Dr E.R. Wallach
Lent Term 2014-15
Materials Science Part II
University of Cambridge
C9 Alloys
Synopsis and reading list
Part II Materials Science and Metallurgy
C9 Alloys (9 lectures)
This course deals with the design and use of metallic alloys with a focus on the development and
control of microstructure, the relationship between microstructure and properties, and applications.
1.
Introduction. Sustainability. Basis of lattice types. Hume Rothery rules for solid solutions.
Summaries of mechanical and physical properties.
Production of light alloys (Al, Ti, Mg).
2.
Aluminium alloys. Phase diagrams. Alloys and tempers. Alloy characteristics. Review of
preciptation hardening, oxidation, corrosion resistance and fatigue.
3.
Titanium alloys. Pure Ti. Alloying Ti. Specific alloys: α, α + β, and β. Superplasticity.
4.
Magnesium alloys. Mg alloys. Heat treatment of Mg alloys.
5.
Anisotropy. Microstructural and crystallographic. Single crystal and polycrystalline materials.
Effect on properties. Examples.
6..
Steels. Review of plain C, alloy steels and cast irons. Commercial steels: high-strength, lowalloy (HSLA), bainitic, dual-phase, transformation induced plasticity (TRIP). Stainless steels.
7.
Copper alloys. Overview copper alloys.
8.
Nickel alloys: overview and superalloys.
9.
Non-destructive testing
Reading list
1. “Light Alloys: from traditional alloys to nanocrystals”, Polmear I. J.,
th
pub. Elsevier Butterworth-Heinemann, 4 edition (2006)
Dept. library: Eb153
2. “Steels, microstructure and properties“, Bhadeshia H. & Honeycombe R.,
rd
pub. Butterworth-Heinemann, 3 edition (2006)
Dept. library: De100
3. “Structure and properties of engineering materials“, Henkel D. & Pense A.W.,
th
pub. Tsinghua University Press / McGraw Hill, 5 edition (2008)
Dept. library: AB222
4. “An introduction to textures in metals“, Hatherley M. & Hutchinson W.B.,
pub. Institution of Metallurgists Monograph 5 (1979)
Dept. library: Mc6b
5. “Bainite in Steels”, Bhadeshia H.K.D.H.,
Dept. library: De96
2nd ed., Institute of Materials. 2001. Chapter 13 “Modern bainitic steels”
Also can download from
www.msm.cam.ac.uk/phase-trans/newbainite.html
6. “Non-destructive testing“, Halmshaw R.,
nd
pub. Edward Arnold, 2 edition (1991)
Dept. library: Ma100
7. Non-destructive testing – excellent website:
www.ndt-ed.org/EducationResources/CommunityCollege/communitycollege.htm
rob.wallach@msm.cam.ac.uk
Page (i)
2014-15
Materials Science Part II
University of Cambridge
C9 Alloys
Lecture 1: Metals and alloys
1. Metals and alloys
1. Introduction
In the periodic table, 87 elements are classified as metals, 61 of which are commercially available1.
The most commonly used metals and alloys are based on Al, Cu, Fe, Ni, Pb, Sn and Zn.
Advantages of metallic materials include:
▪
huge range of alloys and tempers allowing optimisation of properties for diverse applications;
▪
their generally high electric conductivity, thermal conductivity, strength, corrosion resistance;
▪
ease of shaping both by casting and subsequently by deformation processing;
▪
ability to manufacture smart materials e.g. based on their superconducting, optical and
magnetic properties plus biocompatibility.
Major metallic alloy systems, all with distinctive properties, include:
•
steels
– low cost, high strength (over 90% by weight of all metal usage is steel);
•
aluminium alloys – high specific strength, corrosion resistance, specific conductivity;
•
titanium alloys
•
copper
– high electrical & thermal conductivity, easy to form/cast, corrosion resistance;
•
nickel
– high temperature strength and creep resistance (superalloys).
– higher specific strength and higher temperature application;
Applications include:
Metal
Industries
Applications
Steels
Very wide
Automotive, ships, buildings, white
goods, railway lines, reinforced concrete
Aluminium alloys
Aerospace, packaging, sports
equipment, energy, construction
Aircraft, food containers, power cables,
building cladding
Titanium alloys
Biomedical, aerospace
Body implants & medical, military
airframe and engines, spacecraft
Copper alloys
Construction, electronics, coins,
transport
Plumbing, wiring, circuit boards,
electronic components
Nickel superalloys
Aerospace
Aircraft engines and space craft
Plating on steel for corrosion resistance
Challenges for the future:
Sustainability issues are increasingly important because of
•
greater awareness and hence consumer pressure;
•
legislation – Kyoto agreement (ended in 2012 and future very uncertain - Paris, 2015);
•
scarcity as key metals, especially rare earth metals are limited in their abundance.
Sustainability is more narrowly focussed on
•
abundance of different elements and whether or not economic alternatives exist;
•
energy to extract – embodied energy – and to shape;
•
“greenhouse gas” emissions when extracting and shaping (main emphasis normally on CO2).
See Figs. 1.1 and 1.2 on next page.
1 www.ceram.com/materials/metals/
rob.wallach@msm.cam.ac.uk
Page 1 of 61
2014-15
Materials Science Part II
University of Cambridge
C9 Alloys
Lecture 1: Metals and alloys
Figure 1.1 Reserves versus annual world production of various elements [CES software]
Figure 1.2 CO2 emissions versus embodied energy, both for primary production [CES software]
rob.wallach@msm.cam.ac.uk
Page 2 of 61
2014-15
Materials Science Part II
University of Cambridge
C9 Alloys
Lecture 1: Metals and alloys
1.2 Background to metals and alloys
a) Basis of different lattice types for pure metals
As mentioned on page 1, there are 87 known metals in the periodic table. They can be grouped in
various ways, e.g. ferrous, non-ferrous and noble, or classified as shown in Figure 1.3 below.
noble metals - generally unreactive, e.g. silver, platinum, gold and palladium;
alkali metals - very reactive with low melting points and soft, e.g. potassium and sodium;
alkaline earth metals – less reactive, higher melting points and harder than alkali metals, e.g.
calcium, magnesium and barium;
transition metals - hard, shiny, strong, and easy to shape, e.g. iron, chromium, nickel, and copper;
“other” metals – diverse properties, e.g. aluminium, gallium, indium, tin, thallium, lead and bismuth.
Figure 1.3 Periodic table 2.
Question arises of whether the periodic table can explain, in a first-order manner, the different lattice
type for metals since it is the electron distribution, and the resulting energy of a given arrangement
within the material, that determines the lattice type.
A simple approach, by Hume-Rothery, considered the correlation between electronic configuration
and crystal structure.
Crystal structures were shown to be the same for many non-transition metallic elements and alloys
having the same average number of valence electrons per atom, e/a.
This did not extend to the transition metals since, as explained by later researchers, their d electron
orbitals are relatively more localised than the outer-shell s and p electron orbitals; the latter have a
greater influence on the long-range order. It appeared that the ratios e/a are:
bcc structures form in alloys with < 1.5
outer-shell s, p electrons per atom,
hcp structures form in alloys with 1.7 – 2.1 outer-shell s, p electrons per atom,
fcc structures form in alloys with 2.5 – 3.0 outer-shell s, p electrons per atom,
diamond structures in alloys with
4.0
outer-shell s, p electrons per atom.
This fits for elements in the third row of the periodic table on adding an additional electron, namely
from Na, Mg, Al to Si. However, deviations are noted in the fourth row of the periodic table for both
Ca (fcc and not hcp) and Ga (orthorhombic and not fcc).
The majority of transition element metals near their melting temperatures have bcc structures. This
can be explained by the presence of vacant d orbitals which means that electrons from outer-shell
orbitals can be accommodated rather than going into p orbitals.
2 Smell-O-Mints, www.jschilling.net/sw_smellomints.php
rob.wallach@msm.cam.ac.uk
Page 3 of 61
2014-15
Materials Science Part II
University of Cambridge
C9 Alloys
Lecture 1: Metals and alloys
More recent modelling is based on “many-electron” calculations but these tend to be complex and it is
often necessary to make approximations. This is problematic given that the energy differences of
possible crystal structures for a given element often are small, less than 1% of the bonding energy,
One approach is to consider the energy contributions from
a. the bonding or conduction electrons although, in metals, the potential and kinetic energies of
these electrons are substantially, though not totally, structure independent;
b. the Coulomb interaction energy between the positive metal ions and the “sea” of electrons .
b) Hume Rothery rules for solid solutions
The majority of metals used in commercial applications are alloys which are typically are solid
solutions or, alternatively, comprise two (or more) phases.
For substitutional alloys, the following Hume Rothery rules indicate the extent of solid-solubility.
Rule 1:
Atomic size factor or the 15% rule: extensive substitutional solid solubility may occur if the
relative difference between the atomic radii (r) of the two elements is less than 15%, i.e.
rsolute - rsolvent
for solid solubility,
≤ 15%
rsolvent
Conversely, if the difference > 15%, solubility generally is limited.
Rule 2:
Crystal structures of the two elements must be identical for appreciable solid solubility.
Rule 3:
Valency. The solute and solvent atoms should typically have the same valency in order to
achieve maximum solubility. For different valencies, a metal will dissolve a metal of higher
valency to a greater extent than one of lower valency.
Rule 4:
Electronegativities need to be similar for maximum solubility, i.e. a solute and solvent should
be close in the electrochemical series. When the difference in electronegativities increases,
intermetallic compounds tend to form rather than substitutional solid solutions.
Darken-Gurry maps can be used to show the effects on solid-solubility of electronegativity (vertical
axis) and atomic radius (horizontal axis). Ellipses are added to show the effects of the above rules.
Two Darken–Gurry plots are shown in Fig. 1.4 for aluminium3 and Fig 1.5 for palladium4 as solvents.
Atomic radii r (nm)
Figure 1.4
High solubility expected if the solute atom
is within first ellipse (<0.2 difference in X and <7.5%
difference in r from the solvent, Al).
Solubility of at least 5% is expected if the solute atom is
within second ellipse (<0.4 difference in X and <15%
difference in r from the solvent, Al).
Note that Mg has high solid-solubility in Al, as does Cu
Yet neither would have been predicted from the above.
