and A744 are used for general applications and A351 for
pressure-retaining castings. For critical applications, additional
specifications may be necessary. Items which may be addressed
include filler material, interpass temperature, solution heat treating
temperature, quench method, surface condition, nondestructive
examination, etc. [23].
Page 274
D
Superaustenitic Stainless Steels
Austenitic SSTs with alloying element contents (particularly nickel
and/or molybdenum) higher than the conventional 300 series SSTs are
commonly categorized as "superaustenitic" SSTs (see Table 10). In
some cases, they have even been classified as nickel alloys. These
alloys typically contain 1625% Cr, 3035% Ni, Mo, and N; some also
contain Cu. No single element exceeds 50% [24]. The additional
nickel provides added resistance to reducing environments and the
additional molybdenum, copper, and nitrogen boost the resistance to
pitting in chlorides. Even in the cast form, these alloys are fully
austenitic, making them considerably more difficult to cast than the
ferrite-containing austenitic grades. Foundry experience and expertise
is critical in casting superaustenitics.
CK3MCuN and CE3MN are the cast equivalents of Avesta 254SMO
(trademark of Avesta AB) and AL6XN (trademark of Allegheny
Ludlum, Inc.), respectively. They are part of the "6 Mo"
superaustenitic family. These alloys have complete resistance to
freshwater, steam, boiler feedwater, atmospheric and marine
environments. They also have excellent resistance to phosphoric,
dilute sulfuric, and many other acids and salts. They are highly
resistant to acetic, formic, and other organic acids and compounds
[25].
Superaustenitics are particularly suitable for high-temperature,
chloride-containing environments where pitting and SCC are common
causes of failure with other SSTs. Resistance to chloride SCC extends
beyond 250°F (121°C). They also have excellent resistance to sulfide
stress cracking. CK3MCuN will resist pitting in 6% FeCl3 (60,000
ppm Cl) at 104°F (40°C) while the conventional SSTs will pit at
ambient temperature [26]. In some applications, superaustenitic SSTs
can be used instead of nickel base alloys at a lower cost [27].
CN7M, commonly called alloy 20, is the cast equivalent of Carpenter
20Cb3 (trademark of Carpenter Technology). This is the industry
standard alloy for sulfuric acid. CN7MS is a modified version [28].
They have useful resistance over most of the sulfuric acid
concentration range below 160°F (71°C) and below 10% to the
boiling point. They have excellent resistance to chloride SCC.
Although the ASTM specifications permit up to 0.07% carbon, 0.03%
maximum is recommended [29]. CK3MCuN and CE3MN are
superior for chloride environments.
CU5MCuC is the cast version of Incoloy 825 (trademark of Inco
Alloys International) although columbium is substituted for titanium.
Titanium will oxidize rapidly during air melting; columbium will not.
CU5MCuC has corrosion resistance and weldability similar to CN7M.
It has equal corrosion resistance in sulfuric, nitric, and phosphoric
acids, seawater, and other environments. It is also highly resistant to
chloride SCC.
Weld procedures for superaustenitic SSTs must be carefully
developed to preserve the special corrosion properties. Heat input
must be kept to a minimum,
Table 10 Cast Superaustenitic Stainless Steels
ASTM specification and
Wrought
C
grade
equivalent max
Cr
Ni
Mo
A351, A743, A744 Grade 20Cb3
0.07 1922
27.530.523
CN7M
A743, A744 Grade CN7MS Modified
0.07 1820
2225
2.53
20Cb3
A351, A743, A744 Grade 254SMO
0.02519.520.517.519.567
CK3MCuN
A351, A744 Grade CE3MN AL6XN
A494 Grade CU5MCuC
Incoloy 825
Other
elements
Cu 23
Cu 1.52
Cu 0.51,
N
0.180.24
0.03 2022
23.525.567
N
0.180.26
0.05 19.523.53846
2.53.5 Cu
1.53.5,
Cb
0.61.2
Page 276
and interpass temperatures must be in the 250350°F (121177°C)
range. Overmatching weld filler materials are generally used for weld
repairs and fabrication welds [30]. American Welding Society (AWS)
filler metal grades NiCrMo-3, NiCrMo-7, NiCrMo-10, and NiCrMo12 are the most commonly used grades [31,32]. Welding with
matching filler requires re-solution heat treatment after all welding.
Autogenous welding (without filler material) should never be
performed on these materials. AWS 320LR weld filler is normally
used on CN7M.
E
Precipitation-Hardening Stainless Steels
CB7Cu-1 and CB7Cu-2 are the cast versions of 17-4PH and 15-5PH
(trademarks of Armco Steel). These are high-strength, precipitationhardening, martensitic SSTs (Table 11). Although there are many
other wrought precipitation-hardening SSTs, these are the only two
cast alloys coverd by ASTM specifications. Typically these materials
are cast, solution heat-treated, machined, and then aged. CB7Cu-1 is
more commonly cast than CB7Cu-2.
Cast CB7Cu-1 was recently added to the NACE standard MR0175 for
non-pressure-containing, internal valve and pressure regulator
components [15]. It is acceptable for sour service in the H1150 DBL
condition to a maximum hardness of 310 HB (30 HRC). For both
alloys, the higher hardness conditions are quite susceptible to stress
corrosion cracking. SCC resistance improves with increasing aging
temperature and decreasing strength and hardness.
The corrosion resistance of these alloys is similar to CF8 and 304 and
better
Table 11 Cast Precipitation Hardening Stainless Steels
Specification and
Wrought
Other
grade
ASTM A747
Grade CB7Cu-1
ASTM A747
Grade CB7Cu-2
Condition
H900
H1075
H1150
H1150 DBL
equivalent C max
Cr
Ni elements
17-4PH
0.07 15.517.7 3.64.6 Cu
2.53.2,
Cb
0.150.35
15-5PH
0.07 1415.5 4.55.5 Cu
2.53.2,
Cb
0.150.35
Minimum strength,
ksi (MPa)
Hardness
Tensile Yield (HB)
170 (1170) 145 375 min
(1000)
145 (1000) 115 277 min
(795)
125 (860)
97 269 min
(670)
310 max
Page 277
than the 400 series SSTs [33]. CB7Cu-1 and CB7Cu-2 resist
atmospheric attack in all but the most severe environments. They are
resistant to natural water except seawater where pitting can be
expected. They are widely used in steam, boiler feedwater,
condensate, and dry gases.
F
Duplex Stainless Steels
When the chemistry of a stainless steel is adjusted properly, both
ferrite and austenite will be present at room temperature. SSTs with
approximately 50% austenite and 50% ferrite are called duplex SSTs
(see Table 12). The popularity of these materials has increased rapidly
in recent years because they offer superior corrosion resistance and
higher yield strength than the austenitic SSTs with a lower alloy
content. Due to the formation of s phase at elevated temperatures,
duplex SSTs are limited to a maximum service temperature of 500°F
(260°C). The formation of s phase adversely affects both toughness
and corrosion resistance. Use of s-phase formation as a hardening
mechanism is occasionally done but is not recommended.
Welding of duplex alloys can also be somewhat difficult due to the
potential for forming s-phase. Welding filler material containing about
12% more nickel than the casting is normally used when the castings
will be re-solution heat-treated. Filler material with 3% additional
nickel is used when castings are not re-solution heat-treated [34].
Duplex stainless steels have complete resistance to freshwater, brine,
steam, boiler feedwater, atmospheric and marine environments. They
are particularly suitable for high-temperature, chloride-containing
environments where pitting and SCC are common causes of failure
with other SSTs. Duplex alloys have inherently better SCC resistance
than single-phase alloys, since at least one of the phases is generally
resistant to cracking in a given environment. These alloys have good
resistance to urea and sulfuric, phosphoric, and nitric acids [35]. They
are also highly resistant to acetic, formic, and other organic acids and
compounds [36].
Alloy Z 6CNDU20.08M to French National Standard NF A 320-55 is
the cast version of Uranus 50M (trademark of Creusot-Loire). It is the
only cast duplex SST grade which is currently acceptable per NACE
MR0175 for general use [15,37]. Unlike other duplex SSTs, Z
6CNDU20.08M is limited to 2540% ferrite in NACE MR0175, which
means it is only a borderline ''duplex" SST. Its corrosion resistance is
slightly better than CF8M but inferior to the other duplex SSTs.
CD3MN is the cast version of wrought UNS S31803 or 2205. It is
actually listed in ASTM A890 as grade 4A. This is a nonproprietary
duplex SST available from many sources worldwide. With its lower
alloy content compared to other duplex grades, its cost is lower, but
some corrosion resistance is sacrificed.
CD4MCu is a cast duplex SST which has been in use for many years.
It is
Table 12 Cast Duplex Stainless Steels
CD3MN
Wrought
Specifications C
Other
equivalents
and grade
max Cr
Ni
Mo elements
0.0424.526.54.756 1.752.25Cu
Ferralium 255 A351, A890
2.753.25
Grade 1A
S31803, 2205 A890 Grade 4A 0.032123.5 4.56.52.53.5 N 0.10.3
CD3MWN
Zeron 100
Designations
CD4MCu
Z6CNDU20.08M Uranus
50M
A890 Grade 5A 0.032426
6.58.534
0.081923
NF A 32-055
Grade
Z6CNDU20.08M
79
23
N
0.20.3,
Cu 0.51,
W 0.51
Cu 12
Page 279
used for environments which are too corrosive for the commonly used
austenitic SSTs or where SCC may be a problem. It is similar to
wrought Ferralium 255 (trademark of Bonar Langley Alloys Ltd.). Its
corrosion resistance is better than that of CF8M.
CD3MWN is a new duplex recently added to ASTM. It is the wrought
equivalent of Zeron 100 (trademark of Weir Materials, Ltd.) [38]. It
has higher alloy content that the other duplex grades, giving corrosion
resistance nearly as good as that of the superaustenitic alloys.
V
Nickel Alloys
Most cast nickel-base alloys are derived from wrought alloys (Table
13). The nickel-base alloys are considerably more difficult to cast than
the austenitic SSTs. Nickel alloy castings should never be purchased
using a wrought alloy trade name. Foundry selection is critical in
obtaining high-quality, corrosion-resistant castings. To develop the
required technical expertise, a foundry must pour nickel alloys on a
daily basis. Other important factors are dedicated high-alloy patterns,
careful alloy selection, and additional specifications beyond the
normal ASTM requirements. Items covered would include foundry
qualification, heat qualification using a weldability test, raw material
restrictions, heat treating, nondestructive examination, and repair
welding [3941].
A
Commercially Pure Nickel
CZ100 is the cast commercially pure nickel grade. The wrought
equivalent is nickel 200. CZ100 has higher carbon and silicon for
castability and is generally used in the as-cast condition. Its properties
are not affected by heat treatment. Nickel is used for dry halogen
gases and liquids (F2, HF, Cl2, and HCl) and ambient temperature
hydrofluoric acid. Nickel is used for caustics, including sodium
hydroxide and potassium hydroxide, over a wide range of
temperatures and concentrations. Ammonium hydroxide corrodes
nickel rapidly [42].
B
Nickel-Copper
Monel is the Inco trademark of the original nickel-copper alloy
developed in the 1930s. Monel has excellent resistance to organic
fouling and corrosion in seawater. The most common cast grade is
M35-1. Other cast grades are M35-2 and M30C. M25S is high-silicon
nickel-copper alloy with superior wear and galling resistance but is
the most difficult of all to cast. It is also known as S-Monel
(trademark of Inco Alloys). These grades are used in the as-cast
condition, except M25S which does respond to heat treatment. They
are the industry standards for oxygen, dry chlorine, fluorine, and
hydrogen fluoride gases (no water vapor present). They are also used
for hydrofluoric acid, neutral
Table 13 Cast Nickel-Base Alloys
Minimum stren
Specification and
grade
ASTM A494
Grade CZ100
ASTM A494
Grade M35-1
ASTM A494
Grade M35-2
ASTM A494
Grade M30C
ASTM A494
Grade M25S
ASTM A494
Grade CY40
ASTM A494
Grade CW6MC
Wrought
equivalent
Nickel 200
C
maxCr
1
Ni Fe
95a3a
Monel 400
0.35
bal.3.5a
Monel 400
0.35
bal.3.5a
Monel 400
0.3
bal.3.5a
S-Monel
0.25
bal.3.5a
Inconel 600
0.4 1417 bal.11a
Inconel 625
0.062023 bal.5a 810
ASTM A494
Grade CW2M
ASTM A494
Grade CX2MW
ASTM A494
Grade CW6M
ASTM A494
Grade N7M
ASTM A494
GradeCY5SnBiM
aMaximum content
Hastelloy C
0.021517.5bal.2a 1517.5
Hastelloy C22
Chlorimet 3
0.022022.5bal.26 12.514.5 W 2.5 80 (550)
3.6
0.071720 bal.3a 1720
72 (495)
Hastelloy B2
0.071a
Waukesha 88
0.051114 bal.2a 23.5
Mo
Others
Tensile
50 (345)
Si
1.25a
Si 2a
65 (450)
65 (450)
Si 12, 65 (450)
Cb 13
Si
300 HB min
3.54.5 aged condition
70 (485)
Cb
3.15
4.5
70 (485)
72 (495)
bal.3a 3033
76 (525)
Bi 35,
Sn 35
Page 281
and alkaline salts, and sodium hydroxide [43]. Other common uses are
brine and seawater.
C
Nickel-Chromium
CY40 is the cast equivalent of Inconel 600 (trademark of Inco
Alloys). CY40 is a nickel-chromium alloy without the molybdenum
content of most other nickel-chromium alloys. In most environments,
the corrosion resistance of CY40 is poor compared to the nickelchromium-molybdenum alloys. Pitting can occur in moist, humid
conditions, seawater, chloride environments, and salts. CY40 is used
in steam, boiler feedwater, and alkaline solutions including
ammonium hydroxide. The resistance to chloride SCC is good [44].
D
Nickel-Chromium-Molybdenum
Nickel-chromium-molybdenum alloys offer excellent corrosion
resistance and good mechanical properties over a wide range of
environments and temperatures. CW2M, the cast version of Hastelloy
C, is the workhorse of the group. Castings should not be called
"Hastelloy" [45,46]. The properties of the different Hastelloy alloys
vary widely in specific applications. Disaster can result from use of
the wrong grade.
CW2M has excellent corrosion resistance in many chemical process
environments including hydrochloric and sulfuric acids at
temperatures below 125°F (52°C). At low concentrations, the useful
temperature range is much higher. Corrosion resistance is excellent in
organic acids. Contamination by strong oxidizing species, such as
ferric and cupric ions, will not cause the accelerated attack common
with other alloys such as Hastelloy B2. CW2M is resistant to most
forms of SCC including chloride, caustic, and H2S [47]. AWS filler
materials NiCrMo-7 or NiCrMo-10 maintain good as-welded
corrosion resistance [48]. CW12MW is the original Hastelloy C type
of casting grade. Segregation problems inherent with the alloy
resulted in corrosion resistance inferior to wrought C276. CW12MW
has been largely replaced by CW2M. The casting characteristics,
weldability, and ductility are all greatly enhanced.
In addition to CW2M, there are a number of other nickel-chromiummolybdenum casting alloys. Some of the alloys are CX2MW (cast
Hastelloy C22), CW6MC (cast Inconel 625), and CW6M (Chlorimet
3, trademark of Duriron Co.).
E
Other Nickel-Base Alloys
N7M is the cast equivalent of Hastelloy B2. This nickel-molybdenum
alloy has excellent corrosion resistance in all concentrations and
temperatures of hydro-
Page 282
chloric acid. If ferric or cupric ions are present, however, severe attack
will occur. It is also good for sulfuric, acetic, and phosphoric acids
[49].
CY5SnBiM is a proprietary alloy known as Waukesha 88 (trademark
of Waukesha Foundry). Tin and bismuth are added as solid metal
lubricants for improved galling resistance. It is primarily used in the
food industry to prevent galling against SST. Weld repairs are
prohibited. It is not as corrosion-resistant as other nickel-base alloys;
however, it performs well in food industry applications.
VI
Titanium
Titanium is only cast by a few foundries which specialize in titanium
and zirconium castings. Molten titanium reacts instantly with air or
foundry mold sand, causing severe embrittlement or a dangerous
pyrophoric reaction. The entire casting process must be done in a
vacuum. Titanium is melted and poured into graphite molds in an
evacuated chamber. Castings are normally used in the as-cast
condition; however, heat treatment is recommended after welding.
Titanium is routinely welded by paying strict attention to cleanliness
and the use of qualified welding procedures. Grade C3 is the most
common titanium casting grade. The alloyed grades, such as C5, are
slightly less corrosion-resistant in certain environments (Table 14).
Titanium relies on a stable, tightly adherent, oxide film for corrosion
resistance. If the protective oxide film is removed or damaged, the
film will reform instantly when exposed to air or moisture. Fine chips
and shavings from machining must be handled properly to prevent a
pyrophoric reaction. Certain chemicals can also produce violent
exothermic reactions. Titanium is only good for HCl, and Cl2 service
when H2O is present. If H2O is not present, the oxide film will
quickly be destroyed and the titanium consumed [50].
Titanium has excellent resistance to seawater, brine, and other salt
solutions including chlorides, hypochlorides, sulfates, and sulfides. It
has good resistance in chlorine dioxide, a commonly used chemical in
pulp and paper. It has excellent resistance to wet chlorine gas, nitric
acid, molten sulfur, aqueous and anhydrous ammonia, aqua regia,
hydrogen sulfide, dilute caustic, many organic compounds, and most
oxidizing acids [51]. Titanium should never be used in reducing
environments. Titanium should not be used in warm or concentrated
hydrochloric, sulfuric, phosphoric, or oxalic acids. Strong oxidizers
cause rapid attack, such as red-fuming nitric acid, HF, and dry
chlorine.
Pitting corrosion is seldom a problem unless iron is embedded in the
titanium surface. The iron will cause localized, galvanic corrosion.
Titanium has excellent resistance to stress corrosion cracking. It will
solve most SCC problems encountered with SSTs. Cracking may
occur in methanol, chlorinated solvents, liquid metals (cadmium,
mercury, and silver-base brazing compounds) and chloride salts over
500°F (260°C).
Table 14 Cast Titanium Grades
Max. content
(%)
Minimum strength,
ksi (MPa)
ASTM specification and UNS and wrought
grade
equivalent
B367 Grade C2
R50400, Grade 2
O
0.40
Fe
0.20
Tensile
50 (345)
B367 Grade C3
0.40
0.25
65 (450)
B367 Grade C5
R50550, Grade 3
UNS
R56400, Grade 5
Minimum
strength,
Content (%)
ksi (MPa)
Al
V Tensile Yield
5.56.753.54.5 130
120
(895) (825)
Page 284
VII
Zirconium
Zirconium is a highly corrosion-resistant material with characteristics
similar to titanium. In commercial production since the 1950s, its
initial use was for uranium fuel rod cladding. Today zirconium is
widely used in the chemical process industry where its outstanding
corrosion resistance makes it cost competitive with some nickel-base
alloys. Like titanium, zirconium must be melted and poured into
graphite molds in a vacuum. Great care is also required during
welding to prevent contamination. Three grades of cast zirconium are
now covered by ASTM B752 (Table 15). All three grades are used ascast except for postweld heat treatment of weld repairs.
The corrosion resistance of zirconium is outstanding; even better than
titanium. The corrosion resistance of all zirconium alloys is similar.
Seawater, brine, and heavily polluted water will not attack zirconium.
The corrosion rate of zirconium in hydrochloric acid is less than 5
mpy at all concentrations and temperatures to at least 260°F (127°C).
This is better than all metals except tantalum, which has about the
same resistance as zirconium. The corrosion resistance in sulfuric acid
is good at 070% concentrations and to boiling temperatures and
beyond [52]. The same exceptional resistance is shown in phosphoric
and nitric acids [53]. Zirconium should not be used in hydrofluoric
acid where concentrations as low as 0.001% will produce attack.
Zirconium is resistant to virtually all alkaline solutions up to the
boiling point. It is resistant to most organic solutions and is widely
used in the extremely corrosive urea and carbamate areas of urea
plants. Zirconium resists attack in certain molten salts and liquid
metals. It is severely attacked by molten zinc, bismuth, and
magnesium. Zirconium resists pitting, crevice corrosion, and stress
corrosion cracking in most environments which affect steels and
SSTs. SCC has been documented in FeCl3, CuCl2, concentrated nitric
acid, and liquid mercury and cesium.
VIII
Tantalum
Tantalum is another member of the refractory metals family like
titanium and zirconium. Casting is impractical due to the extremely
high melting point, 5425°F (2996°C). Its corrosion resistance is even
better than that of titanium and zirconium. Up to 300°F (150°C) it
resists attack by almost all chemicals.
IX
Copper
Copper alloys are primarily used because of their high thermal and
electrical conductivity, good corrosion resistance, good bearing
surface properties, and other special properties (Table 16) [54]. The
composition of the casting alloys
Table 15 Cast Zirconium Alloys
Minimum
strength,
ksi (MPa)
ASTM specification and Total residual
grade
elements max
ASTM B752
0.4
Hf
max Sn
CbTensile Yield
4.5 0.3
55
40
max
(380) (276)
Grade 702C
ASTM B752
0.4
4.5
0.3
max
Grade 704C
ASTM B752
0.4
4.5
12
Grade 705C
60
40
(413) (276)
23 70
50
(483) (345)
Maximum har
Table 16 Cast Copper-Base Alloys
UNS number and grade
C83600
Leaded Red Brass
C85800
Leaded Yellow Brass
C86300
High Strength Manganese
Bronze
C86500
#1 Manganese Bronze
C87800
Silicon Brass
C90300
Modified G Tin Bronze
Nominal composition
(%)
ASTM specification and CuPbSnZn Others
form
B584
85 5 5 5
Sand Castings
B176
61 1 1 36
Die Castings
B584
61
27 Fe 3, Al 6,
Mn 3
Sand Castings
B584
Sand Castings
B176
Die Castings
B584
Sand Castings
(table continued on next page)
Tensile
30 (207)
55 (379)
110 (758)
58
39 Fe 1, Al 1,
Mn 1
65 (448)
89
14 Si 4
85 (586)
88
8 4
40 (276)
(table continued from previous page)
Nominal composition
Minimum stre
(%)
UNS number and grade ASTM specification and CuPb SnZnOthers
Tensile
form
C90500
B584
88 102
40 (276)
G Tin Bronze
Sand Castings
C92200
B61
88 1.56 4
34 (235)
Valve Bronze
Steam and Valve Bronze
Castings
C93200
B584
83 7 7 3
30 (207)
High Leaded Tin Bronze Sand Castings
C95200
B148
88
Al 9, Fe 3 65 (450)
9% Aluminum Bronze Sand Castings
C95400
B148
85
Al 11, Fe 4 75 (515)
11% Aluminum Bronze Sand Castings
C95500
B148
81
Al 11, Fe 4, 90 (620)
Ni 4
D Nickel-Aluminum
Sand Castings
Bronze
C95800
B148
81
Al 9, Fe 4, 85 (585)
Ni 4,
Alpha Nickel-Aluminum Sand Castings
Mn 1
Bronze
C87300
B584
95
Si 3, Mn 1 45 (310)
Silicon Bronze
Sand Castings
C96400
B369
68
Ni 30, Fe 1, 60 (415)
Cb 1
70-30 Copper-Nickel
Cu-Ni Alloy Castings
Aloy B
Page 288
varies from that of the wrought alloys. The commercially pure copper
alloys are not commonly cast. Copper-base alloys are resistant to
attack in most industrial atmospheres. Even unalloyed copper can be
used in steam, freshwater, and seawater, except when velocities are
high. The corrosion resistance of all copper alloys is a function of
fluid velocity, since they all rely on the formation of a protective layer
[55]. Corrosive attack is generally accelerated by dissolved oxygen,
carbon dioxide, and/or ammonium ions. Ammonia causes stress
corrosion cracking in many copper-base alloys. This phenomenon is
commonly called ''season cracking." Copper-base alloys are resistant
to neutral and slightly alkaline solutions, dry gases, natural gas, and
most other hydrocarbons. They are attacked by hydrogen sulfide and
other sulfur compounds, most acids (especially oxidizing acids), and
strong alkalies.
Brass is a copper-base alloy containing zinc as the main alloying
element. Brass can also contain other alloying elements. Brass
generally has less corrosion resistance in aqueous solutions than the
other copper alloys, although zinc does improve the resistance to
sulfur compounds. Zinc also decreases the resistance to seasonal
cracking in ammonia. Alloys with more than 15% zinc may be
susceptible to a corrosion process called "dezincification." Zinc is
leached out of the surface, leaving a weak, porous copper structure.
Dezincification occurs in freshwater and is more likely to occur in
softened water with high carbon dioxide levels or in water containing
chlorides. High temperature, crevices, and corrosion deposits promote
dezincification. Phosphorus, arsenic, and antimony can be added to
copper to improve resistance to dezincification. Adding tin to brass
improves resistance to both corrosion and dezincification.
Bronze is a copper-base alloy that does not contain zinc or nickel as
the main alloying element. Copper-tin (tin bronze), copper-aluminum
(aluminum bronze), and copper-silicon (silicon bronze) are the main
cast bronze alloys. Tin bronze has good resistance to flowing seawater
and some nonoxidizing acids (except hydrochloric acid). Tin additions
of 810% provide good resistance to impingement attack. Tin bronze is
less susceptible to SCC than brass but has less resistance to corrosion
by sulfur compounds. Silicon bronze has about the same corrosion
resistance as copper, but better mechanical properties and superior
weldability. Actual corrosion rates are influenced less by oxygen and
carbon dioxide contents than with other copper alloys. It has better
resistance to SCC than the common brasses. Silicon bronze is
susceptible to embrittlement in high-pressure steam.
Aluminum bronzes with 512% aluminum have excellent resistance to
impingement corrosion and high-temperature oxidation. They are
resistant to many nonoxidizing acids. Oxidizing acids and metallic
salts will cause attack. Heat treatment is important in alloys with more
than 8% aluminum, since it affects both corrosion resistance and
toughness. Aluminum bronzes are suscepti-
Page 289
ble to SCC in moist ammonia. a-Aluminum bronzes containing no tin
are susceptible to SCC in steam when highly stressed [56].
Copper-nickel alloys (cupronickels) have the best resistance to
corrosion, impingement, and SCC of all copper alloys. They are
among the best alloys for seawater service and are immune to season
cracking.
X
Aluminum
There are many different grades of cast aluminum (Table 17). The
compositions are modified from the wrought grades to improve
casting properties, such as fluidity and hot-shortness, which can lead
to excessive cracking in the mold. Aluminum is a very reactive metal
with a high affinity for oxygen. When bare aluminum is exposed to air
or water, a dense, adherent, uniform aluminum oxide film is rapidly
formed. The film is an amorphous aluminum oxide (Al2O3) with a
thickness of 25 Å to a few hundred Å [57]. The film is relatively inert
and responsible for the good corrosion resistance of aluminum.
Table 17 Cast Aluminum Alloys
Minimum
strength,
ksi (MPa)
Alloy,
ASTM specification and Nominal
temper
form
composition
Si 3, Cu 4
A02080 ASTM B26 Sand
F Temper Castings
A03550 ASTM B26 Sand
Si 5, Cu 1,
T6 Temper Casting
Mg 0.5
Si 7, Mg 0.3
A03650 ASTM B26 Sand
T6 Temper Casting
A03650 ASTM B108 Permanent Si 7, Mg 0.3
T6 Temper Mold Casting
Tensile Yield
19
12
(131) (83)
32
20
(221) (138)
30
20
(205) (140)
33
22
(228) (152)
A03600 ASTM B85 Die Casting Si 9.5,
As-cast
Mg 0.5
A03800 ASTM B85 Die Casting Si 8.5, Cu 3.5
As-cast
A04130 ASTM B85 Die Casting Si 12
As-cast
A07130 ASTM B26 Sand
Cu 0.7, Mg
T5 Temper Casting
0.4, Zn 7.5
A07130 ASTM B108 Permanent Cu 0.7, Mg
T1 or T5 Mold Casting
0.4, Zn 7.5
Temper
aTypical properties for separately cast test bars.
44
(300)a
46
(320)a
43
(300)a
32
(220)
32
(221)
25
(170)a
23
(160)a
21
(140)a
22
(150)
22
(152)
Page 290
Several properties of aluminum can be substantially improved by
surface treatments. These treatments alter the surface film by chemical
and electrochemical methods. Paint adhesion, corrosion resistance,
and wear resistance can be improved. Other properties which can be
altered are surface conductivity, emissivity, plating adhesion, and
decorative appearance. All films are attacked by strong alkalis and
strong acids. Chemical conversion coatings use an aqueous solution of
chromate salts and/or other compounds to react with the base metal
forming a complex surface film about 200 Å thick. Corrosion
protection is provided by both the film and the inhibitive effect of the
chromate and other compounds. The film is soft and gelatinous when
freshly formed. It hardens as it dries, providing some abrasion
resistance.
Aluminum alloys are anodic to most other metals. Therefore, galvanic
corrosion is likely in all but the mildest environments. In industrial
atmospheres, salt spray, and other aggressive environments, aluminum
must be coated to prevent galvanic attack with other metals. Corrosion
resistance improves with increasing silicon and magnesium while
decreasing with increasing copper, iron, and other impurities. Lowcopper alloys provide the best corrosion resistance. In industrial and
marine environments, however, corrosion may be severe. Coatings are
also required for protection.
Alloy A03800 is a copper-silicon type of aluminum die-casting alloy
which is generally recognized as being superior to other aluminum
alloys in all-around performance. It has good castability,
machinability, and adequate corrosion resistance. Alloy A03600 is a
low-copper aluminum die-casting alloy which is used for applications
where increased resistance to corrosion in marine environments is
required. This material is used for die castings on offshore oil
platforms and other corrosive environments. High-quality coatings are
also required for these applications.
Aluminum alloys are not resistant to most mineral acids. Exceptions
are nitric acid above 82% and sulfuric acid above 98% [58, 59].
Aluminum is resistant to ammonium hydroxide but is rapidly attacked
by other alkaline solutions, such as sodium hydroxide and potassium
hydroxide. Aluminum is resistant to most neutral (pH 58.5) salt
solutions at ambient temperature; however, chlorides may cause some
localized attack. Heavy metal salts will attack aluminum. Aluminum
is resistant to most organic compounds and organic acids at ambient
temperature. However, formic-, oxalic-, and chloride-containing
organic acids are corrosive to aluminum.
XI
Magnesium
Magnesium is primarily used for automotive and aerospace
applications. Because magnesium is anodic to all other common
metals, galvanic corrosion with dissimilar materials can be a severe
problem [60]. Magnesium is commonly used
Page 291
in cathodic protection systems as a sacrificial anode. Recently
developed alloys with very low cobalt, copper, iron, and nickel levels
offer improved corrosion resistance. Table 18 lists some cast
magnesium alloys.
Magnesium resists corrosion in freshwater, hydrofluoric acid, pure
chromic acid, fatty acids, dilute alkalis, aliphatic and aromatic
hydrocarbons, pure halogenated organic compounds, dry fluorinated
hydrocarbons, and ethylene glycol solutions. Ambient temperature dry
gases, such as chlorine, iodine, bromine, and fluorine, do not attack
magnesium [61]. It is rapidly attacked by seawater, many salt
solutions, most mineral acids, methanol and ethanol, most wet gases,
and halogenated organic compounds when wet or hot. In coastal
atmospheres, the high-purity alloys, such as M11918, offer better
corrosion resistance than steel and aluminum.
XII
Zinc
About half of the world's zinc consumption is used as a coating on
steel to prevent corrosion in water and natural atmospheres. Cast zinc
is primarily used for small die castings stressed at low levels [62].
Because zine will creep at ambient temperature, the mechanical
properties listed in Table 19 cannot be used for design purposes. The
mechanical properties are listed for comparison between the different
grades only [59]. Zinc has good resistance to neutral pH water, but the
attack is greatly accelerated when oxygen or carbon dioxide are added
or the
Table 18 Cast Magnesium Alloys
Minimum
strength, ksi
(MPa)
Alloy
ASTM
Nominal
and
specification and composition
temper
form
Tensile Yield
16
M11918 ASTM B80 Sand Al 9, Zn 0.7, 34
T6
Casting
Mn 0.3, total (234) (110)
Temper
others 0.3 max
Al 6, Mn 0.2 32
19
M10602 ASTM B94
Die Casting
(220)a (130)a
F
Temper
Al 10, Mn 0.3 20
10 (69)
M10100 ASTM B199
Permanent Mold
(138)
F
Temper Casting
34
10 (69)
M11810 ASTM B403
Al 8, Zn 1,
Investment
(234)
T4
Mn 0.3
Temper Castings
aTypical properties.
Page 292
Table 19 Cast Zinc Alloys
ASTM
Specification
and grade
ASTM B86
Z33520
ASTM B86
Z33523
ASTM B86
Z35531
ASTM B86
Z35541
Typical properties
for comparison only,
not for design purposes
Tensile
strength,
ksi Elong.Hardness
Al
Mg
(MPa) (%) (HB)
3.54.30.020.05 41 (283) 14
82
Cu
0.25
max
0.25
3.54.30.0050.0241 (283)
max
0.751.253.54.30.030.08 47.7
(329)
2.53
3.54.30.020.05 52 (359)
14
76
7
91
7
100
temperature is increased above 120°F (49°C). If the pH is outside the
6.512.5 range, corrosion rates will become excessive.
Zine naturally forms a protective layer of zinc oxide and hydroxide
over a period of time. When the layer has completely formed,
corrosion rates decrease greatly. A white powder will often appear on
fresh zinc surfaces when stored in humid or damp conditions. This can
usually be prevented by dipping in a chromate corrosion inhibitor.
The life of zinc is limited in aggressive industrial environments. A
chromate treatment followed by a reliable paint system is
recommended. The corrosion resistance of the alloys listed in Table
19 are all very similar.
References
1. D. Bush, Technical Monograph 35, Standardizing Metallic
Designations, Fisher Controls, Marshalltown, IA, 1987.
2. Metals and Alloys in the Unified Numbering System, Society of
Automotive Engineers, Inc., Warrendale, PA, 1994.
3. Iron Castings Handbook, 3rd ed., Iron Castings Society, 1981, pp.
803, 506, 128, 410.
4. N.J. Davies, Modern Castings, 28th Census of World Castings
Production, 1993, pp. 3233.
5. J. R. West, Sulphur vs. construction materials, Chem. Eng. (Sept.
1951).
6. Engineering Properties and Applications of the Ni-Resists and
Ductile Ni-Resists, Inco Alloys International, New York, 1976.
7. ASM Handbook, Vol. 13, 9th ed., Metals Park, OH, p. 568.
8. Fontana and Greene, Corrosion Engineering, McGraw-Hill, New
York, 1967, p. 159.
Page 293
9. Steel Castings, Handbook, 5th Edition, Steel Founders Society of
America, Rocky River, OH, 1980, pp. 15.4415.46.
10. ASTM G101, Estimating the atmospheric corrosion resistance of
low-alloy steels, Metal Corrosion, Erosion and Wear, Vol. 3.2,
American Society for Testing and Materials, Philadelphia, 1993, pp.
408413.
11. C.G. Munger, Marine and Offshore Corrosion Control, Mat.
Perform., September 1993, pp. 3741.
12. R. Jones, Carbon and low alloy steels, Processing Industries
Corrosion, (Moniz and Pollock, eds.), NACE, Houston, 1986, p. 385.
13. Corrosion Data SurveyMetals Section, 6th ed., NACE, Houston,
1985.
14. D. Warren, Hydrogen effects on steel, Mat. Perform., January
1987, pp. 3848.
15. API Publication 941. Steels for Hydrogen Service at Elevated
Temperatures and Pressures in Petroleum Refineries and
Petrochemical Plants, American Petroleum Institute, Washington,
DC, 1983.
16. MR0175, Sulfide Stress Cracking Resistant Metallic Materials for
Oilfield Equipment, NACE, Houston.
17. RP0472, Methods and Controls to Prevent In-Service Cracking of
Carbon Steel Welds in Corrosive Petroleum Refining Environments,
NACE, Houston.
18. RP0590, Recommended Practices for Prevention, Detection and
Correction of Deaerator Cracking, NACE, Houston.
19. 8X194, Materials and Fabrication Practices for New Pressure
Vessels Used in Wet H2S Refinery Service, NACE Houston.
20. Avesta Sheffield Corrosion Handbook, Avesta Sheffield AB,
Stockholm, 1994.
21. ASM Handbook, Vol 3, 9th ed., Metals Park, OH, pp. 7893.
22. W.H. Herrnstein, Structure and constitution of cast ironchromium-nickel alloys, Handbook of Stainless Steels (D. Peckner
and I.M. Bernstein, eds.), McGraw-Hill, New York, 1977, pp. 1012.
23. A.H. Tuthill, Practical guide for procurement of quality stainless
steel castings, Mat. Perform., May 1990, pp. 5560.
24. High-tech steels to the rescue, Chem. Eng. (W. Grinthal, ed.),
January 1992.
25. AL-6XN Alloy, Allegheny Ludlum Corp., Pittsburgh, 1991.
26. J.L. Gossett, unpublished work, Fisher Controls, Inc.,
Marshalltown, IA, 1994.
27. B. Wallén, Sewater resistance of a high molybdenum stainless
steel, Proceedings of the Second BSE-NACE Corrosion Conference,
Bahrain, 1981, pp. 140151.
28. C. McCaul, Evaluation of intergranular corrosion susceptibility in
an as-welded high alloy austenitic stainless steel casting, Br.
Corrosion J., December 1991.
29. J.L. Gossett, New and improved, high nickel alloy castings, Paper
322, Corrosion, 88 Conference, National Association of Corrosion
Engineers, Houston, 1988.
30. R.M. Davison and J.D. Redmond, Practical guide to using 6 Mo
austenitic stainless steels, Mat. Perform., December 1988, pp. 3943.
31. AWS A5.11, Specification for Nickel and Nickel Alloy Electrodes
for Shielded Metal Arc Welding, American Welding Society, Miami.
32. AWS A5.14, Specification for Nickel and Nickel Alloy Bare
Welding Electrodes and Rods, American Welding Society, Miami.
33. Product Data, Bulletin No. S-6d, Armco 17-4PH, Armco, Inc.,
Middletown, OH, 1982.
Page 294
34. J.C.M. Farrar, Welding castings of duplex stainless steel, Weld.
Design Fabr., November 1992, pp. 4851.
35. F. Dupoiron, L. Renaud, M. Verneau, and J. Charles, Industrial
Applications and Experience of Duplex and Superduplex Stainless
Steels in Chemical Industries, Paper 386, Proceedings of Corrosion
94, NACE, Houston, 1994.
36. R.M. Davison, H.E. Deverell, and J.D. Redmond, Ferritic and
duplex stainless steels, Processing Industries Corrosion (Moniz and
Pollock, eds.), NACE, Houston, 1986, pp. 427443.
37. NF A 32-055, Weldable Cast Steels for Boilers and Pressure
Vessels, Association Francaise de Normalisation, Paris, 1985.
38. R. Francis, The role of duplex stainless steels for downhole
tubulars, Paper 7319 OTC, Houston, 1993.
39. J.L. Gossett, Two improved high nickel alloy castings: CW2M
and AOD refined CN7M, Mat. Perform., December 1988, pp. 4447.
40. J.L. Gossett, Specifications for obtaining high quality, high nickel
alloy castings: CW2M and AOD Refined CN7M, Mat. Perform.,
January 1989, pp. 6466.
41. J.L. Gossett, Reliable base for high nickel equipment, Chem.
Engin., November 1989, pp. 145148.
42. Nickel Alloys Bulletin, Inco Alloys, Huntington, WV, 1972, pp.
712.
43. E.C. Hoxie, Nickel and nickel-base alloys, Process Industries
Corrosion (Moniz and Pollock, eds.), NACE, Houston, 1986, pp.
463464.
44. Inconel 600 Bulletin, Inco Alloys, Huntington, WV, 1973, pp.
1418.
45. F.G. Hodge, Hastelloy C-4C, an improved nickel base casting
alloy for the CPI, presented London, England, March 1981.
46. F.G. Hodge, Cast alloy resists corrosion by hot acids and
oxidizing agents, Industrial Res. Dev., July 1983.
47. Corrosion Resistance of Hastelloy Alloys, Haynes International,
Kokomo, IN, 1978.
48. J.L. Gossett, New and improved, high nickel alloy castings,
Proceedings of the BHRA Conference, Developments in Valves and
Actuators for Fluid Control, Manchester, England, March 1988, pp.
137154.
49. F.G. Hodge and R. W. Kirchner, An improved Ni-Mo alloy for
hydrochloric acid service, Mat. Perform., August 1976, pp. 4045.
50. E.E. Millaway and M.H. Klineman, Factors affecting water
content needed to passivate titanium in chlorine, Corrosion, April
1972, p. 88.
51. L.C. Covington and R.W. Schutz, Corrosion Resistance of
Titanium, Titanium Metals Corp. of America, Pittsburgh.
52. R.T. Webster and T.L. Yau, Corrosion, February 1986, pp. 1517.
53. Zircadyne Corrosion Properties, Teledyne Wah Chang Albany,
Albany, OR.
54. A. Cohen, Copper and copper-base alloys, Processing Industries
Corrosion (Moniz and Pollock, eds.), NACE, Houston, TX, 1986, p.
483.
55. D.T. Peters and K.J.A. Kundig, Selecting copper and copper
alloys, Adv. Mat. Processes, June 1994, pp. 2026.
56. Aluminum bronze alloys corrosion resistance guide, Publication
80, Copper Development Assoc., Orchard House, Herts, UK, 1982.
57. H.P. Godard, Plenary Lecture 1981, An insight into the corrosion
behavior of aluminum, Mat. Perform., July 1981.
Page 295
58. E.D. Verink and D.B. Bird, Designing with aluminum alloys for
various corrosive environments, Mat. Prot., Feb. 1967, pp. 2832.
59. R.B. Mears, Aluminum and aluminum alloys, The Corrosion
Handbook, (H.H. Uhlig, ed.), John Wiley and Sons, New York, 1948.
60. Fontana and Greene, Corrosion Engineering, McGraw-Hill, New
York, 1967, p. 32.
61. ASM Metals Handbook, Vol. 2, 9th ed., Metals Park, OH, 1979,
pp. 597602, 629637.
62. A.R.L. Chivers, Zinc and zinc alloys, Corrosion, Vol. 1,
Butterworth, Boston, 1976, pp. 4:1504:162.
Page 297
12
Mechanisms of Chemical Attack, Corrosion Resistance,
and Failure of Plastic Materials
Philip A. Schweitzer
Fallston, Maryland
Celluloid, developed by Hyatt in 1868, was the forerunner of plastics.
It was followed by the phenol-formaldehyde resins (Bakelite) in 1907,
nylon in 1938, the polyesters and polyethylene in 1942, the epoxies
and acrylonitrile-butadienestyrene (ABS) in 1947. During the 1940s
these materials found limited usage, but expanded to a much larger
scale usage during the 1950s.
The chemical industry has made use of plastics to perform many
difficult assignments. In many situations the performance of 316ss or
high-nickel alloys can be matched or exceeded at a greatly reduced
cost. Occasionally, problems have occurred. In most cases the
problem is the result of improper selection of the plastic to be used or
in not following the recommended installation procedures. Lack of
knowledge with the consequent nonapplication of proper engineering
techniques is the most general cause of problems.
Plastics are composed primarily of carbon, hydrogen, chlorine,
fluorine, oxygen, silicon, and nitrogen in various combinations. These
elements are held together with single, double, and triple bonds in
both aliphatic and aromatic structures. A wide variation of properties
result from the various combinations and arrangements.
The properties of plastics can be modified with the addition of fibers,
fillers, ultraviolet light stabilizers, colored pigments, and flame
retardants. Polymer blends are also used to produce specific
properties.
Page 298
When a plastic is filled and reinforced it is referred to as a composite
or laminate. Laminates are produced using multiple layers of fiberreinforced resin or layers of different polymers. The latter
construction is also referred to as a dual laminate and is used to take
advantage of specific properties of each polymer.
Plastics are divided into two groups: thermoplasts and thermosets. The
thermoplasts can be melted, cooled, and remelted without destroying
the physical or mechanical properties of the polymer. The thermosets
begin as a powder or liquid which is reacted with a second material or
which through catalyzed polymerization results in a new product
which has characteristics differing from those of either starting
material.
I
Thermoplastics
A general guide to the relative corrosion resistance as exhibited
between the various polymers can be gotten from the periodic table,
specifically from the halogen category. The elements included are
fluorine, chlorine, bromine, and iodine. Since these elements are the
most electronegative in the periodic table, they are the most likely to
form a stable structure by attracting an electron from another element.
Of the halogens, fluorine is the most electronegative; therefore it
combines well with hydrogen and carbon atoms but not with itself.
The carbonfluorine is responsible for the important properties of
polyvinylidene fluoride (PVDF). In addition to the strength of the
bond between the carbon and fluorine, the fluorine also acts as a
protective shield for other, less strong bonds within the polymer chain.
Carbonhydrogen bonds are considerably weaker and carbonchlorine
bonds weaker yet. Plastics such as polypropylene and polyethylene
are composed of the carbonhydrogen bond whereas polyvinyl chloride
contains carbonchlorine bonds.
The chemical resistance of a fully fluorinated plastic such as
polytetrafluoroethylene (PTFE) provides a greater range of chemical
resistance than does plastics with carbonhydrogen or carbonchlorine
bonds.
A
Corrosion of Plastics
Plastics are attacked by chemical reaction or by solvation and do not
experience a specific corrosion rate as do metals. A plastic material is
either completely resistant or it deteriorates rapidly. Chemical attack
of a plastic material can take place in one or more of the following
ways:
1. Disintegration or degradation of a physical nature due to
absorption, permeation, solvent action, or other factors
2. Oxidation, where chemical bonds are attacked
3. Hydrolysis, where ester linkages are attacked
Page 299
4. Radiation
5. Thermal degradation, involving depolymerization and possible
repolymerization
6. Dehydration (relatively uncommon)
Results of such attacks will appear in the form of softening, charring,
crazing, blistering, embrittlement, discoloration, dissolution, or
swelling.
B
Properties of Plastics
As each specific plastic material is discussed, the physical and
mechanical properties will be presented. Table 1 gives the
abbreviations used for the various thermoplasts to be discussed. A
brief explanation will be given for certain physical/mechanical
properties.
1
Deflection Temperatures
The heat deflection (distortion) (HDT) test is one in which a bar of the
plastic is uniformly heated in a closed chamber while a load of 66 psi
or 264 psi is placed in the center of the horizontal bar. The HDT
indicates how much mass (weight) the object must be constructed of
to maintain the desired form, stability, and
Table 1 Abbreviations Used for Plastics
ABS
Acrylonitrile-butadienestyrene
CPVC
Chlorinated polyvinyl
chloride
ECTFE Ethylenechlorotrifluorethylene
FEP
Perfluoroethylenepropylene
HDPE
High-density polyethylene
PEEK
Polyetheretherketone
PES
Polyethersulfone
PFA
Perfluoralkoxy
PA
Polyamide
PB
Polybutylene
PC
Polycarbonate
PF
Phenol formaldehyde
PP
Polypropylene
PPS
Polyphenyl sulfide
PTFE
Polytetrafluoroethylene
PVC
Polyvinyl chloride
PVDC
Polyvinylidene chloride
PVDF
Polyvinylidene fluoride
UHMWPEUltrahigh molecular weight
polyethylene
Page 300
strength rating and provides a measure of the rigidity of the polymer
under load as well as temperature. Table 2 lists the HDT of the more
common plastics.
2
Tensile Strength of Plastics
Tensile strength is a measure of the stress required to deform a
material prior to breakage. It is calculated by dividing the maximum
load applied to the material before its breaking point by the original
cross-sectional area of the test piece. This is in contrast to toughness,
which is the measure of energy required to break a material.
Tensile strength alone would not be used to determine the ability of a
plastic to resist deformation and retain form. Other mechanical
properties such as elasticity, ductility, creep resistance, hardness, and
toughness must also be taken into account. Table 3 lists the tensile
strength of plastics.
3
Fire Hazards
The table of physical and mechanical properties located under the
heading of each thermoplast contains two entries relating to the
potential fire hazard.
The first is the limiting oxygen index percent. This is a measure of the
minimum oxygen level required to support combustion of the
thermoplast. The second is the flame spread classification. These
ratings are based on common tests as outlined by the Underwriters
Laboratories and are defined as follows:
Table 2 Heat Distortion Temperature
of the Common Plastics (°F/°C)
Pressure (psi)
Polymer
66
264 Melt point
PTFE
250/121132/56 620/327
PVC
135/57 140/60 285/141
LDPE
104/40 221/105
UHMW PE155/68 110/43 265/129
PP
225/107120/49 330/160
PFA
164/73 118/48 590/310
FEP
158/70 124/51 554/290
PVDF
248/148235/113352/178
ECTFE
240/116170/77 464/240
ETFE
220/104165/74 518/270
PEEK
320/160644/340
PES
410/210
PC
280/138265/129
Page 301
Flame spread
rating
025
2550
5075
75200
Over 200
Classification
Noncombustible
Fire-retardant
Slow-burning
Combustible
Highly
combustible
As the physical and mechanical properties vary between each
thermoplast, so do the maximum allowable operating temperatures.
Table 4 lists the maximum allowable operating temperatures of the
common thermoplasts.
C
Polyvinyl Chloride
Polyvinyl chloride (PVC) is the most widely used of the thermoplasts.
It is polymerized vinyl chloride, which is produced from acetylene
and anhydrous hydrochloric acid. The structure is as follows:
PVC is stronger and more rigid than other thermoplastic materials. It
has a
Table 3 Tensile
strength of Plastics at
73°F (25°C) at Break
Plastic Strength (psi)
PVDF 8000
ETFE 6500
PFA
40004300
ECTFE 7000
PTFE
FEP
PVC
PE
PP
UHMW
PE
PEEK
PES
PC
25006000
27003100
60007500
12004550
45006000
5600
13,20023,800
12,20020,300
10,000
Page 302
Table 4 Maximum
Operating Temperature of
the Common Thermoplasts
Thermoplast
°F/°C
PVC
140/160
CPVC
180/82
HMW PE 140/60
UHMW PE 180220/82104
ABS
140/60
PP
180220/82104
PB
220/104
ECTFE
300/149
ETFE
300/149
FEP
375/190
PEEK
480/250
PES
390/200
PFA
500/260
PA
212250/100121
PC
210265/99128
PPS
450/230
PTFE
450/230
PVDF
320/160
high tensile strength and modulus of elasticity. Additives are used to
further specific end-uses, such as thermal stabilizers, lubricity, impact
modifiers, and pigmentation.
Two types of PVC are produced: normal impact (type 1) and high
impact (type 2). Type 1 is a rigid unplasticized PVC having normal
impact with optimum chemical resistance. Type 2 has optimum
impact resistance and reduced chemical resistance. Table 5 lists the
physical and mechanical properties of PVC.
Type 1 (unplasticized PVC) resists attack by most acids and strong
alkalies, gasoline, kerosene, aliphatic alcohols, and hydrocarbons. It is
particularly useful in the handling of hydrochloric acid.
The chemical resistance of type 2 PVC to oxidizing and highly
alkaline mediums is reduced.
PVC may be attacked by aromatics, chlorinated organic compounds,
and lacquer solvents. Refer to Table 21 for the compatibility of PVC
with selected corrodents.
Page 303
Table 5 Physical and Mechanical Properties of PVC
Property
Type Type
1
2
Specific gravity
1.45 1.38
Water absorption 24 hr at 73°F/23°C, % 0.04 0.05
Tensile strength at 73°F/23°C, psi
6800 5500
Modulus of elasticity in tension at
5.0
4.2
73°F/23°C × 105 psi
Compressive strength, psi
10,0007900
Flexural strength, psi
14,00011,000
Izod impact strength, notched at
0.88 12.15
73°F/23°C, ft-lb/in.
Coefficient of thermal expansion, in./in. 4.0
6.0
°F × 10-5
Thermal conductivity, Btu/hr/ft2/°F/in. 1.33 1.62
Heat distortion temperature, °F/°C
130/54135/57
at 66 psi
155/68169/71
at 264 psi
Limiting oxygen index, %
43
Flame spread
1520
Underwriters laboratory rating Sub 94
94 VO
D
Chlorinated Polyvinyl Chloride
When acetylene and hydrochloric acid are reacted to produce
polyvinyl chloride the chlorination is approximately 56.8%. Further
chlorination of the PVC to approximately 67% produces chlorinated
PVC (CPVC) whose chemical structure is as follows:
The additional chlorine increases the heat deflection temperature and
permits a higher allowable operating temperature. While PVC is
limited to a maximum operating temperature of 140°F (60°C), CPVC
has a maximum operating temperature of 180°F (82°C). Because of
the higher operating temperature, CPVC finds application as piping
for condensate return lines in areas having corrosive external
conditions. It has also found application in hot water piping. The
physical and mechanical properties are given in Table 6.
The corrosion resistance of CPVC is similar to that of PVC but not
identical. There are some differences. CPVC can be used to handle
most acids, alkalies, salts, halogens, and many corrosive wastes. In
general, it cannot be used in contact with most polar organic materials
including chlorinated or aromatic
Page 304
Table 6 Physical and Mechanical Properties of
CPVC
Specific gravity
1.55
Water absorption 24 hr at 73°F/23°C, % 0.003
Tensile strength at 73°F/23°C, psi
8000
Modulus of elasticity in tension at
4.15
73°F/23°C × 105 psi
Compressive strength at 73°F/23°C, psi 9000
Flexural strength, psi
15,100
Izod impact strength at 73°F/23°C, ft- 1.5
lb/in.
Coefficient of thermal expansion in./in. 3.4
°F × 10-5
Thermal conductivity, Btu/hr/ft2/°F/in. 0.95
Heat distortion temperature, °F/°C
238/114
at 66 psi
212/100
at 264 psi
Limiting oxygen index, %
60
Flame spread
15
Underwriters lab rating, UL 94
VO; SVA5VB
hydrocarbons, esters, and ketones. Refer to Table 21 for the
compatibility of CPVC with selected corrodents.
E
Polypropylene
Polypropylene (PP) is one of the most common and versatile
thermoplastics. It is closely related to polyethylene, both of which are
members of a group known as ''polyolefins." The polyolefins are
composed of only carbon and hydrogen. When unmodified, PP is the
lightest of the common thermoplastics, having a specific gravity of
0.91. In addition to its light weight it has the advantages of high heat
resistance, stiffness, and a wide range of chemical resistance. Within
the chemical structure of PP a distinction is made between isotactic
PP and atactic PP, with the isotactic form accounting for 97% of the
PP. This form is highly ordered:
Atactic PP is a viscous liquid-type PP having a PP polymer matrix.
Polypropylene can be produced either as a homopolymer or as a
copolymer with polyethylene. The copolymer is less brittle than the
homopolymer and is able
Page 305
Table 7 Physical and Mechanical Properties of
Copolymer and Homopolymer PP
Homo- CoProperty
polymerpolymer
Specific gravity
0.905 0.91
Water absorption, 24 hr at 73°F/23°C, 0.02
0.03
%
Tensile strength at 73°F/23°C, psi
5000 4000
Modulus of elasticity in tension at
1.7
1.5
73°F/23°C × 105 psi
Compressive strength, psi
9243 8500
Flexural strength, psi
7000
Izod impact strength, notched at
1.3
8
73°F/23°C, ft-lb/in.
Coefficient of thermal expansion in./in. 5.0
6.1
°F × 10-5
Thermal conductivity, Btu/hr/ft2/°F/in. 1.2
1.3
Heat distortion temperature °F/°C
220/107220/107
at 66 psi
140/60 124/49
at 264 psi
Limiting oxygen index, %
17
Flame spread
Slow burning
Underwriters lab rating Sub 94
94 HB
to withstand impact forces down to -20°F (-29°C), while the
homopolymer is extremely brittle below 40°F (4°C). The physical and
mechanical properties are shown in Table 7.
Although the copolymers have increased impact resistance, their
tensile strength and stiffness are considerably lower, increasing the
potential for distortion and cold flow particularly at elevated
temperatures.
The homopolymers, being long-chain, high molecular weight
molecules with a minimum of random orientation, have optimum
chemical, thermal, and physical properties. For this reason
homopolymer material is preferred for difficult chemical, thermal, and
physical conditions.
Polypropylene is subject to degradation by ultraviolet light. Therefore
if exposed to sunlight an ultraviolet absorber or screening agent must
be used to protect the material. It is not affected by most inorganic
chemicals, except the halogens and severe oxidizing conditions. PP
can be used with sulfur-bearing compounds, caustics, solvents, acids,
and other organic chemicals.