Figure 1.5
Darken–Gurry map with ellipse around the
solvent Pd. The vertical line is tangent to
the ellipse at ±15% of the diameter of Pd.
3 www.synl.ac.cn/org/non/zu1/knowledge/Hume-Rothery-rules.pdf
4 Fang S S, Lin G W, Zhang J L, Zhou Z Q, “The maximum solid solubility of the transition metals in palladium”,
International Journal of Hydrogen Energy, Volume 27, Issue 3, March 2002, pages 329–332.
rob.wallach@msm.cam.ac.uk
Page 4 of 61
2014-15
Materials Science Part II
University of Cambridge
C9 Alloys
Lecture 1: Metals and alloys
It has been suggested in the case of the Pd that the discrepancy between predicted and observed
extents of solid-solubility are due to the fact that the electrons in the d and f orbitals (especially) in
some elements do not behave as if completely delocalised. Such electrons retain some unpaired
electron characteristics, e.g. their spin, which accounts for the magnetic properties of these metals.
This is consistent with the Hume-Rothery explanation for transition metal structures.
For interstitial alloys, the following Hume Rothery rules indicate the extent of solid-solubility.
Interstitial solid solutions are more likely to be formed if
- a solute is smaller than the interstitial sites in the solvent lattice of a solvent;
- a solute has approximately the same electronegativity as the solvent.
In practice, there are very few elements that can form ions which are sufficiently small to fit in
interstitial sites, and so appreciable solubility is rare for interstitial solid solutions.
Possible metal ions that may form interstitial solid solutions are: Li, Na, B.
Plus non-ions H, C, N
Many interstitial solid solutions have a strong tendency to spontaneous ordering and examples of
ordered or partially interstitial solid solutions include Al-Li.
c) Intermetallics
Intermetallic compounds are metallic phases but, unlike the alloys described above, each generally
has a very limited composition range, i.e. intermetallics tend to have a narrow and fixed stoichiometry.
Ordering within the crystal lattice is thus common. One consequence is that typically they are quite
hard and strong.
d) Defects
All metallic crystal lattices have defects - perfect lattices, with all atoms of the same type in identical
positions and no missing atoms, do not exist. Defects are crucial to both mechanical and physical
properties of metals and their presence and distribution can be controlled in order to optimise desired
properties intentionally used to manipulate the mechanical properties of a material.
There are three major categories of crystal defect:5
▪
▪
point defects, where an atom is missing or irregularly placed in a lattice structure, include:
lattice vacancies,
self-interstitial atoms,
substitutional alloying or impurity atoms,
interstitial impurity atoms;
line defects, which are groups of atoms in irregular positions, include:
dislocations;
▪
planar defects, which are interfaces between homogeneous regions of a material, and include:
grain boundaries,
stacking faults
external surfaces.
5 www.ndt-ed.org/EducationResources/CommunityCollege/Materials/Structure/crystal_defects.htm
rob.wallach@msm.cam.ac.uk
Page 5 of 61
2014-15
Materials Science Part II
University of Cambridge
C9 Alloys
Lecture 1: Metals and alloys
1.3 Mechanical properties of metals and alloys - summary
Strength
grain refinement
work hardening
texture strengthening
solid solution hardening
solute pinning
precipitation hardening
ordered structures
two – phase strengthening
intermetallic hardening
oxide dispersion strengthening
Toughness
grain refinement
dislocation motion
control of particles
stress distribution and concentrators
composites
Creep
large grain size
solid solution hardening
coherent precipitates
pinned grain boundaries
Fatigue
small grain size
strengthened surface (ideally in compression – use of shot peening)
tough interior
rob.wallach@msm.cam.ac.uk
Page 6 of 61
2014-15
Materials Science Part II
University of Cambridge
C9 Alloys
Lecture 1: Metals and alloys
1.4 Physical properties of metals and alloys include
Conductivity
electrical (and resistance increases with temperature)
thermal
Magnetic
soft magnets
hard (permanent) magnets
Density
high and three major crystal classes with long range atomic ordering
Melting temperature
huge range from mercury (-39°C) to tungsten (3400°C)
The properties of transition metals strongly depend on the number of d electrons and the strength of
the d-orbital interactions, as is shown below in Fig. 1.6 for cohesive energy, bulk modulus and melting
temperature.
Cohesive energy
Bulk modulus
Melting temperature (°C)
Figure 1.6 Variation in the number of d electrons, Nd, and
cohesive energy, bulk modulus6 and melting temperature.
6 www.synl.ac.cn/org/non/zu1/knowledge/Hume-Rothery-rules.pdf
rob.wallach@msm.cam.ac.uk
Page 7 of 61
2014-15
Materials Science Part II
University of Cambridge
C9 Alloys
Lecture 1: Metals and alloys
1.5 Light alloys - Al, Ti, Mg
Some properties of commercially pure metals
rd
[Polmear (3 edition), Journal of Metals. 54 (2002) 42–48 and Steel World, 2 (1997) 59]
All three alloys have:
- excellent strength to weight ratios
- corrosion resistance due to stable surface oxides – see Ellingham diagram, Figure 1.8.
Hence widespread usage in aerospace and transportation applications.
Figure 1.6 Comparative weights of different alloys for the same stiffness.
th
[Polmear, Light Alloys, 4 edition, p 4, 2006]
Increasingly changing situation for usage of alloys from Al, Ti or Mg due to:
-
composite materials development, e.g. C fibre in epoxy (Boeing, British Aerospace aircraft)
metal matrix SiC in Al or Ti (e.g. car engines7)
ceramic matrix (e.g. Si-SiC nuclear fusion8)
-
cost: although materials cost is small part, generally, of final product price
-
recycling: all three metals are very energy intensive to produce from ores and so recycling is
increasingly promoted, especially for Al alloys.
Figure 1.7 Years of supply (based on known mineral reserves versus recycle rate for various metals.
[Norgate and Rankin, 2002]
7 http://en.wikipedia.org/wiki/Metal_matrix_composite
8 http://composite.about.com/od/aboutcompositesplastics/l/aa030205.htm
rob.wallach@msm.cam.ac.uk
Page 8 of 61
2014-15
Materials Science Part II
University of Cambridge
C9 Alloys
Lecture 1: Metals and alloys
1.6 Production of Al, Ti and Mg [see next page: Ellingham diagram and abundance of elements]
1.6.1 Aluminium production9
Aluminium is produced in a two-stage process:
(i) the ore bauxite - iron and aluminium hydroxides/oxides - is refined to produce alumina Al2O3
Bayer process:
Al2O3.xH2O + 2 NaOH
è 2 NaAlO2 + (x+1) H2O
2 NaAlO2 + 2 H2O
è 2 NaOH + Al2O3.3H2O
(ii) the alumina (aluminum oxide trihydrate) is reduced electrolytically to give metallic aluminum
Hall-Héroult process:
2Al2O3 + 3 C è 4 Al + 3 CO2
Four tons of bauxite produce 2 tons of alumina from which one ton of aluminium is formed.
1.6.2 Titanium production10
(a) Kroll process
(i) reduction of titanium ore rutile TiO2 or ilmenite FeO.TiO2 using coal and chlorine gas at 800°C
TiO2 + 2 Cl2 + 2 C è TiCl4 + 2 CO2
(ii) fractionally distillation of TiCl4 to separate from other chlorides, formed from other metals in
the ore, under an argon or nitrogen atmosphere, and is stored in totally dry tanks;
(iii) reduction of TiCl4 in a batch process using molten Mg (reaction at 1100°C) or Na (550°C)
under argon to yield Ti porous sponge (note that Na approach is used in the UK only).
TiCl4 + 2 Mg è Ti + 2 MgCl2
The Ti is solid at the reaction temperature allowing easy separation from the MgCl2.
(iv) melting of sponge, or sponge plus a master alloy to form an ingot which then is “primary”
fabricated into billet, bar, sheet, strip or tube.
The Kroll process is both time-consuming and expensive, in part because of the batch processing
required for the TiCl4 reduction.
(b) FFC Cambridge process11
A completely different approach, the FFC Cambridge process, is based on the direct electrochemical
reduction of TiO2 to Ti in molten CaCl2. First developed here in the Department between 1996 and
1997, the process is being developed for commercial usage and also has potential for other oxides.
The process typically takes place between 900 - 1100°C, with an anode (typically carbon - graphite)
and a TiO2 cathode (hence which is reduced) in a bath of molten CaCl2.
1.6.3 Magnesium production
(a) electrolytic reduction of MgCl2 at temperatures ~ 700°C. The MgCl2 is obtained from sea
water. the two common ores dolomite MgCO3-CaCO3 and magnesite MgCO3.
(b) the Pidgeon thermic process in which Si or ferro-silicon is used as a reducing agent to extract
Mg from its two common ores dolomite MgCO3-CaCO3 and magnesite MgCO3. The reaction is
highly endothermic. Mg is produced as a vapour which allows its separation.
Si (s) + 2 MgO (s)
Mg (g)
è SiO2 (s) + 2 Mg (g)
è Mg (liq, s)
(~1200 - 1500°C under vacuum)
(temperature < 1090°C, condensation)
From the Ellingham diagram, the reaction is thermodynamically unfavourable but occurs, utilising Le
Chatelier's principle, due to the Mg removal by condensation at temperatures < 1090°C.
9 www.rocksandminerals.com/aluminum/process.htm
10 www.chemguide.co.uk/inorganic/extraction/titanium.html or www.madehow.com/Volume-7/Titanium.html
11 Chen G.Z., Fray D.J. & Farthing T.W., Nature, 407, 361-364 (21 September 2000)
rob.wallach@msm.cam.ac.uk
Page 9 of 61
2014-15
Materials Science Part II
University of Cambridge
C9 Alloys
Lecture 1: Metals and alloys
Si
Ti
Figure 1.8 Ellingham diagram
[www.2classnotes.com/digital_notes_print.asp?p=Thermodynamics_of_Metallurgy]
Figure 1.9 Elemental abundance in earth's upper continental crust
[http://en.wikipedia.org/wiki/Abundance_of_the_chemical_elements]
Figure 1.10 Aluminium production: Hall–Héroult process
[http://en.wikipedia.org/wiki/Hall-Heroult_process]
rob.wallach@msm.cam.ac.uk
Page 10 of 61
2014-15
Materials Science Part II
University of Cambridge
C9 Alloys
Lecture 1: Metals and alloys
Figure 1.11 Titanium production: Kroll process
[www.titaniumexposed.com/titanium-industries.html]
.