It should not be used with oxidizing-type acids, detergents, lowboiling hydrocarbons, alcohols, aromatics, and some chlorinated
organic materials. Refer to Table 21 for compatibility of
polypropylene with selected corrodents.
Page 306
F
Polyethylene
Polyethylene (PE) is produced in various types which differ in
molecular structure, crystallinity, molecular weight, and molecular
weight distribution. The basic structural formula is as follows:
PE is produced by polymerizing ethylene gas obtained from
petroleum hydrocarbons. Changes in the polymerizing conditions are
responsible for the various types of PE.
The terms low, medium, and high density refer to the ASTM
designation based on the unmodified PE. Low-density PE has a
specific gravity of 0.910.925; medium-density PE has a specific
gravity of 0.9260.940; and high-density PE has a specific gravity of
0.9410.959. The densities, being related to the molecular structure, are
indications of the properties of the final product.
The two grades of polyethylene primarily used for corrosion
resistance are high molecular weight (HMW) and ultrahigh molecular
weight (UHMW). The HMW material has an average molecular
weight of 200,000500,000 while the UHMW material has an average
molecular weight of at least 3.1 million. Table 8 gives the physical
and mechanical properties of UHMW polyethylene.
If exposed to ultraviolet radiation from the sun or another source,
photo-or light oxidation will occur. To prevent this it is necessary to
incorporate carbon black into the resin to stabilize it. Other stabilizers
will not provide complete protection. PE does not support biological
growth.
Polyethylene is resistant to a wide variety of acids, bases, inorganic
salts,
Table 8 Physical and Mechanical Properties of
EHMW PE
Specific gravity
0.940.96
Water absorption, 24 hr at 73°F/23°C, % <0.01
Tensile strength at 73°F/23°C, psi
31003500
Modules of elasticity in tension at
1.18
73°F/23°C × 105 psi
Flexural modulus, psi × 105
1.33
Izod impact strength, notched at
0.40.6
73°F/23°C, ft-lb/in.
Coefficient of thermal expansion in./in. 11.1
°F × 10-5
Thermal conductivity Btu/hr/ft2/°F/in. 0.269
Heat distortion temperature, °F/°C
155/66
at 66 psi
250/121
at 264 psi
Flame spread
Slow
burning
Page 307
and many fertilizer solutions. Refer to Table 21 for the compatibility
of HMW and UHMW polyethylene with selected corrodents.
G
Polyphenylene Sulfide
Polyphenylene sulfide is a thermoplastic capable of being used at high
temperatures. It has a maximum service rating of 450°F (230°C). As
the temperature increases there is a corresponding increase in
toughness. Table 9 provides the physical and mechanical properties of
PPS.
Polyphenylene sulfide offers excellent resistance to aqueous inorganic
salts and bases and many organic solvents. It can also be used under
highly oxidizing conditions. Relatively few materials react with PPS
at high temperatures. PPS is resistant to organic solvents except for
chlorinated solvents, some halogenated gases, and alkylamines. It
stress-cracks in chlorinated solvents. Weak and strong alkalies have
no effect on PPS. Polyphenylene sulfide is resistant to weak acids
with the exception of hydrochloric. Strong oxidizing aids such as
sulfuric, nitric, chromic, and 10% perchloric will attack PPS. Refer to
Table 21 for the compatibility of PPS (Ryton) with selected
corrodents.
H
Polycarbonate
Polycarbonate (PC) is also produced under the trade name Lexan (GE
Plastics). Because of this extremely high impact resistance and good
clarity it is widely used for windows in chemical equipment and
glazing in chemical plants. Its exceptional weatherability, corrosion
resistance, and high-impact strength render it highly useful in outdoor
energy management devices, network interfaces, electrical wiring
blocks, telephone equipment, and lighting diffusers, globes, and
housings. Table 10 lists the physical and mechanical properties of PC.
Table 9 Physical and Mechanical Properties of
PPS
Specific gravity
1.34
Water absorption, 24 hr at 73°F/23°C, 0.01
%
Tensile strength at 73°F/23°C, psi
10,800
Modulus of elasticity in tension at
4.86.3
73°F/23°C × 105 psi
Compressive strength, psi
16,000
Flexural strength, psi × 103
1120
Izod impact strength, notched at
0.03
73°F/23°C, ft-lb/in.
Coefficient of thermal expansion in./in.2.73.0
°F × 10-5
Thermal conductivity, Btu/hr/ft2/°F/in. 2.0
Heat distortion temperature, °F/°C at 275/135
264 psi
Limiting oxygen index, %
47
Underwriters lab rating Sub 94
SEO
Page 308
Table 10 Physical and Mechanical Properties
of PC
Specific gravity
1.2
Water absorption, 24 hr at
0.150.2
73°F/23°C, %
Tensile strength at 73°F/23°C, psi × 89.5
103
Modulus of elasticity at 73°F/23°C × 3.24.5
105 psi
Compressive strength, psi × 103
1014
Flexural strength, psi × 103
11.515
Izod impact strength, notched at
416
73°F/23°C, ft-lb/in.
Coefficient of thermal expansion
1.793.9
in/in. °F × 10-5
Thermal conductivity,
1.331.41
Btu/hr/ft2/°F/in.
Heat distortion temperature, °F/°C
285/140
at 66 psi
265/129
at 264 psi
Limiting oxygen index, %
2531.5
Underwriters lab rating, Sub 94
SEOSEI
Polycarbonate is resistant to weak acids and has limited resistance to
weak alkalies. It is resistant to most oils and greases. PC will be
attacked by strong alkalies and strong acids, and is soluble in ketones,
esters, aromatic and chlorinated hydrocarbons.
I
Polyetheretherketone
Polyetheretherketone (PEEK) is suitable for applications which
require mechanical strength with the need to resist difficult thermal
and chemical environments. It has a continuous maximum service
temperature of 480°F (260°C) with excellent mechanical properties
retained to temperatures over 570°F (300°C). Table 11 gives the
physical and mechanical properties of PEEK.
PEEK is insoluble in all common solvents and has excellent resistance
to a wide range of organic and inorganic liquids. Refer to Table 21 for
the compatibility of PEEK with selected corrodents.
J
Polyethersulfone
Polyethersulfone (PES) has a continuous maximum operating
temperature of 390°F (200°C). At room temperature PES is tough,
rigid, and strong with outstanding long-term load bearing properties.
Most of these properties are retained at the maximum operating
temperature. Table 12 lists the physical and mechanical properties of
PES.
PES has excellent resistance to aliphatic hydrocarbons, some
chlorinated hydrocarbons, and aromatics. It is also resistant to most
inorganic chemi-
Page 309
Table 11 Physical and Mechanical Properties of
PEEK
Specific gravity
1.32
Water absorption, 24 hr at 73°F/23°C, 0.5
%
Tensile strength at 73°F/23°C, psi
14,500
Modulus of elasticity in tension at
4.9
73°F/23°C × 105 psi
Compressive strength, psi
17,100
Flexural strength, psi
24,650
Izod impact strength, notched at
1.57
73°F/23°C, ft-lb/in.
Coefficient of thermal expansion in./in.
°F × 10-5
2.6
at 0290°F
6.1
at 290500°F
Thermal conductivity, Btu/hr/ft2/°F/in. 1.75
Heat distortion temperature, °F/°C
320/160
at 264 psi
Limiting oxygen index, %
24
Underwriters lab rating, Sub UL 94
V-O
(1.45)
cals. Hydrocarbons and mineral oils, greases, and transmission fluids
have no effect on PES.
PES will be attacked by strong oxidizing acids, but glass
fiberreinforced grades are resistant to more dilute acids. It is soluble in
highly polar solvents and is subject to stress cracking in ketones and
esters.
PES is not resistant to outdoor weathering and is not recommended
for outdoor applications unless stabilized by incorporating carbon
black or unless painted.
Refer to Table 21 for the compatibility of PES with selected
corrodents.
Table 12 Physical and Mechanical Properties
of PES
Specific gravity
1.51
Tensile strength at 68°F/20°C, psi
12,200
Flexural strength, psi
18,700
Izod impact strength, notched at
1.57
73°F/23°C, ft-lb/in.
Coefficient of thermal expansion
5.5
in./in. °F × 10-5
Thermal conductivity,
1.25
Btu/hr/ft2/°F/in.
Heat distortion temperature, °F/°C at 397/203
264 psi
Limiting oxygen index, %
34
Underwriters lab rating, Sub UL 94 V-O at
0.46
Page 310
K
Polyvinylidene Fluoride
Polyvinylidene fluoride (PVDF) is a crystalline, high molecular
weight polymer containing 59% fluorine. It is similar in chemical
structure to PTFE except that it is not fully fluorinated. The chemical
structure is as follows:
Much of the strength and chemical resistance of PVDF is maintained
through an operating range of -40 to 320°F (-40 to 160°C). It has high
tensile strength and heat deflection temperature and is resistant to the
permeation of gases. Approval has been granted by the Food and Drug
Administration for repeated use in contact with food, as in food
handling and processing equipment. The physical and mechanical
properties are given in Table 13.
PVDF is chemically resistant to most acids, bases, and organic
solvents. It is also resistant to wet or dry chlorine, bromine, and other
halogens.
It should not be used with strong alkalies, fuming acids, polar
solvents, amines, ketones, and esters. When used with strong alkalies
it stress-cracks. Refer to Table 21 for the compatibility of PVDF with
selected corrodents. PVDF is manufactured under the trade names
Kynar (Elf Atochem), Solef (Solvay), Hylar (Ausimont USA), and
Super Pro 230 and ISO (Asahi/America).
Table 13 Physical and Mechanical Properties
of PVDF
Specific gravity
1.76
Water absorption, 24 hr at 73°F/23°C, <0.04
%
Tensile strength at 73°F/23°C, psi
6000
Modulus of elasticity in tension at
2.1
73°F/23°C × 105 psi
Compressive strength, psi
11,600
Flexural modulus, psi
10,750
Izod impact strength, notched at
3.8
73°F/23°C, ft-lb/in.
Coefficient of thermal expansion in./in.7.9
°F × 10-5
Thermal conductivity Btu/hr/ft2/°F/in. 0.79
Heat distortion temperature, °F/°C
284/140
at 66 psi
194/90
at 264 psi
Limiting oxygen index, %
44
Flame spread
0
Underwriters lab rating, Sub 94
94 V-O
Page 311
L
Ethylene Chlorotrifluoroethylene
Ethylene Chlorotrifluoroethylene (ECTFE) is a 1:1 alternating
copolymer of ethylene and Chlorotrifluoroethylene. The chemical
structure is as follows:
This chemical structure provides the polymer with a unique
combination of properties. It possesses excellent chemical resistance,
a broad-use temperature range from cryogenic to 340°F (171°C) with
continuous service to 300°F (149°C), and has excellent abrasion
resistance.
ECTFE exhibits excellent impact strength over its entire operating
range, even in the cryogenic range. It also possesses good tensile,
flexural, and wear-related properties. ECTFE is also one of the most
radiation-resistant polymers. Other important properties include a low
coefficient of friction and the ability to be pigmented. Table 14 lists
the physical and mechanical properties of ECTFE.
ECTFE is resistant to strong mineral and oxidizing acids, alkalies,
metal etchants, liquid oxygen, and practically all organic solvents
except hot amines (aniline, dimethylamine, etc.). ECTFE is not
subject to chemically induced stress
Table 14 Physical and Mechanical Properties of
ECTFE
Specific gravity
1.68
Water absorption, 24 hr at 73°F/23°C, % <0.01
Tensile strength at 73°F/23°/C, psi
4500
Modulus of elasticity in tension at
2.4
73°F/23°C × 105 psi
Flexural strength, psi
Izod impact strength, notched at
73°F/23°C, ft-lb/in.
Linear coefficient of thermal expansion,
in./in. °F at:
-22 to 122°F/-30 to 50°C
7000
No
break
4.4 ×
10-5
5.6 ×
122 to 185°F/50 to 80°C
10-5
7.5 ×
185 to 257°F/85 to 125°C
10-5
9.2 ×
257 to 356°F/125 to 180°C
10-5
Thermal conductivity, Btu/hr/ft2/°F/in. 1.07
Heat distortion temperature, °F/°C
195/91
at 66 psi
151/66
at 264 psi
Limiting oxygen index, %
60
Underwriters lab rating, Sub 94
V-O
Page 312
cracking from strong acids, bases, or solvents. Some halogenated
solvents can cause ECTFE to become slightly plasticized when it
comes into contact with them. Under normal circumstances this does
not affect the usefulness of the polymer since upon removal of the
solvent from contact and upon drying its mechanical properties return
to their original values, indicating that no chemical attack has taken
place. Like other fluoropolymers, ECTFE will be attacked by metallic
sodium and potassium. Table 21 lists the compatibility of ECTFE
with selected corrodents.
M
Ethylene Tetrafluoroethylene
Sold under the trade name of Tefzel by DuPont, ethylene
tetrafluoroethylene (ETFE) is a partially fluorinated copolymer of
ethylene and tetrafluoroethylene. It has a maximum service
temperature of 300°F (149°C). The physical and mechanical
properties are given in Table 15. ETFE is fairly inert to strong mineral
acids, halogens, inorganic bases, and metal salt solutions. Under most
conditions ETFE is resistant to alcohols, ketones, ethers, and
chlorinated hydrocarbons. It is recommended that a test be conducted
before using ETFE with these materials.
Strong oxidizers such as nitric acid, organic bases such as amine, and
sulfonic acids will attack ETFE. Refer to Table 21 for the
compatibility of ETFE with selected corrodents.
Table 15 Physical and Mechanical
Properties of ETFE
Specific gravity
1.70
Tensile strength, psi
6500
Modulus of elasticity, psi × 2.17
105
Elongation, %
300
Flexural modulus, psi × 105 1.7
Impact strength, ft-lb/in.
No
break
Hardness, Shore D
67
Water absorption, 24 hr at
<0.03
73°F/23°C, %
Thermal conductivity,
1.6
Btu/hr/ft2/°F/in.
Heat distortion temperature,
°F/°C
220/104
at 66 psi
160/71
at 264 psi
Limiting oxygen index, %
30
Underwriters lab rating, Sub V-O
94
Page 313
N
Polytetrafluoroethylene
Polytetrafluoroethylene (PTFE) is a fully fluorinated thermoplastic
having the following formula:
It has an operating temperature range of -20°F (-29°C) to 450°F
(232°C). This temperature range is based on the physical/mechanical
properties of PTFE. When handling certain aggressive chemicals it
may be necessary to reduce the upper temperature limit. PTFE is a
relatively weak material and tends to creep under stress at elevated
temperatures. The physical and mechanical properties are given in
Table 16.
PTFE is unique in its corrosion resistance properties. It is chemically
inert in the presence of most materials. There are very few chemicals
that will attack PTFE within normal use temperatures. Materials
which will attack PTFE are the most violent oxidizers and reducing
agents known. Elemental sodium removes fluorine from the polymer
molecule. The other alkali metals (potassium, lithium, etc. ) react in a
similar manner.
Fluorine and related compounds (e.g., chlorine trifluoride) are
absorbed into the PTFE resin with such intimate contact that the
mixture becomes sensitive to a source of ignition, such as impact.
These potent oxidizers should only be handled with great care and
with recognition of the potential hazards.
The handling of 80% sodium hydroxide, aluminum chloride,
ammonia, and certain amines at high temperatures has the same effect
as handling of elemental
Table 16 Physical and Mechanical Properties
of PTFE
Specific gravity
2.132.2
Water absorption, 24 hr at
0.01
73°F/23°C, %
Tensile strength at 73°F/23°C, psi 20006500
Compressive strength, psi
1700
Flexural strength, psi
No break
Flexural modulus, psi × 105
0.71.1
Izod impact strength, notched at
3
73°F/23°C, ft-lb/in.
Coefficient of thermal expansion, 5.5
in./in. °F × 10-5
Heat distortion temperature at 66 250/121
psi, °F/°C
Low-temperature embrittlement, °F/ -450/-268
°C
Page 314
sodium. Slow oxidative attack can be produced by 70% nitric acid
under pressure at 480°F (250°C).
Table 21 provides the compatibility of PTFE with selected corrodents.
O
Fluorinated Ethylene Propylene
Fluorinated ethylene propylene is a fully fluorinated thermoplastic
with some branching but consists mainly of linear chains having the
following formula:
FEP has a maximum operating temperature of 375°F (190°C). After
prolonged exposure at 400°F (204°C) it exhibits changes in physical
strength. It is relatively soft plastic with lower tensile strength, wear
resistance, and creep resistance than other plastics. It is insensitive to
notched impact forces and has excellent permeation resistance except
to some chlorinated hydrocarbons. Table 17 lists its physical and
mechanical properties.
FEP basically exhibits the same corrosion resistance as PTFE, with a
few exceptions, but at lower operating temperatures. It is resistant to
practically all chemicals, the exceptions being extremely potent
oxidizers, such as chlorine trifluoride and related compounds. Some
chemicals will attack FEP when present
Table 17 Physical and Mechanical Properties of
FEP
Specific gravity
2.15
Water absorption, 24 hr at
<0.01
73°F/23°C, %
Tensile strength at 73°F/23°C, psi
Modulus of elasticity in tension at
73°F/23°C × 105 psi
Compressive strength, psi
Flexural strength, psi
Izod impact strength, notched at
73°F/23°C, ft-lb/in.
Coefficient of thermal expansion,
in./in. °F × 10-5
Thermal conductivity,
Btu/hr/ft2/°F/in.
Heat distortion temperature, °F/°C
at 66 psi
at 264 psi
Limiting oxygen index, %
Flame spread
27003100
0.9
16,000
3000
No break
8.310.5
0.11
158/70
129/54
95
Nonflammable
Page 315
in high concentrations at or near the service temperature limit. Table
21 lists the compatibility of FEP with selected corrodents.
P
Perfluoralkoxy
Perfluoralkoxy (PFA) is a fully fluorinated polymer having the
following formula:
Perfluoralkoxy lacks the physical strength of PTFE at elevated
temperatures, but it has somewhat better physical and mechanical
properties than FEP above 300°F (149°C) and can be used at
temperatures up to 500°F (260°C). For example, PFA has reasonable
tensile strength at 69°F (20°C), but its heat deflection temperature is
the lowest of all the fluoroplastics. While PFA matches the hardness
and impact strength of PTFE, it sustains only one quarter of the life of
PTFE in flexibility tests. Refer to Table 18 for the physical and
mechanical properties of PFA.
Like PTFE, PFA is subject to permeation by certain gases and will
absorb selected chemicals. Perfluoralkoxy also performs well at
cryogenic temperatures. Table 19 compares the mechanical properties
of PFA at room temperatures and cryogenic temperatures.
PFA is inert to strong mineral acids, inorganic bases, inorganic
oxidizers, aromatics, some aliphatic hydrocarbons, alcohols,
aldehydes, ketones, ethers, esters, chlorocarbons, fluorocarbons, and
mixtures of those mentioned.
PFA will be attacked by certain halogenated complexes containing
fluorine. This includes chlorine trifluoride, bromine trifluoride, iodine
pentafluoride, and fluorine. It can also be attacked by such metals as
sodium or potassium, particularly in their molten state. Refer to Table
21 for the compatibility of PFA with selected corrodents.
Q
Polyamides
The polyamides (PAs) are more commonly known as nylons. They
are linear molecules with a high degree of crystallinity and have the
following formulas:
Page 316
Table 18 Physical and Mechanical Properties
of PFA
Specific gravity
2.122.17
Water absorption, 24 hr at
<0.03
73°F/23°C, %
Tensile strength:
4000
at 73°F/23°C psi
2000
at 482°F/250°C psi
Modulus of elasticity in tension, psi:
40,000
at 73°C/23°C
6000
at 482°F/250°C
Compressive strength, psi
3500
at 73°C/23°C
60,000
at-320°F/-196°C
Flexural modulus, psi
90,000
at 73°F/23°C
10,000
at 482°F/250°C
Izod impact, notched at 73°F/23°C, No
ft-lb/in.
break
Coefficient of linear thermal
expansion, in./in. °F
7.8 ×
at 70212°F/20100°C
10-5
9.8 ×
at 212300°F/100150°C
10-5
12.1 ×
at 300480°F/150210°C
10-5
Heat distortion temperature, °F/°C:
164/73
at 66 psi
118/48
at 264 psi
Limiting oxygen index, %
<95
Flame spread
10
Underwriters lab rating, Sub 94
94-V-O
Table 19 Comparison of Mechanical Properties of
PFA at Room Temperature and Cryogenic
Temperatures
Temperature
Property
73°F/23°C-320°F/-190°C
Yield strength, psi
2100
No yield
Ultimate tensile
2600
18,700
strength, psi
Elongation, %
260
8
Flexural modulus, psi 81,000
840,000
Izod impact strength, No break 12
notched, ft-lb/in.
Compressive strength, 3500
60,000
psi
Compressive strain, % 20
35
Modulus of elasticity, 10,000
680,000
psi
Page 317
Table 20 Physical and Mechanical Properties of PA
Specific gravity
1.011.17
Water absorption, 24 hr at 73°F/23°C, 0.41.8
%
Tensile strength at 73°F/23°C, psi × 8.312.5
103
Modulus of elasticity in tension at
217
73°F/23°C × 103 psi
Compressive strength, psi × 103
9.712.5
Flexural strength, psi × 103
12.514
Izod impact strength, notched at
0.53.3
73°F/23°C, ft-lb/in.
Coefficient of thermal expansion,
4.55
in./in. °F × 10-5
Thermal conductivity, Btu/hr/ft2/°F/in. 1.21.7
Heat distortion temperature, °F/°C
360/182
at 66 psi Nylon 6
470/243
Nylon 6/6
302/150
Nylon 11
155160/6871
at 264 psi Nylon 6
220/104
Nylon 6/6
131/55
Nylon 11
Polyamides are high-strength thermoplasts. Their average physical
and mechanical properties are shown in Table 20.
The nylons are resistant to weak acids, strong and weak alkalies, and
most common solvents, hydrocarbons, esters, and ketones. They will
be attacked by strong acids. Refer to Table 21 for the compatibility of
PA with selected corrodents.
R
Polybutylene
Polybutylene (PB) is a member of the polyolefin family. The
combination of stress cracking resistance, chemical resistance,
strength, and abrasion resistance makes it a very useful material.
PB is resistant to acids, bases, soaps, and detergents at temperatures
up to 200°F (93°C). It is not completely resistant to aliphatic solvents
at room temperatures and is partially soluble in aromatic and
chlorinated hydrocarbons above 140°F (60°C). It is subject to stress
cracking with even low concentration of chlorine.
II
Compatibility of Thermoplastic Materials with Selected Corrodents
The tables that follow are arranged alphabetically according to
corrodent. Unless otherwise noted, the corrodent is considered pure in
the case of liquids and
Page 318
a saturated aqueous solution in the case of solids. All percentages
shown are weight percents.
Corrosion is a function of temperature. ''R" indicates that the material
is resistant, while "U" denotes that the material is unsatisfactory.
The tables have been taken from Corrosion Resistance Tables, 4th
Ed., Parts AC, by Philip A. Schweitzer, published by Marcel Dekker,
New York, 1995.
Page 319
Page 320
Page 321
Page 322
Page 323
Page 324
Page 325
Page 326
Page 327
Page 328
Page 329
Page 330
Page 331
Page 332
Page 333
Page 334
Page 335
Page 336
Page 337
Page 338
Page 339
Page 340
Page 341
Page 342
Page 343
Page 344
Page 345
Page 347
13
Corrosion of Thermoset Plastics
Dirk L. Pletcher
Zimmer
Warsaw, Indiana
I
Introduction
Thermosetting polymers, or thermosets, are derived from low
molecular weight precursors known as monomers or oligimers. These
precursor materials, when allowed to chemically react or polymerize,
form continuous network polymers of very large molecular weight.
This network is crosslinked to the extent that it is infusible and
insoluble. The applications of thermosetting polymers cover a wide
range of products and industries. Protective coatings, adhesives,
plastics, reinforced plastics or composites, and building materials are
examples of how they may be used. Their excellent corrosion
resistance is often the main reason for selecting a thermoset, whether
it be for metal protection or replacement. This chapter discusses the
basic chemistry of the most commonly used thermosetting plastic
materials. The chemical resistance of each class is reviewed, and the
mechanisms for their chemical attack and failure are discussed.
II
Epoxy
Epoxy- (or epoxide-) based thermosets are perhaps the most widely
used and versatile thermosets. They are applied across many fields of
application i.e., adhesives, coatings, sealants, castings, encapsulents,
tooling compounds, com-
Page 348
posites, and molding compounds, to name a few. This versatility is
due to the wide latitude in properties that can be achieved by the
epoxy formulator. A wide variety of epoxy resins, modifiers, and
curing agents are available, which allows the epoxy formulator to
tailor the epoxy system for each applications requirement. The most
common types of resins, curing agents, and modifiers will be
reviewed, with their consequent affects on corrosion resistance.
A
Epoxy Resins
The epoxide, or oxirane functionality, is a three-membered carbonoxygen-carbon ring, specifically an a or 1,2 epoxide. The simplest 1,2
epoxide is ethylene oxide (1). A common term used in naming epoxy
resins is the prefix ''glycidyl." The terminology for the glycidyl group
(2) is derived from the trivially named glycidol (3) and glycidic acid
(4).
Figure 1 lists the structures of the most common epoxy resins in use
today. In order to achieve a crosslinked network, two or more epoxide
groups per molecule are required. Diglycidyl ether of bisphenol A
(DGEBA) is the workhorse of the epoxy thermoset industry, and is
available in both liquid and solid forms, depending on the molecular
weight of the resin. DGEBA is synthesized by the condensation
reaction between bisphenol A and epichlorohydrin, the molecular
weight of the resin being dependent on the stoichiometry of the
reactants. Comprehensive reviews of the synthesis and chemistry of
epoxy resins are available [1,2].
B
Epoxy Curing Reactions
Epoxy resins are most commmonly cured via an addition reaction
with an active hydrogen-containing material. The most widely used
materials are primary and secondary polyamines, polyamides,
polyamidoamines, and acid anhydrides. Other active hydrogen
components such as phenols and alcohols are used to accelerate the
addition reaction and are used as additives. Primary amines follow the
following reaction sequence [3]:
The reactivity of the amine curing agents follows the general order:
primary
Page 349
Figure 1
Common commercial epoxy resins.
Page 350
aliphatic > secondary aliphatic > cycloaliphatic > aromatic. Primary
amines are often used with ambient temperature cures only, while the
aromatic amines generally require elevated temperature cures.
Anhydrides are quite unreactive with epoxies until the anhydride ring
is opened. Even with catalysts to accelerate the reaction, anhydrides
require an elevated temperature cure. Typical catalysts for the
anhydrideepoxy reaction are tertiary amines, imidazoles, Lewis acid
complexes, and metallic salts [4]:
The second most common method for curing epoxy resins is by using
a catalyst to promote homopolymerization of the epoxy. Anionic
catalysts include metal hydroxides, secondary amines, and tertiary
amines [5]:
The use of cationic catalysts is probably more widely practiced, with
Lewis acid complexes such as boron trifluoride monoethylamine the
most common:
Page 351
Elevated temperature cures are generally required for catalytic
polymerization of epoxies. Figure 2 gives structures of some of the
more common epoxy curing agents.
C
Chemical Resistance
The chemical resistance of epoxy thermosets is very dependent on the
curing agent used to crosslink the epoxy and somewhat less dependent
on the specific epoxy resin used. This is due to the fact that the site for
chemical attack is most often the chemical bond that is formed
between the epoxy resin and curing agent, and the type of bond that is
formed is primarily dependent on the curing agent used. Some
generalizations can be made regarding the stability of these bonds [6]:
1. Ether bonds are stable against most inorganic and organic acids,
and against caustics.
2. Carbonamine nitrogen bonds have good stability in most inorganic
acids and caustics, and poor stability in organic acids.
3. Ester bonds have good stability in organic acids, and fair to poor
resistance to inorganic acids and caustics.
The above generalizations can be affected by the chemical structure of
the curing agent and resin, where steric shielding effects and crosslink
density may tend to offset these chemical bond effects. Other
important affects on chemical resistance include stoichiometry, degree
or extent of cure, and additives or modifiers that may have been
incorporated into the epoxy.
1
Curing Agent Effects
Table 1 gives chemical resistance data for several different curing
agents used with a standard liquid DGEBA epoxy resin. Several of the
general expected trends based on the chemical bond formation during
crosslinking can be observed. The aromatic aminecured epoxy
generally has the best overall resistance to chemical attack, with no
complete failures in any of the exposures. Since the chemical bonds
formed are the stable carbonamine nitrogen type, Eq. (1), they have
good hydrolysis resistance. The aromatic amine also holds up well
against the organic acid, where the aliphatic amine fails. Some of the
poorer performance of the aliphatic amine can be attributed to the
degree of cure and crosslink density compared to the aromatic amine,
since the aliphatic amine specimens were cured only at ambient
temperature, while the aromatic amine required an elevated
temperature postcure. The subject of crosslink density will be
explored further. The BF3MEA catalyticcured epoxy, with primarily
ether bond formation, Eq. (4), holds up fairly well in both acidic and
alkaline environments. It is attacked by the chlorinated solvent, failing
in the flexural testing. The anhydride's susceptibility to an alkaline
environment is seen with severe loss of properties in the caustic
exposure. The ester linkage formed between the epoxy and anhydride,
Eq. (2), is subject to saponification in the presence of very strong base
[7]:
Page 352
Figure 2
Epoxy curing agents.
Page 353
The mechanism for the attack of an alkaline environment on
anhydride-cured epoxy was verified in a study by Hojo and others [8].
The degradation products of MTHPA-cured epoxy castings exposed
to sodium hydroxide solutions were evaluated by infrared
spectroscopy, where the infrared ester peak at 1730 cm-1, disappeared,
with subsequent formation of peaks at 1570 and 1440 cm-1, verifying
the formation of the carboxylate ion.
2
Resin Effects
The effects of resin type on chemical resistance will depend on the
nature of the chemical bonds in the resin backbone and the
functionality of the resin oligimer. The functionality, or number of
epoxy groups per oligimer, will effect the crosslink density of the
network. Figures 3 and 4 show the effects of varying resin type and
crosslink density on chemical resistance. The weight change of
unreinforced resin coupons is shown after 120 days exposure to
several solutions. The resins compared are a Novolac epoxy with
functionality of 3.6, a standard diglycidyl ether of bisphenol A, and
DGEBA flexibilized by substituting 30% of the DGEBA resin with a
long-chain difunctional flexible epoxy resin. Resins in Fig. 3 have
been cured with methylene dianaline, while resins in Fig. 4 were
cured with nadic methyl anhydride [9]. Solvent resistance is seen to be
the most dramatic change related to the resin type and follows an
inverse relationship with the crosslink density of the cured resins. This
can be seen by comparing the glass transition of the cured resins, with
higher crosslink density yielding higher glass transition or heat
distortion point values. The weight change on exposure to acid and
caustic solutions is seen to be more dependent on the type of curing
agent used, with crosslink density and resin type having a much
smaller effect.
III
Unsaturated Polyester
William H. Carothers was the first person to prepare polyesters of
known structures and as early as 1929 devised manufacturing
techniques that are similar to those used today [10]. The next step
toward commercial use of thermoset polyesters was enabled by Ellis,
who discovered in the late 1930s that polyesters' rate of crosslinking
increased around 30 times in the presence of unsaturated monomers.
The large-scale commercializaiton of unsaturated polyesters stems
from the discovery made by the U.S. Rubber Company in 1942 that
adding glass fibers to polyesters tremendously improved their physical
properties. Without reinforcement, the inherent brittleness of the
polyester resins had prevented their widespread use. Radomes for
military aircraft during World War II was the first successful
application of glass-reinforced polyesters. The first glass fiber-
Table 1 Curing Agent Effect on Chemical Resistance of DGEBA Epoxy, Unfilled Castings Af
180-Day Exposure
Change in appearance
Chemical
TETA
mPDA
PA
Exposure @ 54
°C
Distilled water Whitened
No change
No change
Trichloroethylene Blistered, slightly
No change
chipped
25%w
Slight darkening, char- Slightly
greenish
Hydrochloric
like odor
acid
Heavy chalking
No change
6%w Sodium
hypochlorite
25%w Acetic
Chipped, soft, swollen No change
acid
Exposure @ 82
°C
25%w Sulfuric Slight darkening,
Slight dulling
acid
swelling, sugar-like
odor
Considerable
25%w
Considerable
Hydrochloric
darkening and swollen darkening
acid
edges
50%w Sodium Very slight darkening Slight dulling
hydroxide
(table continued on next page)
Chipping, soft, swollen
No change
No change
No change
Considerable darkening
Dissolved; eaten
away; dull, very slight
darkening
(table continued from previous page)
Chemical
40%w Formaldehyde
25%w Chromic acid
28.5% Sodium al.
sulfate
Change in appearance
TETA
mPDA
PA
Swelled, shatteredSlight edge swellingNo change
Heavy chalking Slight chalking
Dulled
Darkening
Slight darkening
Dulled
Very slight dar
Surface dulling
Very slight dar
% Change in flexural strength
% Change in flexural mod
TETA
mPDA PA BF3MEA TETAmPDA
Exposure @ 54°C
Distilled water
-17
-6
-5 (+)7
-39
-1
Trichloroethylene
-17
-19
Failed
Failed -30
-17 Failed
25%w Hydrochloric acid -45
-15
(+)65
-20
-7
(+)1 (+)10
6%w Sodium hypochlorite-10
-3
-10
-19
-49
-8
25%w Acetic acid
Failed
-5
-20
-5
Failed (+)5
Exposure @ 82°C
25%w Sulfuric acid
-47
-25
-2
-12
-87
-7
25%w Hydrochloric acid -33
-20
-53
-12
-46
-4
50%w Sodium hydroxide -6
-6
-94
-5
-3
(+)6 Failed
40%w Formaldehyde
Failed
-25
-48
-18
-60
-6
25%w Chromic acid
-10
-6
-24
-12
-52
-29
28.5% Sodium al. sulfate -33
-4
-8
-10
-56
-20
TETA, triethylenetetramine; PA, phthalic anhydride; mPDA, meta-phenylenediamine; BF
borontrifluoridemonoethylamine complex.
Source: Shell Chemical Technical Bulletin SC:6781.
Page 356
Figure 3
Weight gain of epoxy resins cured with MDA after 120 days immersion at
room temperature. Resins tested were a Novolac epoxy, DEN 438; standard
liquid DBEBA, DER 331; flexibilized system is DER 331 with 30% by weight
DER 732. Unreinforced resin coupons were tested per ASTM D-543 [9].
reinorced boat hulls were made as early as 1946 [11], which is still a
major use for unsaturated polyesters today.
Unsaturated polyesters are the major thermoset used in reinforced
plastics, including those designed for corrosion-resistant service. This
is due not only to their low cost but also to their ease of use. The
uncured resins are typically low-viscosity liquids that cure without
generation of volatiles. They cure extremely rapidly when used in
elevated temperaturecured molding operations. They also can be
cured readily at ambient temperatures where large structures make
elevated temperature cures impractical. Typical applications for
corrosion-resistant service include chemical storage tanks, including
fuels such as gasoline and diesel, chemical-resistant piping, floor
grating, and structural members such as channel and "I" beams.
A
Chemistry
Unsaturated polyester polymers are manufactured via a condensation
reaction between dibasic organic acid or anhydride, and a difunctional
alcohol:
Page 357
Figure 4
Weight gain of epoxy resins cured with NMA after 120 days immersion
at room temperature. Resins tested were a Novolac epoxy, DEN 438:
standard liquid DBEBA, DER 331; flexibilized system is DER 331
with 30% by weight DER 732. Unreinforced resin coupons
were tested per ASTM D-543 [9].
At least one of these components contributes sites of unsaturation to
the oligimer chain. This oligimer, or prepolymer, is then dissolved in
an unsaturated monomer, such as styrene. An initiator, usually a free
radical source such as an organic peroxide, is added to the liquid resin
solution to initiate crosslinking. Metallic compounds and tertiary
amines with certain peroxides may be added to accelerate the reaction
for ambient temperature cures, or the resin may be heat-cured by
using peroxide alone. The crosslinking reaction is a free radical
copolymerization between the resin oligimer and the unsaturated
monomer. The optimum ratio of monomer to oligimer molar
unsaturation has been found to be around 4 to optimize physical
properties of the cured polyester, using styrene as the monomer [12].
This results in a structure with several units of the monomer acting as
a bridge between the oligimer chains. The properties of the cured
polyester are dependent on the types and ratios of components used to
manufacture the oligimers, the manufacturing procedure, and the
molecular weight of the oligimer. A combination of saturated and
unsaturated acids is used to control the crosslink
Page 358
density of the cured polyester, which has an impact on such properties
as brittleness and heat distortion point. The source of unsaturation for
almost all of the polyester resins currently produced is from maleic
anhydride or fumaric acid, with maleic anhydride being the most
common. The most common diol used is propylene glycol, since it
yields polyesters with the best overall properties. Orthophthalic
polyesters based on phthalic anhydride as the saturated monomer are
the lowest cost class of resin. Isophthalic polyesters use isophthalic
acid in place of phthalic anhydride, which increases cost to produce
but improves physical properties and chemical resistance. Figure 5
lists the more common components used in the manufacture of
unsaturated polyesters, and Fig. 6 gives structures of typical polyester
resin oligimers.
B
Chemical Resistance
Polyester polymers are susceptible to hydrolysis of their ester groups,
just as described for the anhydride-cured epoxies in Eq. (5). The
composition of the polyester resin backbone can have a significant
effect on the rate of hydrolysis, with the more chemically resistant
polyesters providing for protection of the ester group from attack by
steric shielding.
Isophthalic-based unsaturated polyesters based on isophthalic acid,
maleic anhydride, and propylene glycol as the diol are the most
common type of polyesters used for chemical service applications,
such as glass-reinforced piping, tanks, and structural members [13].
The standard corrosion grade isophthalic is made with a 1:1 molar
ratio of isophthalic acid to maleic anhydride or fumaric acid, with
propylene glycol. In general, the properties of isophthalic polyesters
are superior to the lower cost orthophthalic polyesters, including not
only chemical resistance but physical properties as well. The
improved corrosion resistance of the isophthalic polyesters has been
attributed to the accessibility of the ester groups to attack, with the
isophthalic providing more steric protection. Another possible reason
relates to the method in which the two different resins are synthesized.
The orthophthalic resins are typically reacted in a one-step process, so
the possibility exists of free phthalic anhydride or low molecular
weight esters contaminating the finished resin. This potential for
contamination is minimized for the isophthalic polyesters, since they
are processed in two steps, with the isophthalic acid being at least half
esterified initially, and then further reacted with the unsaturated acid
in a second step to a high molecular weight prepolymer. Differences
in the rate constants for the esterification of the acid half esters, with
the rate for the second acid group of phthalic acid being more than
five times slower than that of isophthalic acid, increase the potential
for low molecular weight species in orthophthalic polyesters and are
the reason for the typical commercial prepolymers being lower in
molecular weight as compared to most isophthalic-based resins. The
chemical resistance of ASTM C-581 laminates
Page 359
Figure 5
Components of unsaturated polyesters.
has been studied by Amoco and is given for several solutions in Figs.
79. Here the resins were laboratory-prepared, with one of the
orthophthalic resins (HMW) processed by a two-stage cook similar to
the isophthalic and the other a more typical commercial grade low
molecular weight orthophthalic (LMW). Both orthophthalic resins
would be considered unacceptable in these environments [12].
Page 360
Figure 6
Polyester oligimer structures.
Figure 7
Projected retention of physical properties, isophthalic vs.
orthophthalic polyester ASTM C-581 laminates, based on
12-month exposure to 35% HCl at 49°C
(From Ref. 12.)
Page 361
Figure 8
Projected retention of physical properties, isophthalic
vs. orthophthalic polyester ASTM C-581 laminates,
based on 12-month exposure to 25% ethanol at 71°C
(From Ref. 12.)
Figure 9
Projected retention of physical properties, isophthalic
vs. orthophthalic polyester ASTM C-581 laminates,
based on 12-month exposure to 1 N NH4OH at 38°C
(From Ref. 12.)
Page 362
Terephthalic polyesters based on terephthalic acid, the para isomer of
phthalic acid, has also been proposed for use in corrosion-resistant
applications. The properties of cured terephthalic-based polyesters are
similar to the isophthalic polyesters, with the terephthalics having
higher heat distortion points and being somewhat softer at equal
unsaturation levels [12]. Corrosion resistance of the terephthalic
polyesters is fairly similar to that of the isophthalics. Testing by
Amoco indicated that the benzene resistance of comparable
formulated resins was lower for terephthalic vs. isophthalic polyesters
[12]. This trend was also indicated in work by Gillette and Spoo,
where retention of flexural modulus was evaluated for various
terephthalic resins vs. the standard corrosion grade isophthalic [14].
The terephthalic's loss of properties in gasoline was greater than the
isophthalic's at the same level of unsaturation, but as unsaturation
increased, the gasoline resistance reversed, with the terephthalic
performing better. The trend was seen only at unsaturated acid levels
of greater than 50 mol %. This was achieved with a reversal of
performance in 10% NaOH, where the terephthalic with lower
unsaturation was better than the isophthalic but worse at the higher
unsaturation level. This follows a general trend for thermosets, that
being as crosslink density increases, solvent resistance increases. The
reversal in caustic resistance is most likely due to the loss of steric
shielding from substitution of the aromatic acid with aliphatic
unsaturated acid.
Edwards studied the effects of flexibilizing components on the
physical properties and chemical resistance of unsaturated polyesters,
compared to the standard corrosion grade isophthalic [15]. Adipic acid
was incorporated as well as various glycols, and the ratios of both
acids and glycols were varied. Twenty-three resins were screened for
chemical resistance for 30 days, with three selected for a year of
testing in a variety of chemical media. All resins containing adipic
acid with an ether glycol had poor overall chemical resistance. Adipic
acid/diethylene glycolcontaining resins had poor caustic resistance.
Neopentyl glycol improved the caustic resistance of resins when
substituted in part or whole for propylene glycol, even when adipic
acid was used. The shielding effects of (2,2,4-trimethyl-1,3pentanediol (TMPD)) were further studied by Hillman et al. [16].
Resins were synthesized using a 75:25 weight ratio of TMPD to
propylene glycol, with a 1:2 molar ratio of saturated acid to maleic
anhydride. One resin with isophthalic and one with
dimethylterephthalate were compared to the standard isophthalic
corrosion grade resin, as well as to some premium corrosion grade
resins. Retention of flexural properties is given in Fig. 10. The
shielding effects of the methyl groups can be readily seen, with
dramatic improvements in hydrolysis resistance under neutral, acidic,
and basic conditions. The performance of the premium corrosion
grade resins is also quite impressive compared to that of the standard
isophthalic resin. Premium corrosion grade resins are generally
3050% higher in cost but can be well worth it in terms of
performance. Chlorendic anhydride (HET) acidbased polyester resins,
based on chlorendic
Page 363
Figure 10
Effects of TMPD glycol shielding on flexural strength retention of unreinforced
resin castings, after 6 months exposure [16].
anhydride, have excellent resistance to oxidizing environments as well
as being inherently flame-retardant. An additional benefit of the HET
acid resins is the high heat distortion point that is developed, allowing
use at elevated temperature without significant loss of elevated
temperature properties. Figure 11 shows flexural strength retention vs.
temperature of laminates made with several types of corrosionresistant resins. Common applications for the HET acid polyesters
take advantage of their combination of high-temperature performance
and resistance to oxidizing agents, including hot, wet chlorine.
Corrosion-resistant applications include use as chimney liners, flue
gas breeching, chrome plating tanks, pickling tanks, and chlorine
headers [17].
Bisphenol A is incorporated into unsaturated polyesters, typically
replacing a portion of the aliphatic diol, to produce another class of
premium corrosion grade resins. The standard bisphenol A polyesters
are derived from the propylene glycol or oxide diether of bisphenol A
and fumaric acid [18,19]. The aromatic structure contributed by the
bisphenol A provides several benefits. Thermal stability is improved,
and the heat distortion point of the resin is increased, mainly from the
more rigid nature of the aromatic structure. The number of interior
chain ester groups are reduced, so that resistance to hydrolysis and
saponificaiton are increased. Bisphenol A fumarate polyesters have
the best hydrolysis resistance of
Page 364
Figure 11
Flexural strength vs. temperature for glass-reinforced polyester
laminates. Construction of the laminates: V M M Wr M Wr M;
V = veil, M = chopped mat 1 1/2 oz/ft2, Wr = woven roving
24 oz/yd2 [17].
any commercial unsaturated polyesters. They have good resistance to
hot water and steam and are also resistant to acids, bases, and some
solvents. A condensed application guide is compiled in Table 2 for
several grades of corrosion-resistant polyesters. Comprehensive
application guides are available from most resin manufacturers.
IV
Vinyl Ester
Vinyl ester oligimers have unsaturated ester groups only at the
terminal ends of the oligimer, as opposed to polyesters where the
vinyl and ester segments are distributed all along the molecular chain.
The vinyl ester class of resins was developed in the late 1950s and
early 1960s, with several patents granted during the 1960s [20]. The
earliest applications were in dental fillings, with several patents
granted to R. L. Bowen. The need was to improve toughness and
bonding over the acrylic materials that were being used at that time.
Fekete developed vinyl esters for applications in the electrical and
corrosion markets. Bearden solved some of the early stabilization
problems with the vinyl esters that allowed commercialization of the
vinyl esters for the composite corrosion market. These early vinyl
esters were terminated with acrylic endgroups. Beardon, Jernigan,
Najvar, and Hargis found that substituting the acrylic with methacrylic
endgroups improved the chemical resistance of the vinyl ester by
improving shielding of the
Page 365
ester groups. These developments led the way to the widespread use
of vinyl esters in the corrosion-resistant equipment field today.
Contemporary vinyl esters exhibit several advantages over
unsaturated polyesters. In general, they provide for improved
toughness in the cured polymer while maintaining good thermal
stability and physical properties at elevated temperatures. This
improved toughness allows for use in castings as well as in reinforced
plastics. Since they have internal hydroxyl gorups, they have
improved bonding to inorganic fillers and reinforcements. Composites
manufactured from vinyl esters have improved damage resistance,
provided by the toughness of the resin and improved bonding to the
reinforcements. Shrinkage during cure is generally lower, providing
for less internal stress developed in the finished article.
The desirable properties of vinyl esters have steered them into many
diverse applications, including protective floor coatings, liners for
chemical tanks, and composites for many markets, including
automotive, construction, recreational, and corrosion-resistant
structures for the chemical industries.
A
Chemistry
Vinyl esters are derived from epoxy resin oligimers by reaction of the
epoxy groups with an unsaturated organic acid, typically acrylic or
methacrylic acid:
The addition reaction is catalyzed by tertiary amines, phosphines,
alkalis, or onium salts [21]. The oligimer is dissolved in an
unsaturated monomer, typically styrene, along with an inhibitor such
as hydroquinone, to prevent premature reaction of the vinyl groups.
Curing is similar to that of the unsaturated polyesters, with addition of
a free radical source such as an organic peroxide. As with the
polyesters, the vinyl esters can be cured at ambient or elevated
temperatures, depending on the promotors and initiators used. The
basic vinyl esters are derived from the DGEBA epoxies (see Sec.
IIA). Also in wide use are vinyl esters based on Novolac epoxies,
rubber-modified vinyl esters for further improvements in toughness
and elongation, and brominated vinyl esters for flame retardancy.
Structures of some of the more common vinyl esters are given in Fig.
12.
Table 2 Chemical Resistance Guide for Unsaturated Polyesters, Vinyl Esters, and Furan:
Maximum Recommended Exposure Temperature, °F
Bisphenol
Bisphenol Novolac
Chemical
ConcentrationIsophthalic
A
Het acid A vinyl
vinyl
environment
(%)
polyester polyester polyester ester
ester
Acetic acid, glacial
100
NR
NR
LS
NR
100
Acetic acid
10
160
200
210
210
210
Acetone
100
NR
NR
NR
NR
NR
Ammonium
29
NR
100
NR
100
100
hydroxide
Ammonium
10
NR
140
NR
150
150
hydroxide
Analine
100
NR
NR
NR
NR
70
Benzene
100
LS:90
NR
90
NR
100
Carbon
100
LS:90
110
125
150
180
tetrachloride
Chromic acid
30
NR
NR
140
NR
NR
Citric acid
Sat'd.
180
210
220
210
210
Cottonseed oil
100
100
200
100
210
210
Detergents,
100
160
210
160
180
sulfonated
Diethyl ether
100
NR
NR
NR
NR
NR
Dimethylformamide
100
NR
NR
NR
NR
NR
Distilled water
100
160
210
210
180
180
Ethyl acetate
100
NR
NR
NR
NR
70
Ethyl alcohol,
95
LS
110
100
80
100
denatured
Ethyl alcohol
50
90
120
150
100
150
Ethylene dichloride
100
NR
NR
NR
NR
NR
2-Ethylhexyl
100
sebacate
Heptane
100
200
150
200
210
210
Hydrochloric acid
37
100
110
100
150
180
Hydrochloric acid
10
160
210
230
180
230
Hydrofluoric acid
40
90
Hydrofluoric acid
20
NR
100
100
100
(table continued on next page)
(table continued from previous page)
Concentration Isophthalic Bisphenol Het acid Bisphenol A Novolac
Chemical
environment (%)
polyester A polyester polyester vinyl ester vinyl ester
Hydrogen
30
100
150
150
peroxide
Hydrogen
5
150
150
210
180
350
peroxide
Isooctane
100
Kerosene
100
175
210
175
180
180
Methyl
100
90
110
100
NR
100
alcohol
Mineral oils
100
180
210
220
210
250
Nitric acid
50
NR
110
140
Nitric acid
40
NR
140
NR
80
Nitric acid
10
90
200
Nitric acid
5
160
175
210
150
180
Oleic acid
100
180
210
200
210
200
Olive oil
100
180
200
140
210
250
Phenol
5
NR
110
180
120
Soap
90
90
90
solution
Sodium
35
NR
160
100
180
180
carbonate
Sodium
10
LS:160
180
160
180
180
carbonate
Sodium
180
210
250
180
200
Sat'd.
chloride
Sodium
50
NR
210
NR
210
180
hydroxide
Sodium
10
NR
150
NR
180
150
hydroxide
Sodium
5
NR
150
180
150
hydroxide
Sodium
1
LS:125
200
hydroxide
Sodium
5-1/4
120
125
125
180
150
hypochlorite
Sulfuric acid
93
NR
NR
NR
NR
NR
Sulfuric acid
50
150
210
200
Sulfuric acid
25
160
210
250
210
210
Toluene
100
90
NR
90
80
120
Transformer
100
90
210
220
210
300
oils
Turpentine
100
90
200
120
150
210
NR, not recommended; LS, limited service; Sat'd., saturated.
Sources: Data from Ref. 12 and Dow Chemical Bulletin 125-00043-594 SMG.
Page 368
Figure 12
Vinyl ester oligimers.
B
Chemical Resistance
Vinyl esters have good chemical resistance to a wide variety of
chemicals in spite of the fact that they contain ester groups, which as
discussed earlier for unsaturated polyesters, are known to be
susceptible to hydrolysis and saponification. The improved chemical
resistance is due to several factors. As can be seen by reviewing the
structures of typical DGEBA vinyl esters in Fig. 12, the vinly ester
linkage sites are all terminal. This reduces the number of ester groups
available for attack. These ester groups are shielded by the methyl
group of the methacrylic acid most commonly used to synthesize
vinyl esters for corrosion-resistant applications. The aromatic
backbone of the vinyl esters derived from bisphenol A epoxies or
Novolac epoxies, with the addition of an aromatic vinyl monomer
such as styrene, reduces the polarity of the vinyl ester resin as
compared to both unsaturated polyesters and epoxies. This results in
lower affinity for moisture, which decreases the tendency for loss of
properties due to the plasticizing effects of moisture, and reduces the
amount of water available for hydrolysis to occur. The effects of hot
moist conditions on a brominated vinyl ester/carbon fiber laminate
were compared to an aliphatic aminecured DGEBA epoxy/carbon
fiber laminate [22], with results given in Fig. 13.
Page 369
Figure 13
Comparison of physical property changes of graphite-reinforced laminates
with epoxy/aliphatic amine matrix (Tactix 123/H31) vs. a brominated vinyl
ester (Derakane 510C-350). FRTD = flexural strength, room temperature
dry conditions. FHW = flexural strength, hot wet conditions. CRTD,
CHW and SRTD, SHW are compressive and shear, respectively,
under same conditions. Hot wet samples conditioned to constant
weight at 185°F/95% humidity, then tested at 200°F wet [22].
Epoxy composites typically have higher matrixdominated physical
properties than most other thermoset composites, due to their ability
to bond tenaciously to the reinforcing fibers. This advantage is seen to
disappear for the epoxy under hot moist conditions. The percentage
loss of compressive strength and flexural strength, both
matrixdominated properties, are greater for the epoxy than for the
vinyl ester.
The chemical resistance of a rubber-modified vinyl ester, which
provides for cured resin tensile elongation of 1012% vs. 56% for an
unmodified standard bisphenol A vinyl ester, was evaluated by
Kardenetz et al. [23]. The results for flexural strength retention of
unreinforced resin castings in various media are displayed in Fig. 14.
Property retention for the modified vinyl ester is similar to the
standard resin, except for lower solvent resistance to lower molecular
weight solvents such as ethanol and toluene. Advantages of the
modified vinyl ester are improved impact, crack and abrasion
resistance, and improved adhesion to various substrates including
aluminum and reinforced composites.
Solvent resistance is an application area where vinyl esters are
typically found to provide good service. As with other thermosets,
increasing crosslink density generally provides for improved
resistance to solvents. The C-581 lami-
Page 370
Figure 14
Flexural strength retention of rubber modified vinyl ester vs. standard bisphenol
A epoxy vinyl ester resin castings. Exposure for 12 months at 150°F, except for
ethanol and toluene, which were exposed at 77°F [23].
nate resistance of several different vinyl esters and unsaturated
polyesters to toluene, methanol, and an alcohol-containing fuel is
given in Fig. 15. The Novolac vinyl ester has the highest crosslink
density of the vinyl esters, and the bisphenol A vinyl ester the lowest.
Generally, as crosslink density increases, the retention of properties is
higher. Also noted by the authors of this work is the importance of the
laminate appearance after exposure. In general, all of the vinyl ester
laminates retained good appearance in all of the fluids tested, even in
those where flexural properties were lowered. The Novolac epoxybased vinyl esters had the best overall appearance retention, while the
unsaturated polyesters showed some loss of transparency and fiber
prominence, especially in 100% toluene [24].
The use of vinyl ester composites in pulp and paper mill bleaching
environments is common due to their ability to resist the combination
of oxidants and caustics typically found in those applications. In situ
testing of C-581 coupons was conducted by Cowley, where it was
shown that a brominated bisphenol A epoxy vinyl ester had equal or
better performance to a standard bisphenol A epoxy vinyl ester and a
Novolac epoxy vinyl ester, which were already proven to be
satisfactory based on numerous case histories in that
Page 371
Figure 15
Flexural property retention of ASTM C-581 laminates at 100°F after 270 days
immersion. Iso = isophthalic, UPe = unsaturated polyester, V. Ester = vinyl ester [24].
environment. The exposure was to chlorination, chlorine dioxide, and
caustic process streams at elevated temperatures for 12 months [25]. There
is some evidence based on laboratory evaluations of C-581 coupons that
Novolac epoxy vinyl esters perform best in chlorine dioxide at elevated
temperatures, while brominated bisphenol A epoxy vinyl esters perform
best in sodium hypochlorite at elevated temperatures [26]. Also noted in
this work was the observation that in unstable elevated temperature
sodium hypochlorite solutions, the cure system used for the vinyl ester had
an effect on performance. Benzoyl peroxide/dimethyl anilinecured
laminates were preferred over methyl ethyl ketone/cobaltcured systems.
This difference was not seen in exposure to chlorine dioxide. The
reasoning given is that the cobalt accelerates the decomposition of the
sodium hypochlorite to hypochlorous acid, which is more aggressive to
the reinforced laminate. The performance of the vinyl esters in corrosionresistant applications is due not only to their good chemical resistance to a
wide variety of chemicals but to their improved mechanical properties
when compared to polyesters. The higher tensile elongation and improved
bonding to most types of substrates allows for more damagetolerant
structures to be built, which accounts for their popularity as the resin
matrix in corrosion-resistant composites. A condensed application guide is
compiled in Table 2 for several grades of corrosion-resistant vinyl esters.
Comprehensive application guides are available from most resin
manufacturers.