Figure 1.12 Titanium production: FFC process
[www.ornl.gov/sci/propulsionmaterials/pdfs/Emerging_Titanium.pdf]
rob.wallach@msm.cam.ac.uk
Page 11 of 61
2014-15
Materials Science Part II
University of Cambridge
C9 Alloys
Lecture 2: Aluminium alloys
2. Aluminium Alloys
2.1 Typical Phase Diagrams
Since aluminium exists solely with fcc crystal structure, there are no allotropic phase changes - unlike
Fe, C, Si, S all of which can exist individually in several different crystalline forms.
Control of microstructure and hence properties is by alloying, mainly through solid-solution hardening
and precipitation. Up to 70 wt% of zinc can dissolve in aluminium (at temperatures ~ 400°C), followed
by magnesium (17 wt% at 450°C), copper (5.7 wt%) and silicon (1.65 wt%).
Typical eutectic and peritectic phase diagrams are illustrated in Fig. 2.1; these two forms describe the
vast majority of phase diagrams for aluminium alloys.
Figure 2.1 Typical phase diagrams for aluminium alloys, illustrating a eutectic and peritectic.
The solubility of solute in the aluminium matrix phase (either denoted by α or by κ) depends on the
phase with which the α is in equilibrium. In the Al–Cu system, the stable precipitate is CuAl2 but
metastable GP1 zones can form preferentially as they are easier to nucleate (they have lower surface
energy so the activation energy is less). Thus, the free energy curve for GP1 zones is located above
that for CuAl2, as shown in the figure below. The common tangent construction shows that this leads
to a higher solubility of copper in α when it is in equilibrium with GP1 zones at a particular
temperature – see points P and Q. In addition, note that for a given composition, a greater
undercooling is required before GP1 zones can precipitate (see points R and S in the figure below).
R
P
P Q
Q
S
Figure 2.2 The solubility of solute in α is larger when it is in equilibrium with GP1 zones compared
with when it is in equilibrium with CuAl2, as is also shown on free energy (G) diagram.
In the Al–Cu system, the enthalpy of mixing, ΔHM is positive. Hence, at low temperatures, there will
be a tendency for like atoms to cluster, giving rise to a miscibility gap (Fig. 2.3). The enthapy of
mixing is given by:
ΔHM = Na z (1 − x) x ω
where ω = εAA + εBB − 2 εAB
x is concentration of B in binary A,B solution
z is the coordination number
Na is Avogadro’s number
By including the entropy changes when mixing two types of atoms, the thermodynamics of solid
solutions can be described and the free energy of mixing ΔGM is given by:
ΔGM = ΔHM - T ΔSM
= Na z (1 − x) x ω + Na kT {(1-x) ln (1- x) + x ln (x) }
[See www.dopitpoms.ac.uk/tlplib/solid-solutions/thermodynamics.php]
rob.wallach@msm.cam.ac.uk
Page 12 of 61
2014-15
Materials Science Part II
University of Cambridge
C9 Alloys
Lecture 2: Aluminium alloys
(see www.doitpoms.ac.uk/tlplib/solid-solutions/thermodynamics.php)
Figure 2.3 A eutectic phase diagram with a hidden miscibility gap (left) and free energy of
mixing plotted as a function of temperature and enthalpy ΔHM of mixing.
In the above figure:
- ΔHM = 0
corresponds to an ideal solution where the atoms of different species always tend
to mix at random and it is always the case that δµA/ δxA > 0;
- when ΔHM < 0 the atoms prefer unlike neighbours and it is always the case that δµA/ δxA > 0;
-
when ΔHM > 0 the atoms prefer like neighbours so for low temperatures and for certain
composition ranges δµA/ δxA < 0 giving rise to the possibility of uphill diffusion.
The miscibility gap at any temperature can be determined by the usual common tangent construction.
Noting that the regular solution model has symmetry about x = 0.5, the compositions corresponding
to the common tangent construction can, in this special case, be obtained by setting
δΔGM
δx
that is
Na z (1 − 2 x) ω + Na kT ln(
=
0
x
) = 0
1− x
and, in the limit of small x (low solid solubility), this gives
x = exp{–zω /kT}
The solubility therefore changes exponentially with the reciprocal of temperature, and increases as
ΔHM tends to zero. This is illustrated for a variety of solutes in aluminium in Fig. 2.4 below.
Figure 2.4 Solubility of a variety of solutes in aluminium.
Copper has the largest solubility, i.e. the smallest enthalpy of solution. Solid solution strengthening is
useful but it leads only to an increase of about 40 MPa in the strength of commercial alloys.
However, the copper solubility decreases exponentially as temperature falls and this is used to
facilitate precipitation hardening in aluminium alloys.
rob.wallach@msm.cam.ac.uk
Page 13 of 61
2014-15
Materials Science Part II
University of Cambridge
C9 Alloys
Lecture 2: Aluminium alloys
2.2 Aluminium alloys and tempers
Two families:
cast products - use in as-cast condition so cooling rate and use of grain refiners important since
small grain size improves strength and toughness;
- many based on Al- 12wt% Si which is the eutectic composition and hence the
lowest melting point, and also on Al- Zn (for die-casting especially);
wrought products (rolled, extruded, forged);
- final properties rely on appropriate heat-treatment/working;
- alloys are in two major classes of either precipitation hardened or work hardened;
- precipitation alloys: strength controlled by precipitate distribution, size and
coherency;
- worked alloys: strength from final grain size and defect density.
Alloys are classified by their compositions according to a US system:
Main additions
Strengthening
Major applications
1xxx
pure aluminium
work harden
Foil & electrical conductors
2xxx
Al-Cu
Ppte: CuAl2
Al-Cu-Mg for aircraft wing sheet
3xxx
Al-Mn
work-harden
General purpose – cookware
4xxx
Al-Si
“Si needles”
Casting alloy, modified with sodium
5xxx
Al-Mg
work-harden
Very common, corrosion resistant, structural & boats
6xxx
Al-Si-Mg
Ppte: Mg2Si
Extruded products e.g. building artefacts, golf clubs
7xxx
Al-Zn-Mg
Ppte: MgZn2
High strength alloy
8xxx
Other incl. Al-Li
Ppte: δ Li3Al
Low density & increased modulus alloy e.g. aircraft
Possible strengthening mechanisms include:
work-hardening
solid-solution hardening
grain size control
two-phase alloys
precipitates of optimised spacing and coherency
oxide dispersion strengthening.
Temper designation
Aim: to describe the likely strength as a consequence of fabrication history.
Two families of suffixes after the alloy designation.
-Tx for precipitation hardened alloys where x is a number corresponding to a specific treatment,
e.g.
T4 solution treated and then naturally aged
T6 solution treated and artificially aged.
-F, O or Hx for wrought alloys
e.g.
F as fabricated
O annealed
Hx strain hardened and where the value of x denotes extent of hardening
H4 – half hard, H8 – fully hard.
T6 is a very common heat treatment for aircraft alloys (typically 2xxx, i.e. Al-Cu), which after solution
treatment might be aged for 6–8 h at 150–170ºC to obtain the required tensile properties.
More complex heat treatments also used, e.g. better combination of tensile and fatigue properties if
alloys aged for a shorter time at 150–170ºC, followed by some natural ageing at ambient
temperatures.
Natural ageing generally leads to secondary precipitation on a finer scale.
rob.wallach@msm.cam.ac.uk
Page 14 of 61
2014-15
Materials Science Part II
University of Cambridge
C9 Alloys
Lecture 2: Aluminium alloys
2.3 Pure aluminium for electrical applications:
-
highest electrical conductivity at ambient temperature than any other element for a given weight;
-
its weight specific electrical conductivity is about twice that of copper;
-
used extensively for overhead electrical cables;
-
high purity aluminium has a very low yield strength of ~7 MPa and so overhead power
distribution cables are reinforced with steel – see Fig. 2.5 below;
-
in some applications, e.g. electrical motors, the use of aluminium windings increases the
volume of the equipment, so copper is used instead despite the disadvantage in weight.
Figure 2.5 Aluminium overhead power line cable, reinforced with steel wire cores
2.4 Alloy family characteristics
1xxx
These are minimally alloyed and generally used in the annealed condition with yield strengths
σy ≈ 10 MPa. Applications: electrical conductors, chemical equipment, foil and architecturally.
2xxx
Al-Cu precipitation hardened alloys, and used extensively in civil aircraft due to their high
strength to weight ratios. These alloys generally are not fusion weldable, and aircraft are
riveted. Various alloy additions can be made to optimise their properties.
More recently, the addition of Li has been investigated. Lithium as an alloy element in Al alloys
is unique in that it, unusually for most alloying additions, significantly improves stiffness: each
1 wt.% of Li reduces the density of an Al-Li alloy by ~3% and increases the stiffness by ~5%.
The alloy 2090 has a composition of 2.7 wt.% Cu, 2.2 wt.% Li and 0.12 wt.% Zr.
3xxx
These are the Al–Mn or Al–Mn–Mg alloys with moderate strength ductility and excellent
corrosion resistance. The strength, at about σy ≈ 110 MPa, comes from dispersoids which
form in the early stages of solidification. The Mn concentration is restricted to about 1.25 wt%
to avoid excessively large primary Al6Mn particles. Magnesium (0.5 wt%) gives solid solution
strengthening and the Al–Mn–Mg alloy is used in the H or O conditions. Beverage cans
represent the largest single use of either aluminium or magnesium alloys.
A typical 3000 family Al-Mn alloy has the chemical composition Al–0.7Mn–0.5Mg wt.%.
4xxx
Al–Si is a simple binary eutectic with eutectic composition ~11 wt.% Si. Used only for castings
(e.g. aluminium car engine blocks with hyper euctectic composition of ~18 wt.% Si) or for
brazing other aluminium alloys (based on eutectic composition).