Page 372
V
Other Thermosets
A
Phenolic
Phenolic thermosets are the oldest commercial synthetic class of
polymers in use today, dating back to work as early as 1872 by A. von
Baeyer. The commercial value of these materials was realized by Leo
H. Baekeland, who applied for his famous ''heat and pressure" patent
in 1907, which made possible the economic development and
application of phenolic molding compounds [27]. Phenolic resins are
used across a wide spectrum of products, including glues for wood
lamination and abrasive products, molding compounds, coatings,
foundry binders, and laminates.
1
Chemistry
Phenolic resins are typically derived from phenol and formaldehyde,
although other phenols and aldehydes may be used for special end-use
requirements. There are two types of phenolic prepolymers that can be
manufactured, depending on the stoichiometry and reaction conditions
used. Resoles are manufactured using an excess of formaldehyde
under alkaline conditions, and are a mixture of methyl
alcoholsubstituted phenolics and their condensation products, with the
composition depending on reaction conditions and extent of reaction,
Eqs. (8) and (9).
Page 373
Resoles are usually liquid resins which can be cured with heat alone
or by the addition of an acid catalyst. Strong acid catalysts will cure
the resoles at room temperature. Novolacs are synthesized by using an
excess of phenol under acidic conditions:
The novolacs are usually solids which are thermoplastic in nature in
that they may be melted by heating repeatedly without any further
reaction. They are compounded with either hexamethylenetetramine
(HMTA) or paraformaldehyde, which serves to cure the Novolac
when heat is applied, since formaldehyde is generated from either
with application of heat and water. The HMTA decomposition also
generates a significant amount of ammonia, which is liberated during
the curing reaction [28].
2
Chemical Resistance
Phenolic thermosets have excellent resistance to most organic
solvents, especially aromatics and chlorinated solvents. Small organic
polar solvents capable of hydrogen bonding, such as alcohols and
ketones, can attack phenolics. Although the phenolics have significant
aromatic character, the phenolic hydroxyls provide sites for hydrogen
bonding and attack by caustics. They are not suitable for use in strong
alkaline environments. Strong mineral acids also attack the phenolics,
with acids such as nitric, chromic, and hydrochloric causing severe
degradation. Sulfuric and phosphoric acids may be suitable under
some conditions. Certain organic acids, such as acetic, formic, and
oxalic may cause some loss of properties [29]. Figure 16 shows
flexural strength retention of a phenolic laminate made with a
phenolic resole, Cellobond J2018L, and 35% by weight of a chopped
strand mat [30].
B
Furan
Furan resins have excellent chemical resistance, good thermal
stability, and are inherently flame-retardant with low smoke
generation and toxicity. Their main drawbacks are being somewhat
brittle and not being very user-friendly, since they are more difficult
to work with than most other thermosets. Like the phenolics, they are
condensation polymers and generate volatiles when cured.
Page 374
Figure 16
Flexural strength retention of phenolic/glass fiber laminates after exposure at 25°C.
Key to immersion media: Ac10 = 10% acetic acid; HC10, HC30 = 10%, 30%
hydrochloric acid; Ni10, Ni30 = 10%, 30% nitric acid; Su10, Su30 = 10%, 30%
sulfuric acid; AH10 = 10% ammonium hydroxide; H2O = water; EtAl = ethyl
alcohol; Xy = Xylene; CT = carbon tetrachloride; TC = trichloroethylene [30].
1
Chemistry
Furan resin oligimers are derived from furfuryl alcohol, which itself is
derived from the digestion of vegetable waste with sulfuric acid and
steam [31].
Page 375
The furan prepolymers are catalyzed with an acid catalyst, such as ptoluenesulfonic acid, which promotes crosslinking to a ladder-like
structure. This tightly crosslinked structure gives the furans their
excellent thermal stability and chemical resistance.
2
Chemical Resistance
Furan thermosets are well known for their excellent resistance to
solvents and are considered to have the best overall chemical
resistance of all of the thermosets. They also have excellent resistance
to strong concentrated mineral acids, caustics, and combinations of
solvents with acids and bases. They are not recommended for use in
oxidizing media, such as chromic or nitric acids, peroxides,
hypochlorites, and chlorine. A condensed application for furan
laminates is given in Table 3. Extensive application guides are
available from furan resin suppliers [17,32].
VI
Corrosion Mechanisms
A
Thermoset Resin
Corrosion of unreinforced, unfilled thermosets can occur by several
different mechanisms. Understanding the corrosion mechanism can be
quite complicated if there is more than one operating at a time. The
type of corrosion can be divided into two main classes: chemical and
physical.
Chemical corrosion occurs when bonds in the thermoset are broken by
means of a chemical reaction with the plastic's environment. Since
thermosets are most often multifunctional, there may be more than
one type of chemical corrosion occurring at the same time. Chemical
corrosion is usually nonreversible.
Physical corrosion is defined as the interaction of a thermoset with its
environment so that its properties are altered but no chemical
reactions occur. An example would be the diffusion of a liquid into a
thermsoset. Physical corrosion is often reversible, where the original
properties are restored once the liquid is removed. The mechanisms of
both types of corrosion will be further considered in this section.
1
Chemical Corrosion
R. C. Allan studied the corrosion of several types of vinyl ester resin
castings by sodium hydroxide and sodium hypochlorite solutions
using X-ray-induced photoelectron spectroscopy (XPS) [33]. Castings
(1/8 in.) were exposed for 30 days to 5%, 25%, and 50% NaOH
aqueous solutions at 66°C and 100°C. Sodium ion concentrations was
determined at various depths by machining away the surface of the
castings and evaluating by XPS. The sodium ion concentration for all
resins was found to be maximum at the surface, typically between 0.5
and 1.5 Na ions
Page 376
per 100 carbon atoms, and then tapering off rapidly to negligible
amounts at 5 mils depths. The maximum sodium ion concentration
was found to occur with the 25% NaOH solution and was also found
to increase slightly at 100°C for all exposures. However, even at
100°C, detectable penetration depth was still limited to 5 mils or less.
Similar results were seen with the sodium hypsochlorite solution,
where chlorine ion concentration was determined at various depths
after 30 days at 66°C exposure. Chlorine ion concentration was used
since it was found that sodium ions were not detected on the surface
or interior of the exposed castings. The chemical shift of the chlorine
indicated that it was bound to carbon. The following mechanism was
proposed to explain the bound chlorine and lack of sodium ion, with
attack occurring at the hydroxyl on the vinyl ester backbone:
A study was conducted by Hojo and associates [8], whereby corrosion
of anhydride-cured epoxy, aromatic aminecured epoxy, and
unsaturated polyester by sodium hydroxide solutions was evaluated.
Resin castings were exposed at various combinations of temperatures
and concentrations. The anhydride-cured epoxy was found to degrade
by corrosion of the surface, with dissolution occurring in some cases.
Corrosion depth was measured by physically wiping away the
corroded layer, which was easily removed with an acetone-dampened
towel. Corrosion rate was found to follow a linear rate law:
where A = Arrhenius pre-exponential factor, CL = concentration of
solution, and a = order of reaction, with experimental results yielding
The layers underneath the corroded layer were examined by infrared
analysis, with no signs of ester hydrolysis seen. The aromatic
aminecured epoxy was essentially unattacked due to the lack of ester
bonds in the cured epoxy.
A different corrosion mechanism was seen for the unsaturated
polyester where dissolution of the surface did not occur. Rather, the
surface layer was discolored and softened to a rubber-like layer. The
rate of corrosion was found to be diffusion-controlled, and the rate
law parabolic instead of linear:
Page 377
Solving for k2 experimentally and using the Arrhenius relationship for
temperature gives:
The difference in behavior is explained by the crosslinked structures
of the two thermosets. The anhydride-cured epoxy is crosslinked only
at the ester sites, with hydrolysis depolymerizing the resin back to low
molecular weight fragments. The crosslinking of the polyester resin is
through the vinyl unsaturation in the polyester main chain and styrene,
which remains intact. The ester groups along the chain are attacked,
but dissolution into small fragments does not occur due to the stable
carboncarbon crosslinking between oligimer chains.
Caddock, et al. evaluated the diffusion of hydrochloric acid into
several unsaturated polyester thermosets. Cast resin plates and
cylindrical rods were exposed to radioactive tagged HCl; 36Cl was
used to trace movement. Water uptake was traced with tritium (3H),
with both isotopes determined simultaneously by b spectrometry using
a liquid scintillation counter. The water uptake for the polymers was
found to follow Fick's law, with the isophthalic polyester having a
diffusion coefficient of 3 × 10-9 cm2 sec-1 and a saturation level of
1.82% at 20°C. The HCl uptake showed negligible chlorine levels,
indicating that only surface absorption had occurred [34].
2
Physical Corrosion
Physical corrosion occurs when the polymer absorbs a liquid or gas,
resulting in plasticization or swelling of the thermoset network. For a
crosslinked thermoset, swelling caused by solvent absorption will be
at a maximum when the solvent and polymer solubility parameters are
exactly matched. The solubility parameter, d, is defined as the square
root of the cohesive energy density and is a measure of the attractive
strength between molecules:
where
is the energy of vaporization of species i and Vi is the molar
volume of species i [35]. The retention of flexural modulus of
Novolac vinyl ester C-581 laminates vs. solubility parameter is shown
in Fig. 17 [33]. Note that other interactions are important to consider
besides d alone, e.g., polarity, hydrogen bonding, acidbase
interactions, since some solvents with the same solubility parameter
may be more or less aggressive depending on these other factors.
Generally, solvents with smaller molar volumes will be more
aggressive given that other parameters are equal.
The sorption kinetics for thermosets may or may not follow a Fickean
pat-
Page 378
Figure 17
Flexural modulus retention vs. solubility parameter, ASTM C-581
Novolac vinyl ester laminates exposed at ambient temperature for
120 days [33].
tern, whereby diffusion is driven by a concentration gradient alone.
For Fickean sorption, the inital uptake of solvent will be proportional
to t1/2 and usually correlates with reversible plasticization. Fickean
sorption is typical for crosslinked elastomers above their glass
transition, where a log-log plot of uptake vs. time will have a slope of
0.5. For thermosets in the glassy state, positive deviations from Fick's
law are often seen and may result from other phenomena such as
microcrack formation, leaching of unreacted substituents, and
swelling processes by moisture or polar solvents, allowing access to
more polar regions of the polymer [36].
A sharp boundary is often seen between swollen and unswollen
polymer, whereby the boundary advances at a constant velocity. This
results in solvent uptake proportional to t and has been called case II
sorption [37]. For case II sorption, the initial slope of a log-log plot of
uptake vs. time is 1.0, and intermediate slopes between 0.5 and 1.0 are
designated as anomalous sorption.
Thermosets that undergo case II sorption will have a swollen shell and
unswollen glassy core. The stress resulting on a flat specimen has
been approximated to be [37]:
where G = shear modulus of swollen shell, g = linear swelling factor,
n = velocity
Page 379
of swelling front, and b0 = thickness of flat slab. As swelling
progresses, the stress on the glassy core continues to increase until it
reaches the tensile strnegth of the core, where core fracture results. In
some cases the swollen shell is seen to fragment and separate from the
core before the core fracture occurs, typically for very aggressive
solvents where the swelling factor is large.
B
Reinforced Thermosets
Thermoset polymers are often used with fiber reinforcements in
corrosion service applications because reinforcement is usually
required for most structural applications of thermoset plastics.
Degradation of the reinforced plastic or composite material from
corrosion can be complex, since there are now several different
materials that can degrade separately or simultaneously. Additionally,
there are concerns about degradation of the interface between polymer
and reinforcement, which can also have serious effects on the
composite physical properties.
1
Fiberglass Composites
Regester studied the diffusion of water, acids, and salt solutions
through fiberglass laminates [38]. Water was found to permeate freely
through all laminates, which were constructed with various types and
amounts of glass reinforcements and polyester resins. Permeation
rates were measured by exposing the laminate plates on one side and
measuring diffused water at the opposite side. Permeation rate was
highly dependent on resin type, with orthophthalic >> isophthalic >
bisphenol A fumarate. The rate increased with temperature and was
inversely proportional to glass content. Very little effect was seen by
increasing the solution pressure up to 20 atm, with water at 100°C.
The laminates exposed to acid and salt solutions were examined by
several methods to determine rate and depth of ion transport. Ions
were found to penetrate only the surface layers, with increasing
concentrations and temperatures increasing the penetration depth in
most cases. The depth of penetration for sulfate ion was found to be
lower than that of chloride ion, where resin-rich barrier layers were
employed.
Permeability of the composite to moisture is less important than the
absorption at saturation, since it is absorbed moisture that lowers
physical properties of the composite. The glass fibers are generally
treated with a sizing agent to improve bonding to the thermoset and
resistance to the degrading affects of moisture. The importance of
property retention after exposure to moisture can be seen for pultruded
glass fiber rods in Fig. 18, where glass with and without treatments of
silane coupling agents are compared before and after exposure to
boiling water.
The silane coupling agents are dual-functional molecules, with the
silanol portion reacting with available hydroxyl sites on the glass
fiber, or other mineral
Page 380
Figure 18
Effect of silane coupling agents on glass fiberpultruded rods.
Flexural strength dry vs. wet after 4-h exposure to boiling water [39].
filler, leaving the organic functional end of the silane free to react
with the thermoset resin [39]:
It is well known that glass fibers are subject to chemical attack by
basic solutions:
Acidic environments have also been shown to degrade the properties
of glass fibers when exposed directly to corrosive solutions, including
HCl, H2SO4, and HNO3 [40]. The importance of the thermoset resin
in protecting the reinforcements is quite obvious. Laminates are
therefore constructed with corrosion barriers known as surfacing veils,
which are typically 90% resin. The surfacing veils are usually made
from corrosion grade "C" glass, or a synthetic fiber with good
chemical resistance such as thermoplastic polyester. The veil layer is
backed by layers of chopped strand glass mat, which is typically
7080% by weight resin, followed by alternating woven roving and
chopped strand mat for structural
Page 381
strength. The corrosion barrier is typically at least 100 mils thick to
prevent penetration of ionic corrosives into the structural portion of
the composite.
2
Stress Corrosion
The effects of stress on the failure rate of glass-reinforced composites
can be significant. Composites exposed to combinations of acid and
stress have been studied extensively by Hogg [41]. Weakening of the
glass fibers upon exposure to acid is believed to be caused by an ion
exchange process between the acid and glass. Under stress, an initial
fiber fracture occurs, which is specifically a tensile type of failure. If
the resin matrix surrounding the failed fiber fractures, the acid is
allowed to attack the next available fiber, which subsequently
fractures. The process continues until catastrophic failure occurs.
Stress corrosion failures are evidenced by the appearance of the failed
composite, since the fibers are sheared off in a clean, smooth fashion,
as opposed to the fiber tear-out seen in a normal composite failure
mode. The toughness of the resin matrix has been found to be a more
important factor than the chemical resistance of the resin under stress
corrosion. The rate of stress corrosion crack growth was evaluated for
three different resins in unidirectional laminates exposed to aqueous
H2SO4 and HCl. The resins evaluated were a HET acid polyester, a
standard isophthalic polyester, and a flexibilized isophthalic polyester.
The normal expectations based on resin corrosion resistance would be
HET acid > Iso > Flex-Iso, but under stress corrosion conditions the
reverse was found. One possible explanation for these results is that
the tougher, more flexible resins do not crack after the fiber fracture,
so that advancement of the acid to the next fiber becomes diffusioncontrolled, slowing the stress corrosion process [42]. More recent
work studying the stress on the fibers in the fracture area suggests that
more brittle resins transfer higher stress to the fiber at the crack tip,
whereas the more ductile resins are able to yield and modify the stress
level on the crack tip fibers [43].
References
1. H. Lee and K. Neville, Handbook of Epoxy Resins, McGraw-Hill,
New York, 1967.
2. C. A. May (ed.), Epoxy Resins Chemistry and Technology, Marcel
Dekker, New York, 1988.
3. Dow Chemical Technical Bulletin 296-346-1289.
4. Shell Chemical Technical Publication SC:6781.
5. C. A. May (ed.), Epoxy Resins: Chemistry and Technology, Marcel
Dekker, New York, 1988, p. 493.
6. H. Lee and K. Neville, Handbook of Epoxy Resins, McGraw-Hill,
New York, 1967, p. 621.
7. N. L. Allinger, M. P. Cava, D. C. De Jongh, C. R. Johnson, N. A.
Lebel, C. L. Stevens, Organic Chemistry, Worth, New York, 1976, p.
514.
Page 382
8. H. Hojo, T. Tsuda, K. Ogasawara, and T. Takizawa, Corrosion
Behavior of Epoxy and Unsaturated Polyester Resins in Alkaline
Solution, ACS Symposium Series 322, American Chemical Society,
Washington, DC.
9. Dow Chemical Technical Bulletins 296-279-585, 296-311-390X,
296-224-790X.
10. W. H. Carothers, J. Am. Chem. Soc., 51:2560(1929).
11. H. V. Boenig, Unsaturated Polyesters: Structure and Properties,
Elsevier, Amsterdam, 1964, pp. 23.
12. Amoco Chemical Company Technical Bulletin IP-70b.
13. T. B. Brown, P. P. Burrell, R. E. De La Rosa, and D. J. Herzog,
Development of a new terephthalic acidbased polyester resin.
Proceedings of The Society of the Plastics Industry (SPI) Composites
Institute, 48th Annual Conference, 1993, 14-A, p. 1.
14. R. Gillette and K. Spoo, An alternative to orthophthalic and
isophthalic based unsaturated polyesters in corrosion applications,
Proceedings of SPI Composites Institute, 44th Annual Conference,
1989, 19-B.
15. H. R. Edwards, The flexibility and chemical resistance of
unsaturated polyestersmeans of testing and their relative merits,
Proceedings of SPI Composites Institute, 30th Annual Conference,
1975, 6-E.
16. S. L. Hillman, M. Palmer, and J. H. Davis, Corrosion resistance of
new unsaturated polyester formulations based on TMPD glycol with
isophthalic acid or dimethyl terephthalate, Proceedings of SPI
Composites Institute, 34th Annual Conference, 1979, 4-F.
17. Ashland Chemical Company Technical Bulletin 2026-1, 1993.
18. H. V. Boenig, Unsaturated Polyesters: Structure and Properties,
Elsevier, Amsterdam, 1964, p. 99.
19. E. Grazul and D. M. Longnecker, Unsaturated Polyester
Technology (P. F. Bruins, ed.), Gordon and Breach, New York, 1976,
pp. 5, 281.
20. T. F. Anderson and V. B. Messick, Developments in Reinforced
Plastics, Vol. 1(G. Pritchard, ed.), Applied Science Publishers,
London, 1980, p. 30.
21. R. E. Young, Unsaturated Polyester Technology (P. F. Bruins,
ed.), Gordon and Breach, New York, 1976, p. 316.
22. L. T. Blankenship, M. N. White, and P. M. Puckett, Vinyl ester
resins: versatile resins for composites, Proceedings of 34th
International SAMPE Symposium, 1989, p. 239, 240.
23. P. D. Kardenetz, V. B. Messick, and I. J. Craigie, Unique vinyl
ester resin provides solutions to tough problems, Proceedings of SPI
Composites Institute, 40th Annual Conference, 1985, 6-C.
24. J. F. Kamody, A. Damiani, and R. J. Stadelman, The use of FRP
with alcoholcontaining fuels, Proceedings of SPI Composites
Institute, 47th Annual Conference, 1992, 3-D.
25. T. W. Cowley, Use of in situ corrosion testing in pulp mill
applications, Proceedings of SPI Composites Institute, 44th Annual
Conference, 1989, 19-E.
26. L. M. Adkins, D. W. Daniel and D. A. Rust, Factors affecting the
fabrication of corrosion barriers for RP equipment in bleaching
environments, Proceedings of SPI Composites Institute, 49th Annual
Conference, 1994, 6-A.
27. A. Knop, and W. Scheib, Chemistry and Application of Phenolic
Resins, Springer-Verlag, Berlin, 1979, p. 1.
Page 383
28. A. Knop and L. A. Pilato, Phenolic Resins: Chemistry,
Applications and Performance, Springer-Verlag, Berlin, 1985, pp.
2458.
29. Rogers Corporation Technical Literature 9582-096-5.OA.
30. BP Chemicals Technical Literature, Chemical Resistance of
Phenolic Laminates.
31. G. Pritchard, Developments in Reinforced Plastics, Vol. 1 (G.
Pritchard, ed.), Applied Science, London, 1980, pp. 1718.
32. Q. O. Chemicals Technical Bulletin 405-A.
33. R. C. Allen, Some corrosion mechanisms in attack of resin and
resin glass laminates, Polym. Eng. Sci., 19(5):329(1979).
34. B. D. Caddock, K. E. Evans, and D. Hull, The diffusion of
hydrochloric acid in polyester thermosetting resins, J. Mat. Sci,
22:3368(1987).
35. E. A. Grulke, Polymer Handbook, 3rd ed. (J. Brandrup and E. H.
Immergut, eds.), John Wiley and Sons, New York, 1989, p. VII/519.
36. E. G. Wolff, Moisture effects on polymer matrix composites,
Sampe J. 29(3): 11(1993).
37. T. Alfrey, T. F. Anderson, Attack of vinyl ester resins by organic
solvents, Proceedings of SPI Composites Institute, 31st Annual
Conference, 1976, 15-E.
38. R. F. Regester, Behavior of fiber reinforced plastic materials in
chemical service, Corrosion, 25:157 (1969).
39. E. P. Plueddemann, Silane Coupling Agents, Plenum Press, New
York, 1982, pp. 8, 115.
40. H. S. Kliger and E. R. Barker, A comparative study of the
corrosion resistance of carbon and glass fibers, Proceedings of SPI
Composites Institute, 39th Annual Conference, 1984, 5-E.
41. P. J. Hogg, The acidic stress corrosion of GRP, Prog. Rubb. Plast.
Technol., 5(2): 136(1989).
42. P. J. Hogg, J. N. Price, and D. Hull, Stress corrosion of GRP,
Proceedings of SPI Composites Institute, 39th Annual Conference,
1984, 5-C.
43. P. J. Hogg, The effect of acidic environments on stressed
glassreinforced plastics, Conference ProceedingsPhase Interaction in
Composite Materials, Omega Scientific, 1992, pp. 5063.
Page 385
14
Chemical Attack and Failure of Elastomers
Philip A. Schweitzer
Fallston, Maryland
The technical definition of an elastomer as given by ASTM states:
An elastomer is a polymeric material which at room temperature can be
stretched to at least twice its original length and upon immediate release of
the stress will return quickly to its original length.
More commonly, an elastomer is generally considered to be any
material that is elastic or resilient and in general resembles natural
rubber in feeling and appearance. These materials are sometimes
referred to as rubbers.
Elastomers are primarily composed of large molecules that tend to
form spiral threads, similar to a coiled spring, that are attached to each
other at infrequent intervals. As a small stress is applied these coils
tend to stretch or compress but exert an increasing resistance as
additional stresses are applied. This property is illustrated by the
reaction of an elastic band.
The maximum utility of elastomers, either natural or synthetic, is
achieved by compounding. In the raw state, elastomers tend to be soft
and sticky when hot and hard and brittle when cold. Ingredients are
added to make elastomers stronger, tougher, or harder; to make them
age better; to color them; and in general to impart specific properties
to meet specific application needs. Vulcan-
Page 386
izing agents are also added since the vulcanizing process extends the
temperature range within which they are flexible and elastic.
Depending on the application of the elastomer, certain specific
properties may be required. The following examples illustrate some of
the important properties that are required of elastomers and the typical
services that require these properties:
Resistance to abrasive wear: automobile tires, conveyor belt covers,
soles and heels (shoes), cables, hose covers
Resistance to tearing: tire treads, footwear, hot-water bags, hose
covers, belt covers, O rings
Resistance to flexing: auto tires, transmission belts, V belts,
mountings, footwear
Resistance to high temperatures: auto tires, belts conveying hot
materials, steam hose, steam packing, O rings
Resistance to cold: airplane parts, automotive parts, auto tires,
refrigeration hose, O rings
Minimum heat buildup: auto tires, transmission belts, V belts,
mountings
High resilience: sponge rubber, mountings, elastic bands, thread,
sandblast hose, jar rings, O rings
High rigidity: packing, soles and heels (shoes), valve cups, suction
hose, battery boxes
Long life: fire hose, transmission belts, tubing
Electrical resistivity: electrician's tape, switchboard mats, electrician's
gloves, wire insulation
Electrical conductivity: hospital flooring, nonstatic hose, matting
Impermeability to gases: balloons, life rafts, gasoline hose, special
diaphragms, stack linings
Resistance to ozone: ignition distributor gaskets, ignition cables,
windshield wipers
Resistance to sunlight: wearing apparel, hose covers, bathing caps,
windshield wipers
Resistance to chemicals: tank linings, gaskets, valve diaphragms, hose
for chemicals, O rings
Resistance to oils: gasoline hose, oil suction hose, paint hose,
creamery hose, packing house hose, special belts, tank linings,
gaskets, O rings, special footwear
Stickiness: cements, electrician's tape, adhesive tapes, pressuresensitive tapes
Low specific gravity: airplane parts, forestry hose, balloons
Lack of odor or taste: milk tubing, brewery and winery hose, nipples,
jar rings, gaskets, O rings
Acceptance of color pigments: ponchos, life rafts, welding hose
Page 387
Since this chapter deals with the corrosion resistance of elastomers,
future discussions will be limited to those properties which affect or
are affected by the corrosion resistant application of elastomers.
I
Causes of Failure
Chemical deterioration occurs as the result of a chemical reaction
between the elastomer and the medium or by absorption of the
medium into the elastomer. This attack results in a swelling of the
elastomer and a reduction in its tensile strength. The temperature and
concentration of the corrodent will determine the degree of
deterioration. Normally the chemical attack is greater as the
temperature and/or concentration of the corrodent increases. Unlike
metals, elastomers absorb varying quantities of the material they are
in contact with, especially organic liquids. This can result in swelling,
cracking, and penetration to the substrate in an elastomer-lined vessel.
Swelling can cause softening of the elastomer and in a lined vessel
introduce high stresses and failure of the bond. Permeation is another
factor which can cause failure of a lining. When an elastomer exhibits
a high absorption, permeation usually results. However, it is not
necessary for an elastomer to have a high absorption rate for
permeation to occur. Some elastomers, such as the fluorocarbons, are
easily permeated but have very little absorption. An approximation of
the expected permeation and/or absorption of an elastomer can be
based on the absorption of water, for which data are usually available.
All materials tend to be somewhat permeable to chemical molecules,
but the permeability rate of some elastomers tends to be an order of
magnitude greater than that of metals. Though permeation is a factor
closely related to absorption, factors which influence the permeation
rate are diffusion and temperature rather than concentration and
temperature. Permeation can pose a serious problem in elastomer-
lined equipment. When the corrodent permeates the elastomer it
comes into contact with the metal substrate, which is then subject to
chemical attack. This can result in
1. Bond failure and blistering, caused by an accumulation of fluids at
the bond when the substrate is less permeable than the lining or from
formation of corrosion or reaction products if the substrate is attacked
by the corrodent
2. Failure of the substrate due to corrosive attack
3. Loss of contents through lining and substrate as the result of
eventual failure of the substrate
The degree of permeation is affected by lining thickness. For general
corrosion resistance, thicknesses of 0.0100.020 in. are usually
satisfactory, depending on the elastomeric material and the specific
corrodent. Thick linings may be required
Page 388
when mechanical factors such as thinning due to cold flow,
mechanical abuse, and permeation rates are taken into consideration.
Increasing lining thickness will normally decrease permeation by the
square of the thickness. However, this is not necessarily the answer to
the problem since increasing the liner thickness can introduce other
problems. As the liner thickness increases, the thermal stresses on the
boundary increase, which can cause bond failure. Temperature
changes and large differences in coefficients of thermal expansion are
the most common causes of bond failure. These stresses are
influenced by thickness and modulus of elasticity of the elastomer. In
addition, the labor cost of installing the liner increases as the thickness
increases.
Temperature and temperature gradient in the liner also affect the rate
of permeation. Lowering these will reduce the permeation rate. Lined
vessels, when used under ambient conditions such as storage tanks,
provide the best service.
Linings can be installed either bonded or unbonded to the substrate. In
unbonded linings it is important that the space between the liner and
support member be vented to the atmosphere to permit the escape of
minute quantities of permeant vapors and also to prevent the
expansion of entrapped air which could cause collapse of the liner.
Although elastomers can be damaged by mechanical means alone, this
is usually not the case. Most mechanical damage occurs as a result of
chemical deterioration of the elastomer. When the elastomer is in a
deteriorated condition, the material is weakened, and consequently it
is more susceptible to mechanical damage from flowing or agitated
media.
Some elastomeric materials are subject to degradation when placed in
outdoor applications as a result of weathering. The action of sunlight,
ozone, and oxygen can cause surface cracking, discoloration of
colored stocks, serious loss of tensile strength, elongation, and other
rubber-like properties. Therefore the resistance to weathering must
also be taken into account when selecting an elastomer, as well as
other corrosion resistance properties, when the material is to be
installed where it will be subject to weathering.
II
Selecting an Elastomer
When the need arises to specify an elastomer for a specific
application, physical, mechanical, and chemical resistance properties
must all be taken into account. The major physical and mechanical
properties which may have to be considered, depending on the
application are
Abrasion resistance
Electrical properties
Compression set resistance
Page 389
Tear resistance
Tensile strength
Adhesion to metals
Adhesion to fabrics
Rebound, cold and hot
Resistance to heat aging and flame
It should be remembered that these properties may be altered by
compounding, but improvement of one property may result in an
adverse effect on another. Because of this it is best to provide a
competent manufacturer with complete specifications and let that
manufacturer provide an appropriate elastomer.
The primary requirement of the elastomer is that it be compatible with
the corrodent to be handled. Therefore all temperatures and
concentrations of the corrodent to which it will be exposed must be
provided.
Specifications should include any specific properties required for the
application, such as resilience, hysteresis, static or dynamic shear and
compression modulus, flex fatigue and cracking, creep resistance to
oils and chemicals, permeability, and brittle point, all in the
temperature ranges to be encountered in service. Table 1 provides
comparative properties of the more common elastomers and Table 2
provides the operating temperature ranges of the common elastomers.
III
Corrosion Resistance
The concentration and temperature of the corrodent is a determining
factor in the capacity of the elastomer to resist attack by the corrodent.
Another important factor is the composition of the elastomer. It is a
common practice in the manufacture of elastomers to incorporate
additives into the formulation to improve certain of the physical
and/or mechanical properties. These additives may have an adverse
effect on the corrosion resistance of the base elastomer, particularly at
elevated temperatures. Conversely, some manufacturers compound
their elastomer to improve their corrosion resistance at the expense of
physical and/or mechanical properties. Because of this it is important
to know whether or not any additives have been used as the corrosion
resistance charts are applicable only for the pure elastomer.
Keep in mind that there are several manufacturers of each generic
compound. Since each may compound slightly differently, the
corrosion resistance may be affected. When a generic compound is
listed as being compatible with a specific corrodent it indicates that at
least one of the trade name materials is resistant to the corrodent, but
not necessarily all. The manufacturers must be checked. Table 3
provides a cross-reference to generic elastomers, trade names, and
manufacturers.
Page 390
(table continued on next page)
Page 391
(table continued from previous page)
Page 392
Table 2 Operating Temperature Range of Common
Elastomers
Temperature range
°F
°C
Elastomer
Min.Max.Min.Max.
NR, natural rubber
-59 175 -50 80
IR, isoprene rubber
-59 175 -50 80
CR, neoprene rubber
-13 203 -23 95
SBR, Buna-S
-66 175 -56 80
NBR, nitrile rubber, Buna-N
-40 250 -40 105
IIR, butyl rubber
-30 300 -34 149
CIIR, chlorobutyl rubber
-30 300 -34 149
CSM, Hypalon
-20 250 -30 105
BR, polybutadiene rubber
-150 200 -101 93
EA, Ethylene-acrylic rubber
-40 340 -40 170
ABR, acrylate-butadiene rubber
-40 340 -40 170
EPDM, ethylene-propylene
-65 300 -54 149
SBS, styrene-butadiene-styrene
150
65
SEBS, styrene-ethylene-butylene- -102 220 -75 105
styrene
ST, polysulfide
-30 212 -34 100
FA, polysulfide
-30 250 -34 105
AU, polyurethane
-65 250 -54 105
-40 300 -40 149
polyamides
PE, polyesters
-40 302 -40 150
TPE, thermoplastic elastomers
-40 277 -40 136
SI, silicone
-60 450 -51 232
FSI, fluorosilicone
-140 375 -73 190
HEP, vinylidene fluoride
-40 450 -40 232
FKM, fluoroelastomers
-10 400 -18 204
ETFE, ethylene-tetrafluoroethylene -370 300 -223 149
ECTFE, ethylene-105 340 -76 171
chlorotrifluoroethylene
FPM, perfluoroelastomers
-58 600 -50 316
Source: Philip Schweitzer, Corrosion Resistance of
Elastomers, Marcel Dekker, New York, 1990.
Page 393
Table 3 Elastomer Cross-Reference
Manufacturers,a common
Generic name Designation
or trade name
Natural rubber NR
2631
Isoprene
IR
Polychloroprene CR
2631, neoprene (1), Bayprene
(2)
ButadieneSBR
2630, Buna-S, GR-S
styrene
ButadieneNBR
16, 2631, nitrile rubber, Bunaacrylonitrile
N, perbunan (2), Nytek (21)
Butyl rubber
IIR
GR-1, 2630, Kalar (19)
Chlorobutyl
CIIR
2630
rubber
XNBR
Carboxylic16, 2631
acrylonitrilebutadiene
2628, 30, 31, Hypalon (1)
ChlorosulfonatedCSM
poly-ethylene
Polybutadiene BR
2628, 30, 31, Buna-85, BunaDB (2)
Ethylene-acrylic EA
13, 28, Vamac (1)
AcrylateABR
13, 28
butadiene
Acrylic
ACM
13, 28
esteracrylic
halide
EthyleneEPDM
2631
propylenediene
EPT
EthyleneNordel (1), Royalene, EPDM
propylene
(8), Dutral (9)
terpolymer
StyreneSBS
Kraton D (3)
butadiene
styrene
StyreneSEBS
ethylenebutylene-styrene
Polysulfide
ST
Polysulfide
FA
Urethane
AU
Polyamides
Nylon
Polyester
Thermoplastic
elastomers
PE
TPE
Silicone rubber SI
Fluorosilicone
Vinylidene
fluoride
FSI
HFP
Kraton G (3)
27, 28, 30, Thiokol (4)
Blak-stretchy (14), Blak-tufy
(14), Gra-tufy (14)
16, 27, 30, 31, 38, Adiprene (1),
Baytec (2), Futrathane (11),
Conathane (16), Texion (2),
Urane (23), Pellethane (22),
Pure CMC (14)
Nylon (1), Rilson (12), Vydyne
(18), Plaskon (25)
Hytrel (1), Kodar (20)
Duracryn (1), Flexsorb (17),
Geolast (18), Kodapak (20),
Santoprene (18), Zurcon (24)
2729, 32, Cohrplastic (15),
Green-Sil (14), Parashield (13),
Baysilone (2), Blue-Sil (14)
Parashield (13)
Kynar (7), Foraflon (5)
(table continued on next page)
Page 394
Table 3 Continued
Generic name
Fluoro elastomers
Manufacturers,a common
Designation
or trade name
FKM
24, 26, 2831, Viton (1),
Fluorel (6), Technoflon
(9)
ETFE
Tefzel (1), Halon-ET (9)
Ethylenetetrafluoroethylene
ECTFE
Ethylenechlorotrifluoroethylene
Perfluoroelastomers FPM
Halar (9)
Kalrez (1), Chemraz (10),
Kel-F (6)
aList of manufacturers: (1) E. I. Du Pont. (2) Mobay Corp.
(3) Shell Chemical Co. (4) Morton Thiokol Co. (5) Atochem
Inc. (6) 3-M Corp. (7) Pennwalt Corp. (8) Uniroyal, (9)
Ausimont. (10) Green, Tweed, & Co. Inc. (11) Futura
Coatings Inc. (12) Attochem Inc. (13) Parker Seal Group.
(14) Ther Perma-Flex Mold Co. (15) CHR Industries. (16)
Conap Inc. (17) Polymer Corp. (18) Monsanto Co. (19)
Hardman Co. (20) Eastman Chemical Products Inc. (21)
Edmont Division of Becton, Dickinson & Co. (22) Dow
Chemical USA. (23) Krebs Engineers. (24) W. S. Shamban &
Co. (25) Allied Signal. (26) General Rubber Co. (27) Hecht
Rubber Co. (28) Minor Rubber Co. (29) Newco-Holz Rubber
Co. (30) Aldan Rubber Co. (31) Burke Rubber Co. (32)
Unaflex.
Page 395
15
Corrosion Resistance of Specific Elastomers
Philip A. Schweitzer
Fallston, Maryland
Table 1 specifies the corrosion resistance of some of the more
common elastomers in contact with specific corrodents. It must be
remembered that the data in the table are based on ''pure" elastomers.
Compounding by various manufacturers may have an effect on the
corrosion resistance. Consequently, the manufacturer should be
consulted to ensure that this situation does not arise.
Because of the effects of compounding, the corrosion resistance of
elastomers from different manufacturers may vary. Compatibility of
an elastomer as shown in Table 1 with a specific corrodent indicates
that there is at least one manufacturer's material that is suitable.
The corrodents shown are either pure compounds or saturated
solutions unless otherwise specified. Temperatures shown are in °F
and are the maximum for which data are available. The elastomers
may be used up to this temperature. Higher operating temperatures
may be permissible but must be verified with the manufacturer.
A "U" in the table indicates that the elastomer is not compatible with
the corrodent at any temperature; a blank indicates that no data are
available.
Page 396
(table continued on next page)
Page 397
(table continued from previous page)
(table continued on next page)
Page 398
(table continued on next page)
Page 399
(table continued from previous page)
(table continued on next page)
Page 400
(table continued on next page)
Page 401
(table continued from previous page)
(table continued on next page)
Page 402
(table continued on next page)
Page 403
(table continued from previous page)
(table continued on next page)
Page 404
(table continued on next page)
Page 405
(table continued from previous page)
(table continued on next page)
Page 406
(table continued on next page)
Page 407
(table continued from previous page)
(table continued on next page)
Page 408
(table continued on next page)
Page 409
(table continued from previous page)
(table continued on next page)
Page 410
(table continued on next page)
Page 411
(table continued from previous page)
(table continued on next page)
Page 412
(table continued on next page)
Page 413
(table continued from previous page)
(table continued on next page)
Page 414
(table continued on next page)
Page 415
(table continued from previous page)
(table continued on next page)
Page 416
(table continued on next page)
Page 417
(table continued from previous page)
Page 419
16
Aqueous Corrosion of Advanced Ceramics
Eugene L. Liening
The Dow Chemical Company, Midland, Michigan
James M. Macki
Materials Technology Institute of the Chemical Process Industries,
Inc.
St. Louis, Missouri
I
Introduction
A
Overview
This chapter describes important aspects of advanced ceramics
corrosion in aqueous media, their general corrosion resistance
properties, and how to measure behavior and report the results.
Specifically considered are fundamental ways in which ceramic and
metallic corrosion differ, some basic concepts of corrosion behavior,
common forms of corrosion of ceramic materials, and simple
thermodynamic tools for evaluating corrosion of ceramics.
The performance of advanced ceramics in aqueous environments is
not well documented and the corrosion test procedures are not yet
standardized. All published corrosion data for advanced ceramics are
suspect with regard to comparing results from different investigators.
Because of the lack of standard testing procedures and inconsistent
corrosion data documentation, no attempt is made here to provide
tabulated information recommending materials for corrosion service
application. As is discussed below, each generic type of advanced
ceramic has several variations, all of which may differ in corrosion
resistance. Therefore, this chapter briefly summarizes what advanced
ceramics are and some of the prob-
Page 420
lems in measuring their corrosion, and then shows how corrosion
testing should be carried out and how the results should be reported.
B
Background
Early civilizations developed ceramic technologies based on local
clays to fabricate articles for domestic and early trade use. More
recent industrial civilizations used more specialized technologies
based on clays and refined minerals to fabricate articles for industrial
applications such as acid brick, building bricks, pipes, containers, and
so forth. Today a new generation of ceramics is emerging based on
ceramic powders made by controlled chemical reactions to form pure
compounds; these powders are fired to produce high-value, highperformance ceramics for very specific applications for our global
marketplace. These new ceramics are known as advanced ceramics,
fine ceramics, technical ceramics, engineering ceramics, and highperformance ceramics.
The ancient ceramic technologies based on clays and refined materials
are still used extensively for commercial and domestic applications;
however, this chapter deals only with aqueous corrosion of advanced
ceramics. Ceramics are traditionally used for high-temperature and
wear service, and these applications are well documented. Limiting
the scope to advanced ceramics in aqueous service focuses this
chapter on less documented applications where proper corrosion
testing is critical to successful performance.
Examples of advanced ceramics are reaction-bonded silicon carbide,
sintered a-silicon carbide, partially stabilized zirconia, reactionbonded silicon nitride, hot-pressed boron carbide, high-conductivity
aluminum nitride, 99% alumina, etc. Generic terms like silicon
carbide are not sufficient to define advanced ceramics, and even the
above examples are somewhat generic because a range of differently
performing materials fit within each example.
C
Ceramics Are Brittle
The Achilles heel of advanced ceramics is their inherent brittleness
compared to most metals. Civil engineers have used brittle materials
for millennia, but their use by chemical and mechanical engineers has
been, and is, restricted by this lack of ductility. Successful structural
use in chemical environments depends on predicting the probability of
mechanical failure using reliability statistics. This requires designers
to use property distributions, such as Weibull modulus, instead of
average properties. The Weibull modulus approach is based on the
weakest link theory, which assumes that a stressed ceramic fails at the
most severe flaw. Richerson provides an excellent introduction to
ceramic design in chapters 14 and 15 of his book [1].
Using a Weibull modulus and probability of failure leads to design
strength results that may appear contradictory. For example, Parker
[2] shows that for a 1
Page 421
in 106 failure probability, a material with a Weilbull modulus of 20
and a strength of 200 MPa is equivalent to an apparently much
stronger material with a Weibull modulus of 7 and a strength of 700
MPa.
Testing advanced ceramics for structural service in corrosion
environments should consider the effects of the environmental
exposure on the strength and Weibull modulus of the material in
addition to the traditional measurements.
D
What Are Advanced Ceramics?
Traditional ceramics made from clay and/or refined materials are
usually prepared by mixing the powder with water and additives,
forming a shape, drying the part, and sintering at high temperatures to
form a densified part. Advanced ceramics are fabricated by processing
ranging from the traditional methods to very complex processes. The
powder is made into greenware by several processes, including slip
casting, injection molding, tape casting, or cold-pressing; the
greenware is then densified by firing, again using one or more specific
processes. Alternatively, the powder can be densified by combining
the forming and firing operations; for example, in hot-pressing the
powder is pressed in a mold while at the firing temperature. In
addition, the powder can be formed into a complex, but porous, shape
and densified by reaction with a liquid or gas to form additional
bonding ceramic phase in situ. The best process depends on the
application, the material, the shape, and the economics.
The term ''firing" in this chapter generically refers to the process of
densifying an advanced ceramic powder to full density. This
simplification is necessary because there are many processing options
for densifying advanced ceramic powders. These options include
forming, drying, and sintering as with traditional ceramics; combining
the sintering and forming steps by hot-pressing, forming, drying,
sintering, and then pressure-sintering as a final step (hot isostatic
press, or HIP); forming, drying, preheating, and isostatically forging
(rapid omnidirectional compaction, or ROC), etc. The actual process
and processing conditions are important because they determine the
microstructure, which affects corrosion and other properties.
Reaction-sintered or reaction-bonded silicon carbide (RBSC) is an
example of a nontraditional processing method. In this process, silicon
carbide and carbon powders are mixed and formed into a shape. The
shaped part may then be machined to more exact dimensions. The
machined part is fired in a furnace while in contact with liquid silicon
metal. The silicon metal reacts with the free carbon in the part to form
more silion carbide and thereby forms a bonding phase. This is a
common method for making high-performance silicon carbide parts,
but its corrosion behavior is usually controlled by the residual
unreacted silicon metal that is always present after reaction bonding
[3].
Hot-pressed silicon nitride (HPSN) can be processed by hot-pressing a
silicon nitride powder mixed with yttria (Y2O3) powder at 1850°C.
The oxide
Page 422
reacts with the impurity silica (SiO2) that is always present in silicon
nitride to form a glass phase that promotes densification and bonds the
structure together. Again, the nature of the residual glass phase may
be key in determining the corrosion behavior of the ceramic.
E
Typical Physical and Mechanical Properties
Table 1 shows a compilation of physical and mechanical properties of
some common advanced ceramics and, for comparison, of some
metals. The difference among the various versions or grades for a
generic type is illustrated by the three silicon nitride versions, two
silicon carbide versions, and two partially stabilized zirconia (PSZ)
versions.
II
Comparison of Ceramics and Metals Corrosion
Aqueous corrosion of advanced ceramics takes forms that are similar
to those found in metals, so many of the same principles apply. For
example, selective
Table 1 Typical Properties of Selected Advanced
Ceramics, Metals, and Cermetsa
Material
Si3N4
HPSN[4]
RBSN[4]
SSN[5,6]
SiC
S-SiC[4]
SA-SiC[5][7]
Electrical
Flexural
Elastic resistivity
Density strength Weibull modulus(ohm(g/cm3) (MPa) modulus (GPa) cm)
3.2
2.5
3.25
700
235
650
12
10
310 >1011
180 >1011
290 1011
3.1
3.15
400
550
4
10
400 0.1
440 102106
Al2O3
345 >1015
3.97 375
Al2O3[7]
360 1013
3.98 400
Al2O3[6]
ZrO2
PSZ-MS[4]
21
205
5.75 690
PSZ200
6.0
1000
5%Y2O3[6]
Nonceramics
WC/6% Co [6] 15.0 1790
615
4140 Steel [8] 7.8
200
1740b
Cast iron [6]
117
7.2
500
aTypical properties or average of typical property range
reported.
bYield strength of quenched and tempered steel.
Page 423
leaching in ceramics is analogous to selective leaching of zinc from
brass alloys, erosioncorrosion occurs with both advanced ceramics
and metals, etc. However, there are other areas in which ceramics
differ from metals; for example, many ceramics are not electrically
conductive. Some significant differences between ceramics and metals
are particularly relevant to corrosion behavior and need elaboration
before further discussion.
A
Description of the Ceramic
ASTM or AISI specifications usually adequately identify metallic
alloys, i.e., 304 stainless steel is essentially the same regardless of
manufacturer. In the case of advanced ceramics, there are few
standard compositions, and it is often necessary to cite a fair amount
of detail to adequately describe the ceramic.
Not only the composition but also the amounts and type of phases
may be important. Many ceramics contain a mixture of glassy and
crystalline phases that have dramatically different corrosion
resistances. Apparently similar ceramics may have very different
corrosion behaviors because of a few percent more of a particular
phase.
The type of processing is also important in some ceramics. For
example, reaction-bonded silicon carbide has dramatically different
corrosion resistance than sintered silicon carbide in many corrosives.
B
Porosity
Porosity is uncommon in metallic alloys, but 12 vol % voids are
common in advanced ceramics. Porosity may greatly affect the
corrosion resistance of a ceramic by providing greater surface area for
corrosive attack, affecting weight loss test results by retaining
corrosive or corrosion product, and by promoting mechanical failure
by formation of voluminous corrosion product in the pores.
C
Brittleness
The brittleness of ceramics compared to that of metals means that the
geometry of the part is often much more important for ceramics.
Square corners and stress concentrations are more detrimental for
ceramics, and chipping and spalling may be the result of corrosive
action as well as mechanical effects.
D
Electrical Conductivity
While metals are electrically conductive, many ceramics are not. This
property has a large impact on corrosion behavior. Nonconductive
ceramics do not participate in galvanic couples, so contact with
another material does not accelerate or retard corrosion. Conductive
ceramics behave like metals in this regard, and their corrosion rates
could either increase or decrease as the result of contact
Page 424
with another material. Conductive ceramics also are more prone to
corrosion rates within pits and crevices that increase with time as
described below. From a testing perspective, conductive ceramics
have the advantage of being evaluable using the same types of
electrochemical corrosion testing techniques used for metals [9].
A ceramic may contain several phases, only some of which are
conductive. In this case, the galvanic behavior depends on the amount
and continuity of the conductive phases. If the conductive phase is
continuous in the ceramic matrix, the ceramic is capable of galvanic
behavior. If the conductive phase is not continuous (e.g., discrete
conductive particles in a nonconductive matrix), then the ceramic will
tend not to be affected by contact with other materials. In the latter
case, however, there still could be local galvanic corrosion in the
immediate vicinity of the contact, with galvanic corrosion of only
those particles of conductive phase that directly contact the other
material.
The electrical conductivities of many carbides are given by
Kosolapova [10]. Metal carbides generally exhibit at least some
metallic conductivity and nonmetallic carbides generally exhibit lower
semiconductor conductivity. Metallic conductivity is indicated by
positive coefficients of specific resistivity. Boron carbide and silicon
carbide are exceptions to the metallic conductivity generalization.
Both exhibit semiconductor conductivity as indicated by negative
coefficients of specific resistivity. The method of manufacture may
affect the quantitative conductivity of carbides.
III
Generalized Corrosion Behavior of Ceramics
A
Comment on Corrosion Rate Units
The corrosion rates shown in this section are expressed in µmpy
(micrometers per year) because commonly used units, such as mg/cm2
per yr, are misleading. The large differences in density among
ceramics requires that corrosion data be normalized for density for
meaningful comparisons. This is elaborated on later in this chapter, in
the section on corrosion testing and evaluation.
For example, comparing the corrosion rates of cobalt-bound tungsten
carbide (WC/Co) (density of 15 g/cm3) and boron carbide (density of
2.6 g/cm3) by using weight loss per unit area would improve the
apparent corrosion resistance of the boron carbide relative to WC/Co
by a factor of 5 compared to the actual dimensional loss rate. Since
the primary interest is the life of the equipment used, corrosion should
be expressed as loss of dimension per unit time, or penetration rate.
Fontana [11] suggests that a good corrosion rate expression should
involve (1) familiar units, (2) simple calculation, (3) ready conversion
to life in years, (4) penetration, and (5) whole numbers without
cumbersome decimals. He advocated using mils penetration per year
where a mil is 0.001 in.; however, in the SI or metric system, units of
micrometers per year also meet
Page 425
Fontana's criteria. Micrometers per year, or µmpy, is used in this
chapter and earlier data that are referenced are converted to µmpy.
The corrosion rate for ceramics is frequently reported as mg/cm2 per
yr, which converts to µmpy using the expression µmpy = (0.84)
(mg/cm2 per yr)/D, where D is the density in g/cm3.
B
Other Corrosion Expressions
Lay [12] uses an ABC grading system to tabulate the corrosion
performance data for common advanced ceramics in a variety of acid
and alkaline solutions. This tabulation provides useful generalized
information; however, these data cannot be used without other
information because the individual tests supporting each reported
datum were not necessarily performed in the same way, the test
materials are not well defined, the test conditions are not defined, and
so forth.
C
Table Summarizing Published Corrosion Data
Table 2 shows published corrosion data for some advanced ceramics
and a WC/6%Co cermet. These data have been published in several
articles about corrosion, but with the corrosion rates expressed as
corrosion weight loss in mg/cm2 per yr. In this table, the rates are
converted to micrometers penetration per year (µmpy) using the
formula:
Where D is density of the ceramic. The test times were 125300 h of
immersion in the stirred test solutions. Rates with a > prefix were
reported at > 1000 mg/cm2
Table 2 Comparison of Corrosion Performance
Test solution
Corrosion rate (µm/yr)
Si/SiC (RBSC)
12% Si
15
>280
2
2
<1
>280
<1
>280
Wt %
°C
100
98% H2SO4
50% NaOH
100
53% HF
25
100
85% H3PO4
100
70% HNO3
45% KOH
100
25% HCl
70
10% HF + 57%
25
HNO3
Source: Data from Refs. 5, 1316.
Sic
(SASC)
WC/6% 99% No Free
Co
Al2O3
Si
>56
14
<1
<1
16
<1
<1
4
<<1
3
>210
<<1
>56
1
<<1
<1
13
<<1
5
15
<<1
>56
3
<<1
Page 426
per yr and were completely destroyed during the test. Rates with a <
prefix were < 1 µmpy after conversion. The rates with a << prefix
were for SASC-SiC with originally reported rates of < 0.2 mg/cm2 per
yr (< 0.05 µmpy) and no observed corrosion except for that caused by
surface cleaning.
Table 2 shows why corrosion rates should be reported as penetration
rates rather than as weight loss rates. For example, the samples that
were consumed during the test were originally reported as corrosion
rates > 1000 mg/cm2 per yr, but when densities are considered the
rates range from > 56 to > 280 µmpy. Similarly, the relative corrosion
rates in 25% HCl for WC/6%Co and 99% Al2O3 are reversed when
weight loss is converted to penetration:
Corrosion
Corrosion rate
rate
Material (mg/cm2 per yr) (µmpy)
WC/6%Co
85
5
99%
72
15
Al2O3
The rates of 85 and 72 mg/cm2 per yr suggest similar corrosion
resistance, but the 5 and 15 µmpy rates clearly show that the
penetration rate differs by a factor of 3.
D
Additives and Impurities Determine Corrosion Resistance
Single-phase monolithic advanced ceramics generally do not exist.
The pure compound is modified by impurities and by adding materials
that are required for processing or for obtaining the required
properties after firing. These additives may stabilize a specific phase,
promote densification of the powder, control grain growth, improve a
desirable property, etc. (Metastable phases may also be locked into the
structure by processing conditions.) For example, magnesia (MgO),
calcia (CaO), or yttria (Y2O3) is added to zirconia to stabilize hightemperature phases and thereby to improve mechanical properties.
This produces a two-phase microstructure with a stabilized hightemperature phase and a partially transformed low-temperature phase.
Selective degradation of the stabilized phase has been observed in
aqueous environments [17].
A typical example of how the residual intergranular phase controls the
corrosion behavior of an advanced ceramics is RBSC in hot alkali
service. The silicon carbide phase is not affected, but the residual
silicon metal phase is dissolved. However, RBSN is effectively used
in environments where the silicon metal is more stable, such as
hydrochloric acid, nitric acid, sulfuric acid, and hydrofluoric acid [3].
Since intergranular corrosion due to less corrosion-resistant grain
boundary phases is common, the corrosion resistance of advanced
ceramics generally improves as its purity increases. For example, the
grain boundary glass phase in
Page 427
95% Al2O3 alumina renders the ceramic unusable for some hot acid
service, while 99.5% Al2O3 alumina is unaffected in the same service.
Very-high-purity alumina is used to make translucent alumina for
sodium lamps. Translucent alumina has clean grain boundaries and
exhibits corrosion resistance similar to single-crystal alumina
(sapphire) [3].
In contrast to the alumina example above (where the corrosion
resistance is improved by decreasing the amount of less corrosionresistant grain boundary glass phase), the corrosion resistance can be
improved by additions that make the grain boundary phase more
corrosion-resistant. For example, 95% Al2O3 with magnesia instead
of calcia forms grain boundary glass phases that are resistant to acid
attack. In acid service where high-purity 99.5% Al2O3 is partially
attacked, the 95% Al2O3 with magnesia is not attacked [3].
E
Importance of Documenting Specific Material Tested
The terms alumina, zirconia, silicon carbide, boron carbide, silicon
nitride, and aluminum nitride are de facto generic names for advanced
ceramics in the same way that the term stainless steel refers to a large
family of alloys. These ceramic materials are pure compounds with
impurities and additives that affect the processability of the powders
and the properties of the fired ceramic. Therefore, in corrosion testing,
it is important to document the specific ceramic being tested; this
documentation should include as much of the processing history of the
ceramic as possible. The requirement for processing information for
ceramics is similar to stainless steels where the heat treatment of
welded T304 stainless steel is required or where the temper of some
aluminum and titanium alloys is required. The need to fully identify
the ceramic is discussed in more detail in the section on corrosion
testing and evaluation of ceramics.
F
Effect of Environmental Exposure on Performance
Advanced ceramics are used where one or more of their specific
properties are required. In addition to chemical resistance, these
specific properties may be hardness, wear, specific gravity, modulus,
electronic properties, etc. The problem is that exposure to an aqueous
environment may degrade the required property while producing no
obvious chemical corrosion. For this reason, corrosion testing of
advanced ceramics should include testing for the required properties
before and after the corrosion test. Corrosion test results should report
the effect of the corrosion test exposure on the important mechanical
and physical properties.