5xxx
The magnesium concentration is usually maintained to less than 3–4 wt% in order to avoid
Mg5Al8. The strength is in the range σy ≈ 40 – 160 MPa with rapid work hardening during
deformation. Work hardened aluminium alloys tend to soften with age because the
microstructure is not stable even at ambient temperature. Therefore, it is better to excessively
work harden and then to anneal to the required strength and stability. The alloys, which have
excellent corrosion resistance, are used to make the bodies of boats or vehicles. They are
readily welded.
6xxx
This Al-Mg-Si family of alloys relies on Mg2Si precipitation hardening. These alloys are readily
extruded and also anodized. Hence they are used both architecturally as well as for sports
equipment including bike frames.
7xxx
High strength alloy based on MgZn2 precipitation hardening. Tendency to stress corrosion for
greater than (Mg + Zn) > 6 wt.%.
8xxx
Various other alloys including Li–based with δ Al3Li pptes. Low density, increased stiffness
leads to aircraft applications. Problem with poor toughness.
Dispersion strengthened alloys: Molten aluminium is broken up into droplets which are immediately
oxidised on their surfaces. On compaction, the surface oxide breaks up into highly stable
dispersoids in an aluminium matrix. There are alloys with up to 20 wt% of alumina.
Other additions e.g. SiC or Si3N4 also used to provide “metal matrix composites”.
rob.wallach@msm.cam.ac.uk
Page 15 of 61
2014-15
Materials Science Part II
University of Cambridge
C9 Alloys
Lecture 2: Aluminium alloys
2.5 Precipitation
See Part IA and Part IB lecture notes which cover the following key points.
(a) Age hardening involves the rapid cooling of a solid solution from a high temperature to one at
which it becomes supersaturated so that ex-solution of solute (precipitation) begins on ageing.
Note that it is not just the solute concentration which is supersaturated but also the vacancy
concentration and the vacancies act as heterogeneous nucleation sites.
(b) The need to age in a manner to avoid precipitate–free zones. The latter form either due to either
vacancy or solute depletion in the vicinity of grain boundaries.
(c) The role of metastable precipitates in the development of precipitation hardening (more coherent
metastable precipitates, with lower surface energies, decreases the nucleation activation energy).
2.6 Oxidation and corrosion resistance
Aluminium oxide forms extremely readily, as shown by Al2O3 line location on the Ellingham diagram.
The natural alumina film (2–10 nm thick) protects in neutral environments but not in alkaline or in
strong acids (with the exception of concentrated nitric acid which is a strong oxidising agent). Note
that the nature of the oxide changes both with temperature (amorphous to crystalline as the
temperature is raised) and also with composition (spinels can form in Al-Mg alloys)
Anodising: Naturally formed oxide films can be thickened by immersion in hot acid to some 1–2 µm
thickness. Even thicker films (10–20 µm) can be obtained by anodising aluminium. This involves
making the component the anode in dilute H2SO4 solution. The film contains a cellular structure of
open pores; these can be sealed by boiling in water which makes the cells expand by hydration. On
drying the cells remain closed. The cells can be filled with dye before sealing to produce coloured Al.
An increase in the current density and voltage during anodising causes microscopic arcing which
locally induces the oxide to fuse and solidify rapidly. With sufficient arcing, a tenacious, hard and fully
dense alumina coating is formed. This plasma electrolytic oxidation process can be exploited in
making components such as rollers, which require wear resistance.
Corrosion resistance: Zinc in solid solution lowers the Al–Zn electrode potential; such alloys are
therefore used for cladding and as galvanic anodes for sacrificial protection. The presence of
intermetallic compounds in an aluminium alloy reduces corrosion resistance. For example, iron and
silicon compounds are regions where the alumina film is weakened. As a result, pure aluminium
corrodes at a much lower rate than alloys, and hence pure aluminium (or Al-Zn alloys) is often used
to clad aluminium alloys to protect against corrosion. Such cladding can be introduced by rollbonding together the two alloys of interest.
2.7 Fatigue
There are two major difficulties. Coherent precipitates are cut by dislocations; each passage of a
dislocation shears the particle, producing steps at the entry and exit sites, thereby reducing the
particle cross–section on the slip–plane (Fig. 2.6). This makes it easier for a subsequent dislocation
to cut the particle. Slip then tends to focus on particular planes, leading to stress concentrations
which promote fatigue. It is better therefore to have a mixture of fine, coherent and bigger semi–
coherent precipitates so that the danger of localised inhomogeneous slip is reduced.
Fatigue is also initiated at pores in thick aluminium components. This can only be controlled by
careful processing, and by rolling deformation where this is permitted.
Figure 2.6 The effect of a dislocation passing through a coherent precipitate.
rob.wallach@msm.cam.ac.uk
Page 16 of 61
2014-15
Materials Science Part II
University of Cambridge
C9 Alloys
Lecture 3: Titanium alloys
Titanium Alloys
3.1 Pure Titanium
•
Melts at 1670°C but use restricted to < 400°C
•
Density 4.51 g cm
−3
Pure titanium has excellent resistance to corrosion and is used widely in the chemical industries.
Ti forms a passive oxide film and so is resistant to corrosion in oxidising solutions. The corrosion
resistance can be further improved by adding palladium (0.15 wt%), which makes hydrogen evolution
easier at cathodic sites so that the anodic and cathodic reactions balance in the passive region.
Many chemical plants use steel vessels which are protected by titanium alloys; titanium sheet (0.5
mm thick) is frequently a loose liner or may be roll or explosion bonded to the steel. Titanium
condenser tubes are used in power plants and in desalination plants.
Titanium appears ideal, given its high melting temperature, for use in components which operate at
elevated temperatures, especially where large strength to weight ratios are required.
Figure 3.1 Specific strength (yield stress divided by density) versus temperature
However, if titanium alloys rub against other metals at elevated temperatures, the titanium alloy
oxidises extremely rapidly and causes severe damage. This limits its application, in the harsh
environment of aero engines, to regions where the temperature does not exceed ~600ºC.
The world production of titanium (currently by Kroll process) is, therefore, relatively small, hundreds of
thousands of tonnes, which compares with steel at 750 million tonnes per annum. Some 80% of all
the titanium produced is used in the aerospace industries. Car suspension springs could easily be
made of titanium with a great reduction in weight but titanium is not available in the large quantities
needed at the price required for automobile applications. The target price for titanium needs to be
reduced to about 30% of its current value for serious application in mass–market cars. This now is
potentially possible by replacing the traditional Kroll process and subsequent refinement by the FFC
electrolysis of fused salts – see Section 1.6.2.
The crystal structure of titanium at ambient temperature and pressure is close–packed hexagonal (α)
with a c/a ratio of 1.587. Due to the distortion from the ideal c/a ratio of 1.67, slip is possible on the
pyramidal, prismatic and basal planes in the close–packed directions and hence Ti alloys are ductile
at room temperature. However, twinning can occur especially when compressively loaded. At about
890°C, titanium undergoes an allotropic transformation to a body–centred cubic phase which remains
stable to the melting temperature.
3.2 Alloying of Ti
All elements which are within the range 0.85–1.15 of the atomic radius of titanium alloy
substitutionally and have a significant solubility in titanium – as per Hume Rothery rules.
Elements with an atomic radius less than 0.59 that of Ti occupy interstitial sites and also have
substantial solubility (e.g. H, N, O, C).
The ease with which solutes dissolve in titanium makes it difficult to design precipitation–hardened
alloys. Boron has a similar but larger radius than C, O, N and H; it is therefore possible to induce
titanium boride precipitation. Copper precipitation is also possible in appropriate alloys.
rob.wallach@msm.cam.ac.uk
Page 17 of 61
2014-15
Materials Science Part II
University of Cambridge
C9 Alloys
Lecture 3: Titanium alloys
Figure. 3.2. Plot of atomic radius versus Pauling electronegativity for elements. Notice many elements of similar
size to titanium, and B, H, N, O and C all fall in interstitial range of Hume–Rothery rules.
Alloying elements can be categorised according to their effect on the stabilities of the α and β phases:
Al, O, N, Ga, Zr and Sn are α–stabilisers;
Mo, V, Si, W and Ta
are β–stabilisers.
Cu, Mn, Fe, Ni, Co and H are also β–stabilisers but can form a eutectoid in principle. However, the
eutectoid reaction is frequently sluggish (since substitutional atoms involved) and is suppressed.
Figure 3.3. Phase diagrams for Ti alloys
Molybdenum and vanadium have the largest influence on β stability and are common alloying
elements. Tungsten is rarely added due to its high density. Cu forms Ti2Cu which makes the alloys
age–hardening and heat treatable; such alloys are used as sheet materials. It is typically added in
concentrations less than 2.5 wt% in commercial alloys.
Interstitials
Interstitials inevitably cause changes in the lattice parameters. Body centred cubic β Ti has three
octahedral interstices per atom whereas hcp α Ti has one per atom. The latter are therefore larger, so
that the solubilities of O, N, and C are much higher in the α phase.
Hydrogen is the most important interstitial. Titanium is capable of absorbing up to 60 at% of hydrogen,
which can also be removed by annealing in a vacuum. Hydrogen enters the tetrahedral holes which
are larger in b.c.c. than c.p.h. Thus the solubility of hydrogen is larger in β alloys. The enthalpy of
solution of hydrogen in Ti is negative (ΔH < 0).
H2 → 2H
0
ΔG = ΔG + RT ln
so that
a H2
pH2
a H2
⎧ ΔS ⎫
⎧ ΔH ⎫
= exp ⎨ ⎬ exp ⎨−
⎬
pH2
⎩R⎭
⎩ RT ⎭
As ΔH<0, the solubility decreases with temperature (Figure. 3.4). This contrasts with iron which
shows the opposite trend as temperature increases and hence is more susceptible to embrittlement.
rob.wallach@msm.cam.ac.uk
Page 18 of 61
2014-15
Materials Science Part II
University of Cambridge
C9 Alloys
Lecture 3: Titanium alloys
Figure 3.4. Solubility of hydrogen in titanium
Fusion reactors
Because of the decrease in hydrogen solubility with temperature, titanium had been considered as a
candidate material for the first wall of magnetically confined fusion reactors. The hydrogen-based
plasma is not detrimental since at 500ºC and 1 Pa pressure, the Ti does not pick up enough hydrogen
for embrittlement. An additional feature is that Ti resists swelling due to neutron damage.
Energy storage
A large enough concentration of hydrogen induces the precipitation of hydrides. TiH 1.5−2.0 has a cubic–
F lattice and its precipitation causes embrittlement due to a volume expansion of about 18%. There
are regions of hydrostatic tension at crack tips where it forms preferentially, leading to large increases
in the crack growth rate, some 50–fold during fatigue.
The hydride reaction can also be used to store hydrogen reversibly:
2.2 FeTiH1.04 + H2
↔ 2.2 FeTiH1.95
The energy to weight ratio for such a cell is about a tenth that of petrol. However, one problem with
this method of hydrogen storage is that hydride formation is accompanied by a considerable volume
expansion, which in turn can embrittle the alloy. Amorphous alloys of titanium are better in this
respect, since they do form hydrides and yet reversibly accommodate large quantities of hydrogen by
an expansion of the nearest– neighbour distance.
The Zr–Ti Laves phase Ti0.24Zr0.76(Ni0.55Mn0.3V0.065Fe0.085)2.1 has been found to reversibly
−1
accommodate nearly 1.5 wt% of hydrogen, with a battery rating of some 440 mA h g .
3.3 Specific alloys
α –alloys
Aluminium is the main alloying element. The combined effect of the α stabilisers is expressed as:
aluminium equivalent, wt% = Al + 1/3 Sn + 1/6 Zr + 10 (O + C + 2N)
If this exceeds about 9 wt% then there may be detrimental precipitation reactions, generally by the
formation of Ti3X which has an ordered h.c.p. structure.
The presence of a small amount of the more ductile β–phase in virtually all α alloys is advantageous
for heat treatment and the ability to forge. The alloys may therefore contain 0.5 - 2 wt% of Mo, e.g.
Ti − 6Al − 2Sn − 4Zr − 2Mo
where the Zr and Sn give solid solution strengthening.
A near– α alloy has been developed, with good elevated temperature properties (T < 590 ºC):
Ti − 6Al − 4Sn − 3.5Zr − 0.5Mo − 0.35Si − 0.7Nb − 0.06C
The niobium is added for oxidation resistance and the carbon to allow a greater temperature range
over which the alloy is a mixture of α + β, in order to facilitate thermomechanical processing. This
particular alloy is used in the manufacture of aeroengine discs and has replaced discs made from
much heavier nickel base superalloys. The final microstructure of the alloy consists of equiaxed
primary α grains, Widmanstätten α plates separated by the β–phase.
α + β alloys
Most α + β alloys have high–strength and formability, and contain 4–6 wt% of β–stabilisers which
allow substantial amounts of β to be retained on quenching from the β → α + β phase fields, e.g.
Ti − 6Al − 4V
rob.wallach@msm.cam.ac.uk
Page 19 of 61
2014-15
Materials Science Part II
University of Cambridge
C9 Alloys
Lecture 3: Titanium alloys
Al reduces density, stabilises and strengthens α while vanadium provides a greater amount of the
more ductile β phase for hot–working. This alloy, which accounts for about half of all the titanium that
is produced, is popular because of its strength (1100 MPa), creep resistance at 300ºC, fatigue
resistance and castability.
One difficulty with the β phase, which has a body–centred cubic crystal structure, is that like ferritic
iron, it has a ductile–brittle transition temperature. The transition temperature tends to be above room
temperature, with cleavage fracture dominating at ambient temperatures.
Burn–resistant β –alloys
Titanium fires can occasionally occur in aero engines or in titanium-based heat exchangers used in
the chemical industries.
The addition of chromium in concentrations exceeding 10 wt% helps improve the burn–resistance of
titanium alloys. The alloy Ti–35V–15Cr wt% has sufficient chromium to resist burning in an
aeroengine environment to temperatures up to about 510ºC.
Chromium is not effective in binary Ti–Cr alloys as a continuous film of protective oxide does not form.
Quenching from β: martensites
Quenching the β phase leads to the formation of h.c.p. α’ martensite. This is not particularly hard and
there are increasing quantities of retained β in the microstructure as the solute concentration
increases and the MS temperature decreases (Fig. 3.5). The habit plane of the martensite is close to
{3 3 4}β and the crystallographic relationships between the original β and the martensitic α’ are:
(1 1 0)β || (0 0 0 1)α’
and
MF
Mf
[1 1 1] β || ([1 1 2 0]α’
MS
.
Figure 3.5 Martensitic transformation from β
Transformation β to ω
The ω phase is metastable and forms from β in alloys based on titanium, zirconium and hafnium. It is
important because its formation generally leads to deterioration in mechanical properties. There is
also an increase in the electrical resistance as ω forms and, in Ti–Nb alloys, its formation influences
superconduction.
The β to ω transformation is diffusionless, occurs below the T0 temperature and frequently cannot be
−1
suppressed even by quenching at 11,000°C s . It is reversible and diffusionless but is not
martensitic in the classical sense since there is no invariant–plane strain shape deformation.
However, it does involve the coordinated motion of atoms.
Titanium aluminides Ti3Al and TiAl - intermetallics
The most successful of the aluminides has a lamellar structure made up of alternating layers of an
ordered hexagonal (D19) Ti3Al and α2 compound and ordered f.c.t. (L10) TiAl or γ. The γ phase forms
with its most closely packed plane parallel to the basal plane of the α2:
{1 1 1}γ || {0 0 0 1}α2
< 1 1 0 >γ || < 1 1 2 0 > α2
The lamellar microstructure is a direct consequence of this orientation relationship.
Ductility is about 4- 6% at room temperature with the γ aluminide being more ductile. Their densities
2
are about 4.5 g cm and the aluminium makes them resistant to oxidation or burning. The alloys have
been extensively studied for aerospace and automotive turbochargers because of their high strength,
low density and creep resistance. Possible replacement for some Ni superalloy turbine blades.
rob.wallach@msm.cam.ac.uk
Page 20 of 61
2014-15
Materials Science Part II
University of Cambridge
C9 Alloys
Lecture 3: Titanium alloys
3.4 Superplasticity
Superplasticity is ability of polycrystalline material to deform to very large strains without failure.
- strains greater than 200% common and several commercial materials can show over 1000%
while development alloys can exceed 4000%
- need two-phase microstructure which remains stable at processing temperature (T ≥ 0.5 Tm)
- grain size must be (and has to remain) very small, few µm diameter
-6
-2
-1
- strain rate must be low and is range limited (10 to 10 s )
•
m
σ = k [ε ]
where m is strain-rate sensitivity, and
for superplasticity 0.4 < m < 0.9
- consequence is resistance to unstable (or local) necking which otherwise limits elongation.
Many different alloys can be superplastically formed provided the above conditions are fulfilled.
The α + β alloy Ti - 6 Al - 4 V is readily superplastically formed using stresses ~ 10 MPa at ~ 950°C
and controlled strain rates as above
Exploit in simple forming operations such as blow forming / deep draw.
Can also incorporate diffusion bonding (solid-state joining) as part of a combined manufacturing
operation, e.g. to form stiff honeycomb structures. Apply pressure at elevated temperature for time to
promote bonding and then use inert gas at high pressure to cause superplastic shape change.
pressure
stop off so
no bonding
3.4.1 Theory
a) T < 0. 4 Tm
n
True stress - true strain described by σΤ = k εΤ
Load (P) instability, i.e. dP = 0, occurs when
dσ T
=σ
dε T
i.e. when equivalent uniform strain εΤ = n
Note if
dσ T
>σ
dε T
any local neck is stable and overall work-hardening occurs instead.
rob.wallach@msm.cam.ac.uk
Page 21 of 61
2014-15
1
2
Materials Science Part II
University of Cambridge
C9 Alloys
Lecture 3: Titanium alloys
b) T > 0.4 Tm
Additional time dependence, hence
•
m
σ = k' ε n [ε ]
Note that at T < 0.4 Tm
σΤ ≈ k εΤn
m ≈ 0.03 and
If minimal strain hardening occurs at higher temperatures, n ≈ 0
and
Relationship between m and uniform elongation at higher temperatures
•
m
P
A
σ = k [ε ]
=
also
δε
=
δt
•
•
ε
=
1/m
ε = ⎡1⎤
⎢⎣ A ⎥⎦
hence
⎡ 1 ⎤ ⎡δ l ⎤
⎢⎣ δ t ⎥⎦ ⎢⎣ l ⎥⎦
1/m
⎡P ⎤
⎢⎣ k ⎥⎦
⎡ 1⎤ ⎡ δ l ⎤
⎢⎣ l ⎥⎦ ⎢⎣ δ t ⎥⎦
=
- ⎡ 1 ⎤ ⎡⎢ δ A ⎤⎥
⎢⎣ A ⎥⎦ ⎣ δ t ⎦
=
Combining these two equations
- dA = A
dt
=
Hence for m < 1
1
m
1
A
P
k
P
k
1
m
1
m
1
A
(1-m)/m
area reduced more rapidly as the value of A decreases
•
as m => 1
stress increases with ε (viscous material) and stable neck forms in tension
as m => 0
material not strain rate sensitive and local unstable neck forms.
Measurement of m
•
•
Use two different strain rates ε1 and ε 2
Then
m =
d logσ
•
≈
d log ε
Δ logσ
•
=
Δ log ε
•
In practice, m varies with temperature, grain size, and strain rate ε
(
ln σ 2 / σ 1
)
⎛
⎞
ln ⎜ ε 2 / ε 1 ⎟
⎝
⎠
•
•
(which depends on stress σ)
log σ
m
10
-5
10
-3
-1
10
rob.wallach@msm.cam.ac.uk
log
strain rate
Page 22 of 61
2014-15
Materials Science Part II
University of Cambridge
C9 Alloys
Lecture 3: Titanium alloys
3.4.2 Models for superplasticity
Require: two-phase microstructure which remains stable at processing temperature (T ≥ 0.5 Tm);
stable fine grain size which is (and remains) very small, few µm diameter;
equiaxed grain structure with relatively high grain boundary (gb) misorientations;
similar strength of both phases to allow deformation of both with no gb cavitation;
ability for material to undergo dynamic recrystallisation rather than creep during
deformation to avoid build up of dislocation density or networks;
mobile gbs to counteract stress build-up at gb triple points and so facilitate gb sliding.