IV
General Corrosion Chemistry
The corrosion resistance of monolithic advanced ceramics is governed
by the general dissolution chemistry of the chemical compound
forming the ceramic. For
Page 428
example, acidic compounds (SiC) and oxides (SiO2) tend to dissolve
in bases and tend to be stable in acids, whereas basic oxides (MgO)
tend to dissolve in acids and be stable in bases. Corrosion engineers
traditionally use silica-based glass and glass linings for acid service
and avoid it for alkaline service. This is the relative resistance to
attack of some common ceramic compounds (from Lay [3]):
Note that the mixed ceramics, such as ZrSiO4, exhibit corrosion
properties intermediate between the constituents of the spinel, SiO2
and ZrO2. Since advanced ceramics typically contain additives and
metastable phases, the actual resistance of the material can be
different from that of the pure compounds.
A
Silicon Carbide Ceramics
Pure silicon carbide powder is resistant to most acids [18]. The
powder does not dissolve in hydrochloric, sulfuric, hydrofluoric, and
nitric acids; however, some data show that it will react with mixtures
of hydrofluoric and nitric acids. Densified silicon carbide that contains
free silicon metal corrodes in alkaline solutions due to the reaction:
Similarly, free silicon metal is attacked by hydrofluoric acid according
to the reaction:
The free silicon metal is not deleterious to the corrosion of silicon
carbide in sulfuric and hydrochloric acids, and SiC is used extensively
in this service.
B
Boron Carbide Ceramics
Pure boron carbide is insoluble in hydrochloric, sulfuric, and nitric
acids, even after prolonged boiling. However, studies show that the
resistance of hot-pressed boron carbide to sulfuric acid and sodium
hydroxide depends on the additive and impurity content [18]. Strong
oxidizing acids and oxidizing mixtures may cause some corrosion by
oxidizing the free carbon present in the boron carbide.
Page 429
Oxidizing alkaline solutions containing hydrogen peroxide or bromine
also oxidize free carbon.
C
Silicon Nitride Ceramics
Silicon nitride powder is resistant to sulfuric acid, hydrochloric acid,
and nitric acid; but aqua regia, hydrofluoric acid, and phosphoric acid
attack silicon nitride above 100°C [18].
Additives and impurities significantly affect the corrosion resistance
of silicon nitride. For example, the corrosion of hot-processed silicon
nitride in sulfuric acid increases by two orders of magnitude as the
CaF2 content increases from 1% to 20%.
Silicon nitride is highly resistant to alkaline solutions below 50%
concentration. For example, HPSN and RBSN show a mass loss of
only 0.1% after exposure to 20% NaOH at 20°C for 1000 h [18].
D
Aluminum Nitride Ceramics
Aluminum nitride is noted for its combination of high thermal
conductivity combined with high electrical resistance. This leads to
thermal management applications in electronics where it replaces
plastic and alumina in certain applications.
Aluminum nitride hydrolyzes in water according to the reaction
High-purity aluminum nitride is less reactive with water than lower
purity grades [18]. The reaction can also be controlled by surface
treatments. In general, the reactivity of aluminum nitride is only
significant for highly reactive powder and is not a problem for fully
dense aluminum nitride.
E
Aluminum Oxide Ceramics
Pure alumina, Al2O3, exhibits good corrosion resistance to acids and
moderate resistance to alkalis [18]. As with other advanced ceramics,
the corrosion resistance of alumina is dependent on the minor phases
and processing. For acid service, magnesia or silica is added to
alumina to improve the acid resistance of the glassy phase formed
during sintering. This addition must be controlled because silica at
higher concentrations forms silicate phases in the grain boundaries
that are soluble in the acid.
F
Zirconium Oxide Ceramics
Zirconium oxide is thermodynamically more stable than most other
advanced ceramics, but it is seldom used in the pure form. The most
common zirconia
Page 430
ceramics are partially stabilized zirconia (PSZ). Percent levels of
calcia, magnesia, or yttria are added to zirconia to stabilize the hightemperature cubic or tetragonal phase formed during sintering.
"Partially stabilized" means that the additives are insufficient to
completely stabilize the high-temperature phase. The stabilizing
oxides tend to reduce the corrosion resistance in basic solutions.
V
Basic Corrosion Behaviors
From a fundamental point of view, resistance to corrosion comes from
one of three basic behaviorsimmunity, passivation, and kinetically
limited corrosion.
A
Immunity
Immunity refers to the lack of a reaction between a material and its
environment, i.e., the ceramic is inert or thermodynamically incapable
of spontaneously reacting with its environment. This type of corrosion
resistance can be predicted by thermodynamic calculation if the
necessary thermodynamic data are available. Plots of potential vs. pH
have been published by Pourbaix [19] for metals in water showing
regions of immunity. In principle, similar plots could be made for
specific ceramic/environment combinations. Immunity is uncommon
in metals exposed to corrosives, being limited for practical purposes
to precious metals such as gold. Indeed, gold is the only metal that
exhibits immunity in room temperature air. Immunity is more
common with ceramic materials.
B
Passivation
Passivation is the limitation of corrosion by the formation of a
protective corrosion product. It is a special case of kinetically limited
corrosion, which is discussed below. Passive films on metals normally
are tight, adherent, and completely cover the material. Such films are
often so thin as to be transparent and not obvious to an observer. In
some cases passive films become thick enough to display color
patterns from light wave interference. Metallic examples of
passivation include titanium in oxidizing acids and stainless steels in
atmospheric exposure. Passivation can also occur by the formation of
thick and visually obvious corrosion product, as long as it is highly
protective. Such behavior might reasonably be expected in some
ceramic/environment combinations. It is important to recognize that
ceramics that passivate are not inert, even though they may appear so.
C
Kinetically Limited Corrosion
Many ceramics are not inert, do not effectively passivate, and yet
exhibit only low corrosion rates. These ceramics may develop visible
corrosion product that
Page 431
is not highly protective or a corrosion product that is not apparent
because it is soluble in the corrosive. Corrosion in this case is limited
by the slow kinetics of the ceramic/environment system.
Diffusion is a common kinetic limitation to corrosion. Limitation
could be from transport of a corrosive specie through the solution to
the surface of the ceramic. Limitation also occurs by the development
of corrosion product on the surface that impedes access of corrosive to
the ceramic. Activation-limited corrosion occurs when a low
thermodynamic driving force limits the corrosion rate (or, more
precisely, when there is a relatively high activation energy compared
to the driving force for the reaction).
Understanding kinetic limitations to corrosion is important because it
allows one to predict how changes in the environment affect corrosion
behavior. For example, if corrosion is limited by transport of a
chemical specie to the ceramic surface, then corrosion will be
increased by greater velocities and agitation, but will be relatively
unaffected by temperature increases. On the other hand, corrosion
limited by a low thermodynamic driving force will be highly
dependent on temperature changes but will be relatively insensitive to
changes in velocity. Corrosion limited by a partially protective
corrosion product is likely to be accelerated by conditions that
promote removal of the corrosion product, such as impingement or
introduction of flowing solids.
VI
Forms of Corrosion
In this discussion the types of corrosion experienced by ceramics are
grouped into uniform corrosion, various forms of localized corrosion,
and corrosion-assisted cracking. Many forms of ceramic corrosion are
similar to those for metallic corrosion. However, the authors do not
presume the reader to be familiar with the common forms of metallic
corrosion. Important differences in corrosion mechanisms and
behavior between metals and ceramics are emphasized.
A
Uniform Corrosion
Uniform corrosion is one of the most common types of attack and is
characterized by a more or less uniform loss of material. There is a
continuum between uniform and localized corrosion, with the
distinction being either arbitrary or a matter of identifying the
corrosion mechanism as one typically regarded as a localized
phenomenon.
Simple chemical dissolution is the typical mechanism for uniform
corrosion. The corrosion product is nonprotective, being either
soluble, poorly adherent, or a good transport medium for corrosive
species to the ceramic surface. Uniform corrosion has the advantage
of being predictable based on experience or test data. As such,
depending on the type of service, relatively high rates of
Page 432
uniform corrosion may be economically acceptable by designing into
the part an allowance for corrosion.
B
Crevice Corrosion
Crevice corrosion develops within an occluded area of a corroding
material. Once a crevice ''activates" (i.e., begins corroding), corrosion
rates within the occluded area often increase with time. This increase
in corrosivity is caused by one or both of the following factors: (1)
depletion of a passivating component within the occluded area and (2)
increasing acidity within the occluded area. The occlusion may be
geometric as in a deep tight crevice, or may be caused by a deposit of
corrosion product or scale. Here the term "crevice" denotes either
case.
The mechanism of crevice corrosion for ceramics is similar to that for
metallic corrosion, as described in Fontana [20]. However, because of
the nature of ceramics, there are some significant differences. One of
these is that increasing acidity within a crevice area only occurs if the
ceramic is electrically conductive. The reason becomes clear after
reviewing the mechanisms at work.
The most straightforward mechanism is simple depletion of a
passivating specie in the corrosive. This mechanism applies equally to
electrically conductive and nonconductive ceramics. In this case the
corrosion reaction consumes the passivating specie inside the crevice,
making the crevice environment significantly more corrosive than the
bulk solution. The occluded nature of the crevice prevents
replacement of the passivating specie by diffusion, and so the
environment within the crevice remains significantly more corrosive.
As corrosion proceeds, the corroded crevice area grows, and more
corrosion product forms to further restrict diffusion of the passivating
specie into the crevice.
Given this simple mechanism, it follows logically that whether a
given combination of ceramic/environment/crevice activates (i.e.,
begins corroding faster than in the bulk solution) depends upon both
the initial corrosion rate consuming the passivating specie and the
diffusion rate of that specie into the crevice. If the corrosion consumes
the passivating specie faster than diffusion replaces it, then the crevice
eventually activates. It follows from this that crevice corrosion is
more likely when the crevice is more severe (i.e., more occluded), the
passivating specie is less mobile (diffuses more slowly), and the initial
corrosion rate is higher (but not so high as to open the crevice by
corrosion).
In the case of nonconductive ceramics, the crevice corrosion
mechanism stops here. One expects relatively little acceleration of
corrosion rate after the crevice totally loses passivity.
For conductive ceramics, however, an additional mechanism operates
which causes the crevice to become more acidic and corrosive with
time, resulting in increasing rates of corrosion as time passes. This
mechanism may operate on its
Page 433
own or in addition to the mechanism described above. The following
example of a ceramic attacked by an aqueous corrosive illustrates this
mechanism.
Corrosion initially proceeds equally within and outside the crevice by
reaction (1). Cer is any electrically conductive ceramic. Cer+
represents any sort of cationic ceramic corrosion product.
Cer+ reacts within the crevice to form a corrosion product by reaction
(2). The corrosion product is represented by Cer+OH-. (Though a
ceramic corrosion product will typically be more complex than a
simple hydroxyl compound, this illustrates the mechanism.)
Conservation of charge requires that every electron produced by
reaction (1) be consumed by a reduction reaction. Some electrons are
consumed within the crevice by reduction of dissolved oxygen to
hydroxyl ions [reaction (3)] and reduction of H+ to hydrogen [reaction
(4)]. However, some electrons escape from the crevice area by
moving through the ceramic to a surface outside the crevice where
they are consumed by reduction reactions in the bulk solution.
For every mole of electrons that escape the crevice, a mole of H+ (i.e.,
acidity) accumulates within the crevice. This excess H+ is initially
neutralized by hydroxyl ion from oxygen reduction, but dissolved
oxygen within the crevice is soon depleted because of the small
volume of corrosive within the crevice and the restricted diffusion of
dissolved oxygen into the crevice. Once the dissolved oxygen is
depleted, excess H+ accumulates within the crevice, accelerating the
corrosion rate, which accelerates the accumulation of H+. The
accumulating acidity is prevented from leaving the crevice by the
restricted diffusion. This acidity is also a charge imbalance between
the crevice and the bulk solution, and it may electrostatically attract
corrosive anions from the bulk solution into the crevice. In this
manner, the crevice becomes progressively more acidic and corrosive,
and corrosion rates progressively increase with time.
This accumulation of acidity is possible only by the movement of
electrons through the ceramic, from the crevice area to the bulk
solution. That is why this mechanism is only operable with
electrically conductive ceramics. Because of this additional
mechanism, crevice corrosion rates tend to be more severe and
increase more dramatically with time for electrically conductive
ceramics than for nonconductive ceramics.
Page 434
Because of the time required to deplete passivating species and/or to
accumulate acidity within a crevice, there is an initiation period before
a crevice activates. Once activated, however, the crevice may corrode
at high rates. For this reason one must exercise care in drawing
conclusions about rates of crevice corrosion and extrapolating a useful
service life based on an exposure period that includes an initiation
time. Typically, crevice corrosion is dealt with by changing to a
material that is not susceptible to this type of corrosion. In some cases
crevices can be eliminated, but this is often not practical for other than
simple parts.
C
Pitting
Pitting is a form of highly localized corrosion typified by holes in
open (i.e., not creviced) areas of a material. Pits often form as the
result of a corrosive specie damaging a protective film, although pits
also initiate from other types of damage to protective films. Such
damage could occur by mechanical abrasion or thermal effects, or
from microstructural faults such as inclusions or pockets of a
susceptible phase. In a pitting environment, damage to the film is
followed by attack at the damage site, which causes a hole to develop
in the surface. For electrically conductive ceramics the pitting may be
driven by galvanic differences between the active pit and the passive
surface. A good example of pitting corrosion of a ceramic material is
cited by Yoshio and Oda [21], in which silicon nitride developed
classic pit morphology as the result of exposure to high-temperature
water.
Pits have a range of densities and shapes. Pit density (i.e., number of
pits per unit area) ranges from widely scattered pits to pitting that is so
dense the pits begin to overlap. In the latter case, the distinction
between dense pitting and irregular uniform corrosion is somewhat
arbitrary. Individual pits also vary widely in shape, from broad open
pits to deep narrow pits.
Broad open pits with no occluding corrosion products may continue to
corrode at a relatively uniform pace or even slow if corrosion is
limited by the cathodic reaction rate on the passive surface. Occluded
pits, such as deep narrow pits or those forming crusts of corrosion
product, behave for all practical purposes as crevices. The
mechanisms discussed above for crevice corrosion apply in this case,
and one should anticipate similar behavior. As described for crevices,
occluded pits in electrically conductive ceramics are more likely to
experience increasing corrosion rates with time. Open clean pits in
nonconductive ceramics are more likely to experience relatively stable
corrosion rates.
Depending on the application, pitting can be highly detrimental to the
function of a ceramic component even when there is relatively little
actual loss of ceramic. Pitting corrosion rates that are merely
calculated from weight loss data are misleading. Actual measurement
of pit depths by a gauge or microscopy
Page 435
is normally done. Like crevice corrosion, there is an initiation time for
pitting, and any extrapolation of service life based on test samples
needs to take this into account. Pitting may also decrease the effective
strength of a ceramic because the pits are surface defects.
D
Cavitation
Strictly speaking, cavitation is a mechanical phenomenon, not an
aqueous corrosion phenomenon. However, it is discussed here
because it is similar in appearance to pitting, has been reported in
ceramics in connection with corrosion [6], and sometimes occurs
synergistically with corrosion.
Cavitation is the mechanical removal of material caused by the
implosion of vapor bubbles in a liquid. This occurs most readily at
low-pressure areas in equipment handling liquids, such as trailing
edges of impeller vanes, and downstream of valve components that
obstruct flow. Cavitation often has the appearance of clean, open pits,
sometimes elongated in the direction of liquid flow. There is generally
no occlusion of the pits by corrosion product.
Corrosion sometimes promotes cavitation damage of ceramics by
dissolving a matrix phase that holds hard particles. For example,
Parrott [6] reports a case in which "pitting" of a reaction-bonded
silicon carbide seal ring appeared to have been actually caused by
loosening of silicon carbide particles by corrosion of the silicon
matrix, followed by removal of silicon carbide particles by cavitation.
Understanding the role of corrosion in what appears to be cavitation
damage is important because normal fixes for cavitation damage will
not be effective if corrosion is at the root of the problem.
As corrosion may promote cavitation damage, so cavitation damage
may promote corrosion. Ceramics that are not susceptible to corrosion
because of protective corrosion products or passive films may have
the protective layer damaged by cavitation, allowing corrosion pitting
to develop. In such cases, corrosion product is usually found in the pit
bottoms, allowing this to be distinguished from pure cavitation.
E
Erosion and Erosion-Corrosion
Erosion and erosion-corrosion involve the mechanical removal of
either the ceramic, components of a ceramic, or corrosion product of a
ceramic. As such, only erosion-corrosion is truly an aqueous corrosion
phenomenon. However, both are briefly discussed here because they
are similar in appearance and frequently occur together.
Erosion is the direct mechanical removal of a ceramic, or component
of a ceramic, by a fluid or moving particle. Erosion-corrosion is
similar, except that the fluid or moving particle removes a corrosion
product, not the actual ceramic. In this case the ceramic develops what
would otherwise be a protective corrosion
Page 436
product, followed by removal of the corrosion product by erosion,
followed by more corrosion of the newly exposed ceramic. The cycle
repeats, resulting in localized ceramic loss in a flow pattern looking
very much like ordinary erosion.
Although the mechanisms are similar, it is important to recognize the
difference between erosion and erosion-corrosion because the
remedies differ. A more corrosion-resistant ceramic is required to
solve an erosion-corrosion problem, whereas a harder more impactresistant ceramic may solve an erosion problem.
Distinguishing between erosion and erosion-corrosion based on
appearance can be difficult because both appear as localized loss in
areas of impingement or flow, often in a flow pattern or head-and-tail
shape. Both can be very clean, with no corrosion product in the areas
of loss. Often the best clue is to closely examine areas near the
damage for signs of corrosion product indicating that corrosion was
occurring. The presence of such corrosion product away from
flow/impingement areas, and its absence in flow/impingement areas,
suggests erosion-corrosion rather than ordinary erosion.
F
Galvanic Corrosion
Galvanic corrosion refers to the corrosion that occurs when two
dissimilar materials are in electrical contact in a corrosive. Normally
one member of a galvanic couple experiences increased corrosion and
the other decreased corrosion as the result of being galvanically
coupled. When galvanically coupled, electrons generated by corrosion
on one member of the pair flow to the other member where they
participate in a reduction reaction. The member supplying the
electrons experiences increased corrosion and is termed "active"; the
member receiving the electrons experiences reduced corrosion and is
termed "noble." (Although there are some exceptions to this with
metallic materials that have "activepassive" behavior, this is
uncommon for ceramics and the authors are not aware of ceramics
exceptions to the general galvanic behavior described above.)
Two requirements must be met for galvanic corrosion to occur:
1. There must be two materials that are electrically conductive,
galvanically different, and in electrical contact with each other.
2. There must be a corrosive environment which is ionically
conductive and in contact with both materials.
The requirements for electrical and ionic conductivity can be
understood in a general sense by recognizing that a galvanic couple is
fundamentally an electrical circuit in which current flows in a circular
path. Half of the circuit is electron flow through the ceramics from the
active ceramic to the noble ceramic, and the other half is the ionic
current flow through the corrosive between the ceramics. This current
is the galvanic current, and it is a direct measure of the galvanic
Page 437
effect on corrosion. Anything that disrupts the galvanic current circuit
also disrupts the galvanic corrosion behavior.
Requirement 1 means that only those ceramic materials which are
electrically conductive are susceptible to galvanic corrosion.
However, this requires additional explanation. If only one of a pair of
ceramic materials is conductive, then galvanic corrosion does not
occur. If a ceramic is composed of two or more phases and only one
or some them are conductive, then normally the ceramic participates
in a galvanic couple only if the conductive phases are continuous
through the bulk of the ceramic part. For example, this would be the
case if the conductive phase is a continuous matrix and the
nonconductive phase consists of embedded particles. It is important to
realize that in a case like this, as a practical matter, galvanic corrosion
affects both phases because as the conductive continuous phase
corrodes, the nonconductive embedded phase is lost by dropout. If
only a discontinuous embedded phase is conductive, then galvanic
effects are negligible for practical purposes, since only those
conductive particles in direct contact with the other part are affected.
The extent of galvanic corrosion in such a case is usually so limited as
to be insignificant. If the ceramic conducts electrons poorly (i.e., is a
high-resistance conductor), then galvanic effects are reduced in
proportion to the resistance. In essence. Ohm's law applies, and the
more resistance there is in the galvanic current circuit, the more the
galvanic current and galvanic effect decrease.
Requirement 1 also says that the two ceramic materials must be
galvanically different. This is a matter of degreethe more galvanically
different two ceramics are, the greater the galvanic effect will be. As
is the case with metals, two ceramics that are just slightly galvanically
different will experience only slight galvanic effects, while two
ceramics that are greatly galvanically different will experience large
galvanic effects.
Measurements with zero resistance ammeters or high-resistance
voltmeters can easily establish a galvanic series of conductive
ceramics for a given corrosive environment. Such a series allows one
to predict which members of a galvanic couple will experience
increased corrosion (the one toward the active end of the series) and
decreased corrosion (the one toward the noble end of the series). The
distance between any two members of a galvanic couple in the series
also gives a qualitative indication of the severity of the galvanic
couple. The order of ceramics in such a series changes with the
corrosive, particularly for those ceramics which experience
passivation in some corrosives.
The area ratio between two ceramics in a galvanic couple is important.
A large noble member greatly affects a small active member, but a
small noble member only slightly affects a large active member.
Recalling a favorite anecdote from Mars G. Fontana, one can visualize
this by considering two riveted plates in seawater, each containing
copper (noble) and steel (active). A copper plate with steel rivets soon
loses its rivets to galvanic corrosion because of the unfavorable
Page 438
area ratio. The steel plate with copper rivets is unaffected by galvanic
corrosion because there is not enough copper to influence the large
area of active steel. As a mental model, consider that the galvanic
effect is exerted by the noble member on the active member; therefore
a large noble member concentrates a large galvanic effect on a small
active member. Conversely, a small noble member has its galvanic
effect reduced by being spread over a large active member. For
ceramics with both conductive and nonconductive phases, only the
conductive phases are considered in determining area ratios.
Requirement 1 also says that the two ceramics must be in electrical
contact, i.e., that electrons must be able to flow from one to the other.
Galvanic corrosion cannot occur if both ceramics are in the same
ionically conductive solution, but without electrical contact. The
electrical contact does not have to be by direct contact between the
two ceramics. Any electron path from one to the other, regardless how
indirect, will support galvanic corrosion. However, in accordance with
Ohm's law, greater resistance in the electron path reduces the galvanic
effect.
Because of mixed phases in many ceramics, one may encounter the
situation in which there are two galvanically different, electrically
conductive phases in the same ceramic. In such a case (assuming the
other requirements for galvanic corrosion are met) one should expect
the more active phase to experience increased corrosion because of
the galvanic effect and the more noble phase to experience lower rates
of corrosion. As a practical matter, however, the galvanic corrosion
may detrimentally affect even the noble phase as the active phase
around it corrodes. Depending on the morphology and relative
amounts of the phases, the noble phase may be loosened and more
susceptible to wear effects, or may simply drop out if it is present as
discrete separate particles. Also, if the corrosion products of the active
phase are voluminous, then the ceramic may be susceptible to
cracking and spalling from the stresses caused by the corrosion
product growing inside the structure of the ceramic.
Metal-matrix composites (MMCs) are beyond the normal scope of
this discussion but are briefly mentioned here because galvanic
corrosion is a common problem in this type of material [22]. The
metal matrix is an electrical conductor, and if the ceramic
reinforcement is also a conductor, then galvanic corrosion can
potentially occur. Though not always practical, use of a
nonconductive ceramic for the reinforcement eliminates the potential
for galvanic corrosion.
Requirement 2 says that the environment must be corrosive. If an
environment has no corrosive properties for two ceramics separately,
then usually it is noncorrosive to both as a galvanic couple. It is
possible for galvanic coupling to induce corrosion where none
occurred before, but that behavior is generally limited to materials that
exhibit activepassive transitions (i.e., capable of going quickly from
high corrosion rates to very low corrosion rates with modest changes
in corrosive oxidizing power, due to the formation of a passive film).
The
Page 439
common galvanic effect is to increase or decrease corrosion that is
already present in measureable amounts.
Requirement 2 also says that the corrosive must be ionically
conductive. Recalling that galvanic corrosion requires a circuit of
current flow, if the corrosive cannot transmit an ionic current, then the
galvanic circuit is broken and galvanic corrosion cannot occur. Ohm's
law applies again. Progressively greater resistance to ion flow in a
corrosive results in progressively less galvanic current and less
galvanic corrosion. Corrosives that are not ionically conductive do not
support galvanic currents, and galvanic corrosion does not occur in
them.
Requirement 2 also says that the corrosive must be in contact with
both members of the galvanic couple. Clearly, contact with both
members of the galvanic couple is required to complete the circuit of
galvanic current flow. It is worth noting in this context that the
ceramic area contacted by the corrosive is the only area that
participates in a galvanic couple, and that this is the area which must
be used when determining the area ratio between the active and noble
members of a galvanic couple.
The principles and requirements for galvanic corrosion discussed
above lead directly to a variety of methods for mitigating galvanic
corrosion, some of which have been suggested indirectly already.
These include the following:
Use ceramics that are close in galvanic behavior rather than very
different.
Electrically isolate the members of the galvanic couple. For ceramics,
making one member of a couple a nonelectrically conductive ceramic
is a reliable way to do this. If electrically isolating two conductive
ceramics, remember that all electron paths between them must be
severed. It is not adequate to simply eliminate direct contact.
Use area ratios to your advantage. A large active member and a small
noble member will result in little galvanic effect.
Isolate the corrosive from one of the members. However, remember
the area ratio principle. If using a coating, coat the noble member, not
the active member. While it seems intuitively obvious to coat the
active (corroding) member of galvanic couple, doing so concentrates
all the galvanic effect of the large noble member at the small areas of
pinholes and defects on the coated active member. This results in
rapid and extensive local corrosion at these sites. Admittedly, coating
the active member is effective if the coating is 100% reliable and
intact, but it is safer and nearly as effective to coat the noble member
of the couple.
Extend the electrolyte path. This generally means physically
separating the two members of the couple, e.g., with spacers or
intermediate parts that are not galvanically active. Extending the
electrolyte path increases the circuit resistance and decreases the
galvanic current and galvanic corrosion.
Page 440
G
Selective Leaching
Selective leaching occurs with discrete phases, segregated impurities,
or even specific ions. Multiphase ceramics with a phase that is
significantly less resistant to a given corrosive than the other phases
are susceptible to selective leaching of the entire phase. In such a case
the susceptible phase dissolves from the matrix while the neighboring
phases are not attacked. Even though neighboring phases are not
attacked, they may be affected if loss of the leached phase weakens
the overall structure of the ceramic.
Such selective leaching of entire phases is common in ceramic
materials. For example, Shimada and Sato [23] report hydrochloric
and hydrofluoric acids attacking grain boundary phases in silicon
nitride, and sodium hydroxide solutions leaching free silicon from
reaction-bonded silicon nitride ceramics [6].
Other corrosion mechanisms can induce and exacerbate selective
leaching of phases. For example, galvanic corrosion could promote
the leaching of a susceptible phase, eventually leading to pitting or
crevice attack. Tressler [24] reports such a case, in which selective
leaching of certain phases in silicon carbide and silicon nitride
ceramics created pits.
Selective leaching also occurs with specific ions rather than entire
phases. What ions are dissolved is highly specific to the ceramic and
the corrosive. When exposed to hydrochloric solutions, silicon nitride
with yttria, alumina, and aluminum nitride additives has Y and Al ions
leached from the ceramic. However, when exposed to hydrofluoric
acid, silicon and Al ions are leached [23]. Stabilizing oxides in
zirconia also leach away in certain aqueous corrosives [6].
Leaching of small phases or specific ions may affect wearability,
hardness, or other aspects of performance far more than weight loss
corrosion measurements suggest. Removal of material from the
surface weakens it and makes it more susceptible to mechanical
damage of various sorts.
H
Intergranular Corrosion
Intergranular corrosion is related to selective leaching in some cases
but is unique in that intergranular corrosion proceeds along grain
boundaries, sometimes to the extent that entire grains drop from the
ceramic. Grain boundaries are regions of high-energy and
crystallographic mismatch, so that impurities and additives tend to
segregate to grain boundaries. The susceptibility of grain boundaries
to corrosion typically results from corrosion of such impurities or
additives. Sometimes, even in the absence of impurities, the grain
boundaries are preferential sites for corrosion simply because of the
high interfacial energy.
Yoshio and Oda [21] report an example of intergranular corrosion in
which HIP silicon nitride without stabilizing additives intergranularly
corroded in high-temperature water. This grain boundary attack is
attributed to an amorphous silica
Page 441
phase that formed at the grain boundaries. They believe that the
amorphous silica phase originated with impurities in the starting
powder.
I
Corrosion-Assisted Cracking
Corrosion-assisted cracking refers to subcritical (stable) crack growth
caused or assisted by corrosion reactions. It is analogous to metallic
cracking phenomena such as stress corrosion cracking (SCC), liquid
metal embrittlement, and corrosion fatigue. Corrosion-assisted
cracking is specific to the ceramic/environment combination and
occurs in a variety of ceramics, including alumina, boron carbide,
zirconia, and many glasses.
The mechanisms involved vary according to ceramic and
environment, and many are still poorly understood. In various
ceramic/environment combinations, corrosion-assisted cracking has
been related to the presence of secondary phases at grain boundaries
(reaction-sintered silicon carbide), direct oxidation by oxygen (silicon
carbide), chemical adsorption of water and subsequent corrosion
(zirconia, silica, alumina, uranium carbide), and effects of
environmental impurities (silicon nitride, silicon carbide), as well as
other phenomena [6,25].
Stress corrosion cracking is generally recognized as a type of
subcritical crack growth in ceramics. Relative humidity is an
important factor in SCC crack growth for many ceramics, and water
generally increases SCC crack growth rates in ceramics. Even
chemicals similar to water in chemistry, such as ammonia and
hydrazine, are believed to participate in crack-tip reactions. Water
reacts directly with silica and alumina at crack tips to propagate
cracks [26] and has been identified as an active cracking agent for
uranium carbide [27]. Data exist for some ceramic/environment
combinations showing the effect of corrosives on SCC crack growth
rates, e.g., the effect of various acidic and caustic environments on
crack growth rates in soda-lime silicate glass, silica glass, and vitreous
silica [28,29].
Although not widely recognized, many ceramics are susceptible to
fatigue cracking, i.e., subcritical crack growth induced by cyclic
stresses. Laboratories in various parts of the world are beginning to
generate fatigue cracking data for ceramics. The results indicate that
one can measure subcritical fatigue crack growth rates for ceramics,
though understandably they tend to be much more dependent on the
stress intensity than are the growth rates for metallic materials [30].
Corrosion is capable of either enhancing or reducing the fatigue life of
ceramics. The explanation has been similar to that for the effect of
corrosion on ceramic strength, i.e., that corrosion changes the surface
profile, which in turn affects the initiation of cracks [31]. Corrosion
that is highly uniform and smoothes the surface has the effect of
delaying crack initiation and enhancing fatigue life, while corrosion
that roughens the surface and provides stress risers has the effect of
promoting crack initiation and decreasing fatigue life. This
explanation
Page 442
deals only with the initiation stage of fatigue cracking, not with
subcritical crack growth.
However, the effect of corrosion on fatigue is likely to go beyond its
effect on surface profile. For metals it is known that corrosion
increases the growth rate of fatigue cracks, compared to rates in a
noncorrosive environment. This effect is so important that it is
referred to by the separate term ''corrosion fatigue." Given the many
analogies between metallic and ceramic corrosion behavior, it seems
likely that corrosive environments also reduce the fatigue life of
ceramics by increasing subcritical crack growth rates. This is
supported by the above observation regarding the effects of humidity
and water on SCC crack growth rates. Despite growing activity
generating fatigue data for ceramics, the authors know of no data
generated specifically to determine to what degree corrosion increases
subcritical fatigue crack growth rates. With many future ceramic
applications involving exposure to cyclic stress, such as gas turbines
and internal combustion engine components, the question of corrosion
fatigue behavior will be important.
VII
Thermodynamic Aspects of Ceramic Corrosion
The thermodynamic aspects of aqueous corosion of ceramics are
considered only in general terms here. For more detailed discussion of
thermodynamics related to corrosion phenomena, the reader is
referred to Fontana [32], Shrier [33], and Pourbaix [34]. Discussions
directed specifically toward ceramics applications are found in Lay
[35] and Livey and Murray [36].
A
Thermodynamics vs. Kinetics
For corrosion to occur spontaneously (i.e., without input of energy
from outside the reaction), it must be both thermodynamically and
kinetically possible. The kinetic limitations have been discussed
above. Given adequate thermodynamic data, one can calculate
whether a specific corrosion reaction can occur spontaneously. If one
determines on the basis of thermodynamic data that a given corrosion
reaction cannot occur, then one can reliably predict that the particular
corrosion reaction is not a practical problem. However, if one
determines from thermodynamic data that a given corrosion reaction
is able to occur, it still may not be a practical problem because the
reaction is kinetically limited.
Using thermodynamic data to predict corrosion behavior needs to be
done carefully, with full appreciation for the impact of kinetic
limitations. Even so, thermodynamic analysis is a powerful tool for
understanding and predicting corrosion behavior. The major
challenges one faces in thermodynamically evaluating corrosion is
finding adequate and complete thermodynamic data and considering
all the possible corrosion reactions that could be important for
complex ceramic compositions and structures.
Page 443
B
Free Energy
The basic concept in the application of thermodynamics to corrosion
phenomena is DG, the change in free energy as the result of a
corrosion reaction. Free energy can be calculated by the equation
where n is the number of electrons in the reaction, F is the Faraday
constant, and E is the cell potential, as determined by measurement vs.
a reference electrode. The change in free energy must be negative for
a corrosion reaction to occur spontaneously (i.e., without input of
energy from outside the system).
Although Eq. (5) defines the change in free energy, as a practical
matter one more often calculates the change in free energy by
For example, if one proposes the corrosion of a-silicon carbide by
hydrochloric acid to form methylene and dichlorosilylene, the reaction
is
The free energy of formation for each of the compounds in this
reaction is in the JANAF tables [37]. Taking care to keep signs correct
and using free energy values for 25°C and calculating the free energy
of the products minus the free energy of the reactants, one determines
that the change in free energy for reaction (7) is a positive 449 kJ/mol
at room temperature. Because this is a positive change, this corrosion
reaction cannot take place spontaneously at 25°C. Note that this does
not prove that corrosion of a-silicon carbide ceramics parts by
hydrochloric does not occur, only that this particular corrosion
reaction cannot occur spontaneously.
Looking at another case, consider corrosion of boron carbide by
hypochlorous acid to form boron oxychloride and methane:
Calculating the free energy of the products minus the free energy of
the reactants, using values from the JANAF tables for 25°C, one finds
the change in free energy for this corrosion reaction is a negative 1029
kJ/mol at room temperature. Because this is a negative change in free
energy, it is thermodynamically possible for this corrosion reaction to
occur. Recall, however, that kinetic factors may limit the significance
of this reaction.
McNallen et al. [38] and Barsoum [39] provide good examples of the
use of thermodynamics for assessing corrosion.
Page 444
C
Potential-pH Diagrams
Pourbaix [40] outlines a particularly useful application of
thermodynamics for evaluation of corrosion behavior. Though this
work has been for metals, the principles apply as well to electrically
conductive ceramics. This analysis uses the Nernst equation to
calculate regions of thermodynamic stability based on potential and
pH, for all possible corrosion products, for a specific material and
environment. The large number of possible material/environment
reactions necessarily limits consideration to relatively simple
materials and environments.
The results of this type of evaluation are typically represented
graphically on plots of electrochemical potential vs. pH, known as EpH diagrams. These plots show which corrosion products are
thermodynamically stable in various regions of potential and pH.
Pourbaix's treatment considers regions of immunity (i.e., no
corrosion) to be those where there are no thermodynamically stable
corrosion products and only the (metallic) material is stable. Regions
of possible passivation are those where solids are the
thermodynamically stable corrosion products. Regions of corrosion
are those where the thermodynamically stable corrosion products are
soluble ionic species. Analogous thermodynamic analyses can be
performed for ceramics. This approach is especially useful for
explaining the corrosion of RBSN where corrosion of the residual
silicon metal determines the corrosion behavior of the ceramic.
VIII
Corrosion Testing Principles and Procedures
Corrosion testing and evaluation of ceramics are similar in many
respects to that for metallic materials. This discussion addresses
methods for exposing and evaluating samples, and does not attempt to
address more sophisticated techniques such as electrochemical testing
procedures or in situ surface analytical techniques. Considerations that
are common to both ceramic and metallic materials are only briefly
reviewed here, while factors that are a specific or unique concern for
ceramics receive a more in-depth discussion. Readers who are
unfamiliar with basic corrosion testing procedures may benefit from
the more comprehensive discussions of corrosion testing and
evaluation for metallic materials found in Fontana [41], Uhlig [42],
and the Standard Test Method published by NACE International [43].
Corrosion testing and evaluation of ceramics are generally more
complex than for metallic materials. The additional complications
come from factors such as the greater difficulty in adequately defining
the ceramic, the effects of geometry and porosity, and the greater
comprehensiveness of evaluation requirements.
Page 445
A
Types of Corrosion Tests
Corrosion tests can be characterized as laboratory tests of samples or
parts, field tests of samples, and field tests of actual parts. A critical
factor in testing is how well the test reflects actual service conditions.
In general, laboratory tests do this least well, field tests of samples are
better, and field tests of actual parts are best. As one might expect,
these are also in order of generally increasing cost and difficulty. The
appropriate type of test for a particular situation depends on the goals
of the test and the resources available. Many of the same principles
apply for all three types of tests.
B
General Procedure
The three fundamental parts of a corrosion test are conducting an
exposure of sample materials, evaluating the effects of the exposure
on the material, and reporting the results in a way that allows one to
predict and compare performance.
The exposure conditions should simulate the intended service as
closely as possible, using samples that reflect the intended use in all
important respects. This accounts for the preference for field tests and
the use of actual parts, since they reproduce actual service better.
However, this is often not practical, and in many cases laboratory
testing of prepared samples is the appropriate method to accomplish
the goals of the testing program. Important exposure conditions that
are not reproduced in the test need to be noted and taken into account
when reporting results and drawing conclusions.
Evaluating the effects of corrosive exposure on the samples is one of
the areas that is notably more difficult for ceramics than for metallic
materials. Effects of geometry and porosity are generally more
important for ceramics, and they need to be evaluated and accounted
for in the results. Ceramics are also more likely to experience
deterioration of mechanical properties from corrosive exposure, so
that a proper evaluation may require mechanical testing that is not
necessary for metallic materials. The section below on evaluation
elaborates on these factors.
One should report results in a way that allows a meaningful
comparison of performance between different ceramics in the same
corrosive and between the same ceramics in different corrosives. This
includes expressing corrosion rates in appropriate units of penetration
rate, as elaborated on below. The ceramic samples must be adequately
defined. For example, many generically similar ceramics have
important differences in additives, impurities, or processing that
significantly affect corrosion performance. Data reported in the
literature frequently fail to document such pertinent information,
compromising its utility for comparison purposes. Also, the
importance of corrosive effects on mechanical properties
Page 446
complicates reporting because different investigators may evaluate
those effects with procedures that are not comparable. This is
commented on further below.
As mentioned earlier, electrically conductive ceramics can be
evaluated using conventional electrochemical testing techniques.
However, a discussion of those techniques is beyond the scope of this
discussion. For further information, we refer the reader to Fontana
[41] and Liening [44] for metals, and to Divakar et al. [9] for a
discussion of these techniques as they relate to ceramics.
C
Preferred Units for Corrosion Measurement
As mentioned earlier in this chapter, two current problems in
reporting corrosion results for ceramics are the wide variety of units
used, and the use of units that makes it difficult to compare results for
different ceramics or to compare results from different investigators.
From an engineering and practical application perspective for
ceramics, the two main criteria for corrosion performance are loss of
physical dimension and loss of mechanical properties. These should
be reported in units that are meaningful to the potential ceramic user,
allow comparison of various ceramics' performance, and are
comparable with the results of other investigators.
Regarding the loss of physical dimension, units that report corrosion
as a percentage weight loss, or weight loss per unit area per unit time
(e.g., mg/cm2 per yr), do not reflect the amount of dimension lost.
Better units are penetration rates: micrometers (µm or 10-6 m) per
year (µmpy) for the metric system, or mils per year (commonly shown
as "mpy") for the English system (where a "mil" is 0.001 in.). These
units are normalized for ceramic density, exposure time, and exposure
area. The formula to calculate µmpy corrosion rate is
where
W is weight loss, mg
D is density, g/cm3
A is surface area exposed to the corrosive, cm2
T is exposure time, h
To convert µmpy to mpy, multiply by 0.039. The calculate mpy
directly, use
where W, D, A, and T are in the same units as for Eq. (9).
The density to be used in Eqs. (9) and (10) is the apparent density,
based on outside dimensions and the mass of the ceramic, with no
adjustments for porosity or percentage of theoretical density.
Likewise, the surface area to use is
Page 447
the apparent surface area, with no corrections for porosity. As
described elsewhere in this chapter, however, bulk density and
porosity are important ceramic descriptions that should be reported
with the test results.
D
Planned Interval Test Method
One of the difficulties of corrosion testing, particularly with
laboratory testing in small volumes of corrosive, is determining what
in the system is changing when the corrosion rate changes with time.
For example, if a ceramic sample is corroding more slowly toward the
end of an exposure period, is it because the ceramic has become
protected by a corrosion product or because the corrosive in the
solution has been depleted? Similarly, if a ceramic sample is
corroding more rapidly toward the end of an exposure period, is it
because the solution has become contaminated with aggressive
corrosion product or because the ceramic has lost a protective film?
Both solution corosivity and ceramic "corrodibility" may change with
time, and drawing the right conclusions from a test may depend on
being able to determine which is changing.
Wachter and Treseder [45] describe an excellent procedure for
separating the effects of changing solution corrosivity from changing
sample corrodibility. The procedure is called the planned interval test,
and it is also summarized in Fontana [41]. This procedure involves
exposing a minimum of four replicate samples in the same solution.
One is exposed for the entire test duration, the second for only an
initial period at the beginning of the test, the third for all but a final
period at the end of the test, and the fourth for only the final period at
the end of the test.
Comparing the rates of the four samples allows one to determine
several aspects of the corrosion behavior:
The initial rate of corrosion on a fresh sample surface in fresh solution
The rate of corrosion on a corroded sample surface in old solution
The rate of corrosion of a fresh sample surface in old solution
The average rate of corrosion over the entire exposure period for a
single sample
The relative corrosion rates of the four samples allow one to
determine how corrosion rates are changing with time, and whether it
is being caused by changing solution corrosivity or changing sample
corrodibility. Understanding these factors is important in drawing
conclusions about the suitability of a ceramic part for an intended
service.
IX
Test Sample Definition and Preparation
One of the inadequacies of corrosion data in the literature is the poor
definition of the material tested. As mentioned above, consensus
standard definitions of
Page 448
ceramics do not exist as they do for metals and alloys, and generic
descriptions are not specific enough to allow reliable reproduction of
the results by other investigators. While the same ceramic product
from a single manufacturer may be highly reproducible, one should
not assume that the same generic ceramic from a different
manufacturer will exhibit the same corrosion behavior. As such,
corrosion test results on the "same" ceramic may not be reproducible.
Needless to say, nonreproducible results are of limited value to
potential users of ceramic products.
A comprehensive list of everything that should be defined is
impractical because of the differences among ceramics. However, the
items below are expected to affect the corrosion behavior of ceramics
and should be documented with the data.
A
Composition and Phases
One should quantitatively document the ceramic composition,
including the types and amounts of impurities. Even minor phases
should be identified because their corrosion behavior may be
markedly different from the bulk of the ceramic. The mircostructural
morphology of such phases is also important, particularly if they are
continuous (or nearly continuous) around grain boundaries. In such
cases, accelerated corrosion of these phases can lead to grain dropping
and dramatically increased corrosion rates.
B
Porosity
One should document open (interconnected) porosity as part of the
ceramic description because it promotes corrosion by allowing entry
of corrosive to the interior of the ceramic. This increases the amount
of surface area exposed to corrosive and may cause high losses by
spalling and cracking if the corrosion product is voluminous. Open
porosity also complicates weight lossbased corrosion rate calculations
if one cannot effectively remove corrosion product, corrosive, or
cleaning material from the pores. Depending on the circumstances,
one may or may not be able to adequately clean a porous sample, but
in any case the fact of open porosity and the measures taken to
account for it should be documented.
Closed porosity is much less troublesome, but investigators should
take care that pores open to the surface are adequately cleaned and
dried before taking weight loss data. Even a volatile cleaning and
drying solvent such as acetone may take hours to evaporate from
closed porosity that is open to the surface.
The American Society for Testing and Materials (ASTM) Standard
ASTM C-373 for measuring open porosity may be helpful. This
method involves comparing the dry and water-saturated mass of a
sample. Measurement of porosity in a polished cross-section may also
be an adequate approach. The
Page 449
appropriate method will depend on the specific circumstances and
ceramic involved. Whatever method is used, the procedure should be
reported with the results.
C
Sample Preparation
A complete description of sample preparation includes the ceramic's
manufacturing method and the surface preparation. A description of a
ground surface should include the finish (i.e., smoothness of the
grind), the grinding media, and the procedure. Microcracks induced in
ceramics by a surface preparation can be a significant factor in their
corrosion behavior. Surface preparation should reflect the actual
application to the extent practical.
D
Geometry
The effect of stress risers and the susceptibility of square edges to
chipping and spalling from corrosion effects suggest that these should
be avoided if practical. In any case, it may be advantageous to have
the geometry of the sample reflect the geometry of the part being
considered for the actual application. Sometimes the best approach is
to test an actual part, or even portions of an actual part, in the lab or
field. This avoids not just questions about possible effects of geometry
but also questions about other aspects of a ceramic sample that may
not adequately reproduce an actual part being considered for service.
X
Testing Apparatus
Appropriate testing apparatus is largely like that used for corrosion
testing of metals. The reader is referred to Fontana [41] and McGeary
and Lifka [46] for information on common apparatus for laboratory
and field corrosion testing of metals.
A
Laboratory Testing
Laboratory apparatus must be highly resistant to the corrosive so that
it does not contaminate the limited volume of test solution enough to
affect test results. Test samples are normally suspended or held in a
manner that allows access to the entire surface of the sample.
Electrical isolation of ceramic samples is important only for those
ceramics that are conductive. Failure to electrically isolate conductive
ceramics risks unintended galvanic effects. Glass and
polytetrafluoroethylene (PTFE) products are widely used in laboratory
corrosion testing because of their broad corrosion resistance.
Obviously, when these materials are not resistant to the corrosives or
temperatures involved, other materials must be used.
Page 450
B
Field Testing
Field apparatus consists of holders of a wide variety of design that are
sufficiently corrosion-resistant to last at least through the test duration.
Contamination of the corrosive by a corroding holder is generally a
lesser concern for field tests than for laboratory tests because of the
greater volume of corrosive. As a practical matter, however, a field
holder must not only last long enough for an adequate corrosive
exposure of the test samples but must last to the next opportunity to
extract the assembly from its field installation. Failure of a holder not
only risks the integrity of the corrosion data but also may jeopardize
equipment in the field facility, such as pumps and valves.
XI
Test Conditions
Important conditions to record and document include a full
description of the corrosive, such as concentration, temperature, and
low levels of possibly important species. Other important test
conditions to document include velocity or agitation of the corrosive,
aeration, and test duration. Not just the presence of agitation or
aeration but the method used to achieve these conditions should be
reported.
The appropriate test duration depends primarily on the corrosion rate
being measured. The faster the rate, the more quickly there is enough
weight loss for a reliable measurement. Most analytical balances are
capable of ±0.1 mg. Assuming a minimum desired weight loss of 10
mg for an accurate measurement, one can calculate from Eq. (9) a
minimum required test duration as a function of an anticipated
corrosion rate. Rearranging Eq. (9) to do this,
where
T' is an estimate of the minimum required test duration, h
C' is the anticipated corrosion rate, µmpy
The other terms are as described for Eq. (9).
It may happen that one has a criterion that the test last long enough to
demonstrate a corrosion rate below a given threshold level. In a
manner similar to that just described, rearranging Eq. (9) allows one
to do this.
where
T' is an estimate of the minimum required test duration, h,
W' is the reliable precision of the weight loss measurement, mg
C" is the given threshold corrosion rate, µmpy
Page 451
The other terms are as described for Eq. (9).
XII
Cleaning Test Samples
Samples should be cleaned of corrosion product using a method that
does not cause loss of sound ceramic. Cleaning is normally followed
by rinsing in deionized water and drying. Drying is often assisted by
rinsing with a solvent such as acetone. While cleaning is an easy
chore in some cases, in others the corrosion product is highly adherent
and tenacious, and good cleaning is a difficult task.
Mechanical brushing is sufficient to remove corrosion product in
many cases. The brush media and force applied is only limited by the
requirement of not damaging the ceramic.
Cleaning solutions are useful as long as they effectively remove the
corrosion product, do not corrode or damage the ceramic, and do not
leave nonremovable residue that affects the weight loss measurement.
Appropriate cleaning solutions depend on the particular ceramic and
corrosion product. They range from water-base surfactant solutions to
acids or bases that may or may not contain corrosion inhibitors. Note
that not all inhibitors are effective with all materials; if in doubt,
check. Ultrasonic baths are often used with cleaning solutions to
increase effectiveness.
Soft abrasive blasting, such as with corncob or walnut shells or
bicarbonate, may be a satisfactory method of cleaning. Particularly for
the harder nonporous ceramics, this method may be highly effective
for removing tenacious corrosion product without damage to the
ceramic substrate.
Porous ceramics present special problems with respect to cleaning, as
noted above. Trapped corrosion product, corrosive, cleaning solution,
or solvent all affect weight loss and must be either eliminated or
otherwise accounted for.
It may happen that the only practical cleaning method damages the
ceramic, e.g., cleaning by an abrasive blast or a solution that causes
some additional corrosion of the ceramic. In such a case, one should
perform the cleaning procedure on new replicate ceramic samples to
determine the amount of weight loss attributable to the cleaning.
Enough such replicates should be cleaned to determine the range of
weight losses to expect, which then becomes part of the limit on
precision of the results.
XIII
Evaluation of the Ceramic after Testing
Evaluating the results of a corrosive exposure takes several forms. For
metals one typically bases the evaluation on a penetration rate
calculated from weight loss and on microscopy to determine the form
of corrosion. For ceramics one should do both of these, plus additional
evaluations as required to show effects on
Page 452
mechanical properties such as strength and hardness. In addition,
other properties that are critical to the proposed application and may
be affected by corrosion should be tested, such as modulus or
electronic properties. In special cases where the Weibull modulus is
critical to the proposed application, it may be necessary to corrosiontest a large number of replicate samples in order to make a reliable
determination of the effect of corrosive exposure.
A
Weight Loss
Weight loss measurements are straightforward and should be used to
calculate corrosion rate as described in Eqs. (9) and (10). The typical
instrument for measuring weight loss is an analytical balance capable
of ±0.1 mg precision. Since two measurements (before and after the
exposure) are required, the best precision with such an instrument for
an exposure evaluation is ±0.2 mg. The actual precision of the
calculated corrosion rate is also affected by other factors, such as
thoroughness of the cleaning, precision of the surface area and density
measurements, etc.
B
Mechanical Properties
Corrosion not only results in material loss but may degrade
mechanical properties such as strength and hardness [6,31].
Mechanical properties are evaluated by the normal tests used for
ceramics. Tensile strength, flexural strength, and hardness are typical
mechanical properties of interest. The inverse crack length technique
and the four-point bend tests are useful methods for measuring
strength properties. The military standard MIL-STD-1942A is useful
as a guide for flexural testing. Strength units should be intensive (i.e.,
based on cross-sectional area, not purely load) in case corrosion losses
have significantly reduced the load-bearing area.
The corrosion rate quantifies the loss of cross-sectional area; the
mechanical strength test should reflect only the effect of exposure on
the intensive strength properties of the ceramic. Mechanical strength
changes may legitimately reflect the effects of smoothing or
roughening of the ceramic surface by corrosion, so it is usually not
necessary or desired to re-polish the sample after corrosive exposure
unless the investigator requires it for specific purposes. The presence
of such smoothing or roughening may be noted as a mechanism by
which corrosion has affected strength.
Whatever tests one uses to show the effects of corrosive exposure,
they should be conducted both before and after exposure with enough
replicate samples to demonstrate the significance of any changes
found. Nondestructive tests such as hardness may be done on the same
samples that are later subjected to the corrosive exposure. Obviously,
destructive tests must be done on expendable replicate samples.
Page 453
C
Dye Penetrant Testing
Dye penetrant testing is helpful in identifying the depth of localized
corrosion and small corrosion features on a ceramic surface. This
technique is widely used as a nondestructive evaluation (NDE) tool
for metal, particularly welds. Dye penetrant testing supplies are
widely available at welding supply centers. More common dyes such
as methylene blue may also be used.
The procedure to identify the depth of penetration of localized
corrosion involves soaking the ceramic sample in a dye for several
hours or overnight, then cleaning and sectioning the sample. The
depth of dye penetration reveals the depth of corrosive attack.
Ceramics with interconnecting porosity may have to be evaluated with
a noncorroded ''blank" sample to separate effects of corrosion from
those of porosity.
The procedure for identifying surface-localized corrosion features is
relatively simple. One cleans the sample with a solvent, places a
liquid low-surface-tension red dye on the sample and allows it to soak
into cracks and crevices for several minutes, wipes the dye off so that
the sample appears clean, then sprays a white powder-like material
(the "developer") on the surface. The white powder pulls dye from
narrow cracks and crevices where it could not be wiped clean, and
these cracks and crevices show up clearly as red features. While this
procedure does not find microscopic features, it is effective at
highlighting small macroscopic features that a simple visual
examination easily misses.
D
Optical Microscopy
Optical microscopy is normally done with either a common binocular
microscope, or an inverted light microscope (also known as a
metallograph) using polished cross-sections of corroded samples.
These samples are sometimes etched after polishing to highlight
microstructural features of the ceramic. The polished surface is
necessary because of the small depth of focus for inverted light
microscopy, particularly at high magnifications. Binocular
microscopy is useful for identifying larger features of corrosion
morphology and is useful up to about 100×. Inverted light microscopy
is particularly useful for finding microstructural features such as
phases and grain boundaries, and corrosion penetrations along such
features. It is useful up to about 15002000×, depending on the quality
of the sample.
E
Scanning Electron Microscopy
Scanning electron microscopy (SEM) is normally done on the
corroded surface of the sample, either before or after cleaning away
corrosion product, and is helpful in identifying the form of corrosion.
The large depth of focus of SEM does not require a polished or even a
flat surface. A common accessory is energy-
Page 454
dispersive X-ray spectroscopy (EDS), which allows semiquantitative
elemental analysis. The combination of SEM and EDS is a powerful
tool to characterize the forms of corrosion and the mechanisms
involved. The analysis area for EDS ranges from microscopic to the
largest viewing area at low magnification, so that one can quickly
analyze everything from broad areas of corrosion product to tiny
amounts of material inside a pit. This capability allows one to identify
corrosion products, determine if corrosive species have concentrated
in pits, determine if the ceramic surface has been enriched in one
component by the leaching of another, and so forth.
SEM is limited to electrically conductive surfaces, and conductive
ceramics are readily evaluated with SEM. However, nonconductive
corrosion products and nonconductive ceramics "charge up,"
accumulating electrons from the electron beam and generating a
blinding white image. Damage to a nonconductive sample by the
electron beam is possible. Such samples require coating with a
conductive material. The common method to do this is vacuum sputter
coating with carbon or a precious metal such as gold or palladium.
Small desktop units are available from microscopy supply houses for
this purpose. These coatings are usually visible to the naked eye as a
color change on the sample, but are very thin and if done correctly are
essentially invisible to SEM. EDS, however, often detects the coating.
Also, once applied, the coating may be impossible to remove without
damaging the sample.
F
Other Analytical Methods
Other analytical methods are also useful for evaluating effects of
corrosion or deducing mechanisms or corrosion. Microprobe analysis,
Auger electron spectroscopy, X-ray diffraction, and transmission
electron microscopy have all been used successfully [47]. Quantitative
analysis of the corrosive after the test exposure is useful to help
identify not only the extent of corrosion but phases or components
that are preferentially attacked.
G
Dealing with Porosity
Porosity, particularly open porosity, may be very troublesome during
the evaluation phase of a test program. In our discussion above, we
have already recommended using apparent density and apparent
surface area for the corrosion rate calculation, but noting the degree
and type of porosity. This still leaves the difficulty of thoroughly
removing corrosive or cleaning fluids from the sample in order to
obtain a valid weight loss. Sometimes this problem cannot be
satisfactorily resolved, and the only option left is to openly
acknowledge the problem and try to minimize it by using the best
procedure available in a reproducible manner.