Observe: little grain growth and preservation of original small equiaxed grain structure;
shear at grain boundaries leading to mutual displacement of neighbouring grains;
continuous grain boundary migration, sliding and rotation;
low final dislocation density;
detrimental gb cavitation especially if two phases in duplex alloy have different hardness;
texture of material reduced.
Models:
a. Grain boundary sliding
relative shearing of neighbouring grains plus slip and/or diffusion to maintain grain continuity;
rotation and exchange of neighbouring grains (Ashby-Verrall see Fig. 3.6 below).
!
Figure 3.6 Superplasticity model of Ashby-Verrall
[Ashby M.F. and Verrall R.A, Acta Met., 21 p29, 1973]
b. Diffusional flow and creep as in deformation maps
Figure 3.7 Deformation map for copper with grain size of 100 μm
high stress: dislocation core diffusion and glide leading to dislocation climb;
low stress: grain boundary diffusion (Coble creep) or bulk diffusion (Nabarro Herring);
rate controlled by vacancy creation & annihilation at grain boundaries.
rob.wallach@msm.cam.ac.uk
Page 23 of 61
2014-15
Materials Science Part II
University of Cambridge
C9 Alloys
Lecture 3: Titanium alloys
c. Dislocation glide with dynamic recovery and recrystallisation
dislocations glide in a grain and reach the grain boundary and hence dislocation pile up;
pile-up stress results in a higher stress in the adjacent grain and hence slip initiates in it;
dislocations will also climb resulting in more slip;
dynamical recovery and recrystallisation also needed to prevent work-hardening.
3.4.3 Grain refinement methods (not limited to superplasticity)
a. Mechanical working (two phase if for superplasticity)
- ideally similar volume fractions and hardness;
- classical recovery and recrystallisation;
- final grain size determined by amount of initial
cold work plus recrystallisation temperature and
time;
- used for Ti α/β alloys, Al-Cu-Zn eutectics.
b. Mechanical working of other duplex alloys (e.g. eutectics, precipitation hardened alloys)
- typically alloys with < 10% precipitates;
- fine pptes < 0.1 μm inhibit dislocation movement
so subgrains smaller and recrystallises to form
small equiaxed grains with high angle gbs
(hence easier superplastic gb sliding);
- coarser pptes result in localised strain which
promotes high angle gbs on recrystallisaton;
- used for Al-Cu-Zr “Supral” alloys, and Al-Mg-Zr.
c. Thermal cycling through phase transformation
- repeated phase transformations result in fine
grains due to repeated nucleation at gbs;
- often need additional small pptes. to ensure
stable grain structure when superplastic forming;
- used for steels (and α uranium).
d. Phase separation from non-equilibrium phase
- quench to form martensite and then many
available sites to nucleate second phase when
tempering;
- also spinodal decomposition e.g. Zn – 22wt.% Al
e. Casting: - equiaxed zone enhanced if increase chill crystals and dendrite arm remelting;
- grain refine using incoculants, e.g. addition of 0.1 wt.% to Al alloys to form solid TiAl3 by
peritectic reaction while Al alloy remains liquid. Promotes heterogeneous nucleation.
rob.wallach@msm.cam.ac.uk
Page 24 of 61
2014-15
Materials Science Part II
University of Cambridge
C9 Alloys
Lecture 4: Magnesium alloys
4. Magnesium Alloys
4.1 Introduction
-3
Magnesium is the lightest of the structural metals (density of only 1740 kg m , although most
magnesium alloys have slightly higher densities due to the alloying additions needed.
The world production of Mg is relatively small compared to other structural metals such as Al and
steel. However, the alloy is being promoted strongly in China which now produces ~ 50% of the
world’s Mg using the Pidgeon process (see Section 1.6).
About half Mg usage is as an alloying addition to aluminium alloys, e.g. the 5xxx series of alloys such
as 5083 with 4.5 wt.% Mg.
Other metallurgical uses include:
-
pressure die castings, e.g. parts for the car industry to aid fuel economy including
transmission casings cast in AZ91D resulting in 25% weight saving over Al alloys
steering components using AM50A & AM60B alloys (more ductile)
instrument panels, intake manifolds, cylinder head covers, inner boot lid sections
GM Savana & Express vans in the USA use up to 26 kg of Mg alloys.
-
desulphurisation of steel and also removal of bismuth from lead
inoculation of grey cast iron (flake to spheroidal graphite to improve toughness)
sacrificial anodes to protect steel structures from corrosion, e.g. ships, oil and gas pipelines.
4.2 Magnesium alloys
Common magnesium alloys are shown below1.
Alloy designation
Alloying additions (wt.%)
Uses
Reasons for use
AZ91
9.0 Al, 0.7 Zn, 0.13 Mn
General casting
Good castability,
good mechanical properties at
T<150ºC.
High pressure die
casting
Greater toughness & ductility
than AZ91, slightly lower
strength.
Often preferred for automotive
structural applications.
AM60
6.0 Al, 0.15 Mn
AZ31
3.0 Al, 1.0 Zn, 0.2 Mn
ZE41
4.2 Zn, 1.2 RE, 0.7 Zr
AS41
4.2 Al, 1.0 Si
Wrought magnesium
Good extrusion alloy.
products
Specialist casting
General casting
Rare earth RE addition
improves creep strength at
elevated temperatures.
Pressure tight.
Better creep resistance than
AZ91 at elevated temperatures
but lower strength and
corrosion resistance.
Alloy designation - first 2 letters are principal alloy additions;
- subsequent numbers indicate nominal compositions of the 2 major alloy additions.
There are many other specialist alloys containing zirconium, expensive rare earths and even silver.
Current research aims to improve corrosion resistance and high temperature creep resistance of Mg
alloys.
1 http://mg.tripod.com/mggen.htm
rob.wallach@msmcam.ac.uk
Page 25 of 61
2014-15
Materials Science Part II
University of Cambridge
C9 Alloys
Lecture 4: Magnesium alloys
4.3 Major alloying additions
Al
improves castability
solid solution strengthens
precipitation hardening at low T (<120°C)
Mn
controls Fe content by
n
Fe-Mn compound form
increases creep resistance
improves corrosion resistance (removal of Fe)
Zn
improves melt fluidity
weak grain refiner
precipitation hardening
can tend to brittleness
Si
decreases castability
weak grain refiner
increases creep resistance
poorer corrosion resistance
4.4 Heat treatment of Mg alloys
There is relatively low solubility of the above alloying additions in Mg and so alloys are both solidsolution and precipitation strengthened. Heat-treatment of castings may also be specified to reduce
residual stresses and also improve corrosion resistance.
The heat treatment of Mg alloy castings is similar to the heat treatment of Al alloys. However, Mg
develops a negligible change in its properties when allowed to age naturally at room temperatures
following solution heat treatment (T4 temper) and so alloys tend to be artificially aged (T6) for
optimum properties.
Solution heat treatment
Solution heat treatment temperatures for Mg alloy castings are ~ 400ºC and soaking times range from
10 to 18 hours, depending on the alloy and dimensions of the component (may be longer for large
components).
Protective atmospheres of SO2 or CO2 generally are used when solution heat-treating due to the high
affinity of Mg for oxygen. Inert gases also may be used; however, in most instances, these gases
are not practical because of higher cost.
Alloys are air-quenched after solution heat treatment.
Precipitation heat treatment
Precipitation heat treatment temperatures are considerably lower than solution heat-treatment
temperatures and range from 160 to 260ºC with soaking times between 4 and 18 hours. At these
temperatures, a protective atmosphere is not required.
4.5 Wrought alloys
Elastic modulus is relative constant in different directions so preferred orientation (texture) does not
have significant effect.
_
Extrusion at relatively low temperatures orients basal planes & <1010> directions parallel to extrusion
direction. Similarly rolling with basal panes additionally parallel to surface of sheet.
Twinning occurs readily if compressive stresses || basal plane so lower proof stress in compression
than tension. Find ratios vary for different alloys but increases with fine grain size.
If wrought products strengthened by “cold reeling” [in which alternating tension/compression occurs],
find extensive twinning due to compression and reduced tensile strength.
rob.wallach@msmcam.ac.uk
Page 26 of 61
2014-15
Materials Science Part II
University of Cambridge
C9 Alloys
Lecture 5: Anisotropy
5. Anisotropy
5.1 Introduction
The response of many materials, natural and man-made, are not uniform when used in service, i.e.
the material behaves in an anisotropic manner. This is a consequence of the dependence of
properties on a material’s
atomic structure
chemical composition
microstructure and phase distribution
defects
fabrication history, i.e. deformation & heat treatments.
Anisotropy often can be beneficial and hence exploited in high technology applications, e.g.
- liquid crystals used in watches and many other electronic displays,
- piezo-electric effect used in sensors or to generate ultrasound waves,
- preferred microstructure of turbine blades in a jet engine, and
- fabrication of high strength polymers (ropes as well as synthetic spider’s silk).
Anisotropy can be described in two ways:
microstructural
and
crystallographic.
The former can apply to all materials including natural materials such as wood, while the latter clearly
applies only to crystalline materials whether natural (e.g. minerals such as quartz) or man-made
(many metallic alloys).
Figure 5.1 Microstructural anisotropy in wood and extruded commerical purity Al
[See DoITPoMS TLP “The structure and mechanical behaviour of wood” www.doitpoms.ac.uk/tlplib/wood/
plus DoITPoMS Sample 604 www.doitpoms.ac.uk/miclib/]
5.2 Microstructural anisotropy in polycrystalline materials
This includes the distribution and sizes of:
- grains;
- phases;
- precipitates;
- dislocations.
The extent of and also any variations in microstructural anisotropy generally can be characterised
using image analysis to capture an image digitally and then using image processing and an
appropriate statistical package. Hence the relative area fractions of different phases can be readily
assessed as well as orientation as shown using the approach shown in Fig. 5.2.
Micrograph of metal grains
Image processed
Measure no. of intercepts
using grid of parallel lines
Polar plot of mean
intercept length
Figure 5.2 Characterising microstructural anisotropy by measuring number of intercepts
using a grid of parallel lines which is rotated to various orientations.
th
[Russ J.C., “The Image Processing Handbook”, 4 Ed., Taylor & Francis Inc, 2002]
rob.wallach@msmcam.ac.uk
Page 27 of 61
2014-15
Materials Science Part II
University of Cambridge
C9 Alloys
Lecture 5: Anisotropy
Composites can present extremes of microstructural anisotropy, as is shown in Figure 5.3.