Page 455
XIV
Reporting Test Results
Much relevant discussion has taken place earlier in this chapter
regarding reporting results in such a way that ceramics can be
evaluated for practical application in corrosives. This section
reinforces the more important aspects of reporting results.
Corrosion rate units should reflect the penetration rate and should be
normalized for exposed area and ceramic density. Units that do not do
this are of limited usefulness and may not allow comparison with
results from other investigators. The authors discourage units such as
mg/cm2 per year and percenage weight loss. Such units might even be
misleading, as shown in an earlier table. In addition to units such as
mpy or µmpy, the actual weight loss should be reported.
Effects on mechanical properties should be reported with both the
before-exposure and after-exposure values, and (if applicable) a
percentage loss of property. For example, loss of hardness is much
less meaningful and much more difficult to compare with other work
if it is reported only as a percentage loss, or only as "after" values.
Before and after values have more utility, allowing one to take some
account of differences in the ceramics used. The methods used to
determine the property values are important and should also be
reported.
The form of corrosion is important. While a weight loss resulting in
250 µmpy as uniform corrosion may be acceptable, the same weight
loss as pitting may be completely unacceptable. Corrosion-assisted
cracking is unacceptable in most applications but usually involves
very little in the way of weight loss. Where localized corrosion is
encountered, one should attempt to measure the depth of penetration
directly and report those results in addition to the weight loss. One
should also note any occurrence of grain dropping, chipping, or
spalling.
All important aspects of a ceramic's identification should be reported.
Depending on the ceramic, this may include its composition, phases,
impurity levels, degree and type of porosity, method of manufacture,
and surface preparation. The sample's size, shape, and area should
also be reported.
All important aspects of the environmental exposure should be
reported. Depending on the type of exposure, this includes the
solution composition, trace species that may be active, temperature,
aeration or deaeration, velocity or agitation, and volume of solution.
For laboratory tests, the latter is often reported as the volume-to-area
ratio, the area being the total exposed area of all samples in the given
volume of corrosive. This is particularly important for laboratory tests
in which there is depletion of the corrosive.
An adequate description of the test procedure describes the test cell,
the method of supporting the samples, and the methods of providing
heat, aeration, agitation, etc. The test duration and method used to
clean the samples should also be reported.
Page 456
References
1. D. W. Richerson, Modern Ceramic Engineering, 2nd ed., Marcel
Dekker, New York, 1992, p. 662.
2. D. A. Parker, Ceramics technologyapplication to engine
components, Proc. Inst. Mech. Engrs., 199 (A3): 135 (1985).
3. L. A. Lay, Corrosion resistant ceramics, Metals Mat. (J. Inst.
Metals), 3: 241 (1987).
4. J. B. Wachtman, Jr., High performance ceramics: advances in
processing and properties, Advances in Materials Technology for
Process Industries' Needs, NACE, Houston, 1984, p. 146.
5. W. Hof, Hexoloy SIC SAein neuer Werkstoff für RohrbündelWärmetauscher, Chemie-Technik (Heidelberg), 20(12): 18 (1991).
6. S. Parrott, Engineering ceramics for aggressive environments,
Metals Mat. 6(4): 207 (1990).
7. Ceramic Source, Vol. 7, American Ceramic Society, Westerville,
OH, 1991, pp. 269369.
8. Metals Handbook, 9th ed., Vol. 1, American Society for Metals,
Metals Park, OH, 1978, pp. 145152.
9. R. Divakar, S. G. Seshadri, and M. Srinivasan, Electrochemical
techniques for corrosion rate determination in ceramics, J. Am.
Ceram. Soc., 72(5): 479 (1989).
10. T. Ya. Kosolapova, Carbides: Properties, Production, and
Application, translated from Russian by N. B. Vaughan, originally
published by Metallurgiya Press, Moscow, 1968; English edition
published by Plenum Press, New York, 1971, p. 30.
11. M. G. Fontana, Corrosion Engineering, 3rd ed., McGraw-Hill,
New York, 1986, p. 171.
12. L. A. Lay, Corrosion Resistance of Technical Ceramics; 2nd ed.,
HMSO, London, 1991, p. 114.
13. R. D. Kane, Prevent corrosion of advanced ceramics, Chem. Eng.
Prog., 87(6): 77 (1991).
14. Anon; Selecting materials for corrosion resistance, Adv. Mat.
Process., 137(6): 129 (1990).
15. R. W. Lashway, Sintered alpha-silicon carbide: an advanced
material for CPI applications, Chem. Eng., 92(25): 121 (1985).
16. M. C. Kerr, Advanced ceramic heat exchangers utilizing hexoloy
SA, single phase silicon carbide tubes, Ind. Ceram., 9(4): 192 (1989).
17. R. M. Latanison and A.J. Sedriks, Aqueous corrosion resistance,
J. Metals, 39(12): 20 (1987).
18. Y. G. Gogostsi and V. A. Lavrenko, Corrosion of HighPerformance Ceramics, Springer-Verlag, New York, 1992, p. 76.
19. M. Pourbaix, Atlas of Electrochemical Equilibria in Aqueous
Solutions, NACE, Houston, 1974, p. 95.
20. M. G. Fontana, Corrosion Engineering, 3rd ed.,, McGraw-Hill,
New York, 1986, p. 51.
21. T. Yoshio and K. Oda, Aqueous corrosion and pit formation of
Si3N4 under hydrothermal conditions, Ceramic Transactions, Vol. 10;
Corrosion and Corrosive Degra-
Page 457
dation of Ceramics (R. E. Tressler and M. McNallan, eds.),
American Ceramic Society, Westerville, OH, 1990, p. 367.
22. A. Turnbull, Review of corrosion studies on aluminum metal
matrix composites, Br. Corrosion J., 27(1): 27.
23. M. Shimada and T. Sato, Corrosion of silicon nitride ceramics in
HF and HCl solutions, Ceramic Transactions, Vol. 10, Corrosion and
Corrosive Degradation of Ceramics (Richard E. Tressler and Michael
McNallan, eds.), American Ceramic Society, Inc., Westerville, OH,
1990. p. 355.
24. R. E. Tressler, Environmental effects on long term reliability of
SiC and SSi3N4 ceramics, Ceramic Transactions, Vol. 10, Corrosion
and Corrosive Degradation of Ceramics (R. E. Tressler and M.
McNallan, eds.), American Ceramic Society, Westerville, OH, 1990,
p. 99.
25. E. Lilley, Review of low temperature degradation in Y-TZPs,
Ceramic Transactions, Vol. 10, Corrosion and Corrosive
Degradation of Ceramics (R. E. Tressler and M.McNallan, eds.),
American Ceramic Society, Westerville, OH, 1990, p. 387.
26. S. Freiman, Stress corrosion cracking of glasses and ceramics,
Stress-Corrosion Cracking (R. Jones, ed.), ASM International, Metals
Park, OH, 1992, pp. 337344.
27. A. Accary, The behavior of reactive ceramics in atmospheric
environments, Ceramics in Severe environments, Materials Science
Research, Vol. 5 (W. W. Kriegel and H. Palmour III, eds.), Plenum
Press, New York, 1971, pp. 445459.
28. S. M. Wiederhorn and H. Johnson, Effect of pH on crack
propagation in glass, J. Am. Ceramic Society 56: 192 (1973).
29. G. S. White, S. W. Freiman, S. M. Wiederhorn and T. D. Coyle,
Effects of Counter Ions on Crack Growth in Vitreous Silica, J. Am.
Ceram. Soc., 70: 891(1987).
30. R. Dauskardt and R. Ritchie, Fatigue of advanced materials, Adv.
Mat. Process., ASM International, Metals Park, OH, 1993, p. 30.
31. K. L. Weisskopf and K. D. Moergenthaler, On the use of ceramics
in automotive engineering with reference to corrosion problems, High
Temperature Corrosion of Technical Ceramics, (R. J. Fordham, ed.),
Elsevier, Barking, UK, 1990, p. 3.
32. M. G. Fontana, Corrosion Engineering, 3rd ed., McGraw-Hill,
New York, 1986, p. 445.
33. L. L. Shrier (ed.), Corrosion, Vol. 1, Newnes-Butterworths,
Boston, 1976, p. 9-3.
34. M. Pourbaix, Atlas of Electrochemical Equilibria in Aqueous
Solutions, NACE, Houston, 1974, p. 19.
35. L. A. Lay, Corrosion Resistance of Technical Ceramics, HMSO,
London, 1991, p. 73.
36. D. T. Livey and P. Murray, The stability of refractory materials,
Physicochemical Measurements at High Temperatures (J. Bockris et
al., ed.), Butterworth, London, 1959, p. 87.
37. Chase, Davies, Downey, Frurip, McDonald, and Syverud, JANAF
Thermochemical Tables, 3rd ed., American Chemical Society and the
American Institute of Physics, 1985.
38. M. McNallan, M. Van Roode, and J. R. Price, The mechanism of
high temperature corrosion of SiC in flue gases from aluminum
remelting furnaces, Ceramic Transactions, Vol. 10, Corrosion and
Corrosive Degradation of Ceramics (R. E. Tressler and M. McNallan,
eds.), American Ceramic Society, Westerville, OH, 1990, p. 445.
Page 458
39. M. Barsoum, Degradation of ceramics in alkali-metal
environments, NATO Adv. Study Inst. Series B., 199, 241270.
40. M. Poubaix, Atlas of Electrochemical Equilibria in Aqueous
Solutions, NACE, Houston, 1974. p. 29.
41. M. G. Fontana, Corrosion Engineering, 3rd ed., McGraw-Hill,
New York, 1986, p. 153.
42. H. H. Uhlig (ed.), The Corrosion Handbook, John Wiley and
Sons, New York, 1948, p. 953.
43. Laboratory corrosion testing of metals for the process industries,
Standard Test Method TM0169-76, NACE, Houston, 1969.
44. E. L. Liening, Electrochemical corrosion testing techniques,
Process Industries Corrosion: The Theory and Practice, NACE,
Houston, 1986, pp. 85122.
45. A. Wachter and R. S. Treseder, Corrosion testingevaluation of
metals for process equipment, Chem. Eng. Prog., 43: 315 (1947).
46. F. McGreary and B. Lifka, Corrosion testing, NACE Basic
Corrosion Course, NACE, Houston, 1975, p. 15.
47. P. P. J. Ramaekers, Ceramic research in The Netherlands: Initial
experiences and highlights of the Centre for Technical Ceramics,
Kemia Kemi, 16(2): 1989.
Page 459
17
Chemical-Resistant Mortars, Grouts, and Monolithic
Surfacings
Augustus A. Boova
Atlas Minerals and Chemicals, Inc.
Mertztown, Pennsylvania
1
Introduction
The industrialization of America following the turn of the century and
the subsequent expansion of our agricultural industry created the need
for chemical-resistant construction materials. The steel and
metalworking, chemical including explosives, dyestuffs, and fertilizer
industries were the initial industries with severe corrosion problems.
The pulp and paper, petroleum, petrochemical, and automotive
industries followed with similar corrosion problems.
Specifically, the pickling, plating, and galvanizing of metals require
sulfuric, hydrochloric, hydrofluoric, chromic, nitric, and phosphoric
acids. The manufacture of explosives, dyestuffs, fertilizer, and other
agricultural products requires the same acids as well as other corrosive
chemicals.
The pulp and paper industry, from inception of the wood chip into the
digester and the subsequent bleaching of pulp, required similar acids.
Sodium hydroxide, sodium hypochlorite, and chlorine were additional
mandated chemicals for this and other industries.
Rail, automotive, petroleum and petrochemical operations, and food
and beverage sanitation mandates ultimately contributed to additional
industrial corrosion problems.
Early reaction vessels in the process industries utilized lead linings
with or
Page 460
without further protection from various types of brick, tile, porcelain,
and ceramic sheathings. The early jointing materials for installing
these ceramic-type linings utilized siliceous fillers mixed with
inorganic binders based on various silicates as well as mortar based on
litharge and glycerine. The limitations of these mortars and grouts
stimulated research for better setting and jointing materials for
installing chemical-resistant brick, tile, and ceramics. The best brick
or tile in the world for installing floors or tank linings is only as good
as the mortars and grouts used to install it.
In the early 1930s, the first plasticized hot-pour, acid-resistant sulfur
mortar was introduced in the United States. Sulfur mortars were
immediately accepted by industry; however, they have thermal
limitations and lack resistance to alkalies and solvents.
In the late 1930s, the first American acid-, salt-, and solvent-resistant
phenolic mortar was introduced. In the early 1940s, the first furan
mortar was developed and introduced in the United States and soon
became the standard of the industry. It provided outstanding resistance
to acids, alkalies, salts, and solvents.
Additional resin-binder systems have been developed for use as
chemical-resistant mortars, grouts, and monolithic surfacings.
It is important to understand the vernacular of an industry. "Mortar"
and "grout" are terms generally associated with the brick, tile, and
masonry trades. Mortars and grouts are used for ''setting" and joining
various types and sizes of brick and tile.
A mortar can be described as a material of heavy consistency that can
support the weight of the brick or tile without being squeezed from the
joints while the joint is curing. Chemical-resistant mortar joints are
customarily approximately 1/8 in. (3 mm) wide. A mortar is applied
by buttering each unit and is generally associated with bricklayer
trade.
A grout can be described as a thin or "soupy" mortar used for filling
joints between previously laid brick or tile. Grout joints are
customarily approximately 1/4 in. (6 mm) wide. A grout is applied by
"squeegeeing" it into the open joints with a flat rectangular rubberfaced trowel and is generally associated with tilesetting.
Chemical-resistant machinery grouts are also available whose
formulations are similar to those of the tile grouts. Machinery grouts
generally utilize larger aggregates than tile grouts. Resin viscosities
can also vary from those of the tile grouts.
Chemical-resistant monolithic surfacings or toppings are a mixture of
a liquid synthetic resin binder, selected fillers, and a setting agent for
application to concrete in thicknesses ranging from approximately
1/16 in. (1.5 mm) to 1/2 in. (13 mm). Materials applied in thicknesses
greater than 1/2 in. (13 mm) are usually described as polymer
concretes. Polymer concretes are defined as a composition
Page 461
of low-viscosity binders and properly graded inert aggregates, which
when combined and thoroughly mixed yield a chemical-resistant
synthetic concrete that can be precast or poured in place. Polymer
concretes can also be used as a concrete surfacing, with the exception
of sulfur cement polymer concrete. All polymer concretes can be used
for total poured-in-place reinforced or unreinforced slabs. They can
also be used for precasting of slabs, pump pads, column bases,
trenches, tanks, and sumps, to mention a few. By definition,
monolithic surfacings are also polymer concretes.
Chemical-resistant mortars, grouts, and monolithic surfacings are
based on organic and inorganic chemistry. The more popular materials
of each group will be discussed.
II
Materials Selection
The success or failure of a chemical-resistant mortar, grout, or
monolithic surfacing is based on proper selection of materials and
their application.
To select the proper material you must define the problem:
1. Identify all chemicals that will be present and their concentration. It
is not enough to say that pH will be 4, 7, or 11. This tells you it's
either acid, neutral, or alkaline; it doesn't identify whether the
environment is oxidizing or nonoxidizing, organic or inorganic,
alternately acid or alkaline, etc.?
2. Is the application fumes and splash or total immersion? Floors can
have integral trenches and sumps, curbs, pump pads.
3. What are the minimum or maximum temperature to which the
installation will be subjected?
4. Is the installation indoors or outdoors? Thermal shock and
ultraviolet exposure can be deleterious to many resin systems.
5. What are the physical impositions? Foot traffic vs. vehicular traffic,
impact from dropping steel plates vs. paper boxes, etc., must be
defined.
6. Longevityhow long must it last? Is process obsolescence imminent?
This could have a profound effect on cost.
7. Must it satisfy standards organizations such as USDA or FDA?
Some systems don't comply.
8. Are the resin systems odoriferous? This could preclude their use in
many processing plants, such as food, beverage, and pharmaceutical.
Many systems are odoriferous.
Answers to these questions will provide the necessary information to
make a proper selection from the available resin system.
Page 462
III
Chemical-Resistant Mortars and Grouts
Chemical-resistant mortars and grouts are composed of a liquid resin
or an inorganic binder, fillers such as carbon, silica, and combinations
thereof, and a hardener or catalyst system that can be incorporated in
the filler or added as a separate component.
The workability of a mortar or grout is predicated on properly selected
fillers or combinations of fillers, particle size and gradation of these
fillers, resin viscosity, and reactivity of catalysts and hardeners.
Improper filler gradation and high-viscosity resins produce mortars
with poor working properties. Hardener systems must be properly
balanced for application at thermal ranges of approximately 60°90°F
(15°32°C). The higher the temperature, the faster the set; the lower the
temperature, the slower the set. Improper ambient and material
temperature can also have a deleterious effect on the quality of the
final installation, i.e., adhesion to brick, tile or substrate, high or low,
rough or porous, or improperly cured joints.
The most popular fillers used are carbon, silica, and combinations
thereof. Baraytes has been used in conjunction with carbon and silica
for specific end-use application. Irrefutably, carbon is the most inert
of the fillers; consequently in many of the resin and sulfur systems it
is the filler of choice. It provides resistance to most chemicals,
including strong alkalies, hydrofluoric acid, and other fluorine
chemicals.
The general chemical resistance of various fillers to acids, alkalies,
salts, fluorine chemicals, and solvents is enumerated in Table 1. The
most popular resin system from which chemical-resistant mortars and
grouts are formulated follow.
Table 1 Guide to Chemical Resistance of
Fillers
Filler
Combination
Medium, 20% CarbonSilica carbon-silica
Hydrochloric
R
R
R
acid
Hydrofluoric
R
N
N
acid
Sulfuric acid
R
R
R
Potassium
R
N
N
hydroxide
Sodium
R
N
N
hydroxide
Neutral salts
R
R
R
Solvents, conc. R
R
R
R, recommended; N, not recommended.
Page 463
A
Organic
1
Epoxy
The most popular epoxy resins used in the formulation of corrosionresistant mortars, grout, and monolithic surfacings are low-viscosity
liquid resins based on
1. Bisphenol A
2. Bisphenol F (epoxy Novolac)
3. Epoxy phenol Novolac
These base components are reacted with epichlorohydrin to form
resins of varying viscosity and molecular weight. The subsequent
molecular orientation is predicated on the hardener systems employed
to effect the cure or solidification of the resin. The hardening systems
selected will dictate the following properties of the cured system:
1. Chemical and thermal resistance
2. Physical properties
3. Moisture tolerance
4. Workability
5. Safety during use
Of the three systems enumerated, the bisphenol A epoxy has been the
most popular followed by the bisphenol F, sometimes referred to as an
epoxy Novolac resin. The epoxy phenol Novolac is a higher viscosity
resin that requires various types of diluents or resin blends for
formulating mortars, grouts, and some monolithic surfacings. The
bisphenol A resin uses the following types of hardeners:
1. Aliphatic amines
2. Modified aliphatic amines
3. Aromatic amines
4. Others
Table 2 shows effects of the hardener on the chemical resistance of
the finished mortar or grout of bisphenol A systems when exposed to
organic, inorganic, and oxidizing acids as well as aromatic solvents.
Table 3 provides summary chemical resistance of optimum chemicalresistant bisphenol A, aromatic amine cured with bisphenol F resin
systems.
Amine hardening systems being alkaline provide a high degree of
compatibility of these systems for application to a multitude of
substrates, such as poured-in-place and precast concrete, steel, wood,
fiberglass-reinforced plastics (FRP), brick, tile, and ceramics, to name
a few.
The most popular filler used for epoxy mortars and grouts is silica.
Unfortunately, this precludes their use in hydrofluoric acid, other
fluorine chemicals, and hot strong alkalies. Carbon-filled mortars and
grouts are available,
Page 464
Table 2 Types of Epoxy Hardeners and Their
Effect on Chemical Resistance
Hardeners
Modified
Aliphatic
aliphatic
Aromatic
Medium
amines
amines
amines
Acetic
C
N
R
acid, 510%
Benzene
N
N
R
Chromic
C
N
R
acid, <5%
Sulfuric
R
C
R
acid, 25%
Sulfuric
C
N
R
acid, 50%
Sulfuric
N
N
R
acid, 75%
R, recommended; N, not recommended; C,
conditional.
however, with some sacrifice to working properties. Fortunately, their
most popular applications have been industrial and institutional where
by optimum physical properties are required and exposure to elevated
temperatures and corrosives are moderate.
Epoxy systems have outstanding physical properties. They are the
premier products where optimum adhesion is a service requirement.
Minimum physiTable 3 Comparative Chemical and
Thermal Resistance of Bisphenol A,
Aromatic AmineCured vs. Bisphenol F
(Epoxy Novolac)
Medium, R.T.
Bisphenol Bisphenol
A
F
Acetone
N
N
Butyl acetate
C
E
Butyl alcohol
C
E
Chromic acid, 10%
C
E
Formaldehyde, 35%
E
G
Gasoline
E
E
Hydrochloric acid, to
E
E
36%
Nitric acid, 30%
N
C
Phosphoric acid,
E
E
50%
Sulfuric acid, to 50%
E
E
Trichloroethylene
N
G
Max. temp., °F (°C) 160 (71) 160 (71)
R.T., room temperature; C, conditional;
N, not recommended; E, excellent; G,
good.
Page 465
cal and thermal properties are enumerated in Table 4. The values
given are indicative of the differences that can actually be experienced
for the respective hardener systems.
Amine hardening systems are the most popular for ambient
temperature curing epoxy mortars, grouts, and monolithic surfacings.
These systems are hygroscopic and can present allergenic responses to
sensitive skin. These responses can be minimized or virtually
eliminated by attention to personal hygiene and the use of protective
creams on exposed areas of skin, i.e., face, neck, arms, and hands.
Protective garments including gloves are recommended when using
expoxy materials.
The bisphenol F or epoxy Novolac are similar systems to the
bisphenol A epoxy systems in that they use alkaline hardeners and the
same fillers. The major advantage for the use of the bisphenol F is
improved resistance to
1. Aliphatic and aromatic solvents
2. Higher concentrations of oxidizing and nonoxidizing acids
Disadvantages of these systems are that they involve
1. Less plastic with slightly more shrinkage
2. Slightly less resistance to alkaline mediums
The thermal resistance and physical properties are otherwise very
similar to the bisphenol A systems.
2
Furans
The polyfurfuryl alcohol or furan resins are the most versatile of all
the resins used to formulate corrosion-resistant mortars and grouts.
They are used for monolithic surfacings; however, they are not a
popular choice due to their brittleness and their propensity to shrink.
They provide a broad range of chemical resistance to most
nonoxidizing organic and inorganic acids, alkalies, salts, oils, greases,
and solvents to temperatures of 360°F (182°C). Table 5 provides
comparative chemical resistance for furan resin mortars and grouts
with 100% carbon and part carbon/silica fillers.
Of all the room temperature curing resins, furans are one of the
highest in thermal resistance with excellent physical properties. Furan
resins are unique in that they are agriculturally and not
petrochemically based as are most synthetic resins. Furfuryl alcohol is
produced from such agricultural byproducts as corn cob, bagasse, rice,
and oat hull.
The furan resin mortars and grouts are convenient-to-use, twocomponent systems consisting of the resin and a filler. The catalyst or
hardener system is an acid that is contained in the filler. The most
popular fillers are carbon, silica, and a combination of carbon and
silica. The 100% carbon-filled furan resin mortars and grouts provide
the broadest range of chemical resistance because of the
Page 466
Table 4 Minimum Physical and Thermal Properties: Effect of Hardener Systems on
Bisphenol A Epoxy Mortars vs. Grouts
Modified
Aliphatic amines aliphatic amines Aromatic amines
Property
Mortar Grout Mortar Grout Mortar
Tensile, psi (MPa)
1400 1200 (8) 1200 (8) 1000 (7) 1600
(10)
(11)
ASTM Test Method C-307
Flexural, psi (MPa)
2200
2000
2500
2100
3500
(15)
(14)
(17)
(15)
(24)
ASTM Test Method C-580
Compressive, psi (MPa)
8000
7000
8000
7000
9000
(55)
(48)
(55)
(48)
(62)
ASTM Test Method C-579
Bond to Brick or Tile, psi
Brick failed
Brick failed
Brick failed
(MPa)
ASTM Test Method C-321
Max. temp., °F (°C)
140 (60) 140 (60) 135 (57) 135 (57) 160 (71) 160 (71)
Page 467
Table 5 Chemical Resistance Furan Resins
Mortars and Grouts: 100% Carbon vs. Part
Carbon/Silica Fillers
Medium, R.T.
100%
Part
carbon
carbon/silica
Acetic acid, glacial
R
R
Benzene
R
R
Cadmium salts
R
R
Chlorine dioxide
N
N
Chromic acid
N
N
Copper salts
R
R
Ethyl acetate
R
R
Ethyl alcohol
R
R
Formaldehyde
R
R
Fatty acids
R
R
Gasoline
R
R
Hydrochloric acid
R
R
Hydrofluoric acid
R
N
Iron salts
R
R
Lactic acid
R
R
Methyl ethyl ketone
R
R
Nitric acid
N
N
Phosphoric acid
R
R
Sodium chloride
R
R
Sodium hydroxide,
R
C
to 20%
Sodium hydroxide,
R
N
40%
Sulfuric acid, 50%
R
R
Sulfuric acid, 80%
C
C
Trichloroethylene
R
R
Trisodium phosphate
R
C
Xylene
R
R
R, recommended; N, not recommended; C,
conditional.
inherent chemical resistance of the resin and the carbon filler to all
concentrations of all alkalies as well as hydrofluoric acid and other
fluorine chemicals. The advantages of mortars with part carbon and
part silica fillers are slightly improved workability, physical
properties, and cost. Grouts generally utilize 100% carbon filler
because of the superior chemical resistance and flow properties.
The acidic catalysts employed in furan systems precludes their use
directly on concrete, steel, and other substrates that could react with
the acid. This limitation is easily circumvented by using various
membranes, primers, or mortar bedding systems that are compatible
with the substrate.
The process industries use lining systems (membranes) on most
substrates
Page 468
onto which brick and tile are installed to ensure total resistance from
aggressive environments encountered in such applications as
1. Pickling, plating, and galvanizing tanks in the steel and
metalworking industries
2. Absorber towers in sulfuric acid plants
3. Scrubber in flu gas desulfurization applications
4. Floors in wet acid battery and chemical plants
5. Above-grade applications in dairies, food and beverage, and other
processing plants
The versatility of furans is further exemplified by these available
variations:
1. High-bond-strength materials for optimum physical mandates
2. Normal-bond-strength materials for economy and less demanding
physical impositions
3. 100% carbon filled for resistance to all concentrations of alkalies
and most fluorine chemicals
4. Different ratios of carbon and silica for applications requiring
varying degrees of electrical resistance or conductivity
Table 6 provides thermal and physical properties for normal-and highbond-strength, 100% carbon-filled furan mortars and grouts.
Part carbon, part silicafilled furan mortars have physical properties
equal to or slightly better than the normal bond 100% carbon-filled
furan mortar.
Seldom if ever are furan grouts, normal or high bond strength, filled
with
Table 6 Minimum Physical and Thermal Properties: 100%
Carbon-Filled Furan Mortar vs. Grout
Mortar
Grout
Normal High Normal High
Property
bond
bond
bond
bond
Tensile, psi (MPa)
800 (6)
800 (6) 800 (6)
900 (6)
ASTM Test Method C307
Flexural, psi (MPa)
1600 (11) 1600 1600 (11) 1800
(11)
(12)
ASTM Test Method C580
Compressive, psi
5000 (34) 5000 5000 (34) 5000
(MPa)
(34)
(34)
ASTM Test Method C579
Bond, psi (MPa)
150 (1)
800 (6) 150 (1)
700 (5)
ASTM Test Method C321
Max. temp., °F (°C) 350 (177) 350
350 (177) 350
(177)
(177)
Page 469
anything but 100% carbon filler. Consequently, comparative data for
part carbon, part silicafilled grouts will not be provided.
3
Phenolics
The origin of phenolic resins was in Europe dating back to the late
1800s. At the turn of the century the only chemical-resistant mortar
available was based on the inorganic silicates. These materials possess
outstanding acid resistance but little or no resistance to many other
chemicals. The silicates also exhibited significant physical limitations.
After World War I the limitation of the silicates prompted further
investigation of the phenolics. These resins ceased to be laboratory
curiosities and ultimately made their way into a multitude of
applications because of their excellent physical properties. Early
application for phenolic resins was for molding of telephones and
associated electrical applications.
By the 1930s, the chemical process and the steel and metalworking
industries mandated more functional chemical-resistant mortars for
installing chemical-resistant brick. Besides chemical resistance, they
had to have excellent physical properties.
By the mid-1930s, the first chemical-resistant phenolic resin mortar
was introduced in the United States. It met the two most important
mandates of the chemical, steel, and metalworking industries, i.e.,
1. Provide resistance to high concentrations of acids and in particular
to sulfuric acid at elevated temperatures.
2. Provide low absorption with good bond strength to various types of
brick, tile, and ceramics while possessing excellent tensile, flexural,
and compressive properties.
To this day phenolic resin mortars fulfill many of the requirements in
the manufacture and use of the many grades and concentrations of
sulfuric acid.
The steel and metalworking industries continue to use phenolic
resinbased, chemical-resistant mortars for brick-lined pickling,
plating, and galvanizing applications.
Phenolic resins are sufficiently functional to permit use of 100%
carbon, 100% silica, or part silica and part carbon as fillers in phenolic
mortars. Silica fillers are the most dominant for use in high
concentrations of sulfuric acid and where electrical resistance is
required. Carbon fillers are used where resistance to high
concentrations of hydrofluoric acid are required. They are also used as
adhesive and potting compounds for corrosive electrical conductance
applications. Phenolic mortars are similar to the furans in that they are
two-component, easy-to-use mortars, with the acid catalyst or curing
agent incorporated in the powder. Phenolic resins are seldom used to
formulate grouts or monolithic surfacings.
Phenolic resins have a limited shelf life and must be stored at 45°F
(7°C).
Page 470
Phenolic resin mortars, like epoxies, can be allergenic to sensitive
skin. This can be minimized or prevented by exercising good personal
hygiene and using protective creams.
Table 7 provides the minimum physical and thermal properties for
100% carbon and 100% silica-filled phenolic resin mortars. Table 8
provides comparative chemical resistance for phenolic mortars
compared to furan mortars, carbon vs. silica-filled.
4
Polyesters
Chemical-resistant polyester mortar was developed and introduced in
the early 1950s at the specific request of the pulp and paper industry.
The request was for a mortar with resistance to a new bleach process
utilizing chlorine dioxide. Polyester mortars ultimately became the
premier mortar for use where resistance to oxidizing mediums is
required.
Unsaturated polyester resins are also used for formulating tile and
machinery grouts as well as monolithic surfacing. These applications
are formidable challenges for the formulator due to their propensity to
cause shrinkage.
Polyester mortars can be formulated to incorporate carbon and silica
fillers depending on the end use intended. For applications requiring
resistance to hydrofluoric acid, fluorine chemicals, and strong
alkalies, such as sodium and potassium hydroxide, 100% carbon
fillers are required.
Polyester resins are available in a number of types, the most popular
of which are the following:
Table 7 Minimum Physical and Thermal
Properties: 100% Carbon vs. 100% Silica-
filled Phenolic Mortars
Property
Tensile, psi (MPa)
ASTM Test Method
C-307
Flexural, psi (MPa)
ASTM Test Method
C-580
Compressive, psi
(MPa)
ASTM Test Method
C-579
Bond, psi (MPa)
ASTM Test Method
C-321
Absorption, %
ASTM Test Method
C-413
Maximum temp., °F
(°C)
Filler
100%
100%
carbon
silica
800 (6) 400 (3)
1800 (13)
1800
(13)
4500 (31)
6000
(41)
150 (1)
150 (1)
1.0
1.0
350 (177)
350
(177)
Page 471
Table 8 Comparative Chemical Resistance: Phenolic
Mortars vs. Furan Mortars
Furan
Phenolic
Medium, R.T.
Carbon SilicaCarbon Silica
Amyl alcohol
R
R
R
R
Chromic acid, 10%
N
N
N
N
Gasoline
R
R
R
R
Hydrofluoric acid, to 50%
R
N
R
N
Hydrofluoric acid, 93%
N
N
R
N
Methyl ethyl ketone
R
R
R
R
Nitric acid, 10%
N
N
N
N
Sodium hydroxide, to 5%
R
R
N
N
Sodium hydroxide, 30%
R
N
N
N
Sodium hypochlorite, 5%
N
N
N
N
Sulfuric acid, 50%
R
R
R
R
Sulfuric acid, 93%
N
N
R
R
Xylene
R
R
R
R
R.T., room temperature; R, recommended; N, not
recommended.
1. Isophthalic
2. Chlorendic acid
3. Bisphenol A fumarate
The earliest mortars and grouts were based on the isophthalic
polyester resin. This resin performed well in many oxidizing
mediums. It did, however, present certain physical, thermal, and
chemical resistance limitations.
Formulations utilizing the chlorendic and bisphenol A fumarate resins
offered improved chemical resistance, higher thermal capabilities, and
improved ductility with less shrinkage. The bisphenol A fumarate
resins offered significantly improved resistance to alkalies. They
provided essentially equivalent resistance to oxidizing mediums.
All polyester resin systems have provided outstanding chemical
resistance to a multitude of applications in the pulp and paper, textile,
steel and metalworking, pharmaceutical, and chemical process
industries. Typical applications have been brick and tile floors, brick
and tile lining in bleach towers, scrubbers, pickling and plating, and
waste-holding and treating tanks. All the resins provide formulation
flexibility to accommodate carbon and silica as fillers. Carbon-and
silica-filled mortars and grouts are easily mixed and handled for the
various types of installations. They are easily pigmented for aesthetic
considerations. The essentially neutral curing systems provide
compatibility for application to a
Page 472
multitude of substrates, i.e., concrete, steel, fiberglass-reinforced
plastics (FRPs), etc. Properly formulated polyester resin systems
provide installation flexibility to a wide range of temperatures,
humidities, and contaminants encountered on most construction sites.
They are one of the most forgiving of all of the resin systems.
Polyester mortars and grouts have certain limitations that are inherent
in all polyester formulations. They are as follows:
1. Strong aromatic odor that can be offensive for certain indoor and
confined space applications
2. Shelf life limitations that can be controlled by low-temperature
storage (below 60°F [15°C]) of the resin component.
Table 9 provides comparative chemical resistance for the previously
enumerated polyester resins. The physical properties of the respective
systems are of a magnitude that for most mortar and grout, tile and
masonry applications they can be considered essentially equal.
Minimum physical properties for polyester resin mortars utilizing
100% carbon and 100% silica fillers are presented in Table 10.
5
Vinyl Ester and Vinyl Ester Novolac
Chemically these resins are addition reactions of methacrylic acid and
epoxy resin. The chemistry of these resins has prompted their being
referred to as acrylated epoxies. They possess many of the properties
of epoxy, acrylic, and bisphenol A fumarate polyester resins. There
similarity to these resins is why the outstanding chemical resistance
and physical properties of mortars and grouts formulated from these
resins.
The vinyl esters are generally less rigid with lower shrinkage than
many polyester systems. They compare favorably with the optimum
chemical-resistant
Table 9 Comparative Chemical and Thermal
Resistance of Various Polyester Resins
Bisphenol
A
Medium, R.T. IsophthalicChlorendic fumarate
Acids,
R
R
R
oxidizing
Acids,
R
R
R
nonoxidizing
Alkalies
N
N
R
Salts
R
R
R
Bleaches
R
R
R
Max. temp., 225 (107) 260 (127) 250 (121)
°F (°C)
R.T., room temperature; R, recommended; N,
not recommended.
Page 473
Table 10 Minimum Physical Properties:
Polyester Resin Mortars and Grouts
Carbon vs. Silica-Filled
Filler
Property
Carbon Silica
Tensile, psi (MPa)
1500 1800
(10) (12)
ASTM Test Method C307
Flexural, psi (MPa)
3000 4000
(21) (28)
ASTM Test Method C580
Compressive, psi (MPa) 9000 10000
(62) (69)
ASTM Test Method C579
Bond to Brick or Tile, psi 200 (1) 350 (2)
(MPa)
ASTM Test Method C321
bisphenol A fumarate polyester mortars and grouts. The major
advantage of the various vinyl ester systems are
1. Resistance to most oxidizing mediums
2. High concentrations of sulfuric acid, sodium hydroxide, and many
solvents
The vinyl ester resin mortars have supplanted the polyester and have
become the mortar of choice for brick-lines bleach towers in the pulp
and paper industry.
Like polyesters, vinyl ester resin formulations have similar inherent
disadvantages:
1. Strong aromatic odor for indoor or confined space applications.
Isolation of area where installations are being made may be necessary
to prevent in-plant ventilating systems from carrying the aromatic
odor throughout the facility.
2. Shelf life limitations of the resins require refrigerated storage below
60°F (15°C) to extend its useful life.
Table 11 provides comparative chemical resistance for two types each
of polyester and vinyl ester resin mortars and grouts.
B
Inorganic
Inorganic materials predate their organic counterparts, offer fewer
choices, and consequently are somewhat easier to understand. The
most popular materials are
1. Hot-pour sulfur mortars
2. Ambiently mixed and applied silicate mortars
Page 474
Table 11 Comparative Chemical and Thermal Resistance of
Polyester vs. Vinyl Ester Mortars and Grouts
Polyester
Vinyl ester
Bisphenol A Vinyl
Medium, R.T.
Chlorendic fumarate
ester Novolac
Acetic acid,
C
N
N
R
glacial
Benzene
C
N
R
R
Chlorine dioxide
R
R
R
R
Ethyl alcohol
R
R
R
R
Hydrochloric acid,
R
R
R
R
36%
Hydrogen
R
N
R
R
peroxide
Methanol
R
R
N
R
Methyl ethyl
N
N
N
N
ketone
Motor oil and
R
R
R
R
Gasoline
Nitric acid, 40%
R
N
N
R
Phenol, 5%
R
R
R
R
Sodium
N
R
R
R
hydroxide, 50%
Sulfuric acid, 75%
R
C
R
R
Toluene
C
N
N
R
Triethanolamine
N
R
R
R
Vinyl toluene
C
N
C
R
Max. temp., °F
260 (127) 250 (121)
220
230
(°C)
(104) (110)
R.T., room temperature; R, recommended; N, not
recommended; C, conditional.
For years all categories were referred to as ''acid-proof mortars" and
justifiably so because their capabilities are limited to a maximum pH
of 7. They are not intended for alkaline or alternately acid and alkaline
service.
1
Sulfur
The initial application for sulfur mortars was to replace lead for
joining bell and spigot cast iron water lines. Since then sulfur mortars
have been successfully used for installing brick floors, brick-lined
tanks, and joining bell and spigot vitrified clay pipe for corrosive
waste sewer lines in the chemical, steel, and metalworking industries.
Sulfur mortars are plasticized to impart thermal shock resistance.
They utilize 100% carbon or 100% silica fillers.
Due to their outstanding resistance to oxidizing acids, the 100%
carbon-filled sulfur mortar is the mortar of choice for installing carbon
brick in the nitric/hydrofluoric acid picklers in the specialty steel
industry.
Both the 100% carbon and the 100% silica mortars have found
widespread
Page 475
use in the plating, galvanizing specialty, and carbon steel industries.
Advantages to using sulfur mortars when compared to some resin
mortars are as follows:
1. Resistance to oxidizing, nonoxidizing, and mixed acids
2. Ease of use
3. High early strength; "cool ituse it"
4. Resistance to thermal shock
5. Unlimited shelf life
6. Prefabrication and in-place construction
7. Economy
The many advantages of sulfur mortars make it ideal for applications
such as setting of anchor bolts and posts, capping of concrete test
cylinders, "proving" of molds for castings and hubs for grinding
wheels.
A disadvantage of sulfur mortars is flammability. When sulfur mortar
masonry sheathings are being installed, melting equipment is placed
outdoors and molten materials moved to the point of use, thus
minimizing potential for flammability problems.
Installation of brick linings on vertical surfaces with sulfur mortars
utilizes the concept of dry-laying brick by placing the brick on sulfur
mortar spacer chips, papering the face of the brick to dam the joint,
and pouring behind the brick to fill the joints. Horizontal surfaces are
poured until the mortar comes up into the joints approximately 1/4 in.
from the top of the brick. A final "flood pour" is made over the entire
surface, thus ensuring full flush joints. As soon as the joints cool, the
installation is ready for service. Approximately 95% of the
compressive value of sulfur mortars is attained 5 min after the mortar
has solidified.
Table 12 provides minimum thermal and physical properties for 100%
carbon and 100% silica sulfur mortars. Table 13 provides comparative
chemical resistance to environments most commonly experienced in
industries where sulfur mortars are used.
2
Silicates
These mortars are most notable for their resistance to concentrated
acids, except hydrofluoric acid and similar fluorinated chemicals, at
elevated temperatures. They are also resistant to many aliphatic and
aromatic solvents. They are not intended for use in alkaline or
alternately acid and alkaline environments. This category of mortars
includes
1. Sodium silicate
2. Potassium silicate
3. Silica (silica sol)
The major applications for these mortars have been in the construction
of sulfuric acid plants and the brick lining of stacks subjected to
varying concentra-
Page 476
Table 12 Minimum Physical and
Thermal Properties: 100% Carbon or
100% Silica-filled Sulfur Mortars
Filler
100% Carbon or
Property
silica
Tensile, psi (MPa)
400 (3)
ASTM Test Method
C-307
Flexural, psi (MPa)
1000 (7)
ASTM Test Method
C-580
Compressive, psi
4000 (28)
(MPa)
ASTM Test Method
C-579
Bond to Brick, psi
150 (1)
(MPa)
ASTM Test Method
C-321
Max. temp., °F (°C)
200 (93)
Table 13 Comparative Chemical
Resistance: 100% Carbon vs. 100%
Silica-filled Sulfur Mortars
Medium, R.T.
100%
100%
carbon
silica
Acetic acid, to 10%
R
R
Aqua regia
N
N
Cadmium salts
R
R
Chromic acid, to
R
R
20%
Gold cyanide
R
R
Hydrochloric acid
R
R
Hydrofluoric acid
R
N
Iron salts
R
R
Magnesium salts
R
R
Methyl ethyl
N
N
ketone
Mineral spirits
N
N
Nickel salts
R
R
Nitric acid, to 40%
R
R
Nitric/hydrofluoric
R
R
acid
Phosphoric acid
R
R
Silver nitrate
R
R
Sodium hydroxide
N
N
Sodium salts
R
R
Sulfuric acid, 80%
R
R
Toluene
N
N
Trichloroethylene
N
N
Zinc, salts
R
R
R.T., room temperature; R,
recommended; N, not recommended.
Page 477
tions of sulfur and nitrogen oxides at elevated temperatures. Their
upper thermal capabilities approach those of refractory mortars.
The sodium and potassium silicate mortars are available as twocomponent systems, filler, and binder, with the setting agent in the
filler. Sodium and potassium silicates are referred to as soluble
silicates due to their solubility in water. They are not affected by
strong acids; however, this phenomenon precludes the use of many
formulations in dilute acid service. However, this disadvantage
becomes an advantage for formulating single-component powder
systems. All that is required is the addition of water at the time of use.
Obviously, as the name of these materials implies, the fillers are pure
silica.
The sodium silicates can be produced over a broad range of
compositions of the liquid binder. These properties and new hardening
systems have significantly improved the water resistance of some
sodium silicate mortars. These formulations are capable of resisting
dilute as well as concentrated acids without compromising physical
properties.
The potassium silicate mortars are less versatile in terms of
formulation flexibility. They are, however, less susceptible to
crystallization in high concentrations of sulfuric acid so long as metal
ion contamination is minimal.
Potassium silicate mortars are available with halogen-free hardening
systems, thus eliminating the remote potential for catalyst poisoning
in certain chemical processes.
The silica or silica sol type of mortars are the newest of this class of
mortars. They consist of a colloidal silica binder with quartz fillers.
The principal difference compared to the other mortars is total
freedom from metal ion that could contribute to sulfation hydration
within the mortar joints in high-concentration sulfuric acid service.
The workability and storage stability is comparable in the sodium and
potassium silicates. The silica materials are harder to use, less
forgiving as to mix ratio, and highly susceptible to irreversible
damage due to freezing in storage.
Table 14 provides thermal and physical properties for the three types
of silicate mortars. The chemical resistances of the various silicate
mortars are very similar. Table 15 points out the subtle differences
between the respective mortars. Silicate mortars will fail when
exposed to mild alkaline mediums, such as bicarbonate of soda. Dilute
acid solution, such as nitric acid, will have a deleterious effect on
sodium silicates unless the water-resistant type is used.
IV
Chemical-Resistant Monolithic Surfacings and Polymer Concretes
The chemistry of monolithic surfacings is an exploitation of the resin
systems used for mortars and grouts. Additional systems will be
included.
To reiterate, monolithic surfacings are installed at thicknesses of 1/16
in. (1.5
Page 478
Table 14 Minimum Physical and Thermal Properties of
Various Silicate Mortars
Potassium
Property
Sodium Normal Halogen- Silica
free
Tensile, psi (MPa)
400 (3) 700 (5) 700 (5) 400 (3)
ASTM Test Method
C-307
Flexural, psi (MPa) 500 (3) 1400
1800 (12) 900 (6)
(10)
ASTM Test Method
C-580
Compressive, psi
2000 (14) 3000
5000 (34) 3500
(MPa)
(21)
(24)
ASTM Test Method C-579
C-396 C-396
C-396
Bond, psi (MPa)
150 (1) 150 (1) 200 (1) 150 (1)
ASTM Test Method
C-321
Max. temp., °F (°C) 2100
1700
1650
1500
(1149) (927)
(900)
(816)
Table 15 Comparative Chemical Resistance: Silicate Mortars
Sodium
Potassium
Medium, R.T.
Normal Water- Normal Halogenresistant
free
Acetic acid, glacial
G
G
R
R
Chlorine dioxide, water
N
N
R
R
sol.
Hydrogen peroxide
N
R
N
N
Nitric acid, 5%
C
R
R
R
Nitric acid, 20%
C
R
R
R
Nitric acid, over 20%
R
R
R
R
Sodium bicarbonate
N
N
N
N
Sodium sulfite
R
R
N
N
Sulfates, aluminum
R
R
R
R
Sulfates, copper
G
G
R
R
Sulfates, iron
G
G
R
R
Sulfates, magnesium
G
G
R
R
Sulfates, nickel
G
G
R
R
Sulfates, zinc
G
G
R
R
Sulfuric acid, to 93%
G
G
R
R
Sulfuric acid, over 93%
G
G
R
R
R.T., room temperature; R, recommended; N, not recommended;
G, potential failure, crystalline growth; C, conditional.
Page 479
mm) to 1/2 in. (13 mm). Polymer concretes are installed at thicknesses
greater than 1/2 in. (13 mm). These materials are formidable corrosion
barriers. Monolithic surfacings are not intended to replace brick floors
in heavy-duty chemical or physical applications. However, they are
economical corrosion barriers for a broad range of applications.
The most popular monolithic surfacings are formulated from the
following resins:
1. Epoxy including epoxy Novolac
2. Polyester
3. Vinyl ester including vinyl ester Novolac
4. Acrylic
5. Urethane, rigid and flexible
Chemical-resistant polymer concretes are formulated from some of
the same generic resins. The more popular resins used are
1. Furan
2. Epoxy, including epoxy Novolac
3. Polyester
4. Vinyl ester, including vinyl ester Novolac
5. Acrylics
6. Sulfur
The major advantages to be derived from the use of chemical-resistant
monolithic surfacings and polymer concretes are as follows:
1. These formulations provide flexibility, giving aesthetically
attractive materials with a wide range of chemical resistances,
physical properties, and methods of application.
2. These formulations provide high early development of physical
properties. Compressive values with some systems reach 5000 psi (35
MPa) in 2 h and 19,000 psi (133 MPa) as ultimate compressive value.
3. Most systems are equally appropriate for applications to new and
existing concrete including pour-in-place and precasting.
4. Systems offer ease of installation by in-house maintenance
personnel.
5. Systems offer economy when compared to many types of brick and
tile installations.
6. Systems are available for horizontal, vertical, and overhead
applications.
Furan polymer concrete is inherently brittle and in large masses have
a propensity to shrink. They are used when resistance to acids,
alkalies, and solvents such as aromatic and aliphatic solvents are
required. They have been
Page 480
successfully used in small areas in the chemical, electronic,
pharmaceutical, steel, and metalworking industries.
Polyester, vinyl ester, and acrylic polymer concrete have strong
aromatic odors that can be offensive to installation and in-plant
personnel. Fire codes, particularly for acrylics, must be scrutinized to
ensure compliance.
Sulfur cement polymer concrete is flammable; consequently, the
potential for oxides of sulfur would preclude its use for most indoor
applications. Sulfur cement polymer concrete is not recommended as
a concrete topping.
Polymer concretes are not to be misconstrued with polymer-modified
portland cement concrete. Polymer concretes are totally chemicalresistant, synthetic resin compounds with outstanding physical
properties. Polymer concretes pass the total immersion test, at varying
temperatures for sustained periods. Polymer-modified portland
cement concrete can use some of the same generic resins as used in
polymer concretes, but with different results.
The major benefits to be derived from polymer-modified portland
cement concrete are as follows:
1. They permit application of concrete in thinner cross-sections.
2. They provide improved adhesion for pours onto existing concrete.
3. They lower absorption of concrete.
4. They improve impact resistance.
5. They improve resistance to salt but not to aggressive corrosive
chemicals such as hydrochloric or sulfuric acid.
The success of monolithic surfacing installations is very much
predicated on the qualifications of the design, engineering, and
installation personnel, be they in-house or outside contractors.
The following fundamental rules are important to the success of any
monolithic surfacing installation:
1. Substrate must be properly engineered to be structurally sound, free
of cracks, and properly sloped to drains.
2. New as well as existing slabs must be clean and dry, free of laitance
and contaminants, with a coarse surface profile.
3. Ambient slab and materials to be installed should be 65°85°F
(18°29°C). Special catalyst and hardening systems are available to
accommodate higher or lower temperatures, if required.
4. Thoroughly prime substrate before applying any monolithic
surfacing. Follow manufacturer's instructions.
5. Thoroughly mix individual and combined components at a
maximum speed of 500 rpm to minimize air entrainment during
installation.
6. Uncured materials must be protected from moisture and
contamination.
Page 481
Monolithic surfacings are unique, versatile materials used primarily as
flooring systems. They are available in a variety of formulations to
accommodate various methods of application. They employ many of
the installation methods that have been successfully utilized in the
portland cement concrete industry. The most popular methods are
1. Hand-troweled
2. Power-troweled
3. Spray
4. Pour-in-place/self-level
5. Broadcast
Hand-troweled applications are approximately 1/4 in. (6 mm) thick
and are suggested for small areas or areas with multiple obstructions
such as piers, curbs, column foundations, trenches, and sumps. The
finished application is tight, dense, and with a high-friction finish.
Topcoat sealers are recommended to provide increased density and
imperviousness with a smooth, easy-to-clean finish. High-and lowfriction finishes are easily accomplished predicated on end use
requirements.
Power trowel installations are the fastest, most economical method for
large areas with minimum obstructions. Minimum thickness is 1/4 in.
(6 mm). Appropriate sealers are available to improve density of finish.
Spray applications are also ideal for areas where corrosion is
aggressive. Spray applications are applied, minimum 1/8 in. (3 mm)
thick in one pass on horizontal surfaces. On vertical and overhead
areas including structural components, the material can be sprayapplied 1/163/32 in. thick (1.52.4 mm) in one pass and without slump.
The mortarlike consistency of the material can be varied to control
slump and type of finish. Floors installed in this manner are dense,
smooth, safe finishes for people and vehicular traffic.
Pour-in-place and self-level materials are intended for flat areas where
pitch to floor drains and trenches in minimum. They are intended for
light-duty areas with minimum process spills. The completed
installation is 1/83/16 in. (35 mm) thick with a very smooth, highgloss, easy-to-clean, aesthetically attractive finish.
Broadcast systems are simple to install, economical, aesthetically
attractive floors applied in thicknesses of 3/321/8 in. (23 mm). Resins
are squeegee-applied to the concrete slab into which filler and colored
quartz aggregates of varying color and size are broadcast or sprinkled
into the resin. After the resin has set, the excess filler and quartz
aggregate is removed by vacuuming or sweeping for reuse. The
process is repeated until the desired floor thickness is achieved. The
finished floor is easy to clean and outstanding for light industrial,
laboratory, cleanroom, and institutional applications.
Page 482
Table 16 Comparative Chemical Resistance
1-A = bisphenol A epoxyaliphatic amine hardener
1-B = bisphenol A epoxyaromatic amine hardener
1-C = bisphenol F epoxy (epoxy novolac)
2-D = polyester resinchlorendic acid type
2-E = polyester resinbisphenol A fumarate type
3-F = vinyl ester resin
3-G = vinyl ester novolac resin
1
2
Medium, R.T.
ABCD E
Acetic acid, to 10%
RRRRR
Acetic acid, 1015%
CRCRR
Benzene
CRRRN
Butyl alcohol
RCRRR
Chlorine, wet, dry
CCCRR
Ethyl alcohol
RCRRR
Fatty acids
CRCRR
Formaldehyde, to 37%
RRRRR
Hydrochloric acid, to 36%
CRRRR
Kerosene
RRRRR
Methyl ethyl ketone, 100%
NNNNN
Nitric acid, to 20%
NNRRR
Nitric acid, 20% to 40%
NNRRN
Phosphoric acid
RRRRR
Sodium hydroxide, to 25%
RRRNR
Sodium hydroxide, 2550%
RCRNR
Sodium hypochlorite, to 6%
CRRRR
Sulfuric acid, to 50%
RRRRR
Sulfuric acid, 5075%
CRRRC
Xylene
NRRRR
R. T., room temperature; R, recommended; N, not
recommended; C, conditional.
Chemical Resistance
3
F G
RR
CR
RR
NR
RR
RR
RR
RR
RR
RR
NN
RR
NC
RR
RR
CR
RR
RR
RR
NR
Monolithic surfacings and polymer concretes mirror image the
chemical resistance of their mortar and grout counterparts. Refer to
corrosion tables previously presented and Table 16 which follows.
Table 16 provides comparative chemical resistance for the most
popular resins used as monolithic surfacings and polymer concrete.
Those systems without chemical-resistant mortar and grout
counterparts are as follows:
Page 483
1. Acrylic
2. Urethane, rigid and flexible
Acrylic monolithic surfacing and polymer concretes are installed in
thicknesses of 1/81/2 in. (313 mm) and 1/2 in. (13 mm) and greater,
respectively. These flooring systems are an extension of methyl
methacrylate chemistry popularized by Rohm and Haas Co.'s
Plexiglas and Dupont's Lucite. They are intended for protection
against moderate corrosion environments.
The principal advantages for their use are as follows:
1. They are the easiest of the resin systems to mix and apply by using
pour-in-place and self-leveling techniques.
2. Due to their outstanding weather resistance, they are equally
appropriate for indoor and outdoor applications.
3. They are the only system that can be installed at below freezing
temperatures, 25°F (-4°C), without having to use special hardening or
catalyst systems.
4. They are the fastest set and cure of all resin systems. The
monolithics will support foot and light-wheeled traffic in 1 h whereas
the thicker cross-section polymer concrete will also support foot and
light-wheeled traffic in 1 h while developing 90% of its ultimate
strength in 4 h.
5. They are the easiest to pigment and with the addition of various
types of aggregate can be aesthetically attractive.
6. They are equally appropriate for maintenance and new-construction
applications. They bond well to concrete. They are ideal for
rehabilitating, manufacturing, warehouse, and loading dock floors to
impart wear resistance and ease of cleaning.
As previously indicated, a disadvantage inherent in acrylic systems is
the aromatic odor for indoor or confined space applications.
The urethane systems are intended to be monolithic floors with
elastomeric properties installed in thicknesses of 1/81/4 in. (36 mm).
Similar to the acrylic systems, the urethanes are intended for
protection against moderate to light corrosion environments. Standard
systems are effective at temperatures of 10 to 140°F (-24 to 60°C).
High-temperature systems are available for exposure to temperatures
of 10 to 180°F (-24 to 82°C). Many of the urethane systems are
capable of bridging cracks up to 1/16 in. (1.6 mm).
The monolithic urethane flooring systems offer the following
advantages:
1. They are easy to mix and apply using the pour-in-place, self-level
application technique.
2. Systems are available for indoor and outdoor applications.
Page 484
3. The elastomeric quality of the systems provides underfoot comfort
for production line flooring applications.
4. Because of their being elastomeric, they have excellent sounddeadening properties.
5. They have outstanding resistance to impact and abrasion.
6. They are excellent waterproof flooring systems for above-grade
light-and heavy-duty floors. They are equally appropriate for
maintenance and new-construction applications.
7. They are capable of bridging cracks in concrete 1/16 in. (1.5 mm)
wide.
Urethane materials are demanding systems during installation. Mix
ratio of components, temperature, and humidity controls are
mandatory for successful installations.
The comparative chemical resistances of acrylic and urethane systems
are presented in Table 17.
The physical properties of acrylic systems are substantially different
from those of the urethanes. The acrylic flooring systems are
extremely hard and should be considered too brittle for applications
subjected to excessive physical abuse
Table 17 Comparative Chemical Resistance:
Urethane vs. Acrylic Systems
Urethane
Medium, R.T.
AcrylicStandard
Hightemperature
Acetic acid, 10%
G
G
C
Animal oils
G
G
N
Boric acid
E
E
E
Butter
G
F
N
Chromic acid, 510%
C
C
C
Ethyl alcohol
N
N
N
Fatty acids
F
F
N
Gasoline
E
N
N
Hydrochloric acid,
F
C
C
2036%
Lactic acid, above 10% F
C
C
Methyl ethyl ketone,
N
N
N
100%
Nitric acid, 510%
G
C
F
Sulfuric acid, 2050%
G
C
C
Water, fresh
E
E
E
Wine
G
G
F
R.T., room temperature; E, excellent; G, good; F,
fair; C, conditional, N; not recommended.
Page 485
such as impact from large-diameter steel pipe, steel plate, and heavy
castings. The inherent flexibility and impact resistance of urethanes
offer potential for these types of applications.
Table 18 provides physical and thermal properties for the various
acrylic and urethane flooring systems.
Precast and poured-in-place polymer concrete has been successfully
used in a multitude of indoor and outdoor applications in various
industries.
Acrylic polymer concrete has been used for precast trenches and
covers, delta bus supports, and insulators. Ease of pigmenting to
match corporate colors, good weather resistance, resistance to
airborne SO2, SO3, and NOx, and dielectric properties have made
acrylics particularly attractive in the electric utility industry.
The chemical, steel, electronic, automotive, and pharmaceutical
industries have taken advantage of the ease of mixing and placing, as
well as the high early strength and chemical resistance properties, of
various polymer concretes.
V
Chemical-Resistant Mortars, Grouts, and Monolithic Surfacings
Carbon steel and reinforced concrete are outstanding general
construction materials. They have an enviable record of success in a
multitude of industries and applications. Unfortunately, steel and
concrete will corrode when oxygen and water are present. Weather
and chemicals are catalysts that accelerate the corrosion process.
Table 18 Minimum Physical and Thermal Properties Acrylic
Monolithic Surfacing and Polymer Concrete Standard and HighTemperature Urethane Monolithic Surfacings
Acrylics
Urethanes
Polymer
High-
Property
Monolithicconcrete
Tensile, psi (MPa) 1000 (7) 1050 (7)
ASTM Test
Method C-307
Flexural, psi
2500 (17) 2600 (18)
(MPa)
ASTM Test
Method C-580
Compressive, psi 8000 (55) 9500 (66)
(MPa)
ASTM Test
Method C-579
Bond to Concrete Concrete Concrete
fails
fails
Max. temp., °F
150 (66) 150 (66)
(°C)
Standard temperature
650 (5) 550 (5)
1100 (8) 860 (6)
2500 (17) 1500 (10)
Concrete Concrete
fails
fails
140 (60) 180 (82)
Page 486
The cost attributable to corrosion in the United States is estimated to
be in a range of $9 billion to $90 billion. This figure was confirmed
from a study initiated several years ago by the National Institute of
Standards and Technology (formerly National Bureau of Standards).
The range includes corrosion attributable to chemical processes, to
corrosion of highways and bridges from de-icing chemicals, to
atmospheric corrosion of a steel fence. The economic losses
attributable to corrosion are confirmed by various technical
organizations, such as NACE International and others.
Chemical-resistant mortars, grouts, and monolithic surfacings are used
for protecting steel and concrete in a host of applications. Chemicalresistant mortars and grouts for installing all sizes and shapes of brick,
tile, and ceramics provide the premier thermal, physical, and
chemical-resistant sheathing for protecting linings and membranes
applied to steel and concrete. Typical applications include the
following:
1. Below, on, and above grade floors
2. Pickling, plating, storage, and chemical process tanks and towers
3. Waste holding and treatment tanks
4. Dual containment for outdoor process and storage vessels,
including leak detection systems
5. Stacks and scrubbers from incineration and treatment of toxic waste
fumes
6. Above-and below-grade trenches, pipelines, sumps, and manholes.
Polymer concretes can be used as fast-set, totally chemical-resistant,
poured-in-place slabs or as a topping for refurbishing of existing
concrete slabs. They are highly chemical-resistant materials of
exceptional physical properties that can also be used for a multitude of
precast applications, such as curbs, piers, foundations, pump pads,
stair treads, trenches and covers, sumps, and manholes.
Chemical-resistant, nonmetallic construction materials are formidable,
economical corrosion barriers for protecting steel and concrete when
compared to most alloys.
Nonmetallic construction systems can incorporate in their design dual
containment and leak detection to meet all federal and state
environmental mandates.
The pursuit of clean air and water will continue unabated.
Environmental laws will continue to be stringent. Recycling, waste
treatments, and incineration will require close attention to corrosivity
of all processes.
The agricultural industry, producers of various fertilizers and
agricultural chemicals, relies on brick-lined floors and tanks in the
production of sulfuric and phosphoric acids. Chemical storage and
waste treatment facilities require protection from aggressive
chemicals and waste byproducts.
The pharmaceutical, food, and beverage industries are plagued by
corrosion from chemicals and food acids, as well as corrosion from
acid and alkaline
Page 487
cleaning and sanitizing chemicals. The Food and Drug Administration
and U.S. Department of Agriculture will be unrelenting in upholding
sanitation mandates to ensure the health and welfare of the population.
The health care industry is under severe cost containment pressure.
The best cost containment in the food, beverage, and pharmaceutical
industries is their continued high standards of excellence in
maintaining standards of cleanliness. Maintaining these standards is
not without cost from corrosion to concrete and steel.
Chemical-resistant mortars, grouts, monolithic surfacings, and
polymer concrete are proven solutions to a host of these types of
corrosion problems.
Page 489
18
Glass Linings
Donald H. De Clerck
Rush, New York
I
Introduction
The selection of a material for corrosion service should be predicated
on the accurate determination of the relationship between the two
main selection factors, performance vs. cost. As costing data are,
unfortunately, greatly complexed by several other interrelated
subselection factors, e.g., design, availability, fabrication, it is difficult
to define realistic/usable guidelines. This is best done by direct
contact with a reputable manufacturer once the performance
suitability has been given the initial screen approval.
Consequently, this chapter will deal primarily with the performance
factors associated with the rather unique class of multimaterial or
composite materials called ''glass linings." When evaluating the
performance characteristics of any material, it is often instructive to
first ask: Why was the material originally developed? How is it made?
The answers often provide useful information as regards the reasoning
behind certain operating limitations along with the rather specific
maintenance requirements for the equipment.
Page 490
II
Why Glass Linings?
Glass on metal composite systems is far from being a recent
development. Archaeological discoveries have shown that glass
bonded to gold jewelry (enameling) dates back to at least 400 B.C. and
possibly before. In this case, the glass coating functioned merely as a
then very unique and therefore rare aesthetic complement to the gold.
Enameling continued on primarily as an art form until the early 1800s
when cast iron sanitary ware was first coated. Here the glass coating
not only provided a certain amount of corrosion resistivity, which
effectively maintained the clean-looking aesthetics of the piece,
especially when compared to the rust-prone, dark-colored casting, but
also provided an ease of sanitary cleaning that was unknown up to that
time. Although some enameled castings were used for containment of
mildly corrosive chemicals, the very nature of the casting process
drastically limited the available design to simple forms, mostly of the
open-top, nonjacketed varieties. Carbonaceous and sulfurous
outgassing from the castings also made these coating very prone to
integrity-compromising pinholes.
The development of glass linings on steel, as we generally know them
today, had its real beginning as an outgrowth of an invention
associated with the brewing industry. In the early 1880s it was
determined that application of a vacuum to the beer fermentation
process not only accelerated the fermentation time but, more
importantly, drastically improved the consistent quality of the final
brew. Up to that time, the major brewing problem was the tremendous
variation in the taste of the beer from batch to batch. The application
of vacuum effectively overcame this serious problem but presented
another one, i.e., what to use for the containment material. A large
vessel size is closely tied to material tensile strength, easy fabrication,
sanitary maintenance, suitable corrosion resistance, vacuum operation,
and low costall important characteristics. At the time, three materials
and one composite were available. Each was considered from the
positive (+) and negative (-) performance/cost standpoints, as shown
in the following table.
Comparison of 1880s Containment Materials
WoodSteelGlass Glassed cast
iron
Large size
+
+
Material strength
+
+
Design/fabrication
+
Corrosion resistivity,
+
+
santitariness
Vacuum operation
+
+
+
(table continued on next page)
Page 491
(table continued from previous page)
Wood Steel Glass
Glassed cast iron
Fragility
+
+
+/Cost
+
+
-a
aIn the 1880s, glass articles were labor intensive and,
therefore, relatively expensive because they were
individually blown. The glass itself, however, was
inexpensive.
It was evident that the composite approach in which the positive
characteristics of one material is balanced against the negative
characteristics of the other would be the best approach. If the cast iron
could be replaced by the lowcost, high-strength, easily fabricated but
corrosion-prone/unsanitary steel and then coated with the inexpensive,
corrosion-resistant, sanitary but fragile glass, a suitable composite
could be formed. Thus, the glassed steel composite was born along
with the time-honored, although strengthwise exaggerated, definition
of "the strength of steel combined with the corrosion resistance of
glass."
There were relatively few noteworthy developments in the glassed
steel composite from the 1880s up to the start of World War II. At the
war's start, however, the need for critical chemicals, most often of a
highly corrosive nature and frequently sticky, increased dramatically
along with the need for suitable equipment to process them.
Considerable research effort in the areas of metallurgy, fabrication,
especially the very troublesome welding procedures and glass
composition resulted in the development of glass-steel composites that
matched the required wartime needs. It was during these times that the
importance of characteristics other than corrosion resistivity of the
composite, e.g., thermal loading and mechanical stressing, was
recognized. It was also during these times that the first attempts were
made to define the actual use limits of the composite along with
establishing acceptable operational/maintenance practices including
proper gasketing and equipment repair. Since that time, research
efforts have progressed to the point where today glassed steel
represents one of the preeminent materials choices for equipment used
in the chemical process industry (CPI).
III
How Is Glassed Steel Made?
The processing of glassed steel equipment involve several interrelated
steps:
Customer/vendor Communication
Metal selection
Metal fabrication
Glass selection
Glass making
Application of glass to metal
Page 492
Glass fusion firing
Final assembly
Note: Inspections, especially from the standpoint of quality, are an
integral part of all the nonselection steps and will be covered where
appropriate.
A
Customer/Vendor Communication
As there are several composite systems currently available from most
vendors, it is imperative that, at the inquiry stage when any new
chemical campaign/recipe is to be made, a technical dialogue be
established between vendor and customer to ensure that the correct
system has been selected in order to meet all of the anticipated
operational requirements. The high processing temperatures required
to fuse the glass to the metal coupled with the inherent differences
between the glass and metal also necessitates special design
considerations compared to the usually more familiar alloy
fabrication. These design restrictions, especially in regard to metal
thicknesses, jacket configuration, and geometry, may differ quite
widely from the initial customer desires. Based on price, delivery, and
ease of replacement, it is highly recommended to stay with the
manufacturer's standard designs. Vendor literature may be most
helpful in providing general guideline information.
B
Metal Selection
There are several subcriteria that must be met before a metal can be
selected for glassing. Four of the most important are as follows:
Code acceptability. The vast majority of glassed steel vessels
manufactured, repaired, or reconditioned today conform to the criteria
set forth in the latest edition of the ASME Boiler and Pressure Code,
Section VIII, Division I. This Code regulates the type of metal,
thicknesses, joining techniques, designs, and so forth that can be used
for the myriad of pressure/temperature requirements of the CPI. The
exact pressure/temperature limits for each piece of equipment is
clearly spelled out on the Code nameplate affixed to the equipment. It
should be noted that the Code has no jurisdiction over the glass lining
itself. Glassing parameter variation is governed by the quality
standards of the manufacturer coupled with the specific end
requirements and quality specifications of the consumer.
Thermal expansion. The main factor contributing to the mechanical
strength of the glassed steel composite is the residual strain caused by
the differential thermal expansions between the steel and glass. The
fragility of most non-tempered, pure, standalone glass systems may be
traced to their low tolerance level to the application of tensile stresses.
Glass is extremely weak in tension; extremely strong in compression.
By judicially choosing the correct difference between the metal and
glass expansions, a large degree of residual compressive strain may be
readily built into the fragile glass, thereby drastically improving its
Page 493
Figure 1
Thermal expansion relationship between glass and steel.
overall strength. As most metals have greater thermal expansions than
glasses, it becomes only a matter of degree as to the amount of the
residual compressive strain that can be achieved. Too little strain
results in a weak composite; too much in a system that is very
sensitive to shear fracture, especially on critical radii-type geometries.
Figure 1 may be helpful in showing this expansion relationship.
Figure 1 (left) shows separate bars of steel and glass at some
temperature, above which the glass, while rigid, is incapable of
transferring or accepting any stress, e.g., 900°F. As the individual bars
gradually cool to room temperature, the metal bar, having the greater
expansion/contraction characteristics, contracts to the greater degree
(Fig. 1, middle). The difference between the heights of the two bars is
indicative of the thermal expansion difference and is directly related
to the degree of residual compressive strain that can be built into the
actual glassed steel composite system as depicted in Fig. 1 (right). As
will be shown later, when the composite is stressed beyond this strain
limit, glass damage will most often occur.
Transformation temperature. Depending on composition, some metals
can go through a crystal structure transformation that may involve a
prohibitive change in volume expansion. If this occurs below the
temperature where the glass becomes rigid, glass cracking may occur.
Higher carbon, manganese, and chromium steels are prone to this
problem. Due to the greater ease for structural change in the molten
state, weld rod selection and the actual welding itself are especially
critical to the final glass lining performance.
Outgassing. A pinhole-free glass coating is a prerequisite for most
campaigns carried out in the CPI. Metals that contain prohibitive
amounts of outgassing constituents, e.g., carbon and sulfur, are
unacceptable. This is a difficult
Page 494
requirement from the carbon standpoint as carbon contributes directly
to the strength characteristics of the steel. The steels acceptable from
the glassing standpoint are termed "low-carbon steels" and usually
contain carbon in amounts less than 0.20 wt %. Proper storage of weld
rod materials is also a Code requirement that helps prevent hydrogen
outgassing in the welds.
Based on the above criteria, there are several Code-acceptable steels
that can be used for glassing. In the United States, two types are
commonly used: ASME SA-285, a coarse-grained steel and ASME
SA-516, a finer grain material. These two steels have an upper Code
temperature rating of 650°F; a lower of -20°F. Note that Code does
make an interpretable provision for some lower temperature operation
with these steels. A recommended conservative interpretation can
extend the SA-285 to -40°F; the SA-516 to -75°F, provided that the
original design pressure is reduced by a factor of 0.4. For operation
below these temperatures, a different substrate metal should be used,
e.g., 316 stainless steel, in conjunction with a compatible glass
coating. Recently, the Code gave approval to an excellent glassing
steel in which the carbon is stabilized with titanium. This steel has
drastically improved the glassing quality of the previously difficult-tocoat convex radii, e.g., nozzles, agitators, baffles, dip pipes. For
piping-type fittings, castings are the most economical choice. Cast
iron, with its low strength and prohibitive outgassing characteristics
mentioned previously, has largely been replaced by cast steel.
C
Metal Fabrication
In fabricating a vessel (Fig. 2), flat steel plates are first trimmed to
size and beveled for welding. The plate is then rolled to form a
cylinder, called "the shell," and the long seam welded. The top and
bottom parts of the vessel, termed "top and bottom heads" are usually
formed at the steel mill. It is extremely important to maintain the same
relative thicknesses between head and shell in order to minimize
sectional heat-up and cool-down rates that may cause stressing
problems during the glassing cycles. The forming of the head nozzles
termed "swaging'' is a procedure unique to glassed steel manufacture.
As the aforementioned residual compressive strain is critically
proportional to the radius of curvature, all nozzle radii must be made
as generous as possible. To accomplish this, the nozzle locations are
first determined and pilot holes drilled at the centerpoints. The area
around the hole is then heated to red heat and an appropriately sized
pin forced into the hole from the inside of the head. This flares out the
metal to the correct geometry. After swaging, the nozzles are welded
on, the heads trimmed and beveled, fitted up to the vessel, and
circumferentially welded. This is essentially the manufacturing
procedure for a single-shell vessel, e.g., storage type, prior to glassing.
If the vessel is to be jacketed, additional steps are required. There are
currently two main types of jacket design:
Page 495
Figure 2
Schematic of typical jacketed glassed steel vessel.
Conventional, in which an outer shell extending from just below the
top head weld line to the bottom outlet nozzle, is welded to the main
vessel. A dual set of jacket nozzles, located 180° apart, allows for
entrance and exit of heat transfer fluids. Most jackets and internal
shells are Code rated from approximately 100 psig to full vacuum
operation. This design is by far the most economical and the one used
by the vast majority of companies in the CPI. The economics are
traceable to a non-labor-intensive fabrication coupled with relatively
short furnace cycles for the glass fusion procedure. The requirement
for high temperatures to effect the fusion, i.e., approximately 1600°F,
makes furnace time extremely expensive. Shorter times can be
realized only if the vessel can be fired without the jacket attached
thereby preventing uneven heating/cooling-or shielding-type effects.
This can be accomplished by welding a quarter round pipe, called the
"sealer ring," around the vessel circumference near the top head to
shell weld line and a ring, called the diaphragm collar, around the
bottom outlet nozzle (see Fig. 2). These are so designed to allow for
the jacket to be safely welded on, after glassing, at a point far enough
removed from the thermally sensitive glass coating.
Half pipe, in which a "half pipe" is welded around the inner shell in
Page 496
serpentine fashion. The piping can be zoned to allow for sectional
heating/cooling or to prevent cross-contamination of heat transfer
fluids, e.g., one coil dedicated to steam alternating with another
dedicated to brine. Higher pressures, e.g., to 450 psig, are readily
attainable. The potentially higher velocity, directional flow of this
design enhances heat transfer, especially on cooling. However, the
need for the added labor-intensive welding coupled with the
required longer slow-heat/slow-cool firing cycles makes this design
more costly, and therefore far less popular, than the conventional
design.
During these fabrication steps, there are a number of Code-required
checkpoints at which compliance is assured by a Code representative
(not employed by the manufacturer). Special attention is paid to mill
records, dimensional fitup, and welding. All welds are checked by
both dye penetrant and ultrasonics. Vessels for lethal service must
also have all welds X-rayed. The manufacturer also has a quality
assurance checklist that is often stricter than that used by the Code.
This is needed in order to uncover any defects prior to the glassing
operation. Defect removal and correction can be extremely costly
during the glassing cycles and therefore must be minimized. Once the
vessel has been fabricated and inspected, it next is grit-blasted to
white metal. Aluminum oxide is frequently used as the blast media
and provides a clean, roughened surface to which the glass will
eventually be bonded. The blast step is closely timed with the
subsequent glassing cycle startup to reduce possible surface
contamination, e.g., rusting.
D
Glass Selection
While the metal substrate provides the needed strength characteristics
and base thermal expansion for the composite, the majority of end-use
requirements must be controlled via adjustments of the glass
composition.
Requirements include exact differential thermal expansion match;
adhesion of glass to metal; resistance to corrosion-, thermal-,
mechanical-, electrical-type stressing influences; and reduced product
adherence. Most glass systems may be viewed as a three-dimensional
network-type structure comprising one or more oxide groups.
Network formers. These acidic-type oxides form the backbone for the
glass structure. The elements that make up this group are usually
coordinated to four oxygen atoms, allowing them to readily form
chain and/or network glass structures. A common analogy is
frequently made to a "house of cards" where the network formers are
the cards. Silicon dioxide (SiO2) is the premier network former and is
obtained from relatively inexpensive beach sand. It is usually present
in glasses in amounts exceeding 50 wt %. By itself, it is a low thermal
expansion, high-melting, high-viscosity, extremely corrosion-resistant
glass. The commercial pure glass system is termed "fused silica."
Another useful network former is boron
Page 497
oxide (B2O3), which is usually added to an SiO2 glass in amounts less
than 15 wt %. This material maintains the low thermal expansion and
excellent corrosion resistance of the SiO2 system but drastically
lowers the melt temperature allowing the resulting glass to be readily
formed. This is the basis for the excellent corrosion-resistant,
thermally stable borosilicate laboratory glassware.
Network modifiers. These base-type oxides do not enter into the
network forming structure, i.e., they cannot form glasses by
themselves but reside interstitially in the holes formed by the network
"cards." As the name suggests, these oxides modify the properties of
the network glass. The alkali oxides (K2O, Na2O) drastically reduce
the glass melt temperature, increasing thermal expansion but
detracting from the overall corrosion resistivity. The alkali earth
oxides (BaO, CaO, MgO) generally improve corrosion resistivity. It is
the correct combination of these two groups that results in the most
common of all glass systems, the soda-lime-silicate or window glass
system. Here the difficult-to-form high-silica glass is fluxed with
approximately 15 wt % soda (Na2O) to form the low-melting but
corrosion-susceptible "water glass." Corrosion resistivity is added
back through the addition of approximately 10 wt % lime (CaO).
Several of the transition element oxides such as cobalt (which gives
the dark blue color to many glasses), manganese, and nickel, are
extremely important to glass steel composites as they assist greatly in
promoting metal to glass adhesion. Zinc oxide (ZnO) is another useful
modifier oxide that enhances water corrosion resistivity.
Intermediates. These materials are amphoteric in chemical response
and can act as either network formers or modifiers depending on
concentration and the nature and amounts of the other constituents.
Aluminum oxide (Al2O3) and titanium dioxide (TiO2) both enhance
general corrosion resistivity; zirconium dioxide (ZrO2) enhances
alkali resistance.
The cover coat systems currently used for glass linings are complex
arrays of up to 15 oxides taken from the above three groups and built
around the framework of the silica network (> 60 wt %). Oxide
concentrations are frequently optimized against property improvement
via computer statistical programs to balance the very specific glassing
requirements of the manufacturer v. the demanding requirements of
the CPI. In addition to the compositional approach to property
alteration, improvements can also be realized through metal substrate
change, glass thickness variation, application techniques, cooling
rates, layering, secondary phase addition, e.g., porosity, crystal
addition/ precipitation. Some of the glassed steel systems currently
available are as follows:
1. Standard. This most economic system is optimized for the greatest
use versatility by the CPI. It is the glass usually specified and
represents a good balance between chemical and physical property
serviceability. Glass thickness can be further optimized for specific
requirements, e.g.,
Page 498
thin for better heat transfer, thicker for severe corrosion situations.
These systems are generally recommended for the temperature range
of -20 to +450°F. Thermal loading differentials, i.e., the maximum
allowable cold-to-hot or hot-to-cold change, are in the range of 260°F
(note that no safety factor is included). The most common colors are
blue and white (note that color variation does not reduce any of the
chemical or physical properties of the glass). The white coloration is
especially useful for ensuring the completeness of a dark product
cleanout. White-on-blue or blue-on-white calibration markings are
available.
2. Low-temperature service. Here the use of a stainless steel substrate
in conjunction with a special glass lining can extend the lowtemperature operation down to -200°F. However, as is frequently the
case when one glass property is improved, others may be reduced. For
this case, some reduction in corrosion resistivity and thermal
differential loading capabilities results.
3. High-temperature service. These systems can extend the use range
up to 650°F, the upper Code limit for the low-carbon steels usually
specified. In addition, the thermal loading differentials can be
increased to 360°F with no reduction in corrosion resistivities. As is
true for most systems, costing is increased over the standard.
4. Low product adherence. These systems employ a special surface
layer that is largely free of many microscopic surface imperfections
that tend to enhance mechanical or keyhole-type product adherence.
Other properties remain unchanged. It is important to note that the
surface remains effective only until the onset of a corrosive etch.
5. Glass/crystal. These glass/crystal linings were commercially
developed in the late 1950s. The addition of crystals to the glass again
shows the benefits associated with the composite approach to material
formulation. The analogy with reinforced concrete may be instructive.
Concrete, like glass, is extremely weak in tension but extremely
strong in compression. By inserting metal bars in the concrete, any
applied tension-type stresses can then be effectively transferred over
to the strong, ductile metals bars, thereby improving significantly the
mechanical strength of the concrete. For the glass/crystal system, the
metal bars are replaced with crystals that are considerably stronger
than the glass. There are two common approaches to getting the
crystals into the glass structure both of which, depending on many
variables, can improve impact resistance (up to 2×), abrasion
resistance (up to 4×), and heat transfer (up to 39%). Some small
improvements in thermal loading differentials, i.e., approximately
30°F, were also noted origi-
Page 499
nally (more recent developments in the standard glassed steel
composites have essentially negated this initial advantage):
a. Direct addition of relatively glass-insoluble particles or fibers,
e.g., quartz (SiO2), to the glass system prior to application and
fusion to the metal. This approach allows for good control over the
particle type, size, and shape. The desired homogenous distribution
of the crystalline phase(s) throughout the glassy phase is, however,
not easily controlled as it is dependent on the size of the base glass
particles prior to being fused. Colors are usually dark blue or white.
b. In-situ precipitation of the crystals from the glass. Over the
years, this approach has usually given consistently slightly higher
impact results and is currently more widely used. These systems
require special compositions that allow nucleation and crystal
growth to take place within the glassy matrix. As the nucleation
step is highly dependent on direct radiant energy fusion, these
systems are used primarily for accessory items. If vessels are to be
coated, a costly extra heat-treat furnace cycle is required. Colors are
usually light blue or white.
Some additional comments are needed to further categorize these
unique crystal-glass systems. First, while a twofold improvement in
impact resistance may initially sound impressive, i.e., from 9 in.-lb to
18, this does not represent a great deal of significance as regards the
types of high-stress scenarios frequently found in the CPI. However,
these systems exhibit another characteristic that is truly outstanding,
i.e., their ability to drastically retard crack propagation. Once a crack
has formed in a pure glass system, like a car windshield, the crack will
continue to propagate, especially under the continuing influence of
motion stressing, temperature change, and water. While the metal
backing itself drastically retards this effect in the glassed steel
composite, the added crystals act as great supplementary "crack
stoppers." The end result is that damage can be drastically reduced to
a size whereby a repair will have some long-term effectiveness. As
convex geometries are much more prone to damage from mechanical
stressing, these systems find special purpose on accessory items, e.g.,
agitators, baffles, dip pipes.
Also, the addition of the crystals to the glass results in the formation
of interphase boundaries. These, like grain boundaries of metals, can
and frequently do lead to preferential corrosion reactivity. The result
is a more rapid loss of nonstick fire-polished surface compared to the
standard glass system. To counter-balance this effect, some
manufacturers put a top coat of the standard glass over the
glass/crystal layers. This layer detract little from the overall crackstopping ability of the subcoats but drastically improves the fire polish
retention properties of the surface.
Page 500
E
Glass Making
Once the glass composition has been finalized in the laboratory, the
raw materials, usually in the form of oxides or carbonates, are
carefully weighted out and thoroughly dry-mixed. Batch size is
usually based on a final glass weight of 500 lb. The mixed materials
are then charged to a rotary gas-fired smelter where it is fused into a
glass at temperatures of 25003000°F over soak periods of up to 4 h.
The rotary swing of the smelter coupled with the gas evolution from
both the carbonates and the combined and uncombined water helps
ensure glass homogeneity. The molten glass is then poured into cold
water causing it to harden and shatter into small friable globules
called "frit." The frit is then dried prior to further processing into one
of two possible application materials:
Dust. The frit is charged to a ball mill to form a powder.
Slip. The frit is charged to the ball mill along with water and
appropriate suspension agents to form a sprayable material with the
consistency of house paint.
During the various stages of the frit manufacture, a wide variety of
assurance checks are carried out to ascertain the overall quality of the
material, e.g., density, fusion button, X-ray fluorescence, thermal
gravimetric analysis, and slip consistency.
F
Application of Glass to Metal
Over the years, three application methods have evolved. Which
method to use was dependent on several interrelated factors, e.g.,
economics, design, thickness, requirements, and anticipated service.
Straight spray. As the name implies, the enamel slip is used directly in
spray equipment similar to that used to spray a house or car. Coating
thicknesses approximate 0.010 in. or 10 mils. After each application,
the coating must be thoroughly dried, subsequently fused to either the
metal substrate or the previous coat(s), and inspected prior to
application of the next coat. Total coating thickness is dependent on
the initial requirements established by customer and vendor. These
coatings can achieve good coating thickness control but tend to
sometimes highlight largely aesthetic, subsurface coating
discontinuities. The materials added to assist in suspension of the
glass particles also cause the aforementioned two-phase interface
corrosion effect. Consequently, they tend to lose fire polish faster that
those applied by the other methods. The microstructure is
characterized by a large number of relatively small bubbles (Fig. 3).
Spray dust. This method originally used the spray method
immediately followed by bagging on the dust. The dust readily
adheres to the wet spray layer and coating thicknesses can be
increased to 1520 mils. As mentioned previously, furnace time is
costly, so this thicker coat application method can be quite
Page 501
Figure 3
Schematic cross section of one ground coat/one cover spray-dust system.
economical. Coatings applied by this method exhibit greater thickness
variations than those applied by straight spray but do not show any
subsurface lining phenomena or fire polish problems. Newer
equipment allows for the spray and dust to be applied from the same
nozzle assembly, i.e., dual spray, spray floc. The microstructure of the
dust portion of the coating is characterized by relatively fewer but
larger sized bubbles compared to the straight spray method (Fig. 3).
Hot dust. In this very old but most economical method, the dust is
vibrationally dredged or sifted onto the preheated (approximately
1600°F) surface. The dust is continually applied until it no longer
melts and then the piece is returned to the furnace for reheating.
Depending on the heat retention of the piece, coating thicknesses of
2025 mils can be achieved before furnace recharging. However, the
piece is never returned to room temperature until the final coating
thickness has been realized, thereby greatly reducing energy
requirements and processing time. Large thickness variations and
difficulty in carrying out quality measurements during application are
major problem areas. This method also requires dredge accessibility,
so that it is limited to the smaller clamp top (open) vessels or equally
accessible pieces of equipment. The shift to larger, closed-top vessel
design coupled with the prohibitive economics of maintaining two
widely different application lines has led most manufacturers away
from this application method. Its main current use is in the coating of
castings, e.g., pump internals, valves, and fittings.
G
Glass Fusion/Firing
For both the straight spray and spray-dust methods, the coated piece
after thorough drying is charged into a high-temperature furnace to
fuse together the
Page 502
individual glass particles to form the high-integrity glass lining.
Temperatures vary from 1400°F to 1650°F depending on glass
composition. Firing times vary widely dependent on equipment size,
design, and mass. Both electric and gasfired furnaces are used
successfully. Each has its own counteradvantages to the others
disadvantages, i.e.:
Electric. Better zoned heat control, lower degree of convective
contamination, few combustion products
Gas. Faster heat-up with less metal-to-glass reaction time, more
sensitive defect detection atmosphere, relatively inexpensive
Competent manufacturers, through compositional adjustments and
innovative procedures, can routinely process high-quality vessels
using either type of firing method. In the United States where clean,
inexpensive gas is readily available, it is the logical choice; in Europe,
where the gas is of extremely poor quality, electric firing must be
used.
Each standard glass system is composed of two very specific glass
types (Fig. 3):
1. Ground coat (GC). This coating may be likened to a primer coat
used on a house or car and is usually applied by the straight spray
method. Two coats, each of approximately 10 mils, are usually
applied to vessels; one coat to the overall thinner-coated accessory
items. The mandatory addition of the color-inducing transition metals
to these compositions in order to promote adhesion results in colors
ranging from dark blue to gray to black, depending on the degree of
interaction with the substrate metal. The function of the ground coat is
threefold:
a. To assist in dissolving the iron oxides as they continually form at
the metal interface. Failure to do this will result in a non-adherent
system.
b. To provide the proper bubble structure. During the firing
operations, various gases are evolved, some of which, e.g.,
monatomic hydrogen (H0), are relatively soluble in the metal at
these high temperatures. As the temperature decreases, the
solubility is reduced and the gas tends to leave the metal and
accumulate at the metalglass interface. There, two atoms readily
combine to form molecular hydrogen (H2) with a large increase in
stereochemical volume. A proper bubble structure is required to act
as reservoirs for the gas, thereby preventing prohibitive pressure
buildup and the ensuing very serious glass fracturing known as
"fishscaling."
c. To provide the proper stress balancegradient bridge between the
high-thermal-expansion metal and lower expansion cover coat
glass.
In order for these three critical functions to be met, compositional
Page 503
adjustments must be made to these ground coat compositions that
detract greatly from their chemical resistivities as compared to the
cover coat systems. Failure to recognize this important fact can cause
serious operational problems that will be discussed later in the
"Inspection" section. From the quality standpoint, the ground coat
application steps are also extremely important. The glass-to-metal
fusion reaction is very sensitive to any type of surface or slightly
subsurface contamination. It is important to identify these problems at
this early, economically acceptable stage and to take the appropriate
corrective action, e.g., grind out the defect/contamination and recoat.
This will ensure a high-integrity final coating system.
2. Cover coat (CC). As the name implies, this coating "covers" over
the ground coat(s) and provides the good corrosion resistivity, low
product adherence, and adequate thermal/mechanical properties
needed for the many requirements of the CPI. Again, interapplication
inspections are critical and careful quality records are maintained by
most manufacturers. Three main checks are carrier out:
a. Visual. This check differentiates between aesthetic and
potentially service lifereducing defects. At or near the final coating
stage, these defects should be carefully reviewed by both vendor
and customer to determine the realistic need for possible corrective
actions, e.g., another coat, metal repair, complete recoat, v. the
possibility of various unfavorable consequences, e.g., prohibitive
vessel distortion, thicker, and poorer thermally conductive glass
system, longer processing time. It should be carefully noted that
more glass thickness is not always the best solution!
b. Thickness. Glass thickness is important from both positive and
negative standpoints:
Positive. For corrosive situations, greater glass thickness directly
equates to longer service life. Also, as dielectric breakdown
strength is also a function of thickness, resistance to electrical
breakdown will also increase with glass thickness.
Negative. Unfortunately, the strength-imparting residual
compressive strain goes down as thickness is increased. This may
be compounded by the aforementioned convex geometry problem.
Also, as the thermal conductivity of glass is very low, increased
thicknesses will result in reduced heat transfer.
Balancing out these effects, most manufacturers, with allowances for
statistical variation, usually keep thicknesses for corrosion-resistant
glasses within the range of 3590 mils. Deviations are, however,
allowed for in certain circumstances. For instance, if severe corrosive
attack is anticipated, mechanical/thermal loadings are minimal, and
Page 504
heat transfer is of little concern, glass thicknesses on concave
geometries may be maximized in accordance with good glassing
practice; for very low corrosion systems with high heat transfer and
good product release requirements, e.g., some polymerization
reactions, very thin glass would be the choice.
c. Voltage testing (\factory only). Customers usually require
quantitative assurance that the glass system is of sufficient integrity
to withstand the anticipated corrosive service. This assurance is
most readily provided by using a high-voltage tester that will
dielectrically break down thin glass. As most chemical glasses have
breakdown strengths of approximately 500 V/mil, the testing also
provides some indication of minimum glass thickness, i.e., 10 kV
would equate to a thickness of at least 20 mils. Most manufacturers
provide a test voltage to match the service, e.g.:
Final test
Voltage
Possible use
voltage (kV) designation
0
Visual Polymerizers
57
Low
Low corrosion, good heat
transfer, metal-free product
1012
Medium Storage vessels
1520
High High corrosion rate reactions
If an electrical contact is found in the early coating stages, it is a
relatively simple matter to grind out the defect and add another coat(s)
of glass. However, if this occurs near the total thickness limit for the
coating, two alternative approaches must be reviewed:
1. Grind out the defect area and install a metal repair plus. These
plugs are usually 5/83/4 in. diameter and most often made out of
highly corrosion-resistant tantalum metal which is used in conjunction
with a polytetrafluoroethylene (PTFE) sealing washer. There are
quality limits to the number of these repairs allowed. Vessels below
500 gallons are most often ''plug-free"; above, one plug per
approximate 1000 gallons. Improvements in total manufacturing
quality control in recent years have significantly reduced the need for
these factory repairs, especially below 3000 gallons. Quality figures
of over 90% plug-free are now quite common. Although sometimes
commanding a premium, the specification for plug-free equipment
should always be discussed with the vendor for severe services, e.g.,
bromine. Both plugged and plug-free equipment carry the same
general warranty against workmanship
Page 505
and/or material defects within a reasonable time, usually not
exceeding a year from the date of installation or 18 months from the
date of shipment.
2. Complete glass removal and recoating. This is a decision that must
be weighed very carefully. Removal of the glass is not only time
consuming from a delivery standpoint but the additional firings may
lead to Code-affecting distortion of the equipment. This is coupled
with the fact that there is no guarantee that the equipment will have a
need for fewer plugs after the second processing cycle. Also, on total
balance, the addition of repair plugs in slight excess of the usually
acceptable quality limit no longer equates to either reduced
operational conditions or service life for most campaigns found in the
CPI. New-type repairs, coupled with the associated installation
procedures and updated, more exacting methods of inspection should
allay past concerns.
H
Final Assembly
Several steps are involved in final assembly:
Fitup and welding of the jacket, if required.
Hydrotest of vessel and/or jacket according to ASME Code
requirements.
Assembly of accessory items, e.g., agitator, drive, baffle, covers,
gaskets. (Note: Avoid the option of field assembly if at all possible.
The manufacturer has the correct equipment to do an efficient, safe
job and can also proof-run in the entire system to assure its complete
functioning prior to shipping.)
Customer review prior to shipment. The vessel should be carefully
inspected, appropriate records reviewed, and proof runs witnessed.
Operation manuals pertinent to the equipment, especially as regards
installation, operation, and maintenance, should be completely
understood. (Note: For customers who are unfamiliar with glassed
steel equipment, it may be prudent to have the manufacturers service
representatives assist in the relatively complex installation and startup
steps.)
IV
Equipment Installation
Due to the great number of possible designs and the special
requirements associated with each, information on both installation
and hookup of glassed steel equipment is best left to the review and
careful implementation of the specific manufacturer's literature. The
main caution is to always use personnel who are completely familiar
with the manufacturer's stress-sensitive glassed steel equipment, e.g.,
do not assume that a rigger familiar with alloy equipment will install
glassed steel properly.
Page 506
V
Equipment Operation
In the operation of any CPI-type equipment, there is one theme-type
word along with the efforts required to implement it that most often
spells the difference between long-term, profitable service and shortterm failure: Prevention. It has been shown conclusively at many
customer plants that the vast majority of glassed steel equipment
failures can easily be prevented if, and only if, careful attention is paid
to some simple rules and guidelines. This section will review these in
some detail.
Rule #1: The Rule of the Do's. There are three simple "do's" that if
applied to any process equipment operation, i.e., not necessarily
glassed steel, will mean that you "don't" have to worry about longterm serviceability:
1. Use patience. The statement "haste makes waste" applies perfectly
to the operation of glassed steel equipment and should be emblazoned
on the minds of anyone having association with glassed steel
equipment.
2. Use common sense. If some proposed action runs counter to your
common sense feelings, postpone the action until more substantiating
data (see the #3 "do") are acquired.
3. Acquire knowledge regarding all facets of the equipment operation.
Vendor literature and seminars, pertinent technical articles, trade
journals (most are free), and technical societies all provide a wealth of
useful information. In this regard, it is extremely important to note
that the literature, recommendations, and, most importantly,
replacement parts form one vendor may not pertain to or work on
another vendor's equipment. It is also extremely important to think
every action through. Before any action is initiated, one must be able
to say "I know completely the consequences of this action." If this
cannot be said, then the necessary efforts must be expended, including
possible testing, to fulfill this requirement.
Over the years, the research groups of several vendors have
contributed greatly to the knowledge base associated with the correct
operation of their equipment. More specifically, during the late 1970s
and early 1980s, a considerable amount of both laboratory research
and collaborating field trials was carried out in an attempt to
accurately define the major causes of glassed steel damage. Additions
to the original list have continued to be made up to the present and it
is certain that more will be added in the future as recipes and
campaign parameters are continually altered. The results of these
studies are summarized in Fig. 4. (Note the "etc.," indicative of the
above-mentioned open-ended nature of damage analysis.) As
indicated, the damage can be divided into physical and chemical
causes. Contrary to a popular misconception, the physical causes far
outnumber the chemical in the ratio of 9:1. This misconception is
largely traceable
Page 507
Figure 4
Process causes for glassed steel damage.
to the widespread use of inadequate inspection procedures that do not
pinpoint the exact onset of damage. By the time of inspection,
corrosive activities have completely masked any indication of the
original damage, thereby leading to the corrosion cause
misinterpretation.
When discussing operational limits for glassed steel, perhaps the most
important word is located at the bottom of Fig. 4, i.e., "combinations."
It is unfortunate that most of the causes of damage to materials (not
only glassed steel) are additive. Consequently, one may be operating
well within the thermal loading guidelines supplied by the
manufacturer but the additional stress loading from, for example,
vibration or hookup torquing may lead to damage. This combinationtype problem makes it imperative that a complete review of the
process be made by personnel completely familiar with such
interactions, thereby ensuring that adequate safety factors are assigned
to the overall operation.
Another very important fact when discussing combination-type
effects, especially as they may pertain to damage analysis, is the
frequent tendency of glassed steel systems to exhibit delayed or latent
fracturing of the glass. In carrying out a damage analysis, an attempt
is usually made to relate an observed damage appearance or signature
to a specific operational cause. Unfortunately, glass, after being
stressed, may not show immediate visual damage that could easily be
related to the previous campaign. The stress relief mecha-
Page 508
nisms can take considerable time before a microscopic crack
propagates to the visual stage. It often takes the combined effect of
another stress factor(s) to make this happen, e.g., application of a
temperature differential after impact. This fact often makes damage
analysis very difficult, especially in multi-campaign situations.
Over the years, there have been several approaches used to define the
limitations of glassed steel. The most effective approach is to first
analyze each of the damage types as shown in Fig. 4 and then to
discuss the ways to avoid the cause(s) for the damage.
In first discussing the physical causes for glassed steel damage,
reference should be made to Fig. 1 (middle), where it was pointed out
that the strength of the glassed steel composite is traceable to the
differential thermal expansions between glass and steel. When
operational stresses balance out this difference, the glass goes into the
low-strength tension mode and fracture may occur. There are five
main factors that may alone or in combination contribute to this
possible reduction in residual compressive strain:
1. Geometry. As was mentioned in the fabrication section, the degree
of strength-inducing compression in the glass goes down as the radius
of curvature is reduced. In other words, convex radii are considerably
weaker than either flat or concave geometries. Based on both field
surveys and vessels returned for reconditioning, over 90% have at
least some problems associated with a convex geometry, e.g., nozzles,
swage areas, agitators, baffles. Rule #2: Keep all stress loadings on
convex geometries to the minimum. For example, this means not
standing on manway flanges, agitator blades, or bottom outlet swage
areas or using a torque wrench to tighten up on the sensitive nozzles,
i.e., those with a very small convex flange radius.
2. Glass thickness. The residual compressive strain goes down as glass
thickness is increased. That is the reason for the upper thickness limit
as discussed previously. Thickness also enters into a most critical
combination with geometry and provides reason why the glass is
usually thinner on convex geometries.
3. Temperature. As might be inferred by reference to Fig. 1,
increasing temperature reduces the compressive cushion.
Consequently, as operational temperatures are increased, the
allowable thermal loading differentials must be reduced accordingly.
4. Tensile stresses. As these counterbalance directly the compressive
strain in the glass portion of the composite, they should be avoided or
at least minimized to within allowable limits.
5. Electrical. As mentioned previously, the dielectric breakdown
strength for the glass is approximately 500 V/mil of glass thickness.
Page 509
VI
Mechanical Considerations
A
ImpactGeneral
Based on considerable laboratory study, the standard glassed steel on
a flat geometry with an approximate coating thickness of 55 mils will
fracture to 5 kV stabilized electrical contact when subjected to an
impacting energy of 9 in.-lb, i.e., a 1-lb steel ball dropped from a
height of 9 in. As expected, lower energies are required for fracturing
the more sensitive convex radii geometries. Impact-type failures may
result from the interaction of solids, liquids, or vapors either with the
glass surface or, in the case of solids, also from the metal backside.
Solids. The effect of solids is primarily related to the relation among
density, impacting area, mass, and velocity of the system. A lowdensity material, even though it has a relatively large total mass and
therefore a larger impacting area, may have little effect, whereas a
denser particle of smaller total mass and impacting area may cause
considerable damage. It is often the change in momentum (mass ×
velocity) that causes the problem, i.e., a moving agitator hitting a
stationary particle or a moving particle, e.g., a bolt, hitting the
stationary bottom head or sidewall of the vessel. Initial damage is
frequently found on the leading edges of the agitator blades (high
velocity) and directly opposite the agitator on the baffle and sidewall.
For higher density materials, the damage is usually found more
toward the bottom of the vessel. Other common causes for this type of
damage are the use of nonresilient scrapers for product removal or
poor sampling equipment/procedures.
Liquids. Excepting erosion-related problems, the impacting problem
associated with liquids is most often related to the use of high-
pressure fluids to clean the glass surface from product buildup, the socalled liquid blast or jet cleaning. The preferred method for product
removal is to use a suitable solvent that will not etch the glass, is
nontoxic, nonexplosive, etc. If no solvent can be found, then the use
of high-pressure fluids is acceptable provided that Rule #3 is followed
closely. This rule has several component parts:
Use the lowest pressure needed to remove the product, i.e., start at
low pressure and work up. In no case should the nozzle tip pressure
exceed 2000 psig.
Keep the nozzle tip to glass distance greater than 12 in.
Do not leave the nozzle tip at a set point, i.e., continually move the
nozzle over the surface.
For pressures in excess of 1000 psig, avoid directing the stream
toward the sensitive convex radii, e.g., agitator blades.
Avoid impinging on repair or fracture areas.
Filter all fluid streams, especially if the stream is recycled.
Protect the manway with an appropriate resilient liner.
Page 510
Follow all safety procedures.
Vapors. The fact that vapors or, more correctly, the collapse of vapor
bubbles, aka cavitation, can cause damage to glass has been one of the
more recent additions to the damage list. Over many years prior to this
revelation, there was a considerable number of damage situations that
could not be explained using the then-available knowledge base.
However, the gradual realization of the large energy releases
associated with the implosion of vapor bubbles led to several critical
research studies that clearly identified cavitation as the cause for
damage. Calculations indicated that, depending on the hydrostatic
head, a steam vapor bubble upon collapse could generate over 100 in.lb of energy (remember that glass breaks at 9 in.-lb). The actual
bubble collapse can be effected via three mechanisms:
Condensation, e.g., steam sparging into cold water; a vapor bubble(s)
formed by an exothermic addition of a chemical and its collapse at a
cooled vessel wall. As the energy of the bubble collapse is directly
related to the bubble diameter, a reduction in bubble size is highly
beneficial. Consequently, the use of small-holed spargers directed
away from the vessel sidewall and loading rates below the steaming
level can completely eliminate the problem. For condensable vapor
additions to a reactor, the coaddition of an inert, noncondensible
vapor, e.g., nitrogen, will effectively prohibit total bubble collapse.
Pressure buildup, e.g., high-pressure-leading blade surface v. the low
pressure on the blade backside. If low-boiling, high-vapor-pressure
components are present, vapor bubbles can readily form at the
backside and collapse at the front. Higher-pressure vessel operation
and agitator speed reduction are possible corrective solutions.
Chemical reaction, e.g., ammonia in an acidic solution. Rate of
addition and agitation shear control can help reduce the problem.
B
ImpactAppearance
Despite the impact source, i.e., solid, liquid, or vapor, all impact
damage possess the same distinctive characteristics:
The epicenter, where the initial impact force was focused.
The surface fracture outline is scalloped often resembling a clover leaf
pattern and frequently quite symmetrical about the epicenter.
The glass fractures off in a series of plateaus that step down toward
the epicenter. This is the most distinguishing characteristic of impact
damage.
Electrical contact at the recommended 5-kV stabilized voltage is
obviously dependent on the magnitude of the impacting force.
If the impact is from the metal side, e.g., some heavy object hitting
the top
Page 511
head, the plateaus are reduced in size and the area more funnelshaped. For this case, there is always electrical contact.
C
StressGeneral
There are a great number of tension-type stresses that can offset the
strength-imparting residual compressive strain of the glassed steel
composite. Among the most common are pressure, vibration, and
nozzle loadings.
Pressure. Overpressure is a serious problem from the standpoints of
both safety and potential glass damage. For a new vessel, pressures in
the range of 1.5 to 2 times the Code-rated pressure will usually cause
glass damage. For older vessels, where corrosion may have reduced
the metal thickness, lower values may apply. Obviously, the Code
rating should never be exceeded and, in fact, most users operate at
1015% lower to provide for a cushion-type safety factor. A common
cause for overpressure is a clogged pressure relief nozzle combined
with a pressure-generating reaction. Periodic inspections and possibly
a larger vent size can usually overcome the problem. Vessel
overpressure will initially show damage in the top head area; the
jacket, in the middle of the sidewall.
Poor storage procedures can also cause pressure, inducing ice to form
in agitators, baffles, and jackets giving the same identical damage
appearance/signature as if they were overpressured. Inverting or
plugging accessories and opening the appropriate drainage nozzles on
vessels in storage will solve the problem. If at all possible, store
equipment in dry, nonfreezing areas. It is also important to vent
liquid-filled thermocouple baffles to avoid pressure buildup during
reaction heat-up.
Vibration. The problems with vibration usually result from
steam/water hammer, improper baffle positioning, misalignment of
the agitator, and malfunctioning drives/pumps. The solutions are fairly
straightforward: use a professional mixing "T" for hot water from
steam makeup; blow down the vessel prior to admitting steam; all
directional baffles, e.g., "h," "d," fin, should be in line with the
agitator shaft; consult the vendor literature for agitator misalignment
and problems with either the drive assembly or the pumps. Expansion
joints/bellows should interrupt all stressing influences prior to
accessing the glassed steel equipment. It should be noted that
vibration by itself usually will not cause glass damage. The problem is
largely combinational with other loading stresses.
Nozzle loading. Nozzles are without doubt the most trouble-prone part
of glassed steel equipment. This is traceable to the aforementioned
sensitive convex radii associated with the swage and, more
importantly, the flange area. This makes the loading limits extremely
critical. There are two important factors associated with nozzle
loading, i.e., the axial and/or bending momenttype stresses associated
with the appendages attached to the nozzle and the actual gasketing of
the nozzle:
1. Axial and bending moment loading limits. These must be checked
with
Page 512
the specific manufacturer, i.e., equipment made outside the United
States may conform to different coding standards e.g., the European
DIN. Approximate values for the axial and bending moment limits are
100 lb/in. and 200 in.-lb/in. of nozzle diameter, respectively. Rule #4:
Keep all nozzle loadings to the minimum. For complex hookups, it is
recommended that a finite element analysis be done to ensure
minimum nozzle loadings. The use of hangers and expansion joints is
mandatory, with the latter placed as close to the glassed steel
equipment as possible. Special attention should be placed on the very
sensitive bottom outlet nozzle (BON), i.e., if only one expansion joint
is available, use it there!
2. Gaskets. If one were to set up a focal point for a preventive
maintenance program for glassed steel equipment, it would have to
start with gaskets. This is based on two balancing facts: (a) Improper
gasket selection and installation represent by far the leading cause for
glassed steel damage. A most common damage sequence is as
follows: Incorrect gasket selection and/or installation; Flange area
leakage; Retightening to stop leakage without use of a torque wrench;
Flange damage due to over-stressing; Costly repair, including
downtime and possible reglass. (b) The corrective solutions are simple
and relatively inexpensive.
Gasket selection is somewhat straightforward as regards the
recommended general service type. Most requirements are met by the
CRT-type gasket, for which the letter designations are as follows:
C = compressible. These two layers, usually 1/8 in. thick and placed
on either side of the reinforcing ring (R), are made from either aramid
or graphite fibers. These materials should have a compression ratio of
40% with a recovery of 20%. The compressibility of the insert is
needed to take up any unevenness (both tilt and waviness) in the
flange surface, i.e., even though the surface is shiny, it is usually far
from flat. The recovery value is important from the standpoint of
gasket reuse.
R = reinforcing ring. This is usually made from 304 stainless steel but
other materials are available to meet specific needs. The corrosion
problems with regular carbon steels make them unacceptable. The
rings contain several carefully engineered, i.e., height, width, and
number, concentric corrugations that both assist in recovery and more
importantly help prevent gasket blowout or vacuum pull-in.
T = PTFE envelope. The most common design is the slit or V type
with the bottom of the V facing the product side. This completely
shields the compressible layers and reinforcing ring from the
corrosive environment within the vessel/nozzle. The wide corrosion
resistivity of PTFE makes it a suitable choice for most situations
found in the CPI. For specific
Page 513
problems, e.g., polymerizations or high temperature, other envelope
materials are available, e.g., aluminum, stainless steel, tantalum.
There are also several other envelope designs available to suit
specific needs, e.g., double envelope, which completely prevents
environmental access to the compressible layer/ring sandwich =
good for exterior washdowns or external corrosive atmospheres;
channel, with a square cut that minimizes product buildup at the
flange joint. Note that all tape, stickers, and the like must always be
removed from both the envelope and inserts prior to installation.
Rule #5: When connecting anything to a glassed steel flange, a
properly specified CRT-type gasket must always be used.
When choosing a gasket supplier, always go with the specific
recommendations of the manufacturer of the equipment. If suppliers
other than those recommended are chosen, make sure that their
specifications are at least as good as those recommended by the
manufacturer. Always weigh the merits of a cheap gasket purchase vs.
the huge expense associated with either nozzle repair or possible
equipment replacement.
Of equal importance to the correct gasket selection is the use of proper
installation procedures. These include shimming, positioning,
torquing, and the use of the correct number of bolts/clamps.
Shimming. The compressible characteristics of the CRT gasket will be
able to take up to 1/16 in. of flange unevenness for most pressure and
temperature processes. Note: The allowable gap limit decreases with
increase in pressure/temperature; consult the manufacturer's
recommendations. Any gap in excess of these values must be properly
shimmed using the same type of insert material as present in the
gasket proper. On new vessels, shimming is seldom necessary for
nozzle diameters less than 12 in. However, on reglassed vessels,
which may have increased heat distortion, all nozzles should be
checked. This is also true for column sections and pipe of all sizes. Of
special importance is the shimming of the main cover assembly on
vessels of approximately 2000 gallons or less. Here all shimming
must be done on the main flange rather than the agitator access nozzle
which is very susceptible to misalignment problems. Again, the
specific manufacturer's literature should be consulted for details on
the proper procedures to follow. For new vessels assembled at the
manufacturer's plant, the shimming records should be reviewed prior
to delivery and the information filed for possible future reference.
Positioning. This is important from two interrelated standpoints:
closure and reuse.
Closure. In order to minimize leakage, the gasket must be positioned
to completely cover the glassed flange area without any excessive
interior or external protrusion. If this cannot be done, then the gasket
is of
Page 514
incorrect size and must be replaced. The use of tape to hold the
gasket in place is acceptable provided it is removed prior to final
assembly.
Reuse. There is no reason why a properly specified gasket cannot be
reused. How many times is a function of the operational and
reinstallation parameters. The visual criterion is the appearance of the
reinforcing ring corrugations showing through the PTFE envelope.
For gasket reuse, it is mandatory that the gasket be placed in the exact
same position that it was in previously. The gasket, upon original
installation, took on a compressible set corresponding to the
irregularities of the glass surface. If the gasket is flipped or
misaligned, leakage will occur. Several gasket suppliers provide tab
extensions on the reinforcing ring of manway gaskets that can be
turned over the rim of the cover, thereby ensuring exact repositioning.
This design is well worth the extra cost.
Torquing. All nozzle gasket damage is ultimately related to the
stressing associated with effecting leak free joint closure. To ensure
that this does not occur, a rule and a very important tightening
requirement must be satisfied:
Rule #6: Always use a calibrated torque wrench when working on
glassed steel equipment.