Fracture
stress
(MPa)
Fibre orientation (°)
Figure 5.3 Variation of fracture stress with fibre orientation for a composite
[http://aluminium.matter.org.uk/ Anisotropy module, page 5]
For composites, the analysis is relatively straightforward using the rule of mixtures, i.e.
σL = Vf σfb + (1 − Vf) σmb
where σL = fracture stress in longitudinal direction
σfb = fracture stress of the ceramic fibre
σmb = fracture stress of the metal
Vf = volume fraction of the ceramic fibres
Certain types of microstructural anisotropy in metallic alloys can be treated rigorously, e.g. the
variation of yield strength σY with the mean grain size d by the Hall-Petch equation:
σ Y = σ i + k d −1/2
where σi is the lattice friction stress (i.e. stress to move dislocations other than near gb pile-ups)
k is a constant.
However, since microstructural anisotropy in metallic alloys generally is less easy to describe than
that in composites, predictions of properties arising from it are more difficult than for composites.
5.3 Crystallographic anisotropy in single crystals
5.3.1 Transport properties such as diffusion, thermal and electrical conductivities
Single crystals have regular long-range arrangements of atoms/ions. However, the properties will not
be the same in every direction, unlike a gas, an amorphous solid, or even a polycrystalline solid.
Mechanical and physical properties generally show an angular variation with respect to the major
crystal axes in a single crystal and also reflect the inherent symmetry of the crystal. This is the basis
of the piezoelectric, pyroelectric and ferroelectric effects in perovskites, as was introduced in Part IA.
The variation in behaviour is described by Neumann’s principle1 (of symmetry), namely:
the symmetry elements of any physical property of a crystal must include the symmetry
elements of the point group of the crystal.
The variation of a property such as diffusion or thermal conductivity in a single crystal is described
using a second rank tensor. Diffusion in a material is described using Fick’s first law where the flux Ji
is given by:
dC
Ji = - Dij
dx j
where
dC
is the concentration gradient and Dij is the diffusion coefficient.
dx j
This can be illustrated geometrically by a representation surface, the shapes of which2 are
- sphere for cubic lattices (in which the lattice parameters a = b = c),
- a biaxial ellipsoid for hexagonal or tetragonal lattices (in which the lattice parameters a = b ≠ c),
- a triaxial ellipsoid for orthorhombic lattices (a ≠ b ≠ c).
1 As introduced in Course C4 Tensors in Section 1.9.
2 DoITPoMS TLP “Tensors” www.doitpoms.ac.uk/tlplib/tensors, section “The effects of crystal symmetry”.
rob.wallach@msmcam.ac.uk
Page 28 of 61
2014-15
Materials Science Part II
University of Cambridge
C9 Alloys
Lecture 5: Anisotropy
-
A representation surface for hexagonal or tetragonal lattices is shown below in Fig. 5.4.1 2
Ji
or grad C
Figure 5.4 Representation surface for self-diffusion in magnesium
[modified from DoITPOMS TLP “Anisotropy” www.doitpoms.ac.uk/tlplib/anisotropy]
If the concentration gradient is in the same direction as one of the principal axes (x, y or z) then the
resulting flux will be in the same direction.
However, if it is an any other angle to the principal angles, the consequence of the differences in the
magnitudes of Dy and Dz, is that the resulting flux is not in the same direction as the concentration
gradient.
The direction and magnitude of the flux may be calculated using direction cosines but also are shown
conveniently using the above representation surface. Using the representation surface, a line is
drawn to show the direction of the concentration gradient relative to the two principal axes as then:
-
the direction of the resulting flux is given by the normal to the tangent to the ellipse at the point
where line for the concentration gradient crosses the ellipse;
-
the (magnitude of the diffusion coefficient)
chosen direction.
–1/2
is proportional to the length of the radius in that
Consider diffusion in, say, an hcp Mg single crystal. A unit cell has two lattice parameters a and c,
and the ratio c/a ideally is √(8/3) or 1.633. This suggests that diffusion within the close-packed basal
plane (0001) should be higher than that normal to the basal plane, i.e. in the c direction or [0001].
-12
2
-1
This has been confirmed experimentally; the measured values were Dx or Dy = 1.85 x 10 m s and
-12
2 -1
Dz = 1.0 x 10 m s at a temperature of ~425°C.3 Hence the representation surface would be
similar to that in Fig. 5.4 above.
5.3.2 Mechanical properties: crystal orientation and texture hardening in single crystals
Plastic deformation by dislocation movement occurs when the applied force/s F on a body of crosssectional area A results in a shear stress τ in a slip system which exceeds the critical resolved shear
stress CRSS.4 A slip system is a combination of a close (or closest) packed plane and a close
packed direction in that plane.
The resolved shear stress on a close packed plane and in a close packed direction is given by:
F
τ =
cosφ cosλ
A
where ϕ is the angle between the tensile force direction and the normal to the slip plane and λ is the
angle between the tensile force direction and slip direction in the slip plane.
The product cosϕ cosλ is the Schmid factor.
1 As introduced in Pt II Course C4 Tensors in Section 8.1 “The representation surface for second rank tensors”.
2 DoITPoMS TLP “Anisotropy” www.doitpoms.ac.uk/tlplib/anisotropy, section “Anisotropy ellipsoid”.
3 McKie D and McKie C, “Crystalline Solids”, Thomas Nelson, p 367 1974.
4 Part IA Course D and Pt II Course C6 Crystallography in Section 6 “Deformation and texture”.
rob.wallach@msmcam.ac.uk
Page 29 of 61
2014-15
Materials Science Part II
University of Cambridge
C9 Alloys
Lecture 5: Anisotropy
(a) fcc and bcc single crystals
There are 12 independent slip systems in fcc metals of the form
number in bcc of the form
.
and the same
If a tensile stress is applied to a single crystal, two approaches can be used to find on which slip
system the resolved shear stress will be maximum: OILS rule or Diehl’s rule.1 The latter uses a
stereographic projection as shown in Fig. 5.5 for an fcc or bcc single crystal on which twenty-four slip
systems are shown. The approach to finding the slip system with the maximum resolved shear stress
is implicit in the figure and has been covered elsewhere.2
Note that the three corners of each stereographic triangle are of the form 100, 110, 111.
Figure 5.5 Stereographic projection for fcc or bcc crystal showing 24 triangles corresponding to the 24 slip
systems. [From Pt. II Course C6 Crystallography, Section 6]
The actual location of the tensile axis with respect to the particular (111) plane normal and [110] slip
direction affects the Schmid factor, and also can affect the number of slip systems that can operate
initially. Accordingly, the magnitude of the applied force F required for plastic deformation, varies with
the location of the tensile axis as represented on the stereogram.
From the expression for the Schmid factor, the load needed for plastic deformation will be minimum if
both angles ϕ and λ are 45°. Conversely, a higher load would be needed if the tensile axis was at
some other orientation, e.g. if it was near the <110> or <111> corners of a stereographic triangle.
This is texture hardening and applies to all crystal systems, not just in fcc which was used above to
introduce the concept.
(b) hcp single crystals and work softening
Texture hardening is very evident in hcp single crystals in which there are only three slip systems,
namely the three close packed directions in the only close-packed plane, the basal plane (0001).
Hence if a force is applied at 90° to the (0001) plane, there can be no resolved shear stress in the
plane and so the yield stress is infinitely high (unless slip took place on a different plane).
As the angles between the tensile axis and slip plane normal and slip direction both are decreased
from 90°, the load needed for plastic deformation decreases until it reached its minimum value when
both angles ϕ and λ are 45°. As the angles decreased further below 45°, the load again increases
and becomes infinite if the applied force is in a direction parallel to the (0001) plane.
This can be observed in practice and the decrease in the necessary load is called work softening.
An example for a cadmium single crystal is shown in Fig. 5.6. Initially, the Schmid factor is high due to
the orientation of the tensile axis with respect to the single crystal. As deformation proceeds, the
tensile axis rotates towards the slip direction or, if the axis along which the load is applied is fixed, then
the slip system rotates during deformation. In either case, the Schmid factor decreases and so the
load for plastic flow also decreases.
1 DoITPoMS TLP “Slip in Single Crystals” www.doitpoms.ac.uk/tlplib/slip, section “Slip geometry”.
2 As introduced in Pt II Course C6 Crystallography in Section 6 “Deformation and texture”.
rob.wallach@msmcam.ac.uk
Page 30 of 61
2014-15
Materials Science Part II
University of Cambridge
C9 Alloys
Lecture 5: Anisotropy
Figure 5.6 Schematic load-extension curves for tensile deformation of two cylindrical cadmium crystals.1
In Fig. 5.6, the solid line shows the behaviour of a crystal showing work or geometric softening, where
the load decreases once plastic deformation commenced. The dotted line is for a sample in which no
geometric softening occurs, i.e. the tensile axis is at a different angle to the operative slip system. The
rise in load in both cases as the extension increases is due to “normal” work hardening.
(c) Variation of other mechanical properties, e.g. Young’s modulus
Given the different packing densities in different directions in single crystals, it is not surprising that
certain properties will reflect this. As an example, the anisotropy of Young’s modulus for an aluminium
single crystal has been estimated by calculation to increase by ~ 15% when the tensile direction is
rotated from the [100] to the close-packed [110] direction, as is shown below.
Figure 5.7 Effect of single crystal orientation on the Young modulus of Al.2
5.3.3 Magnetic anisotropy in single crystals
Magnetic anisotropy can occur as a consequence of:
- magnetocrystalline anisotropy where atomic lattice structure affects the ease of magnetisation;
- shape anisotropy where a magnetising field applied to a non-spherical particle is not the same in
2
all directions;
- magnetoelastic anisotropy where an applied stress can alter magnetic response.