Use the proper tightening procedure. There are two acceptable
procedures both of which start after the original bolt/clamp tightening
has been made by hand:
Use the opposite/alternate procedure, i.e., 12 o'clock to 6; 3 to 9;
etc., going up in 15 ft-lb increments until the recommended value is
reached.
Use the progressive procedure in which consecutive bolts/clamps
are tightened, again in 15 ft-lb increments.
It is important to note that all bolts/clamps must be retightened once
either after one batch has been run or 24 hours has passed, whichever
occurs first.
Correct number of bolts/clamps. In order to correctly balance out all
of the point loadings associated with the individual bolts/clamps, it is
mandatory that all of the bolts/clamps recommended by the
manufacturer be used. Failure to do so will result in leakage, uneven
stress distribution, and eventual glass damage. Also be aware that all
bolts and clamps conform to Code requirements. If replaced, they
must meet the original requirements. Under no circumstances should
the bolts/ clamps of one manufacturer be used on another
manufacturer's equipment.
D
StressAppearance
Stress damage is most often shown by a series of roughly parallel
surface lines that often show electrical contact at a stabilized 5 kV.
These lines are sometimes
Page 515
hard to see unless outlined by product buildup or the use of sensitive
identification powders. An important damage identification criterion
is the fact that the cracks always occur perpendicular to the major
applied stress, whether mechanical or thermal.
E
AbrasionGeneral
Because of the number of interrelated variables, e.g., type particle (as
it relates to hardness), particle size and distribution, shape, density,
concentration, velocity (erosion), and solvent, it is impossible to
quantify the use limits as they pertain to abrasion. This is where
laboratory studies, using the specific recipe and associated campaign
parameters, are extremely valuable. As a first-screen type of
evaluation, the use of the following Moh hardness scale for minerals
is sometimes useful. This is essentially a pecking order type of listing
in which a material lower in the table with a higher number will
scratch any material above it. Parentheses denote additional guideline
materials to the original list or a clarification comment.
Material
Talc
Gypsum
(Fingernail)
Calcite (penny)
Fluorite
Apatite
(Knife, glass)
Orthoclase
(Steel file)
Quartz (beach sand)
Topaz
Corundum (aluminum
oxide)
Moh
scale
1
2
2.5
3
4
5
5.5
6
6.5
7
8
9
Diamond
10
With glass at 5.5, any material lower in the table or higher in Moh
rating will scratch it. The Moh hardnesses for many other materials
can be found in handbooks. It is again worth reemphasizing that
particle hardness is only one of several interacting factors that may
contribute to abrasion damage. Keeping the particle size small, the
density light, the velocity low, the solvent ''lubricating," etc., can
greatly help to reduce the potential damage. An important note is in
regard to the combined effect of abrasion with corrosion/velocity. The
byproducts of many corrosion reactions with materials, including
glass, actually assist in reducing further corrosion by way of a barrier
effect. If this barrier is removed
Page 516
via abrasion or velocity, the corrosion rate, rather than tapering off,
will continue at the initial high rate.
F
AbrasionAppearance
In the initial stages, pure abrasion damage usually consists of a series
of circular scratches or tracking lines on the bottom head and lower
sidewalls of the vessel. The leading sides of the agitator blades and
the baffle will also show these tracking lines and often show
considerable greater loss in fire polish or, in more serious cases, glass
removal, compared to the lagging sides. There is no initial electrical
contact at 5 kV stabilized.
G
SpallingGeneral
This is the classic case of "delayed" fracturing that is also directly
related to the aforementioned "combination effect." In a few highly
isolated cases, a high stress point, e.g., a large subsurface bubble, may
occur during the glassing cycle that does not show up in any of the
quality assurance checks or during the time up to final inspection and
shipping. This may be likened to a hand grenade with the pin still
intake. The addition of another stress factor, e.g., thermal or
mechanical loading, is analogous to removal of the pin. This type of
damage is extremely sensitive to the secondary stressing influence
and, consequently, usually shows up quickly after batch cycling has
been initiated. The probability that this is a damage cause falls
exponentially as service time increases. This, obviously, is the
manufacturer's responsibility.
H
SpallingAppearance
Like impact, there is always an epicenter where the stress riser existed
along with some plateauing. The plateaus, however, are usually
rougher than those for impact. Initially there will always be some
evidence of the stress riser, e.g., the remnants of a large bubble (now a
large pit), some inclusion contamination. If considerable corrosion has
taken place since the stress relief occurred, this visual evidence may
be somewhat masked. As geometry is also a very large combinational
factor, this damage is frequently observed on or near convex radii.
VII
Thermal Considerations
A
ThermalGeneral
As noted earlier, the residual compressive strain of the glassed steel
system decreases with increase in temperature. Consequently, special
care must be taken in establishing the thermal operational limits.
When discussing limits, it is first most important to emphasize the
mandatory need to clearly identify the glassed
Page 517
steel system. Most manufacturers have developed many systems over
the years and may also have several current systems all with different
thermal limiting characteristics. Data on the specific system must be
obtained from the manufacturer. Rule #7: Never attempt to crossreference any data, parts, or the like from two different
manufacturers. Most of the standard systems developed since the
early 1980s will maintain the same temperature differentials of
approximately 260°F from the lower allowable temperature of -20°F
up to 250°F. In defining the differential, it is always best to use the
glass wall temperature as the baseline temperature and calculate up or
down from it. In going from a glass temperature of 250°F to the upper
limit for the standard systems of 450°F, a gradual reduction in the
differential must be made in order to maintain suitable compression in
the composite. Following is a very general guideline-type table
containing several averaging assumptions that may help define this
relationship:
Glass wall
temperature (°F)
-20 to 250
300
350
400
450
Maximum allowable
differential (°F)
260
225
195
170
150
Realize that most of the thermal data contained in the manufacturer's
literature applies only to vessels of standard design and to volumes up
to 4000 gallons. For other cases additional safety factors (over the
standard) of close to 25% of the allowable differential may need to be
applied.
Due to the combination effect, a standard safety factor must always be
applied to the final differential determination. A value of 15% is
usually recommended, e.g., if the calculated differential is 200°F, then
200°F, × 0.85 = 170°F would be the safe operating differential. (Note:
The limits for standard vessels using conventional jackets also apply
to the half-pipe design).)
It is most important to pay extremely close attention to the
manufacturer's thermal loading limits and procedures as the damage
resulting from either miscalculation or poor operating judgement is
usually of such a delayed chipping nature and of such a large size as
to make field repairs very unreliable.
The previous temperature limit table also shows a very subtle but
important characteristic of the glassed steel composite, i.e., while the
strength of the glass goes down with increase in temperature, the
reverse statement, that the glass gets stronger as the temperature
decreases, is also true. It should be remembered that the lowtemperature limit for the composite of -20°F is not based on a glass
Page 518
limitation but on the metal. This fact provides reason why liquid
nitrogen at -320°F can be safely admitted to hollow agitator shafts to
allow a shrink-fit closure of a separable blade system.
When evaluating process equipment for potential use in the CPI, one
must be clearly cognizant of the so-called blinder effect. Most times
the focus of the evaluation is on the main piece of equipment and how
it will stand up to the recipes and associated campaigns. This can be a
very costly mistake. One must recognize that the glassed steel vessel
comprises many components each with its individual chemical and
physical limitations, e.g., the PTFE envelope of gaskets, possible
different glass systems between the vessel and accessory items, lower
agitator seal, sight glass, possible repairs, associated pumps, valves,
piping, expansion joints, dip pipes, condensers. It is mandatory that all
materials be closely reviewed (acquisition of knowledge) and the
"weak link" for operation defined. A case in point is the allowable
differential thermal limits for glassed steel vessels as compared to
several types of available glassed pipe/fittings:
Standard steel
vessel
260°F
Cast
steel
230
Ductile
iron
180
Gray cast
iron
100
Another major problem associated with establishing operating thermal
limits is the accurate determination of critical vessel content
temperatures. The temperature of addition chemicals at an upper
liquid level will not register immediately on the sensing probe usually
located in the bottom of the baffle. This time difference can be
critical. This is frequently compounded by the fact that a glassed,
integral tipped probe is used. The poor heat transfer of the glass
coupled with the other series type of thermal resistances of this design
results in extremely poor thermal response. It is recommended that the
faster response resistance-type devices be used for all temperature
critical campaigns. From the standpoint of safety, it is further
recommended that a dual-probe unit be used. It is also recommended
that a portable-type probe be employed to assist in mapping vessel
temperatures, i.e., defining hot/cold spots.
From the damage standpoint, thermal loading can be subclassified in
two ways: thermal shock and thermal stress.
1
Thermal Shock
In this case, the thermal differential, i.e., hot to cold, cold to hot, is not
only applied instantaneous but also relatively uniformly over the
surface of the glassed steel. There are several possible thermal
differential scenarios among glass wall
Page 519
temperature, materials added to the vessel interior, and materials
added to the jacket. Of these, two are most likely to cause thermal
shocktype damage:
Addition of a cold material to the interior of a hot-walled vessel
Addition of a hot material into the jacket of a cold-walled vessel
In the past, it was the manufacturer's approach to assign different
thermal limits to these two possibilities as contrasted with more liberal
limits for other less severe loading situations, e.g., addition of a hot
material to the interior of a cold vessel, addition of a cold material into
the jacket of a hot vessel. This led and continues to lead to
considerable confusion, especially for multiple loading situations. The
worst case scenario should always form the basis for calculations and
the most conservative result used, regardless of heat input scenario.
Again, it is always strongly recommended to use the manufacturer's
data for the exact glassed steel composition employed and to verify
with them the accuracy of any calculations made. Note that jacket
media have considerable variance as regards heat transfer film
coefficients which, in turn, cause corresponding changes in the
allowable loading limits. Rule #8: Never guess when thermal loading
is in question.
2
Thermal Stress
In this case, the thermal differential may or may not be instantaneous
and is always applied in a nonuniform or area-distinct manner. The
major causes for this type of damage are directly related to how the
equipment is piped up and the manner in which materials are added to
it. Generally, for conventional jacket designs (Fig. 2), steam should be
admitted through a top jacket nozzle(s) and trapped off at a bottom
jacked nozzle(s) nearest the bottom outlet nozzle. Cooling media
should be admitted through the side and bottom jacket nozzles
opposite to the side where the steam is admitted. The use of agitating
nozzles for coolants less than 20 centipoises viscosity is highly
recommended. These must be directionally oriented so that a
consistent flow pattern opposite to the direction of agitation is
obtained. This will provide the best heat transfer. There are three
types of common damage associated with thermal stressing:
Ladder. As the name implies, this damage takes on the appearance of
a series of parallel, in-line (top-to-bottom) crack groups resembling
the rungs of a ladder. These cracks are always located below one or
more of the jacket nozzles. The common cause for this damage is the
backflow of cooler material through the nozzle and down the backside
of the hot glassed steel wall. This backflow is most often initiated by
the formation of a partial vacuum after steam shutoff coupled with a
malfunctioning check valve. The horizontal cracking perpendicular to
the vertical cooling stress matches exactly the stress identification
criterion as described in the stressappearance section. To help
counteract the problem, sev-
Page 520
eral manufacturers have now increased both the size and deflection
angle of the impingement baffles located on the most troublesome
top steam inlet/water overflow nozzles (Fig. 2). These effectively
direct any backflow out toward the jacket wall. These have been
available since the mid 1980s on both new and
reconditioned/reglassed vessels. However, the simplest and most
effective approach is to install a vacuum breaker in the sealer ring
vent nozzle.
BON, or bottom outlet nozzle. Of all of the vessel nozzles, the bottom
outlet is by far the most critical. This is because of the confluence of
several strength-reducing factors, i.e., the jacket-to-nonjacket
transition, the bottom convex swage geometry, and the potential
gravitational (tension) loading off the nozzle (Fig. 2). The initiating
stress is most often thermal. The most common scenario is a batch
startup with the vessel filled with cold material. Full-pressure steam is
then applied to initiate heatup. The steam must first drive out the cool
condensate that did not completely trap out from the previous run.
This sets up a hot steamcold condensate demarcation ring area that
tends to put the glass in tension. Once the steam reaches the
diaphragm collar, just outboard of the BON, the previously fast
temperature equilibration process slows down dramatically. Here hot
steam is next to the cold-vessel reactants. The result may be circular
glass cracking approximately opposite the collar weld. As
equilibration gradually proceeds to the sensitive swage area, an axial
cracking around the swage radius, termed "cat's claw fracture,"
usually occurs. There are frequently two opposing fractures 180°
apart. This type of stressing is greatly compounded by gravitational
loading. The solution relates back to Rule #1, i.e., patience. The
proper procedure is to stage the steam admission, i.e., start with 20
psig, hold for 2 min, then increase the steam in 15 psig increments
holding each for 1 min.
Exo/endo. This damage relates to the addition of a chemical(s) to the
vessel interior that reacts with the host chemistries in either a heatgenerating (exothermic) or heat-absorbing (endothermic) manner. The
leading cause for the damage is related to the nonuse or misuse of dip
pipes. If a dip pipe is not used or is of insufficient length, the addition
chemicals can contact the liquid surface and spread rapidly to the
cold/hot sidewall where the damage occurs. For additions of these
chemicals, it is strongly recommended that the dip pipe be of
sufficient length to allow for subsurface entry. However, problems
may remain even with subsurface entry, i.e., if the dip pipe is openended, a relatively large stream of material can enter the vessel.
Agitation forces may then push this stream toward a different
temperature area within the vessel causing glass damage. As with
cavitation, it is always recommended that a sparger be used, i.e., one
with holes directed away from the vessel sidewall, when
Page 521
adding chemicals to the vessel interior. It should also be
remembered to alter the chemical addition rate to prevent streaming
type effects. Rule #9: Always use a properly designed sparger for
addition of vessel chemicals. A corollary-type problem is
associated with the use of heating mantels/tapes on single-shelled
vessels. For these situations, the manufacturer must be consulted.
3
Welding
A frequently asked question is, "Can glassed steel be welded in the
field?" The answer is an emphatic yes, provided that certain
restrictions are adhered to. If the equipment was originally Coderated, a certified Code welder must be used in conjunction with an
inspector from the local jurisdiction. The same thermal limits as for
process conditions still apply. As welding differentials are difficult to
determine, two guidelines may prove helpful: never weld more than a
1/2-in.-long bead in any one area; allow the area to cool to the point
where the hand can be comfortably placed on the glass next to the
weld. Consult the manufacturer for more specific data that pertain to
their designs. Note: the use of water in the jacket to cool down (or
speed up) the process can be risky and should be avoided.
B
ThermalAppearance
Although some specific appearance signatures have been reviewed,
e.g., ladder, it is best to treat the overall appearance from a more
classical standpoint. There are three process damage stages of
increasing severity associated with either the gradual, progressive
increase in the thermal loading or the continued "same-temperature"
loading above the critical point:
First craze. This most often appears as a spider weblike cracking on
the glass surface that often may be difficult to see. Fortunately, as was
the case for stress, the damage is often shown up by a product film or
by application of an electrostatic-sensitive test powder. The initial
crack depth is usually less than 10 mils and there is no electrical
contact at 5 kV stabilized. If the major originating stress was from the
metal or backside, the initial cracking will occur at the metalground
coat interface. At this stage there will be no powder outline or
electrical contact. However, as the cracks propagate, they will
eventually reach the surface, at which time both powder outlining and
electrical contact will be observed.
First chip. As stress loading continues, the cracks propagate further
until localized stress relief causes a chip(s) to spall off. Chip size and
depth are dependent on the thermal loading parameters. For surface
loading there is usually no electrical contact; if backside, there always
is.
Total or catastrophic. As thermal loading continues, so does the glass
fracturing. This eventually results in a highly roughened surface
composed of many hills and valleys that most often shows electrical
contact.
Page 522
Both the first-chip and total damage situations must be completely
avoided, especially the backside type. As the fracture areas are usually
relatively large and the glass continues to stress relieve, the possibility
of making reliable field repairs is quite remote. However, if during a
thorough inspection a nonelectrical contact, first-craze pattern is
observed, backing off on the temperature differentials by 3040°F may
extend the service life significantly.
VIII
Electrical Considerations
A
ElectricalGeneral
Electrical damage is usually traceable to either process recipes that
tend to build up static electricity or to inspection procedures that use
prohibitively high or unstable electrical test voltages. The end results
vary greatly in the degree of severity from both the safety and glass
integrity standpoints.
In terms of safety, two possibilities exist:
Explosion. If a process situation occurs in which a spark of sufficient
energy is generated in an ignitable environment, an explosion will
result. This is another important instance of expert advice being
mandatory. The physics departments of most colleges and universities
may be especially helpful in this regard. Their usual first-step
approach is to study the time-honored fire triangle, which relates the
three requirements needed to cause an explosion, i.e., energized spark,
fuel, and oxygen. Removal of any one of these requirements will
effectively eliminate the explosion possibility. Most experts first
recommend the reduction of the oxygen to a level below the explosion
limit via the use of a proper inerting atmosphere, e.g., nitrogen. Note
that this approach may not be totally effective for chemistries that
contain oxygen as part of their molecular makeup, e.g., nitrated
materials.
Personal shock. This occurs most often in testing. Glass, especially if
the humidity is low, is an excellent retainer of electrical charge. In
testing the glass surface with the recommended 5 kV stabilized, a
condenser-type charge can be readily built up on the glass when the
highly portable, direct current (DC) instruments are used. If the circuit
is closed between the glass surface, the human body, and a lowvoltage ground, a spark will discharge between body and ground.
While this very low current will not directly cause bodily harm, the
surprise aspect has caused backstepping-type tripping and ladder
falloff accidents. (Note: This charge buildup phenomena is only
associated with DC systems. Alternating current (AC) systems do not
have this problem. The DC problem is easily overcome by
periodically wiping off the charge during testing with the unit turned
off.)
Page 523
In terms of glass integrity, again there are two possibilities:
Instrument damage. Charge generation by either testing or process
causes may result in serious instrument damage. All instruments, e.g.,
resistance-type temperature, pH, and fault detection sensors, should
first be adequately fused and be fitted with appropriate voltage
interruption circuitry. Also, for testing, stay at least 1/2 in. away from
any sensitive portions of the instrument probes; for anticipated
process charging, the unit(s) should be turned off.
Glass damage. For this case, the dielectric breakdown strength of the
glass, i.e., 500 V/mil, has been exceeded. Again, this can be caused by
either testing or process operation.
Testing. There have been several previous references to the use of a
"5-kV stabilized" test voltage. The 5-kV maximum is based on studies
that indicated that repeated testing at voltages higher than this can
progressively lead to glass breakdown. If the glass surface is clean of
product films and/or water, this voltage level is more than sufficient to
reveal either direct metal contact or thin (<10 mils) glass. A
wandering, corona-type spark is usually indicative of a product/water
problem in contrast to the sharp-definition, blue/white spark of a true
damage point. The stabilized voltage requirement ties directly to the
maximum 5 kV. Testers without a stabilization circuit have been
known to spike to over 30 kV. Rule #10: Always field-test glassed
steel at 5 kV stabilized.
Process. This is another combination-type situation in which certain
vessel chemistries lead directly to the physical dielectrical breakdown
of the glass. Like most electrostatic phenomena, the problem is related
to the interfacial removal of electrical charge from one phase to the
other. There are a number of these interfacial phase possibilities most
of which act in conjunction with the solid phase vessel wall, i.e.,
liquids in the vessel, particles in the liquid/vapor, two-phase liquids or
emulsions, vapors dispersed in the liquid. There are many factors that
may contribute to electrostatic buildup during process operation. The
major factor relates directly to the electrical characteristics of the
liquid, especially the fact that most problem liquids have very low
specific conductivities, e.g., hexane, the xylenes, toluene, benzene.
Consequently, if charge is generated in one volume of liquid, it cannot
readily be discharged in a continuous manner to some lower voltage
point. Thus, the charge is continually built up to the point where
sufficient energy exists to cause a high-energy spark to be formed that
may dielectrically break down the glass. As this is a localized,
volume-sensitive situation, a grounding point away from the charged
volume will not necessarily be effective. A single grounding point or
even an area will only be effective if the liquid reaches a conductivity
level where it will allow transfer of charge between the generation
point(s) and the ground. While on the subject of grounding, it is most
important to note that discharge-type problems, including explosions,
have been
Page 524
observed in completely grounded metal vessels. This again relates to
the point-to-point or volume-to-volume type relationships that may
exist with many low specific conductivity fluids. An analogy may be
made with heat lightning in which the discharges are cloud-to-cloud
rather than cloud-to-ground. The obvious cure for this problem is
either to use liquids with greater conductivities or to make additions
that will increase it. Quite often, surface-active agents, e.g.,
quaternary ammonium compounds, have been shown to be effective
even in very small amounts.
Other factors that tend to increase charge buildup include anything
that increases the interfacial area, e.g., smaller size of the secondary
phase, or that increases the rate of electron transfer, i.e., agitation
speed. Consequently, increasing particle size and lowering agitation
speed would be suggested corrective actions.
B
ElectricalAppearance
Like thermal loading, static discharge damage may occur in stages. In
fact, static discharge through glass is a thermal loading phenomenon,
i.e., hot sparkcold glass. The first-stage damage will obviously always
contact at 5 kV stabilized provided that a high-resistivity product has
not been redeposited in the discharge path. The initial damage area is
highly dependent on the spark energy level but most often is less than
1/8 in. in diameter and is usually funnel-shaped. If the campaign is
dedicated, i.e., consistent recipes used, the glass corrosion will be very
slight as the mostly organic-type chemistries that cause the problem
have very little corrosive effect on the glass. There is frequently a
circular blackbrown ring around the contact area indicative of organic
material decomposition under the influence of the high-temperature
spark discharge. Under certain flow situations, dendritic or branchlike decomposition lines may also be observed. As more or greater
energy discharges occur at the same point, the damage area grows and
starts to take on many of the characteristics of thermal loading
damage. High-velocity areas, i.e., agitator blade tips and leading
surface/edges, the leading baffle surface/edges, and the vessel
sidewall opposite the agitator, are especially susceptible to this type of
damage.
IX
Chemical Resistance
In discussing possible materials and the associated equipment for use
in the CPI, the most frequently asked question is, "How long is it
going to last?" When most people ask this question, it is in direct
reference to their major concern of material service life based on
corrosion resistivity. This is largely an outgrowth of working with
metals and alloys, in which corrosive activity is indeed the major
cause of reduced service life. However, for glassed steel, this
approach is a classical instance of the aforementioned "blinder effect,"
whereby the evaluation screen is
Page 525
much too narrow. With reference to Fig. 4, it will be remembered that
approximately 90% of the problems relate not to chemical but to
physical problems. There are two counteractive reasons for this
interesting difference between metals/alloys and glassed steel:
1. The greater sensitivity of the glassed steel to physical influences.
While the composite glass on metal approach greatly increases the
strength of the glass, it never approaches that of most metals/alloys
and consequently is more susceptible to physical damage.
2. The greater corrosion resistivity of glassed steel to the majority of
chemicals found in the CPI. This statement is predicted on the
optimization of the selection factor criteria of good performance and
low cost.
A better appreciation of reason #2 may be obtained by referring to
Fig. 5, which shows the relative corrosion resistance of glassed steel
compared with other materials commonly found in the CPI. Three
broad material groupings are represented: glassed steel, metals/alloys,
organic-based. This chart represents different types of acidic-corrosive
environments. The vertical dashed line in the center separates
oxidizing and reducing environments. Increase in material band length
to either the left or right indicates an increase in oxidizing or reducing
power. The middle horizontal dashed line separates chloride-from
non-chloride-containing solutions. The chloride concentration
increases with height in the upper portion. Each material will
generally be resistant to environments below its band. As observed,
glassed steel forms a corrosion resistivity umbrella over the other two
material groups. This is based on the following:
Metals. The major reason for the corrosion resistance of the metals is
the formation of a low free energy surface oxide film on the surface. It
is this oxide barrier film, not the metal itself, that screens out the
corrosive activity. Corrosion may be defined as material deterioration
due to interaction with its environment. The driving force for this
interaction is most often the tendency to form the lowest energy
products. If the material surface is already at a low-energy state, there
is little tendency to interact further via corrosion. However, if the
oxide film is chemically reduced, scratched, abraded, or in some way
stressed off, high corrosive reactivity will again resume until lowenergy products are reformed. In this process, more material will
deteriorate.
The reason for the excellent corrosion resistivity of most glasses
should now be obvious, especially if one refers back to the glass
selection section. The complete makeup of the glass not just the
surface, is low-energy oxides. Consequently, it is only in a relatively
few, very well-defined chemical situations, that glass is corroded in a
manner that is prohibitive to long-term serviceability.
Organic-based. The pure forms of these materials usually possess
surface energies lower than that of glass. In fact based on chemical
deterioration alone,
Page 526
Figure 5
Materials selection chart.
several polymeric systems are superior to glass. The composite type
materials, e.g., fiber-reinforced, are usually more limited in chemical
resistivity. However, other ''weak links" show up with these polymerbased materials, in the form of permeation, temperature, high velocity,
pressure/vacuum, heat transfer, and oxidation, that may cause a
reduced overall performance compared to that of glassed steel. These
weak links are factored into the data shown in Fig. 5.
Prior to discussing the specific limitations of glassed steel, it may first
be instructive to review five of the most important and most
frequently asked questions about corrosion principles which have the
most direct bearing on defining these limitations:
1. How do materials, in general, corrode?
2. How does glass corrode, especially in relation to metals?
3. What does glass corrosion look like?
4. What are the most critical parameters governing glass corrosion?
Page 527
5. How can meaningful test data be obtained?
A
Corrosion of Materials in General
Corrosive activity for any material is mainly related to an interfacial
reaction among a solid and the contacting liquids, vapors, or other
solids. This portion of the discussion will address only the most
common solidliquid interfacial situation as depicted in Fig. 6. When
most liquids come in contact with a solid, a relatively immobile film
is formed on the surface. For corrosion to initiate and then continue,
the following seven progressive steps must occur. The bracketed items
are some of the more important controlling parameters associated with
that particular step.
1. Transfer of corrosive material(s) from a point in the main liquid
stream (X) to the diffusion film (fluid velocity, density, viscosity,
concentration of the reactive corrosive species).
2. Diffusion of the species through the film (temperature, density,
viscosity, diffusion rate constant, concentration of reactants, area).
3. Absorption of the species on the material surface (concentration of
the reactants, temperature, and density).
4. Chemical reaction at the surface, i.e., the actual material
deterioration (area, concentration of reactants and products,
temperature, free energy, kinetic rate constant, surface tension).
5. Desorption of the corrosive products formed at the surfacethe
reverse of step 3.
6. Diffusion of the products through the filmthe reverse of step 2.
7. Corrosive product transfer away from the film back to (X)the
reverse of step 1.
It is extremely important to note that reduction or elimination of any
of these steps through effective adjustments of the controlling
parameters will very effectively reduce or possibly eliminate the
corrosion process. Make special note
Figure 6
The solidliquid interface.
Page 528
that temperature and concentration are associated with five of the
seven steps and therefore should always receive special process
altering attention. Rule #11: Always carry out a detailed parameter
analysis for each campaign to determine possible corrosion
ratelowering adjustments.
B
Corrosion of Glass
While some general analogies can be made to metal corrosion, the
mechanism(s) of attack are completely different. Metals can corrode
in any of nine ways, i.e., uniform, environmental cracking (e.g., stress,
embrittlement), intergranular, pitting, crevice, galvanic, selective, aka
dezincification, erosion, biological. The majority of these are
dependent on electron transfertype reactions that readily transform the
initial high-energy metal into lower energy forms, e.g., oxides.
Glasses because of their low electrical conductivity within the normal
temperature use range cannot readily transfer electrons. Therefore
other corrosion mechanism explanations must be sought. Sufficient
research activity has now been carried out to indicate that these
mechanisms are highly complex. However, for the majority of
applications, a simple explanation using the house-of-cards structure
mentioned in the glass selection section may be appropriate.
There are two fundamental mechanisms for glass corrosion: one
related to the removal of the species contained in the holes formed by
the cards, i.e., the modifiers and intermediates, and the other to the
removal of the cards or the glass formers, i.e., SiO2. Acids (fluorine
and phosphorous compounds excepted), small ions, and the first stage
of water attack involve removal of the "hole" materials by a diffusioncontrolled ion exchange mechanism in which the small ion from the
acid, e.g., hydrogen, exchanges for the larger modifier/intermediate
ion. The disproportionate size exchange causes microscopic stress
relief cracking in the card structure that eventually leads to a dulling
of the glass, i.e., loss of gloss or fire polish, aka LOG. Due to the
diffusion-controlled aspect of this attack, it is self-limiting up to the
point where the stressed barrier layer is removed via velocity, thermal,
etc., influences, thereby exposing fresh glass to further ion exchange.
Corrosive attack by alkalis, fluorine, and phosphorous-containing
compounds and the second stage of water attack involve removal of
the actual network-forming cards through a regenerated dissolutiontype reaction. The "hole" materials then simply fall out in the process.
Since the entire glass structure is continuously destroyed with the rate
directly dependent on the corrosion product solubility, these types of
corrosives should be used with great caution in glassed steel
equipment. Note that no mention is made of not using glassed steel for
these recipes. They can be used provided that the appropriate use
restrictions are closely followed.
Page 529
C
Appearance of Glass Corrosion
Glass corrosion, like other types of glassed steel damage, most often
proceeds in stages:
Stage 1Glass corrosion most often starts out as a dulling of the glass
surface or LOG as referenced above. This may be caused by either the
ion exchange or dissolution type of corrosion reactions and is usually
quite uniform in general appearance. The degree of etch, as regards
topographic relief, is a direct function of the corrosive parameters,
e.g., alkalis frequently cause a sandpaper-type roughness. One of the
most interesting and, unfortunately, most potentially damaging
phenomena associated with glass corrosion is that the glass may
corrode with only a very slight or even nondiscernible LOG. This may
be likened to the chemical polishing of metals. Consequently, the
visual appearance of the glass may be a very poor indictor of the
actual corrosive rate. This fact necessitates special inspection
requirements that will be discussed later.
Stage 2The corrosive upset of the glass surface leads directly to
localized stress differences which, in turn, lead to varying degrees of
stress relief and, eventually, glass chipping. This chipping is most
frequently a combinational effect related to a product buildup, e.g.,
polymerizations and crystallizations. When polymer/crystals are
deposited in the etched surface areas, they usually bond very
tenaciously. Then even slight variations in temperature are sufficient
to set up thermal expansiontype chipping. The end result may be a
very progressive type of damage situation.
Another type of chipping should also be mentioned. This is termed
"fatigue-type chipping" and is usually associated with vessels that
have been in service for relatively long times. Frequently, the surface
still has good fire polish or low LOG, but has a number of relatively
small surface chips that cannot be completely explained by physical
causes. The reason for the problem relates to the important fact that
glass is not a static type of material but is continually changing under
the influence of both physical and chemical influences. Over time,
stresses build up and lead eventually to chipping-type damage.
One of the most serious types of glass corrosion damage is pitting,
which may occur at any time and cannot be classified as a specific
progressive stage type damage. For these cases the use of uniform
corrosion rates for estimation of service life is meaningless, i.e., if all
of the deterioration in a low-corrosion-rate situation is focused on
several small pitted areas, the glass integrity may easily be lost
contrary to what the rate data may indicate. This type of attack is both
recipe-and glass compositionsensitive. Chemical species that attack
the glass-forming portion of the network, e.g., fluorine-based
chemistries and alkalis, may
Page 530
cause this type of damage. Older glassed steel systems and
glass/crystal composites are more prone to this type of attack. Also,
for some of the older systems, other halogen-based acids, e.g., HCl,
may cause pitting.
D
Parameters Governing Glass Corrosion
In going over these parameters, it may be advantageous to use Fig. 7
as a reference. This is a somewhat typical corrosion rate diagram used
by the glassed steel equipment vendors. It plots chemical X's
concentration vs. temperature for a series of constant corrosion or
isocorrosion results.
Type corrosive. From what was said in the glass corrosion section it
should be apparent that different chemical species by themselves have
a widely varying but largely predictable effect on glass corrosivity.
The predictability factor goes down in proportion to the number of
chemical species in the mix. This is another instance of the
"combinational effect." When chemistries are mixed, catalytic (rates
greater than predicted from looking at the individual diagrams),
inhibitive (rates lower), or constant (maximum rate corresponds to the
most corrosive member, e.g., the HCl in aqua regia), effects on the
single member rates may occur. A relatively common catalytic effect
is the alternating use of acid followed by alkali, then back to acid, etc.
For this case, the self-limiting barrier ion exchange layer is
subsequently removed by the alkali, which also breaks down
Figure 7
Typical isocurve corrosion chart.
Page 531
the network structure exposing fresh glass to the next exchange
reaction, etc. Note closely that the data shown in Fig. 7 are for only
one chemical species, i.e., X. With the exception of some chemical
purification campaigns, e.g., crystalizations, it is uncommon in the
CPI to process just one species at a time. Consequently, for mixed
chemistries, the data represented by Fig. 7 should be viewed with
great caution. Rule #12: Always view literature corrosion data as a
guideline only. For complex chemistries, meaningful testing under
conditions that closely approximate the actual process is strongly
recommended. Vendors, as part of their damage analysis studies, also
frequently have information on mixed chemistry campaigns that often
approximate a proposed campaign.
With the exception of materials that attack the glass formers, or
"cards," most glass systems have suitable chemical resistivity up to
250°F and, in some cases, up to 450°F, e.g., high concentrations of
sulfuric acid.
Temperature. As most corrosion processes increase significantly with
temperature, this is an extremely important control parameter. Every
effort should be expended to keep the temperature as low as process
requirements permit. This is especially true for materials that attack
the glass formers, e.g., alkalis.
Concentration. This is a highly variable parameter frequently related
to the degree of ionization of the chemical species. For many
individual systems, a midrange concentration is the most corrosive
(Fig. 7). Materials that attack the glass formers usually exhibit
increased corrosion rates as the concentration increase.
Time. Corrosion rate data similar to those in Fig. 7 are usually
reported in mils per year or mm per year and relate to a continuous 24
hours-a-day, 365 days-a-year service exposure. This timing factor
may have a significant bearing on material performance acceptability.
If the proposed campaign is less than continuous, the appropriate
adjustments must be made, e.g., the yearly corrosion rate is 9
mils/year but the actual anticipated operation is only for 8 hours a day
or one-third of the year. The actual rate will then be 9 mils/year × 1/3
= 3 mils/year. This takes the glass out of a marginally acceptable
category at 9 mils/year into a totally acceptable one at 3.
Velocity. This is an extremely important parameter whose effect is
very seldom included in any data presentation, regardless of material
type. In fact, the majority of U.S. and international standard tests are
based solely on the static mode. With corrosion rate increases of over
20× possible with the addition of the velocity parameter, it should be
obvious that use of purely static data can lead to very misleading
conclusions regarding the suitability of a material. All meaningful
testing programs must include this parameter.
Phase. Like Fig. 7, most data presentations are based on the use of
liquids. There are three other phase situations that must also be
considered, i.e., vapor, condensing vapor, and interfacial. Condensing
vapor and interfacial situations
Page 532
frequently show corrosion rates many times higher than those of the
liquid. Noncondensing data are most often lower than the liquid data.
Contaminants. Most corrosion data were obtained using very pure
reagent grade chemicals. This contrasts with many of the situations
found in the CPI, where economics dictates the use of practical,
technical, or even regenerated grades. These often have percentages of
a secondary species that can drastically alter the corrosive activity.
Even parts per million (ppm) amounts of certain species, especially
when mixed with other critical primary materials, can cause
unacceptable deterioration. These contaminants often bring out a
"weak link" relationship in the material.
E
Obtaining Meaningful Test Data
From what has been reviewed previously in this section, it should now
be apparent that literature data must be used for first-screen or
guideline purposes only. Before a final material selection is made, all
"I think" references must be replaced with "I know." There are several
ways to accomplish this:
Equipment vendor. As mentioned previously, most vendors have been
in business for many years and have accumulated a library of possibly
very useful data. Although it may not be possible to exactly duplicate
the intended recipe/campaign, valuable corollary data that add to the
overall selection knowledge base can usually be obtained.
Consultants. These must chosen very carefully. Do not rely on a
materials engineer who has little knowledge of process engineering or
vice versa. Also, many materials engineers are very focused in their
area of expertise and often bias their recommendations accordingly.
The best pick is usually a chemical engineer with a wide-spectrum-use
base of CPI-type materials including experience with glassed steel.
The consultant should be completely independent without vendor ties
that might prejudice selection. All references should be checked out
and publications closely reviewed to assure the best technical fit.
Existing data. Most potential users of glassed steel equipment are
unaware that they may already have very valuable data available from
their own laboratories. Most recipes and associated campaigns are
developed and fine-tuned in a laboratory using borosilicate-type
glassware. These glasses, because of their rather unique compositional
makeup, have corrosion resistivities usually lower than the majority of
currently available standard glassed steel systems. This is especially
true for chemical systems that attack the glass former portion of the
network, e.g., alkalis. Consequently, if laboratory use involving
realistic recipes over an extended time span indicates no perceptible
attack on the glassware, the odds are extremely high that no effect will
be seen on the glassed steel equipment, provided that the appropriate
parameter adjustments are taken into account, e.g., velocity.
Page 533
Testing. Even when the best data inputs are obtained from the above
sources, there usually remains a few, critical, unanswered questions.
The only way to get the answers is through a meaningful test program.
Such a program must involve careful involvement of the parameters
discussed above, which are not included in most data presentations,
i.e., mixed species, use of actual chemicals, velocity, critical phase,
and contamination. In addition, some other facts must also be
considered:
For largely safety reasons associated with the unfamiliarity of the
possible chemistries involved, most vendors prefer not to do the actual
testing. As mentioned previously, however, vendors can provide
considerable help including guidance in setting up the actual test
program along with the analytical evaluation of the test samples.
Another approach is to use the services of a commercial testing
facility. As was true for consultants, this selection must be done
carefully, especially in view of the fact that glassed steel testing is not
that commonplace.
Most suppliers of glassed steel equipment will provide gratis test
samples. It must be established that these samples are truly
representative of the production equipment. It is not difficult to coat
test samples with glass systems that have extraordinary corrosion
resistivities but that cannot be coated on larger pieces. Also with the
several systems usually available from any vendor, it's important to
make sure that the samples are of the correct composition. Even for
glassed steel samples coated with the identical composition as the
production piece, it is almost impossible to get exact representation.
This is related to the fact that the heat/cool cycles, due to the large
differences in mass, cannot be exactly duplicated. As glassed steel
systems at high temperatures are dynamic from the standpoint of
possible compositional change, e.g., dealkalization of the surface via
volatilization, different corrosion results may be expected. However,
for corrosive situations involving the ion exchange attack mechanism,
this may actually be a plus as the increased dealkalization will give
better long-term rate results, i.e., the production corrosion rate will be
lower than that predicted from the laboratory testing.
Most vendors supply either test dumbbells or disks. The use of
dumbbells should be avoided, if possible. This is based on the
important fact that the majority of campaigns are involved with heat
transfer across the glassed steel wall of a jacketed vessel. This heat
flux and associated "skin temperature" effects may have an important
effect on the overall corrosion rate. With the dumbbell design, heat
transfer through both metal and glass cannot take place. Therefore,
they should only be used for isothermal evaluations, e.g., constanttemperature storage vessels. Rule #13: Avoid the use of dumbbell-type
test samples for jacketed vessel evaluation.
Use of the disks allows for heat flux situations to be approximated
including the very important condensing vapor phase. It is
recommended that the disks be used in conjunction with standard test
procedures, e.g., ISO, DIN. Information on
Page 534
these procedures are available from the vendors. As most of the
standard tests are run at the boiling point or below, special units are
required for higher temperatures. Again, the vendor can provide
useful information on the most current equipment and the associated
procedures.
When carrying out any corrosion test, inhibition effects must be
eliminated if meaningful data are to be obtained. These usually result
from the accumulation of dissolved corrosion products in the liquid
from either the test sample(s) or the test chamber. From the mass
action standpoint this will reduce the corrosion rate, often quite
significantly. To overcome this problem, it is always recommended
that a volume of corrodant to sample surface area ratio of at least 10:1
be used.
If possible, increase the test temperature above that anticipated for the
intended process. In many short-duration tests, the onset of a potential
problem may be marginally difficult to determine. The increased
corrosion activity associated with even a slight increase in
temperature will often provide more defining information.
It is highly recommended that a sufficient number of samples be
tested to provide statistical assurance that the final value(s) is
accurate. Draw-type or planned intervaltype testing is also
recommended to define the most corrosive stage(s) in any campaign.
Make sure that all possible weak link materials are included in the test
program.
While the general corrosion information reviewed previously has
made some mention of specific problem chemistries, an in-depth
analysis may still prove helpful.
X
Effect of Specific Corrodents
A
Halogen Compounds
Fluorides. As ionized fluoride chemistries (not just the fluoride ion) in
low-pH solutions can rapidly cause a regenerative, constant rate,
network formertype deterioration of the glass, they must be used with
the extreme caution. One of the main reasons for this very rapid attack
is that the products of corrosion are usually extremely soluble in the
corroding liquid. Therefore, a barrier layer that might retard continued
corrosive activity cannot be built up. Fluoride is the only ion species
that, even at low ppm levels, can cause prohibitive damage to glass. It
is also one of the species that causes the very unpredictable pinholetype damage. Depending on the recipe makeup, damage can occur in
all phase areas, i.e., liquid, vapor, condensing vapor, and splash zone.
Note that some acids, such as HCl and H2SO4, may drive fluorides
out of solution causing greater vapor phase attack. Operation at higher
pH (this must be balanced against possible alkali attack, discussed
below), in nonionizing solution, e.g., organics, or at low temperature
may drastically reduce the corrosive effect. There are also a variety of
inhibitors that can effectively reduce or eliminate the problem, e.g.,
calcium, iron, aluminum
Page 535
ions, colloidal silicas. The most effective are the colloidal silicas,
which act as a "getter" for the fluoride ion. Amounts of 20% over the
stoichiometrically determined amount of fluoride ion present in the
recipe are usually recommended. The most critical characteristic of
the colloidal silicas is surface area and materials should always be
chosen that maximize this factor. Rule #14: Use fluorine containing
chemicals with caution. Rule #15: Analytical data must always be
obtained on all proposed reactants to ascertain potential corrosive
reactivity.
Chlorides. Hydrochloric acid (HCl) is no doubt the most troublesome
acid in the CPI, especially at elevated temperatures and from the
standpoint of economic materials selection. Most of the current lines
of glassed steels provide good serviceability below approximately
300°F, i.e., a corrosion rate less than 8 mils/year. Caution should be
used with regenerated acids as they may contain fluoride ion
contamination. Some powdered chlorides, such as AlCl3, react
exothermically with water and may cause formation of localized hot
spots or cavitation bubbles in the reaction mix. There is also an
indication that the synthesis of alkyl chlorides using HCl and H2SO4
may lead to prohibitive damage. Some chlorides in combination with
Cl2 and benzene ring organics have been known to attack the tantalum
repairs in glassed steel equipment. Also of importance is the fact that
HCl, along with most other mineral acids, will embrittle tantalum
above 300°F.
Bromides. On the basis of individual comparison, these chemicals
corrode glass at rates considerably less than their chloride
counterparts. The main "weak link" problem with these materials is
with regard to tantalum metal, a very frequent repair material for
glassed steel equipment. In combination with primary alcohols,
deterioration can take place relatively quickly. This is really an
interesting combination effect as tantalum is very resistant to both
bromine and the alcohols on an individual basis. It is believed that an
intermediate compound is formed that leads to the problem. The effect
decreases as the molecular weight of the alcohol increases, i.e., methyl
is considerably worse than ethyl.
Iodides. Similar but smaller overall corrosion effect compared to the
bromides.
B
Sulfur Compounds
Sulfuric acid (H2SO4). This is the only common mineral acid that can
be used in most glassed steel systems at the maximum operating
temperature for standard vessels of 450°F with good service life
(concentrations above 70 wt %). There are, however, three weak link
problems not related to the glass:
At high concentrations, small amounts of sulfur trioxide (SO3) may
be present that are extremely corrosive to tantalum. Obviously, the
same would be true for larger amounts of SO3, i.e., oleum.
There has been some recent indication that fluorinated polymeric
materials, e.g., the PTFE envelope of CRT gaskets, in higher
temperature H2SO4
Page 536
service, may degrade causing a very localized fluoride species
attack. Damage has been observed in the gasket areas of nozzles
where liquid washing is minimal and, consequently, where the
fluoride species concentration can build up.
High concentrations of the acid are oxidizing and not compatible with
the furfuryl alcoholbased fillers commonly used with glassed steel
repairs. For these cases, silicate-based fillers are recommended.
Chlorosulfonic acid (ClHSO3). The problem with this acid is its
tendency to maintain an equilibrium mixture with SO3 which, as
stated above, readily attacks tantalum. There is also evidence that this
acid in combination with other compounds may decompose to form
SO3.
Other. The use of any sulfur-based material should be carried out with
caution in tantalum-containing glassed steel equipment. While SO3 is
a definite problem, there is increasing evidence that other sulfur-based
species may also cause problems. Testing must always be done to
establish serviceability.
C
Nitrogen Compounds
These usually do not represent a corrosion problem for glassed steel
systems up to approximately 300°F. However, three weak link,
nonglass problems exist:
1. The PTFE of gaskets and repairs may be permeated by these
compounds.
2. Amines present in many of the tapes used to hold CRT gaskets in
place can plasticize the PTFE drastically altering its properties.
3. The organic fillers used under metal repairs are attacked by many
nitrated compounds and, more generally, by any oxidizing species.
D
Phosphorous Compounds
Phosphorous compounds are very interesting in that some possess a
mutual solubility for many glass compositions, especially at high
concentrations and temperatures. Consequently, these use areas must
be approached very cautiously. One of the major problems associated
with these materials is the extremely detrimental combinational effect
exhibited with fluorinated species, especially the fluoride ion. Other
than the furnace grades, this often is a difficult problem to circumvent
as the majority of phosphorus chemistries are derived from the
fluorine-bearing mineral apatite. It is very difficult to economically
remove the fluorine chemistries completely from the mineral
processing steps. The use of the lowest possible operating
temperatures and the addition of colloidal silica are usually very
effective in extending service life. Rule #16: Use phosphoruscontaining chemicals with caution.
Page 537
E
Organics
Excepting the aforementioned physical pinholing problem due to
possible electrostatic discharge, these materials present no corrosion
problem to most glassed steel systems up to 350°F. However, acetic
acid (C2H4O2) at temperatures above 300°F has, on special occasion,
caused damage to the PTFE envelopes of CRT gaskets and the sealing
washers of repairs. Many monomers, e.g., vinyl chloride, will also
readily permeate PTFE and then polymerize within the polymer
structure, causing large-scale deterioration of the PTFE, i.e., the
socalled popcorning effect.
F
Alkaline Compounds
As these materials directly attack the glass network structure in a
regenerative manner, their use in glassed steel equipment should be
carried out with extreme caution. Corrosion rates in excess of 10
mils/year for pH systems above 12 and temperatures above 150°F
may be expected. For the majority of these systems, the old standby
temperature relationship for metals applies rather well, i.e., the
corrosion rate doubles for each 18°F (or 10°C) rise in temperature.
Consequently, as is generally true with all corrosion-involving
processes, it is strongly recommended that temperatures be kept as
low as possible. Velocity is also very important and should be
minimized. As there is a critical anion effect with most alkalis, it is
especially important not to extrapolate literature data. Testing is
highly recommended. As was mentioned earlier in the type corrosive
section, an alternating acid/alkali service may be troublesome. As for
fluorine attack, there are several effective inhibitors against this alkali
damage, i.e., ions such as calcium, iron, zinc, and titanium; the
colloidal silicas; and most organic materials, especially those that are
surface-active. A case in point regarding organic inhibition is the fact
that organic materials that possess an alkaline nature, e.g., amines,
while corroding glass do so at rates far lower than inorganic materials,
e.g., hydroxides, at the same pH level. Dip pipes are highly
recommended for use with alkali additions. Rule #17: Always use a
dip pipe, preferably with sparger, when admitting alkaline materials
to glassed steel equipment.
G
Salts
The corrosiveness of any salt, excepting the fluorides, is largely
related to the residual pH and the active anodic or cathodic species,
e.g., NaCl is a neutral salt and the most aggressive part of the solution
would be the water (discussed below); AlCl3 is an acid salt with the
chloride ion as HCl, dominant.
Salts with small cations, e.g., Li+, Mg2+, Al3+, in aqueous media
should be used with caution at temperatures above 150°F. These
cations may enter into an ion exchange reaction much more severe
than the general acid attack described
Page 538
earlier. While the initial corrosion rate is usually small, very damaging
surface stresses can rapidly be set up causing progressive surface
chipping.
H
Water
Water corrosion of glass is a classical case of what is termed ''a getter
reaction." Pure water consists of only hydrogen (H+) or hydronium
(H3O+) and hydroxide (OH-) ions compared to the multi-ionic
makeup of most glasses. Consequently, there is a considerable mass
actiontype driving force to transfer ions from the glass to the water in
order to establish a lower energy ionic equilibrium. The purer the
water, the greater is the transfer rate and consequently the corrosion
rate. This is why the very pure condensing water vapor on the top
head of a vessel may be so troublesome compared to the
"contaminated" water in the liquid phase. The easiest solution to the
problem is a direct outgrowth of understanding the mechanism of the
corrosive attack. Water attack is a two-step process, i.e., the leaching
of alkali ions from the glass surface via ion exchange followed by the
direct attack on the network former structure by the alkali. If the alkali
can be neutralized at the condensation stage, corrosion will effectively
be cut off. This may be easily accomplished by introducing a volatile
acid, e.g., HCl, to the recipe. Other stage-interrupting approaches
include continually washing down the top head with a liquid spray or
running the vessel flooded with an expansion stand pipe. Another
useful approach is to reduce or eliminate the top head condensation
through the use of protected insulation or heat tracing. It is most
important that the insulation be protected against any possible acid
spill that might lead to an even worse type of damage, i.e., fishscaling
(to be discussed later). If heat tracing is used, care must be taken to
prevent prohibitive heat differentials. It should be noted that the more
recent glass compositions supplied by several manufacturers now
have exceptional resistivity to water vapor attack.
I
Other
Another frequently encountered problem relates to the use of
dissimilar metals in glassed steel equipment. Repairs, usually made
from the relatively noble tantalum metal, are present in much glassed
steel equipment. The introduction of another less noble alloy material,
e.g., stainless steel, in conjunction with an electrolytic solution may
set up a galvanic cell that will corrode the less noble, anodic
component and possible embrittle the cathodic tantalum. As the
anodic corrosion is usually over a significantly larger area than the
repair area, the overall loss of material will be slight. However, the
main problem is with the noncorroded but possibly embrittled
tantalum. Once embrittled, the tantalum is susceptible to fracture
under the influence of relatively minor stresses, e.g., agitation forces.
Another variation of this problem involves the use of metal powders
in the recipe,
Page 539
which may also cause tantalum embrittlement. Rule #18: Avoid the
use of dissimilar metals in glassed steel equipment.
Another galvanic cell problem related to repairs pertains to possible
leakage of the repair in electrolytic solutions. For this case, the noble
tantalum may be coupled to the nonnoble steel. This is a very active
cell that can rapidly cause embrittlement of the tantalum and corrosion
of the steel. When sufficient steel in the thread area of the repair has
been removed, the repair may actually fall out, thereby exposing the
remaining steel to the direct influence of the corrosive environment.
Rule #19: Get professional assistance when installing repairs.
J
SpillageGeneral
Fishscaling is one of the most serious problems associated with the
operation of glassed steel equipment. It results from acidic corrosion
of the metal opposite to the glass coating. This occurs most frequently
around nozzle areas, especially on the top head and the area opposite
the jacket. The nozzle problems relate directly to manway spillage or
improper gasketing; the jacket, to improper cleaning recipes and/or
procedures.
The cause of the problem is directly related to one of the products of
the acidic corrosion of steel, i.e., monatomic hydrogen gas (H0). This
acts in the same manner as the monatomic hydrogen formed during
the firing operation (see the ground coat portion of the glass
fusion/firing section). The resulting damage is extremely serious on
three counts:
1. It completely destroys the integrity of the glass coating, thereby
exposing the steel directly to the corrosive environment.
2. The damage is progressive over time, making it extremely difficult
to repair effectively.
3. The metal corrosion may lead to operational safety considerations
or to either pressure downrating or costly reconditioning expenses
upon reglassing.
Rule #20: Fishscale damage must be avoided.
Several recommendations can be made:
Follow closely the gasketing principles outlined previously.
If spillage/leakage occurs, wash off and neutralize immediately.
Use an appropriate protective paint on the nonjacketed areas and keep
it well maintained.
Use dump chutes when loading corrosives through the manway.
Protect all insulation with protective aprons or coatings. The seepage
of corrosives through porous insulation and the resulting long-term
contacting with the steel can be devastating.
Page 540
Rule #21: use only jacket-cleaning media and procedures approved
by the manufacturer. The need for cleaning should be minimal if
suitably treated heat transfer media are used. Rule #22: Always obtain
professional assistance to treat steam, water, brine, etc., heat transfer
fluids.
Due to some relatively recent problems, the last recommendation
requires clarification. The need for jacket cleaning most often is
predicated on a loss of heat transfer capability. A frequently asked
question in when to do the cleaning. For this determination, it is
strongly recommended that a heat/cool run check be made when the
equipment is new. This is simply done by noting the time it takes to
heat the vessel-inhibited water between two temperatures, e.g.,
70190°F and then cooling over the same range. When this time
increases to whatever level is critical for the specific operation,
corrective actions should be taken in the following sequence:
Make sure no changes have been made in the piping system and, more
importantly, that the hookup was originally done according to the
manufacturer's recommendations. Especially important is the
recommendation to install a valve in the diaphragm drain located at
the bottom of the diaphragm ring near the BON (Fig. 2). To avoid
sludge buildup, which in addition to promoting corrosion can also
enhance the BON thermal problem reviewed previously, this valve
should be periodically opened. A better recommendation is to pipe the
diaphragm drain directly into either the trap or coolant lines, thereby
effecting a nonmanual, more continuous flush.
Check the heating/cooling source, e.g., boiler, to ensure that the
output has not changed and meets the vendor specifications.
Check all lines for closed valves, leakage, insulation removal, etc.
If steam is used for the heating media, check the trap for adequate
discharge.
Check the vent system for proper relief of possible vapor buildup.
Check the inlet impingement baffles for blockage and/or
erosion/corrosion.
If these checks fail to uncover a problem, the use of a fast-flush
procedure may next prove highly beneficial, especially for loosely
adherent buildups. With both the diaphragm drain and trap nozzle
open, admit water through the top inlet nozzle staying within the
Code-allowable pressure/temperature limits. If possible, reverse the
flushing. This procedure usually removes considerable amounts of
loosely adhering material.
If the above steps are ineffective, a chemical cleaning is indicated.
The more common types of jacket deposits and the associated
recommended cleaning recipes/procedures follow:
Algae-based. A commercial sodium hypochlorite solution (1015 wt
%) is
Page 541
readily available from swimming pool distributors. Two gallons of
this solution is usually added to approximately 75 gallons of jacket
water and the mix circulated for 4 h at ambient temperature. (Note:
This solution will decompose at higher temperatures).
Organic. For oily-type deposits, recirculating hot water at 150°F used
in conjunction with a good emulsifying detergent has proven
effective.
Hard water deposits, e.g., calcium and magnesium salts. This is a
potential problem area as acidic media, which can cause fishscaling,
are usually required for removal. Fortunately, the deposit buildup is
usually quite small and does not present a very serious blockage
problem. For these cases, only inhibited acids may be used and the
degree of inhibition must be closely monitored. Strict adherence to the
vendor's procedure must also be followed. Always use the lowest
temperature and shortest contact time. The jacket must be completely
neutralized after treatment.
Iron oxides. These deposits are the most common and most
troublesome. The troublesome part is traceable to the fact that the
deposits can readily build up to the point where blockage between the
jacket and shell/head takes place. If these deposits cannot be removed
within the allowed cleaning time (usually less than 24 h), the acid,
which has been absorbed by the deposit and cannot easily be
neutralized, may remain in contact with the glass-backed steel for
prohibitively long times leading eventually to possible fishscale
damage. At present, there is no economically feasible ways to
establish the degree of the blockage. Consequently, it is very
important to abide by the following Rule #23: Never use acidinhibited cleaners for iron oxide(s) removal. Fortunately, there are
now some essentially neutral pHtype cleaners that not only clean but
also tend to passivate the metal against further oxidation. While the
cleaning cycle is usually extended over the inhibited acids, their
inherent safety is an extremely strong recommendation. The glassed
steel vendors should be consulted for specific product
recommendations.
K
SpillageAppearance
The fishscale nomenclature fits exactly the general appearance. While
the initial damage is usually the size of a small perch scale, i.e., 1/8 in.
diameter, continued acid exposure can lead eventually to "whale
scale"sized damage. The bottom head area, where blockage is most
likely, the top head, under insulation, and the outside perimeter of
nozzles where corrosives have leaked to the metal are
Page 542
common problem areas. Provided that product has not accumulated,
the damage areas always contact at 5 kV stabilized.
XI
Inspections
The need for a review of inspection procedures is based on the
difficulties associated with some of the above-discussed damage
interpretations, e.g., the LOG v. corrosion rate problem, especially as
regards defining the exact cause for the damage along with estimation
of the equipment service life. Once the damage cause has been
determined precisely, the appropriate corrective actions can then be
taken. Another important corollary fact is that early identification of
damage equates directly to the installation of highly reliable repairs
that are both small and simple. Rule #24: For glassed steel repairs,
always apply the KISS principle, i.e., Keep It Small and Simple.
To do an effective job of inspecting glassed steel equipment, five
steps must be carried out in the following order:
1. Mapping and marking. The equipment must be gridded or mapped
in a consistent fashion. This will allow thickness readings to be
matched between inspection dates giving valuable insight into highdeterioration areas and also allow for the important calculation of the
exact deterioration rates. Useful inspection sheets and the associated
procedures can be obtained from the vendor.
2. Visual/microscopic observations. In order to establish a possible
damage progression, all visual and microscopic data must be recorded
exactly on the inspection sheet.
3. Thickness measurements. This is an extremely important
measurement that is often neglected in glassed steel inspection. It was
noted earlier that there is no relationship between the rate of corrosive
attack and the LOG. Also, once the glass has been etched, it is
impossible visually to ascertain the glass thickness and therefore
estimate the service life of the equipment. This can only be done by
using a calibrated thickness meter. In this regard, a frequently asked
question is, What is the minimal allowable glass thickness for
containment of corrosive chemicals? The answer is tied closely to the
ground coat composition, which possesses very poor corrosion
resistivity. Consequently, when the ground coat thickness has been
reached, processing with corrosive species must cease. As was
mentioned in the glass fusion/firing section, two layers of ground coat,
each approximating 10 mils in thickness, are applied to most vessels.
This would mean that the minimal service thickness should be 20
mils. Unfortunately, this idealized situation does
Page 543
not occur. In Fig. 4, under "Chemical," is the word pullthrough. This
refers to the fact that the layers of ground coat, because of a
combination of viscosity and surface tension effects, do no lay flat. In
cross-section, they resemble a mountain range with some of the peaks
approaching just under 30 mils. In providing for a safety factor, the
30-mil value is commonly used as the minimum thickness allowed for
corrosive service containment. Once this level has been reached, two
options remain, i.e., use the equipment in less corrosive environments
or reglass. As the glass thickness loss is usually extensive, repairs are
not an option.
4. Electric testing. As the ground coat layers possess a dielectric
strength similar to that of the cover coats, i.e., 500 V/mil, electric
testing alone cannot identify when they are reached. This adds more
support to the thickness measurement requirement. As was pointed
out previously, only a 5 kV stabilized system should be used. It is
extremely important to carry out the thickness measurements first,
followed by the electric testing. This prevents the possibility of any
residual charge that may have built up on the glass during electric
testing from discharging through the sensitive circuitry of the
thickness meter.
5. Accurate record keeping. The previous steps, even when carried out
with great care, are largely meaningless in carrying out a complete
damage analysis unless brought together via a consistent, periodically
updated record keeping format.
Rule #25: Periodic, five-step inspections must be carried out on all
glassed steel equipment.
Another method for vessel inspection that is gaining great favor in the
CPI, especially for those companies adhering to the KISS principle, is
the use of continuous-monitoring, early-warning, fault detection
instrumentation. This system is essentially a current measuring system
based on the galvanic cell principle that detects continuity from the
vessel contents to the metal opposite the glass coating. Note carefully
that this system will not break down thin glass or give any indication
of progressive damage of a nonelectrical contacting nature. It is,
therefore, not a replacement for the five-step inspection procedure but
a supplement to it. The circuit includes a cathodic tantalum alloy
sensing probe located either on the head of a glassed flush valve or on
the inside diameter of a glassed spacer ring (both located in the lowliquid-level BON area) in series with a current detecting instrument
and then with the anodic, grounded metal of the vessel. For the system
to function, the specific resistance of the vessel solution must be less
than 200 ohm-cm. When the electrolyte contacts any of the substrate
metal via any means, e.g., glass fracture, glass repair, or gasket
leakage, the circuit
Page 544
is complete and current will register on the meter. Older systems
could monitor up to 10 vessels on a single detector unit but were
limited to a direct visual indication on the meter. Newer units monitor
only six vessels but have a preamplifier that allows for both "in-thevicinity" audio and/or visual outputs. Both units have potentiometric
compensators to balance out effects from tantalum repairs. For
nonconducting recipes, it is recommended that a suitably conductive
solution be used periodically to check out complete vessel integrity. A
portable, single-channel unit is available for these intermittent
situations. Rule #26: An early-warning fault detection system should
always be used with chemistries that are extremely corrosive to steel,
e.g., bromine.
Page 545
19
Cathodic Protection
Philip A. Schweitzer
Fallston, Maryland
When dissimilar metals are in physical or electrical contact (the latter
via a conductive electrolyte) such as by process fluid or soil, galvanic
corrosion can take place. The galvanic corrosion process is similar to
the action of a simple DC cell in which the more active metal
becomes an anode and corrodes, whereas the less active metal
becomes a cathode and is protected. It is possible to predict which
metals will corrode when in contact with others based on the galvanic
series shown in Table 1.
All metals or alloys have certain built-in properties that cause them to
react as an anode or a cathode when in contact with dissimilar metals
or alloys. Whether a particular material will react as a cathode or an
anode can be determined from their relative positions in the galvanic
series. The further apart the two materials are from each other in the
galvanic series, with all other factors being equal, the greater the rate
of corrosion. The material closest to the anodic end will be the one to
corrode. For example, if tin and zinc were in contact the zinc would
corrode, whereas if tin and copper were in contact the tin would
corrode.
The rate of attack is also affected by the relative size of the material
and the specific electrolyte present. A small anode area in contact with
a large cathode are will result in a rapid severe attack. Conversely, a
large anode area in contact with a small cathode area will lessen the
rate of galvanic attack since the same total
Page 546
Table 1 Galvanic Series
Anodic End
Magnesium
Magnesium alloys
Zinc
Aluminum 5052
Aluminum 6061
Monel
Silver solder
Nickel (passive)
Inconel (passive)
Ferritic stainless
(passive)
Cadmium
Austenitic stainless
(passive)
Aluminum AA2017
Titanium
Iron and carbon steel
Lead-tin solder
Copper steel
Lead
46% chromium steel
Tin
Ferritic stainless (active) 400 Nickel (active)
series
Austenitic stainless (active) Inconel (active)
18-8 series
Hastelloy C (active)
Hastelloy C (passive)
Brasses
Silver
Copper
Graphite
Bronzes
Gold
Cupro-nickel alloys
Platinum
Cathodic End
emf driving force of corrosion will be spread out over a larger area. In
addition, the higher the degree of ionization of the electrolyte, the
greater the rate of attack.
Galvanic corrosion can also take place when metals having the same
analysis have different surface conditions and an electrolyte is present.
In general, the formation of a corrosion cell is induced by the
nonuniformity of the surface condition, such as by defects in the
surface oxide film, localized distribution of elements, and the
difference in crystal face or phase. These nonuniformities of surface
cause the potential difference between portions of the surface and
thereby promote the formation of a corrosion cell.
Galvanic corrosion can be stopped by means of cathodic protection,
which is an electrochemical technique. It can be applied to metals
immersed in water, buried in soil, or in contact with electrolytes in
process application. Cathodic protection consists of a cathodic current
flowing through the metalelectrolyte interface favoring the reduction
reaction over the anodic metal dissolution. The entire structure works
as a cathode.
This electrochemical technique was developed by Sir Humphry Davy
in 1824. The British Admiralty had blocks of iron attached to the hulls
of copper-sheathed vessels to provide cathodic protection.
Unfortunately, cathodically protected copper is subject to fouling by
marine life, which reduced the speed of vessels under sail and forced
the Admiralty to discontinue the
Page 547
practice. Unprotected copper provides a sufficient number of copper
ions to poison fouling organisms. However, the corrosion rate of the
copper had been appreciably reduced.
In 1829 Edmund Davy was successful in protecting the iron portions
of buoys by using zinc blocks, and in 1840 Robert Mallet produced a
zinc alloy that was particularly suited as a sacrificial anode. As steel
hulls replaced wooden hulls the fitting of zinc slabs to the steel hulls,
to provide cathodic protection, became standard practice.
In 1950 the Canadian Navy determined that the proper use of
antifouling paints in conjunction with corrosion-resistant paints made
cathodic protection of ships feasible and could reduce maintenance
costs.
There are two methods by which cathodic protection can be
accomplished. One is by coupling the structure with a more active
metal, such as zinc or magnesium. This produces a galvanic cell in
which the active metal works as an anode and provides a flux of
electrons to the structure. The structure then becomes the cathode and
is protected whereas the anode is destroyed progressively, and is
called a sacrificial anode.
The second method is to impress a direct current between an inert
anode and the structure. The structure receives the excess of electrons
which protect it. About 19101912 the first application of cathodic
protection by means of an impressed electric current was undertaken
in England and the United States. Since that time the general use of
cathodic protection has been widespread. There are thousands of miles
of buried pipelines and cables that are protected in this manner. This
form of protection is also used for water tanks, submarines, canal
gates, marine piping, condensers, and chemical equipment.
I
Sacrificial Anodes
In cathodic protection, the structure to be protected must receive a
cathodic current flow so that it operates as a cathode. The need for an
external DC current to accomplish this can be eliminated by selecting
an anode constructed of a metal which is more active in the galvanic
series than the metal to be protected. A galvanic cell will be
established with the current direction as required. These sacrificial
anodes are usually composed of magnesium or magnesium-based
alloys. On occasion, zinc or aluminum has been used.
Magnesium is much more active than steel; it has a greater tendency
to ionize and its potential is more active than iron. The open-circuit
potential difference between magnesium and steel is about 1 V. This
means that one anode can protect only a limited length of pipeline.
This low voltage can have an advantage over higher impressed
voltages in that the danger of overprotection to some portions of the
structure is less and because the total current per anode is limited, the
danger of stray-current damage to adjoining metal structures is
reduced.
Page 548
Magnesium rods have also been placed in steel hot water tanks to
increase their life. The greatest degree of protection is afforded in
''hard" waters since the degree of conductivity is greater than in "soft"
waters.
II
Sacrificial Anode Requirements
To provide cathodic protection, a current density of a few
milliamperes (mA) is required. In order to determine the anodic
requirement it is necessary to know the energy content of the anode
and its efficiency. With this information it is possible to determine the
size of the anode required, its expected lifetime, and to determine the
number of anodes required.
The three most common metals used as sacrificial anodes are
magnesium, zinc, and aluminum. The energy content and efficiency of
these metals are shown below:
Practical
energy
Theoretical Anodic constant
energy content efficiency (A h/lb)
Metal
(A h/lb)
(%)
(PE)
Magnesium
1000
50
500
Zinc
370
90
333
Aluminum
1345
60
810
Zinc is more economical to use than magnesium, but because of the
relatively small cell voltage it produces, it is primarily useful under
special circumstances, such as to protect ships in seawater or to
prevent the corrosion of systems with low current requirements.
Although magnesium is more expensive than zinc and is consumed
faster than zinc or aluminum, it does provide the largest cell voltage
and the largest current. Care must be taken not to use aluminum in
environments having a pH of ³ 8 since alkaline conditions will
produce a rapid self-corrosion of aluminum.
In determining anodic requirements to provide cathodic protection
several calculations are required. The number of pounds of metal
required to provide a current of 1 A for a year is calculated as follows:
For magnesium this would be
Page 549
The number of years (YN) for which 1 lb of metal can produce a
current of 1 mA is determined from the following equation:
For magnesium this would be
The current density requirements for cathodic protection is on the
order of a few mA. The life expectancy (L) of an anode of W lb,
delivering a current of 1 mA, is calculated as follows:
For magnesium this would be
which is based on 50% anodic efficiency. Since actual efficiencies
tend to be somewhat less, it is advisable to apply a safety factor and
multiply the result by 0.75.
The current required to secure protection of a structure and the
available cell voltage between the metal structure and sacrificial
anode determine the number of anodes required. This can be
illustrated by the following example:
Assume that an underground pipeline has an external area of 200 ft2
and a soil resistivity of 600 ohm-cm. Field tests indicate that 6 mA/ft2
is required for protection. To provide protection for the entire pipeline
(6 mA/ft2) (200 ft2) = 1200 mA is required. Magnesium anodes used
in this particular soil have a voltage of -1.65 V or a galvanic cell
voltage of
Therefore the resistance is
As the number of anodes are increased the total resistance of the
system decreases. Each anode that is added provides a new path for
current flow, parallel to the existing system. The relationship between
the resistance of the system and the number of anodes is shown in the
Sunde equation:
Page 550
where
R = resistance in ohms
P = soil resistivity in ohm-cm
N = number of anodes
L = anode length (ft)
d = diameter of anode (ft)
S = distance between anodes (ft)
Figure 1 shows typical plotting of the results of this equation.
Different anodic shapes will have different curves.
III
Impressed Current Systems
For impressed current systems the source of electricity is external. A
rectifier converts high-voltage AC current to a low-voltage DC
current. This direct current is impressed between buried anodes and
the structure to be protected.
Figure 1
Plot of the Sunde equation.
Page 551
It is preferred to use inert anodes, which will last for the longest
possible times. Typical materials used for these anodes are graphite,
silicon, titanium, and niobium plated with platinum.
For a given applied voltage, the current is limited by electrolyte
resistivity and by the anodic and cathodic polorization. With the
impressed current system it is possible to impose whatever potential is
necessary to obtain the current density required by means of the
rectifier.
Electric current flows in the soil from the buried anode to the
underground structure to be protected. Therefore the anode must be
connected to the positive pole of the rectifier and the structure to the
negative pole. All cables from the rectifier to the anode and to the
structure must be electrically insulated. If not, those from the rectifier
to the anode will act as an anode and deteriorate rapidly whereas those
from the rectifier to the structure may pick up some of the electric
current, which would then be lost for protection.
A
Current Requirements
The specific metal and the environment will determine the current
density required for complete protection. The applied current density
must always exceed the current density equivalent to the measured
corrosion rate under the same conditions. Therefore as the corrosion
rate increases, the impressed current density must be increased to
provide protection.
Factors which affect current requirements are as follows:
1. Nature of the electrolyte
2. Soil resistivity
3. Degree of aeration
The more acid the electrolyte, the greater will be the potential for
corrosion and the greater will be the current requirement. Soils that
exhibit a high resistance require a lower cathodic current to provide
protection. In areas of violent agitation or high aeration, an increase in
current will be required. The required current to provide cathodic
protection can vary from 0.5 to 20 mA/ft2 of bare surface.
Field testing may be required to determine the necessary current
density to provide cathodic protection in a specific area. These testing
techniques will only provide an approximation. After completion of
the installation, it will be necessary to conduct a potential survey and
make the necessary adjustments to provide the desired degree of
protection.
B
Anode Materials and Backfill
Although it is generally preferred to use inert anodes, it is also
possible to use scrap iron. Scrap iron is consumed at a considerably
faster rate than graphite or other inert anode materials. The advantage
of scrap iron is its lower initial cost and lower operating cost because
its power requirements are less. In areas where
Page 552
replacement poses a problem the cost of the use of the more inert
anodes outweighs the reduced cost of the scrap iron.
Platinum clad of 2% silver-lead electrodes have been used for the
protection of structures in seawater and are estimated to last 10 years,
whereas sacrificial magnesium anodes have a life of 2 years.
Since the effective resistivity of soil surrounding an anode is limited
to the immediate area of the electrode, this local resistance is usually
reduced by using backfill. The anode is usually surrounded by a thick
bed of coke mixed with three or four parts of gypsum to one part of
sodium chloride. The consumption of the anode itself is reduced
somewhat since the coke backfill carries part of the current. Backfill is
not required when the anode is immersed in a river bed, lake, or
ocean.
C
Testing for Completeness of Protection
Once the system has been installed it must be tested for completeness
of protection. The preferred method is to take potential measurements.
By measuring the potential of the protected structure, the degree of
protection, including overprotection, can be determined. The basis for
this determination is the fundamental concept that cathodic protection
is complete when the protected structure is polorized to the opencircuit anodic potential of local action cells.
The reference electrode is placed as close as possible to the protected
structure to avoid or minimize an error caused by internal resistance
(IR) drop through the soil. For buried pipelines a compromise location
is directly over the buried pipe at the soil surface because cathodic
protection currents flow mostly to the lower surface and are minimum
at the upper surface of the pipe buried a few feet below the soil
surface.
The potential for steel is equal to -0.85 V vs. the copper-saturated
copper sulfate half-cell, or 0.53 V on the standard hydrogen scale. The
theoretical open circuit anodic potential for other metals may be
calculated using the Mernst equation. Several typical calculated
values are shown below:
E°
Metal (V)
Iron 0.400
Copper-0.337
Zinc 0.763
Lead 0.126
Solubility
product
M(OH)2
1.8 × 10-15
1.6 × 10-19
4.5 × 10-17
4.2 × 10-15
OH2
OVs Cu-CuSO4
scale reference electrode
(V)
(V)
-0.59
-0.91
0.16
-0.16
-0.93
-1.25
-0.27
-0.59
Overprotection of steel structures to a moderate degree does not cause
any problems. The primary disadvantages are waste of electric power
and increased
Page 553
consumption of auxiliary anodes. When overprotection is excessive,
hydrogen can be generated at the protected structure in sufficient
quantities to cause blistering of organic coatings, hydrogen
embrittlement of the steel, or hydrogen cracking.
Overprotection of systems with amphoteoric metals (e.g., aluminum,
zinc, lead, tin) will damage the metal by causing increased attack
instead of reduction of corrosion. This stresses the need for making
potential measurements of protected structures.
IV
Use with Coatings
It is advantageous to use insulating coatings with sacrificial anodes or
impressed current systems when supplying cathodic protection. These
coatings need not be pore-free because the protective current flows
preferentially to the exposed metal areas, which require the protection.
Coatings are useful in distributing the protective current, in reducing
total current requirements, and in extending the life of the anode.
Compared to a bare pipeline the current distribution in a coated
pipeline is greatly improved, the total number of anodes required is
reduced, and the total current required is less. In addition, one anode
can protect a much longer section of pipeline. For example, one
magnesium anode is capable of protecting approximately 100 ft (30
m) of a bare pipeline, whereas the same anode can provide protection
for approximately 5 miles of a coated pipeline.
In a hot water tank coated with glass or an organic coating, the life of
the magnesium anode is increased and more uniform protection is
supplied to the tank. Without the coating the tendency is for excess
current to flow to the side and insufficient current flows to the top and
bottom.
Because of these factors cathodic protection is usually provided with
coated surfaces.
V
Economics
The installation of cathodic protection systems has made it
economically feasible to transport oil and high-pressure natural gas
across North America by
1. Guaranteeing there will be no corrosion on the soil side of the pipe
2. Permitting the use of thinner-walled pipe
3. Eliminating the need for external corrosion allowance
4. Reducing maintenance costs
5. Permitting longer operating periods between routine inspections
and maintenance periods
The cost of the cathodic protection system is more than recovered as a
result of the above savings. Similar savings and advantages have been
realized on other types of installations where cathodic protection
systems have been installed.
Page 555
20
Corrosion Inhibitors
Philip A. Schweitzer
Fallston, Maryland
Corrosion of metallic surfaces can be reduced or controlled by the
addition of chemical compounds to the corrodent. This form of
corrosion control is called inhibition and the compounds added are
known as corrosion inhibitors. These inhibitors will reduce the rate of
either anodic oxidation or cathodic reduction, or both. The inhibitors
themselves form a protective film on the surface of the metal. It has
been postulated that the inhibitors are adsorbed into the metal surface
either by physical (electrostatic) adsorption or chemosorption.
Physical adsorption is the result of electrostatic attractive forces
between the organic ions and the electrically charged metal surface.
Chemosorption is the transfer, or sharing of the inhibitor molecule's
charge to the metal surface, forming a coordinate-type bond. The
adsorbed inhibitor reduces the corrosion rate of the metal surface
either by retarding the anodic dissolution reaction of the metal, or by
the cathodic evolution of hydrogen, or both.
Inhibitors can be used at pH values of acid from near neutral to
alkaline. They can be classified in many different ways according to
1. Their chemical nature (organic or inorganic substances)
2. Their characteristics (oxidizing or nonoxidizing compounds)
3. Their technical field of application (pickling, descaling, acid
cleaning cooling water systems, and the like).
Page 556
The most common and widely known use of inhibitors is their
application in automobile cooling systems and boiler feedwaters.
I
Inhibitor Evaluation
Because there may be more than one inhibitor suitable for a specific
application, it is necessary to have a means of comparing the
performance of each. This can be done by determining the inhibitor
efficiency according to the following correlation:
where
Ieff = efficiency of inhibitor, %
RO = corrosion rate of metal without inhibitor present
Ri = corrosion rate of metal with inhibitor present
RO and Ri can be determined by any of the standard corrosion testing
techniques. The corrosion rate can be measured in any unit, such as
weight loss (mpy), as long as units are consistent across both tests.
II
Classification of Inhibitors
Inhibitors can be classified in several ways as indicated previously.
We will classify and discuss inhibitors under the following headings:
1. Passivation inhibitors
2. Organic inhibitors
3. Precipitation inhibitors
A
Passivation Inhibitors
Passivation inhibitors are chemical oxidizing materials such as
chromate (
) and nitrate ( ) or substances such as Na3PO4 or
NaBrO7. These materials favor adsorption on the metal surface of
dissolved oxygen.
This is the most effective and consequently the most widely used type
of inhibitor. Chromatics are the least expensive inhibitors for use in
water systems and are widely used in the recirculation-cooling
systems of internal combustion engines, rectifiers, and cooling towers.
Sodium chromate, in concentrations of 0.040.1% is used for this
purpose. At higher temperatures or in freshwater that has chloride
concentrations above 10 ppm higher concentrations are required. If
necessary, sodium hydroxide is added to adjust the pH to a range of
7.59.5. If the concentration of chromate falls below a concentration of
0.016% corrosion
Page 557
will be accelerated. Therefore it is essential that periodic colorimetric
analysis be conducted to prevent this from occurring.
Recent environmental regulations have been imposed on the use of
chromates. They are toxic and on prolonged contact with the skin can
cause a rash. It is usually required that the Cr6+ ion be converted to
Cr3+ before discharge. The Cr3+ ion is insoluble and can be removed
as a sludge whereas the Cr6+ ion is water-soluble and toxic. Even so,
the Cr3+ sludge is classified as a hazardous waste and must be
constantly monitored. Because of the cost of conversion of the
chromate ions, the constant monitoring required, and the disposal of
the hazardous wastes, the economics of the use of these inhibitors are
not as attractive as they formerly were.
Because most antifreeze solutions contain methanol or ethylene glycol
the chromates cannot be used in this application due to the fact that
the chromates have a tendency to react with organic compounds. In
these applications borax (Na2B4O710H2O) to which has been added
sulfonated oils to produce an oily coating and mercaptobenzothiazole
are used. The latter material is a specific inhibitor for the corrosion of
copper.
Nitrites are also used in antifreeze-type cooling water systems because
they have little tendency to react with alcohols or ethylene glycol.
Since they are gradually decomposed by bacteria they are not
recommended for use in cooling tower waters. Another application for
nitrites is as a corrosion inhibitor of the internal surfaces of pipelines
used to transport petroleum products or gasoline. Such inhibition is
accomplished by continuously injecting a 530% sodium nitrite
solution into the line.
At lower temperatures, such as in underground storage tanks, gasoline
can be corrosive to steel as dissolved water is released. This water, in
contact with the large quantities of oxygen dissolved in the gasoline,
corrodes the steel and forms large quantities of rust. The sodium
nitrite enters the water phase and effectively inhibits corrosion.
The nitrites are also used to inhibit corrosion by cutting oilwater
emulsions used in the machining of metals.
Passivation inhibitors can actually cause pitting and accelerate
corrosion when concentrations fall below minimum limits. For this
reason it is essential that constant monitoring of the inhibitor
concentration be performed.
B
Organic Inhibitors
These materials build up a protective film of adsorbed molecules on
the metal surface which provides a barrier to the dissolution of the
metal in the electrolyte. Since the metal surface covered is
proportional to the inhibitor concentrates, the concentration of the
inhibitor in the medium is critical. For any specific inhibitor in any
given medium there is an optimal concentration. For example, a
concentration of
Page 558
0.05% sodium benzoate, or 0.2% sodium cinnamate, is effective in
water that has a pH of 7.5 and contains 17 ppm sodium chloride or
0.5% by weight of ethyl octanol.
The corrosion due to ethylene glycol cooling water systems can be
controlled by the use of ethanolamine as an inhibitor.
C
Precipitation Inhibitors
Precipitation inhibitors are compounds that cause the formation of
precipitates on the surface of the metal, thereby providing a protective
film. Hard water that is high in calcium and magnesium is less
corrosive than soft water because of the tendency of the salts in the
hard water to precipitate on the surface of the metal and form a
protective film.
If the water pH is adjusted in the range of 56, a concentration of
10100 ppm of sodium pyrophosphate will cause a precipitate of
calcium or magnesium orthophosphate to form on the metal surface
providing a protective film. The inhibition can be improved by the
addition of zinc salts.
III
Inhibition of Acid Solution
The inhibition of corrosion in acid solutions can be accomplished by
the use of a variety of organic compounds. Among those used for this
purpose are triple-bonded hydrocarbons; acetylenic alcohols,
sulfoxides, sulfides, and mercaptans; aliphatic, aromatic, or
heterocyclic compounds containing nitrogen; and many other families
of simple organic compounds and of condensation products formed by
the reaction between two different species such as amines and
aldehydes.
Incorrect choice or use of organic inhibitors in acid solutions can lead
to corrosion stimulation and/or hydrogen penetration into the metal. In
general, stimulation of corrosion is not related to the type and
structure of the organic molecule. Stimulation of acid corrosion of
iron has been found with mercaptans, sulfoxides, azole and triazole
derivatives, nitrites, and quinoline. This adverse action depends on the
type of acid. For example, bis(4-dimethylamino-phenyl)
antipyrilcarbinol and its derivatives at a 10-4 M concentration
inhibited attack of steel in hydrochloric acid solutions but stimulated
attack in sulfuric solutions. Much work has been done studying the
inhibiting and/or stimulating phenomena of organic compounds on
ferrous as well as nonferrous metals. Organic inhibitors have a critical
concentration value, below which inhibition ceases and stimulation
begins. Therefore it is essential that when organic inhibitors are used
constant monitoring of the solution should take place to ensure that
the inhibitor concentration does not fall below the critical value.
Page 559
IV
Inhibition of near Neutral Solutions
Because of the differences in the mechanisms of the corrosion process
between acid and near-neutral solutions, the inhibitors used in acid
solutions usually have little or no inhibition effect in near-neutral
solutions. In acid solutions the inhibition action is due to adsorption
on oxide-free metal surfaces. In these media the main cathodic
process is hydrogen evolution.
In almost neutral solutions the corrosion process of metals results in
the formation of sparingly soluble surface products such as oxides,
hydroxides, or salts. The cathodic partial reaction is oxygen reduction.
Inorganic or organic compounds as well as chelating agents are used
as inhibitors in near-neutral aqueous solutions. Inorganic inhibitors
can be classified according to their mechanisms of action:
1. Formation and maintenance of protective films can be
accomplished by the addition of inorganic anions such as
polyphosphates, phosphates, silicates, and borates.
2. Oxidizing inhibitors such as chromates and nitrites cause selfpassivation of the metallic material. It is essential that the
concentration of these inhibitors be maintained above a ''safe" level. If
not, severe corrosion can occur as a result of pitting or localized attack
caused by the oxidizer.
3. Precipitation of carbonates on the metal surfaces forming a
protective film. This usually occurs due to the presence of Ca2+ and
Mg2+ ions usually present in industrial waters.
4. Modification of surface film protective properties is accomplished
by the addition of Ni+2, Co2+, Zn2+, or Fe2+.
The sodium salts of organic acids such as benzoate, salicylate,
cinnamate, tartrate, and azelate can be used as alternatives to the
inorganic inhibitors, particularly in ferrous solutions. When using
these particular compounds in solutions containing certain anions such
as chlorides or sulfates, the inhibitor concentration necessary for
effective protection will depend on the concentration of the aggressive
anions. Therefore the critical pH value for inhibition must be
considered rather than the critical concentration. Other formulations
for organic inhibition of near-neutral solutions are given in Table 1.
Chelating agents of the surface-active variety also act as efficient
corrosion inhibitors when insoluble surface chelates are formed.
Various surface-active chelating agents recommended for corrosion
inhibition of different metals are given in Table 2.
Page 560
Table 1 Organic Inhibitors for Use in Near-Neutral
Solutions
Type of
metal
Inhibitor
protected
Organic phosphoruscontaining compounds, salts Ferrous
of amino-methylenephosphonic acid,
hydroxyethylidenediphosphonic acid,
phosphenocarboxylic acid, polyacrylate, polymethacrylate
Zinc, zinc
Borate or nitrocinnamate anions (dissolved
alloys
oxygen in solution required)
Acetate or benzoate anions
Aluminum
Heterocylic compounds such as benzotriazole and Copper
its derivatives, 2-mercaptobenzothiazole, 2and
mercaptobenzimidazole
copperbased
alloys
V
Inhibition of Alkaline Solutions
All metals whose hydroxides are amphoteric and metals covered by
protective oxides that are broken in the presence of alkalies are
subject to caustic attack. Localized attack may also occur as a result of
pitting and crevice formation.
Organic substances such as tannions, gelatin, saponin, and agar-agar
are often used as inhibitors for the protection of aluminum, zinc,
copper, and iron. Other materials which have also been found
effective are thiourea, substituted phenols and naphthols, b-diketones,
8-hydroxyquinoline, and quinalizarin.
VI
Temporary Protection with Inhibitors
Occasions arise when temporary protection of metallic surfaces
against atmospheric corrosion is required. Typical instances are in the
case of finished metalTable 2 Chelating Agents Used as Corrosion Inhibitors in
Near-Neutral Solutions
Type of metal
Chelating agent
protected
Alkyl-catechol derivatives, sarcosine
Steel in
derivatives, car-boxymethylated fatty amines, industrial
and mercaptocarboxylic acids
cooling
systems
Azo compounds, cupferron, and rubeanic
Aluminum
acid
alloys
Azole derivatives and alkyl esters of
Zinc and
thioglycolic acid
galvanized
steel
Oximes and quinoline derivatives
Copper
Cresolphthalexon and thymolphthalexon
Titanium in
derivatives
sulfuric acid
solutions
Page 561
lic materials or of machinery parts during transportation and/or
storage prior to use. When ready for use the surface treatment or
protective layer can be easily removed.
It is also possible to provide protection by controlling the aggressive
gases or by introducing a vapor phase inhibitor. This latter procedure
can only be accomplished in a closed environment such as sealed
containers, museum showcases, or similar enclosures.
Organic substances used as contact inhibitors or vapor inhibitors are
compounds belonging to the following classes:
1. Aliphatic, cycloaliphatic, aromatic, and heterocyclic amines
2. Amine salts with carbonic, carbamic, acetic, benzoic, nitrous, and
chromic acids
3. Organic esters
4. Nitro derivatives
5. Acetylenic alcohols
VII
Summary
Corrosion inhibitors are usually able to prevent general or uniform
corrosion. However, they are very limited in their ability to prevent
localized corrosion such as pitting, crevice corrosion, galvanic
corrosion, dezincification, or stress corrosion cracking. Additional
research work is being undertaken in the use of inhibitors to prevent
these types of corrosion. The importance of these studies is realized
when it is taken into account that only approximately 30% of all
failures due to corrosion in chemical plants result from general
corrosion. The remaining 70% are due to stress corrosion cracking,
corrosion fatigue, pitting, and erosion-corrosion. Attack on metals by
general corrosion can be predicted and life spans of the equipment
determined and/or the corrosion rates reduced by use of inhibitors.
This is not the case with other types of corrosion.
The use of inhibitors can be advantageous in certain cases. However,
before using inhibitors it is essential that the efficiency of the inhibitor
to be used be determined to ensure that inhibition will take place.
Page 563
21
Painting for Protection
Walter M. McMahon
La Habra Heights, California
A chapter on plant maintenance painting would be of epic size if it
covered every industry specifically. How many possibilities are there?
Petroleum refining, pulp and paper, metal processing, metal refining,
sewage treatment, power production, water treatment, food
processing, and brewing are only the start of a long list. It is hoped
that at least some of the following will have application to readers'
concerns and problems.
Before going further, let us consider why companies have
maintenance painting programs. Common reasons are as follows:
corrosion control, safety, efficiency, public relations, and employee
morale. When corrosion control is not some part of the motivation,
painting is much simpler, infinitely cheaper, and probably not for
maintenance purposes. There are five "W" words other than "why"
that we should consider in our planning:
1. "What" are we going to paint?
2. "Where" is it located?
3. "When" is the best time to do it?
4. "Who" will do the painting?
5. "Which" products will be used?
In most instances the "what" are steel surfaces at some temperature
that
Page 564
may or may not fluctuate. However, galvanizing, aluminum, concrete,
plaster, wallboard, wood, and even stainless steel are often part of
coatings' maintenance projects.
"Where" is of extreme importance because it is the environment, i.e.,
the sum of precipitation, air temperatures, relative humidities, and
chemical and radiation exposures. It can also govern timing, methods
of staging, and application techniques.
"When" is not just anytime. Exterior painting must take the weather
into consideration. Interior painting must be scheduled so that there is
minimum interference with production or other operations.
"Who" will do the painting is a major decision. Although plant
personnel can sometimes be used, particularly on minor projects, the
hiring of an experienced and reputable contractor is almost always a
better choice. Selection of a contractor should not be conducted on the
basis of the lowest bid. Some checking of their histories in terms of
job completion, payment of bills, lawsuits, and customer satisfaction
is recommended.
We have now arrived at "which" products to use. Let us start with a
simple, nontechnical definition of paint. Paint is a liquid that upon
application converts to a solid film with protective and cosmetic
properties. Although mixtures more or less meeting this definition
were known to exist several thousand years before Christ, modern
paints have their origin in the Industrial Revolution. Prior to the 1870s
there were not a lot of structures whose values could be enhanced by a
coat of paint. The Industrial Revolution created extensive markets for
decorative and protective coatings, respectively. The rapid bursts of
scientific and technical development following in the 20th century
brought opportunities which could not have been imagined by earlier
paint makers. Modern manufacturing processes, power production,
transportation, product distribution, and living styles created the need
for a wide variety of specialized paints. However, superficially they
are the same as the concoctions of the earliest paintmakers, i.e.,
colorants dispersed in a liquid which converts to a dry film after its
application to a surface. In trade jargon, "dispersing pigment in
vehicle makes paint!" Breaking this down further, we have
Nonvolatile vehicle. The film former, or binder, which is made up of
primary resins, plasticizers, and minor amounts of nonvolatile
additives other than pigment.
Volatile vehicle. Solvents, diluents, water.
Pigment. Colorants, opacifiers, nonsoluble fillers.
Although the roles of the pigment and the nonvolatile vehicle are
obvious, the need for the volatile vehicle is not always crystal clear to
the paint customer or to environmental regulation agencies. Without
the volatile component, paint is a solid, semisolid, or very viscous
liquid that cannot be applied by brush, roller,
Page 565
or spray gun. Few people other than paint chemists give much thought
to the consistencies of paint. Although this characteristic is often
spoken of as a paint's viscosity, a better term is rheology, which
includes viscosity as well as other important properties of liquids. The
most meaningful comments on the rheologies of paint come from
painters. Such words as wet, dry, sagging, running, thin, building,
cobwebs, overspray, leveling, and stiff are used. An answer to the
question, "where would you put this stuff on a scale which has water
at the top, mayonnaise in the middle, and heavy cup grease at the
bottom?" is much more instructive than the technical jargon of the
paint laboratory.
The most desirable rheology for a paint is similar to that of
mayonnaise. When pressure is applied to such a product, it thins down
and flows. When the pressure is removed, the liquid gradually
recovers its original puffiness. This characteristic is called thixotropy.
Almost all paints are somewhat thixotropic, but high-build-per-coat
products are extremely so. Without built-in thixotropy, a wet paint
film on a vertical surface would sag to the ground or floor. If the
rheology is adjusted to approach that of water, the product flows
through spray guns and off brushes or rollers much more easily, but
the films are nothing more than a series of runs and sags. On the other
hand, thick mixtures go on in uneven bands and globs. In the 1950s
and 1960s, paints were never too thick because the painter was able to
increase the volatile vehicle ratio of a paint by blending in a liquid
called by various names, e.g., thinner, reducer, acid, or something
more colorful. This thinner is sometimes of the same composition as
the original volatile vehicle but is more likely to be a combination of
two or three solvents compatible with a variety of paints. During the
1960s environmentalists became alarmed about the amount of these
volatile organic compounds (VOCs) being released in our
environments. The 1970s and 1980s brought governmental
regulations.
The above suggests that the paints used in the 19601990 era will not
make it into the 21st century. Their passing is not due to a failure to
perform but rather to our concern with both personal and
environmental health.
The popular paints for plant maintenance prior to 1990 were alkyds,
vinyls, epoxies, polyurethanes, and, before 1980, chlorinated rubbers.
Most of these products, by today's standards, had high VOC levels.
Furthermore, they were commonly thinned down prior to spray
application. The practice provided the optimum rheology for
obtaining uniform, relatively pinhole-free films. Although there can
be little argument with the federal government's desire to restrict the
release of harmful and smog-producing chemicals into the
atmosphere, the ensuing regulations devastated product lines. Since
the late 1980s the federal guideline for VOCs has been 420 g/L, i.e.,
3.5 lb/gallon of product as applied. Complying paints would have to
have volume solids of at least 50%. The traditional products with
established histories had volume solids around 40% and frequently
less. In spite of the difficulty in formulating to meet these
requirements, the industry was able to come up with some complying
products. How-
Page 566
ever, we have been told that there is a strong sentiment in the U.S.
Environmental Protection Agency for VOC limits of 350 g/L, i.e., 2.9
lb/gallon. Conforming products will have to have volume solids of
60% or more as applied. This will be the final nail in the coffin of
many of the products that have worked well for plant maintenance.
Two exceptions are very high solids products with VOCs approaching
zero and water-based coatings, respectively.
A few more points must be made about nonvolatile vehicles before we
can understand how there can be paints with very low VOCs.
There are five ways in which liquids can convert to solid films having
toughness and integrity.
1. A solution of a relatively high molecular weight resin with the
desirable properties in a suitable solvent. Apply the solution to the
surface to be protected and allow the solvents to evaporate, leaving a
film of the resin. This is the lacquer technique used for vinyls,
acrylics, chlorinated rubbers, polyethers, and certain alkyds. All of
these solutions have very low-volume solids and very high VOCs.
2. Solutions of resins which are capable of reacting with something in
the environment such as oxygen or water to convert to more complex
structures with the desired physical properties. Examples are alkyds,
some polyesters, and moisture-cured polyurethanes. Their solutions
require between 25% and 50% less solvent than lacquers but still have
VOCs in excess of 3.5 lb/gallon.
3. Combinations of resin/chemicals or resin/resin which coreact when
mixed together just prior to application. The reactants are separately
packaged. Examples are polyamine/epoxies, polyamide/epoxies, and
polyurethanes. Some of these types can be formulated to meet the 3.5
lb/gallon VOC standard.
4. Ultrahigh solids versions of the above in which the reactants are
low-viscosity liquids not requiring solvents for the development of
acceptable spray properties. Their VOCs approach zero. These types
of products have been growing in popularity since the late 1970s.
They are mastics and are applied in relatively thick coats.
5. Water-based products which usually contain a small amount of
VOC.
Most coatings experts in the 1970s were sure that by 1990 solventbased coatings would be dinosaurs. In spite of the more stringent
regulations of volatile organic compounds during that 20-year period,
by 1995 water-based coatings still did not dominate the market. There
are at least three reasons why the experts' predictions have not come
to pass. First of all, in spite of the guideline published in response to
the Federal Clean Air Act, regulation and enforcement is for the most
part conducted on a state or local basis. In some localities air pollution
is not the same concern as it is in Los Angeles County, California.
However, the
Page 567
noose gets tighter each year. A second deterrent to water-based
coatings has been their price: they cost more than most people expect.
Finally, several water-based products were over promoted in the past,
and the memory of their mediocre performances lingers.
VOC regulations apply to house paints as much as they do to those
used in plant maintenance. Fortunately, in painting our homes and
most commercial buildings that do not have expanses of structural
steel, the need for the barrier properties of paint is negligible.
Numerous tiny holes in the dry film, quite common in water-based
paints, do not detract from performance and sometimes serve a
positive purpose.
Pinholes are a problem in just about all of the protective coatings used
on steel. This is the major reason for an inhibitive primer as the first
coat of a paint system. When condensation settles on a steel surface,
the electrochemical process of rusting begins. The more electrically
conductive the dew, the more rapid the reaction. A perfect protective
coating film should isolate steel or other susceptible substrates from
aggressive moisture. Since very few perfect films are created in
maintenance painting, formulators have for over a century tried to
compensate for pinholes and other skips by adding chemical rust
suppressors to primers for steel. Examples are red lead, lead chromate,
and zinc chromate. Both lead and chromate compounds have recently
been added to the list of chemicals that are potential hazards to plant
and animal life. Zinc phosphate and more sophisticated chemicals
have been substituted for them. Another method of suppressing
rusting of steel surfaces is to use zinc metal dust as a sacrificial
pigment in the primer. Zinc is chemically more reactive than iron, the
major component of steel, and will corrode preferentially in an
aggressive environment.
Although zinc metal had been used since the early 18th century as a
cladding over steel to prevent corrosion in the process known as
galvanizing, the concept of using the metal as a pigment was not
exploited until the late 1940s.
Two techniques were developed. In one, the zinc dust was dispersed
in a resin system such as epoxy polyamide, chlorinated rubber,
acrylic, or polyether. These products are known in the trade as zincrich primers or organic zincs. The other route was considerably more
imaginative, and its explanation requires a brief chemistry lesson. The
resins and solvents that have already been discussed are categorized as
organic chemicals, i.e., their structures include the element carbon. All
animal and plant life and their derivatives are organic. The only
chemical structures containing carbon that are not classified as
organic are diamond, carbon dioxide, and graphite. All of the
chemical structures that do not contain carbon, i.e., glass, rock, air,
water, metals, and minerals, are said to be
Page 568
inorganic. One of these, window glass, is a mixture of sand, lime, and
lye, manufactured by bringing the mixture above its melting point,
pouring it into sheets, and allowing the molten mixture to cool down.
If the lime is replaced by more lye, the glass will dissolve in water to
make a sodium silicate solution, "waterglass." An engineering firm in
Australia in the 1940s found ways of using water solutions of sodium
silicates to replace solvent solutions of organic resins to make the first
inorganic zinc primer. During the ensuing decades other inventors
created variations on this theme. One of these was really unusual in
that its binder was based on an alcohol solution of an organic
compound, ethyl silicate. It has been the most popular of the inorganic
zinc primers since the 1960s. The film loses its organic compounds
during cure. Products falling into this category are known as solventbased inorganic zinc primers in the trade.
Inorganic zinc primers and, to a lesser extent, organic zincs have been
effective in moderating corrosion in extremely aggressive geographic
environments such as the southeastern and Gulf coasts. Their pricing
is tied to the current price of zinc in the metals market.
Paint formulators have used a variety of additives to reduce the
occurrence of pinholes. The major source of the problem is air
entrained in the product during manufacture as well as during the
thorough mixing just prior to application. Although most of the air
escapes from the paint without causing problems, a significant
quantity remains as tiny bubbles until the film is in its final stages of
drying. As the film takes a set, the stresses caused by its shrinkage
seem to force the air out, leaving holes. Since the film can no longer
flow at this point, the defect remains.
Air adhering to rough or porous substrates can also cause pinholes.
Problem surfaces include concrete, inorganic zinc primers, and cast
iron. Most of the displaced air escapes with the solvent in the wet
film, but there is always a large nuisance quantity that breaks through
the film when it can no longer flow. However, very slow-drying
paints, such as the long oil alkyds, can repair themselves.
In many environments the presence of pinholes in film does not result
in rampant corrosion. This surprise is probably due to the high surface
tension of water, causing it to form droplets rather than flowing on
hydrophobic resin surfaces. In tank lining work, where pinholes
cannot be tolerated, a so-called holiday detector is used to find
discontinuities in the film. This device is simply an open electrical
circuit consisting of a battery and buzzer wired to a clamp as one pole
and a brine-saturated sponge as the other. After the clamp is grounded
to the steel tank, the sponge is rubbed over the lining. If the circuit is
completed by the brine electrolyte flowing through a pinhole or other
holiday, the buzzer sounds. This procedure does not work for pinholes
unless a wetting agent is added to the brine to reduce its surface
tension.
Poor wettability of the surface may also explain why films pigmented
with aluminum flakes have much better water holdout than the same
formulas with
Page 569
conventional pigmentation. This great improvement in barrier
properties is usually attributed to the ability of the flakes to form a
shield by overlapping. However, other flakes such as glass, mica, and
graphite are not as effective. The success of the aluminum flake may
be due in part to the stearic acid applied to its surface during
manufacture to prevent clumping. Although stearic acid is known to
be very hydrophobic, its incorporation into coatings was a matter of
serendipity rather than planning.
In spite of the abilities of aluminum and inhibitive pigments to make
paints better corrosion barriers, the major roles of pigments are to
provide color and opacity, and sometimes to lower costs. Both
synthetic and naturally occurring compounds of specific color are
used alone or in combinations to get whites, blacks, grays, pastels,
reds, yellows, greens, or blues in shades ranging from the mundane to
the exotic. One attribute that these pigments have in common with
corrosion inhibitors is extremely high cost. To reduce that problem
and to increase a paint's solids content, frequently inert or filler
pigments are added to paints. Fine particles of silica, silicates, barytes,
or other rocks, if used with discretion, help paint formulators to meet
solids and cost goals.
Some of the colorants that had been used in making yellows, oranges,
reds, and greens contain lead, a problem already encountered in the
discussion of inhibitive pigments. The toxic nature of lead compounds
has curtailed their use. The substitutes, complex organic compounds,
are even more expensive, cause rheology problems, exhibit
deficiencies in opacities, and frequently fail to match the shades
attainable with the outlawed pigments. Happily, in plant maintenance
the usual customer choices for the coating of extensive areas are
whites, grays, and pastels. Along with color, another cosmetic
property of paint affected by pigment is gloss. Although resin
compatibility can be a factor, the ratio of pigment to resin in the dry
film usually governs its degree of gloss. Resin-rich surfaces have a
high gloss because they are optically smooth and can reflect light in
the same fashion as a mirror. This quality is gradually lost as pigment
concentrations are increased. A flat or matte surface scatters rather
than reflects light. Where light reflectivity increases operational
efficiency, high gloss is desirable. However, it can be a nuisance if
personnel in the area are adversely affected by glare. One should also
keep in mind that gloss accentuates surface imperfections and defects
whereas flatness deemphasizes them.
Long before the first dollar is spent for maintenance painting, a
thorough survey must be made of the existing paint systems in the
plant. Not too many years ago companies employed individuals, if not
departments, to perform this service and keep records of the findings.
Unfortunately, in the 1980s many companies in order to reduce fixed
costs eliminated or severely curtailed such operations. The
managements that made these decisions need to be convinced that
paint is another asset to be kept track of.
If there is no in-house expert, step 1 is to designate someone. Since
Page 570
management of paint programs in not taught in our universities, all of
the existing experts gained their knowledge on the job. Just about
everyone has held a paint brush, roller, or perhaps a spray gun at some
time. These experiences can be built on because the basic, ''clean the
surface, protect the surrounding area, apply the paint, tidy up!" is the
same for all painting. Our expert, experienced or recently dubbed,
should contact three or four of the major paint companies for
assistance in making a sector-by-sector inspection of all painted
surfaces in the plant. (Caution: This is not a request for the help of the
local paint salesman.) All of the major paint companies employ
technicians who have been trained to evaluate and report on the
condition of in-place paint. Most of these people are also competent in
making recommendations for repairs. Although the survey can be
made with one all-inclusive group, a far better procedure is for the
plant expert to accompany each of the paint companies'
representatives on separate tours. Remember that each survey needs to
be thoroughly planned and cleared with operating units.
After the condition and recommendation reports have been submitted,
they should be compared. Although there may be some variance, these
reports tend to have high agreement, at least so far as the current
condition of the coatings is concerned. Plant facility condition surveys
are also, for a fee, available from a number of consulting firms.
The observations in the reports will include terms like rusting,
thoroughly bonded, dead, intercoat delamination, heavy chalking,
light chalking, intact, and adherent, or sometimes references to visual
standards. Where a coating system has failed, its removal and
replacement will be recommended. If the failed coating was
pigmented with lead compounds or with zinc dust, its special disposal
will be an extra expense. However, before getting into remedies, let us
discuss why a paint may no longer be serviceable.
Exterior paints, particularly in areas of high exposure to sunlight,
begin to fail immediately after their application. Symptoms are first a
change in color, followed by a loss of gloss accompanied by a
superficial dust. Gradually the dust layer thickens, and eventually it is
the only remnant of the paint film. The paint is "dead." The same
paint, when used indoors or when shielded from the sun, does not
exhibit this phenomenon. The process, called chalking, attacks the
surfaces of organic polymers when they are exposed to solar or
nuclear radiation. Back in our discussion of zinc-rich primers, organic
chemicals were defined as molecules having the element carbon in
their structures. An organic polymer is a long chain of repeating
chemically bonded units that contain carbon. Examples are
polyethylene, polyvinyl chloride, polyacrylics, polyurethane, and
polyisoprene (natural rubber). Alkyds are polyesters whereas cured
epoxies are usually copolymers of polyamide and epoxy resins. Longchain polymers give a paint film both tensile and compressive
strengths. When radiation destroys the bonds between atoms of a
polymer, the long chains are broken down to shorter, less complex
units that do not contribute to desirable physical properties. Our skins
Page 571
are organic polymers, and we all know how they, particularly very
pale ones, react to the sun's ultraviolet light radiation in the summer.
The deterioration of paint exposed to ultraviolet light, sometimes
spoken of as "aging," proceeds more rapidly with some paints than
with others. In spite of the good chemical resistance of many epoxy
resin formulations, their films must be protected from sunlight to give
worthwhile service in southern climates. Alkyds, vinyls, and
chlorinated rubbers, while better outdoors than epoxies, have never
matched the performance of acrylic lacquers. The latter, by the way,
were also the first automotive paints to give decent service. Since the
1970s the best weathering paints have been aliphatic acrylic
polyurethanes and water-based acrylics. Recently, another weather
resistant product has come to market. Described as an "engineered
siloxane" in the manufacturer's literature, this product is the result of
blending the chemistry of epoxy compounds with that of silicon
oxygen polymers.
The fact that silicon and oxygen atoms can combine to form polymers
was touched on in the discussion of inorganic zinc primers. In such
polymers
takes the place of the
bonding found in organic coatingsresins. Where the inorganic
structure replaces the organic, greater resistance to both radiation and
heat is attained. One group of resins, called silicones, have a high
content of
chains. As a result, films based on silicone resins age more slowly and
can be used at higher temperatures than those made from straight
organic polymers. This is also true of the silicates used for inorganic
zinc primers. Both the latter and aluminum flakepigmented silicone
resins have performed very well on surfaces approaching 800°F
(427°C). Some organic-based products will make it to 300°F but most
fail between 250°F (121°C) and 300°F (149°C). One variant of the
"engineered siloxane" is said in its product literature to be satisfactory
in services up to 2000°F (1093°C).
Underfilm corrosion is another common reason for the failure of paint
films
Page 572
over steel. Admitting some oversimplification, we can state that steel,
acting as a battery, corrodes by one of the following electrochemical
processes:
In examining steel that is rusting through or under a coating, the
inspector employs a lens and a pocket knife. Rusting through, usually
the result of pinholes along with inadequate film thickness, is
characterized by considerable areas of well-bonded coating around
rust spots. Where rusting is going on under an intact film, two
possible causes stand out. The first is inconsistent surface preparation
whereas the second is insufficient barrier properties of the film to
water or acid. When a coating film fails over steel, by blistering or
peeling, a debate usually takes place between the contractor and the
coatings supplier as to whether poor surface preparation or poorquality paint was the cause.
Surveying painted structures may also bring attention to intercoat
delamination. This problem is frequently caused by insufficient
cleaning of an old film before the application of a refresher coat.
Although promotional literature sometimes implies that a product will
stick to any kind of surface condition, a potential user must employ
common sense. If the surface of the old paint is dirt, the new coat will
bond to the dirt and not the underlying paint film. When one is going
over old paint, loose chalk should be considered to be dirt.
At times intercoat delamination occurs even when the old paint had
been scrupulously cleaned. Surfaces can take on chemical
characteristics that interfere with the adhesion of refresher costs.
Abrading the old film before overcoating it usually helps. However, at
times special tiecoats are needed between the new and old film.
Finally, errors of choosing products that are not compatible are
frequently made. The only good way to find out if the new paint if
going to stick to the old is to run a test. This is rarely done in practice.
Page 573
Along with exterior steel structures, manufacturing and processing
plants can have galvanized, stainless steel, concrete, and brick
surfaces. Although much of the time these are not painted, some
environments are aggressive toward them. When painting is strictly
cosmetic, the best recommendation is thorough washing, rinsing, and
drying followed by a water-based acrylic system. This
recommendation also applies to steel structures with numerous layers
of existing, intact paint. Water-based acrylics are also very suitable
directly over steel in facilities with low rates of corrosion.
Galvanizing comes in several grades, most of which are paintable.
However, at least one galvanizing process results in a surface that
does not accept paint, even after vigorous abrasions or other
treatment. For the other grades, new galvanizing often must be
degreased before painting is attempted. Because zinc is a very reactive
metal, it forms oxide and carbonates rapidly while cooling from the
molten state. This thin oxide/carbonate layer is protective and keeps
the underlying zinc from wasting itself in pointless chemical
reactions. When sheets of galvanized steel are stored in stacks until
their fabrication into shapes has been accomplished, an environment
can develop between the sheets in which the protective layer is
destroyed. The exposed fresh zinc reacts very rapidly and heaps of
white corrosion product form. To prevent this deterioration
manufacturers of sheet galvanized steel give their product a light
application of oil. Where the environment is so severely corrosive that
a more resistant system than water-based acrylic is required for new
galvanizing, a vinyl butyral wash primer has traditionally been used
for adhesion. This product has very low solids, very high VOCs, and
may not be available. If it is not, the manufacturer of the other paints
in the system must come up with an effective, environmentally
acceptable substitute. Above all, do not accept the old wives' tale that
treatment of the galvanizing with vinegar makes it paintable!
Painting weathered galvanizing is much less of a problem. Zinc
surfaces are neutralized by years of exposure to the weather, and
epoxy paints bond very well to them. Frequently, galvanizing suffers
pinpoint breakthroughs and rusting after long exposures in some
climates. Zinc-rich epoxies are excellent primer coats in these
instances.
Aluminum, like zinc, is very chemically reactive and is useful
commercially only because it too has a tight protective layer of
oxide/carbonate on its surface. Where alkaline salt fallout destroys the
protective layer and then the underlying metal, aluminum structures
should be painted. Epoxies are a good first coat, followed by an
aliphatic polyurethane where protection from the sun is needed.
Often exterior plant areas have numerous concrete structures that are
attacked by acid fallout. Application of epoxy paints with an aliphaticacrylic polyurethane for weathering usually takes care of such
problems. If we have concrete block rather than poured concrete, our
situation is more difficult. Blocks are extremely porous, and the
surfaces must be treated with "block filler" before
Page 574
being painted. There are a variety of such products available. Some
are 100% solid epoxy compounds that must be worked into the block
with a trowel. Sprayable products, usually vinyl or acrylic latexes
highly loaded with sand or other inert filler pigment, are also on the
market. Although contractors object, these block fillers should be
back-rolled while wet with a long-nap paint roller to get smooth
surfaces.
When the structures to be painted are indoors, solar radiation, the
biggest foe of outdoor paints, is no longer a factor. However, without
the bleaching action of the sun, some films such as epoxies and alkyds
gradually yellow and darken. Although rain can be ignored, high
humidities that produce dews