B induced field
H applied field
Figure 5.8 Magnetocrystalline anisotropy: variation .in applied field H required to achieve the
same induced field B for two different crystallographic axes, i.e. easy and hard directions.3
1 DoITPoMS TLP “Slip in Single Crystals” www.doitpoms.ac.uk/tlplib/slip, section “Slip in HCP metals 4”.
2 http://aluminium.matter.org.uk/ - choose anisotropy section
3 DoITPoMS TLP “Ferromagnetic Materials” www.doitpoms.ac.uk/tlplib/slip, section “Domains”.
rob.wallach@msmcam.ac.uk
Page 31 of 61
2014-15
Materials Science Part II
University of Cambridge
C9 Alloys
Lecture 5: Anisotropy
Magnetocrystalline anisotropy occurs because the ease of aligning electron spins is affected by the
crystallographic direction in a unit cell. The terms “soft” and “hard” directions are used to differentiate
between the magnitudes of the applied magnetic field H required to induce a field B of a given
magnitude. This is shown in Fig. 5.8. Two consequences arise.
Firstly, magnetocrystalline anisotropy energy can be minimised by forming domains such that the
electron spins are aligned in easy crystallographic directions. However, in any domain walls, there
must be a change in the direction of the magnetisation and the electron spins will not all be aligned
along easy axes. Hence magnetocrystalline energy is minimised in materials with large domains and
correspondingly few domain walls.
Secondly, the hysteresis loop on reversing an applied field will be larger if magnetisation takes place
along hard directions rather than soft. There will be a higher energy associated with the larger
hysteresis loop and this means lower efficiencies (and more heat evolution) in applications such as
motors or transformers – see Fig 5.9.
B
H
Figure 5.9. Hysteresis loops for hard and soft magnetic materials
[www9.dw-world.de/rtc/infotheque/electronic_components/inductors.html]
Hard and soft directions in iron, nickel and cobalt are as follows:
Easy
Intermediate
Fe
bcc
<100>
<110>
Ni
fcc
<111>
<110>
Co
hexagonal
<1000>
Hard
<111>
<100>
<10 0>
5.4 Crystallographic anisotropy in polycrystalline metals and alloys
In a polycrystalline material with sufficient equiaxed grains in random orientations, the properties
would be expected to be isotropic due to the averaging of properties associated with individual grains
or crystallographic directions.
However, this is not common in practice due both to microstructural anisotropy (Section 5.2) and also
due to crystallographic anisotropy. The latter arises as a consequence of preferred crystallographic
orientations of many grains with respect to the overall shape of the body, e.g. in the rolling or
extrusion direction of a body. This texture is measured using X-ray diffraction to provide either pole
figures or orientation distribution functions.1
Texture in a polycrystalline material can be beneficial in that it can lead to optimisation of properties
for different applications, e.g. by increasing yield strength, improving formability, minimising magnetic
hysteresis losses. However, it also can be a liability if a material is “loaded” in service inappropriately
in a “wrong” direction.
1 See Part II Course C6 Crystallography, Section 6 “Deformation and texture”.
rob.wallach@msmcam.ac.uk
Page 32 of 61
2014-15
Materials Science Part II
University of Cambridge
C9 Alloys
Lecture 5: Anisotropy
5.4 Origins of crystallographic anisotropy in polycrystalline metals and alloys
(a) Solidification textures
The preferred orientation of dendritic and columnar grains is <100>. This is exploited for cast
magnetic materials which can be solidified in a temperature gradient to ensure a <100> fibre texture,
i.e. in grains grown parallel to the casting direction.
AlNiCo is a magnetic alloy fabricated in this way1 using resin-bonded sand moulds. Magnetic
properties are optimised after casting by heating them above their Curie temperature, and cooling in
an applied magnetic field at a controlled temperature rate. This takes advantage of both the <100>
fibre texture arising from the casting process and then the easy <100> magnetisation direction. The
final microstructure comprises iron and cobalt-rich precipitates in a Ni-Al matrix and the Curie
temperature of these magnets is ~ 800°C which is one of the highest Curie temperatures for any
magnetic material. In practice, AlNiCo magnets are used up to temperatures of ~ 500°C.
The composition of AlNiCo alloys is 8–12% Al, 15–26% Ni, 5–24% Co, balance Fe. The alloys also
may contain up to 6% Cu and up to 1% Ti.
Note that sintering is now an alternative production route for many magnetic alloys including AlNiCo.
In a different approach, a samarium-cobalt alloy was solidified in a magnetic field of several Tesla in
order to ensure the c-axis of the hexagonal alloy was aligned with the solidification grain structure.2
This orientated material was for bulk anisotropic permanent magnets.
(b) Deformation textures
During the deformation of a single crystal, the tensile axis or direction rotates towards the operative
slip direction or, if the direction along which the load is applied is fixed, then the slip system rotates
during deformation. This is to optimise the Schmid factor.
When a polycrystalline alloy is plastically deformed, the individual grains will try to rotate in a similar
fashion and there will be a resulting preferred orientation of grains relative to the tensile direction.
Deformation textures are described in terms of
- fibre textures [u v w] developed in uni-axial process such as extrusion or wire drawing
- sheet or rolling textures {h k l} <u v w> in which planes {h k l} lie parallel to the rolling plane and a
direction of the type <u v w> is parallel to the rolling direction.
Different textures arise in different alloys or in the same alloy deformed at different temperatures.
Examples:
- fcc pure metals or alloys with high stacking fault energies {112} <11 >
- fcc alloys including those with low stacking fault energies {110} <001>
- bcc alloys include {100} <011>
Note that a possible way to minimising a rolling texture is to consider cross-rolling, i.e. changing the
orientation of a sheet by 90° between rolling passes. This can be successful for some fcc alloys but
has the opposite effect in some bcc in which the rolling texture then becomes more pronounced.
(c) Annealing or recrystallisation textures
Consider the differences between a deformed and recrystallised alloy.
deformed
annealed
dislocation density
high
low
grain shape
elongated
equiaxed
yield strength
high
lower
texture
strong
strong and maybe more intense
1 www.fecrco.com/cast-alnico.html and also www.duramag.com/alnico.html
2 Legrand B.A. et al, “Orientation by solidification in a magnetic field. A new process to texture SmCo
compounds used as permanent magnets”, Journal of Magnetism and Magnetic Materials, 173, pp 20-28, 1997.
rob.wallach@msmcam.ac.uk
Page 33 of 61
2014-15
Materials Science Part II
University of Cambridge
C9 Alloys
Lecture 5: Anisotropy
During recovery of a heavily worked alloy, dislocations realign to form subgrains with low angle grain
boundaries within the original but now deformed, strongly elongated grains. During subsequent
recrystallisation, the subgrains with favoured orientations grow and so the resulting annealed alloy
has a more intense and different texture than the original deformed alloy.
Examples:
- fcc alloys with high stacking fault energies include {100} <001>
- bcc alloys {111} <1 0>
so-called cube texture
which is particularly advantageous when deep drawing to form cans
The variation with yield stress and texture in a recrystallised Al-Mg alloy is shown in Fig. 5.10. Note
that the highest value of the yield stress is for the sample cut parallel to the rolling direction. The true
strain to failure (elongation) is much greater for the 45° which has a value approximately 25% greater
than for the sample cut parallel to the rolling direction.
Figure 5.10. Uniaxial stress-strain curve for recrystallised alloy 5754-O
[http://aluminium.matter.org.uk/ - choose anisotropy section]
(d) Transformation or inheritance textures
When a diffusionless, displacive or martensitic phase transformation occurs, there will normally be a
crystallographic relationship between the original and new phases. The Bain model proposed for the
transformation of fcc austentite to bct martensite would suggest the following relationships:
[0 0 1]fcc || [0 0 1]bcc
[1 1 0]fcc || [1 0 0]bcc
[1 1 0]fcc || [0 1 0]bcc
Other variants include the Kurdjumov-Sachs relationship in which
{111}fcc || {110}bcc
and <101>fcc || <111>bcc
However, experimentally observed orientation relationships are irrational and not as simple as
,
above.1 2
Figure 5.11. Orientation relationships between parent (fcc) and martensite (bcc or bct) phases for
(a) Bain, (b) Nishiyama–Wassermann and (c) Kurdjumov–Sachs paths.
Blue atoms indicate a bcc unit cell. The red arrows indicate part of the motion initiating the transformation. The
2
dashed arrows indicate the invariant direction which is shared by the parent and martensite phases.
1 See Pt II Course C6 Crystallography, Section 8 “Crystallography of martensitic transformations”.
2 http://iopscience.iop.org/1367-2630/11/10/103027/fulltext/
rob.wallach@msmcam.ac.uk
Page 34 of 61
2014-15
Materials Science Part II
University of Cambridge
C9 Alloys
Lecture 5: Anisotropy
5.5 Applications of texture in practice.
A few examples where texture is beneficial include:
(a) strengthening of alloy sheet – see Section 6;
(b) drawing of sheet, e.g. drink cans in steel and aluminium alloys – see below;
(c) piezoelectric, pyroelectric and ferroelectric devices (as in Pt. IA, course B);
(d) iron-silicon transformer core steel – see below.
(b) Drawing of sheet, e.g. beverage cans in steel and aluminium alloys1
Drink cans are extremely thin and are made by extensive deformation by deep-drawing. The texture
of the alloy sheet prior to drawing is controlled and is chosen both to
- maximise the deformation while avoiding instability (local necking) in the thickness of the can;
- minimise the formation of so-called 'ears' on the final deep drawn cup - see Fig. 5.12.
Aluminium drink cans are made using two alloys:
- can body from highly formable 3104 H19 temper, nominally 1% Mn, 1% Mg, balance Al;
- can end from 5182 due to its higher strength, nominally 4.5% Mg, 0.3% Mn, balance Al.
The full process for the can body involves blanking, cupping and finally drawing and ironing the sidewalls. The can end is made by blanking, drawing, curl forming, riveting and production of the score
line for the easy open end.
After manufacture, the can body and can end are transported to a filling plant where the beverage is
put into the can and the two components are attached using a folded seam and a small amount of a
sealing compound.
a
b
c
Figure 5.12. (a) Stages in the drawing and ironing of a drink can.
(b & c) minimal and maximum 'earing' in can bodies as a consequence of anisotropy in the sheet deformation.
(d) Iron silicon transformer steel
The ferrous alloy used for sheet from which electrical transformers are made has to be magnetically
soft in order to minimise energy losses as the magnetic field reverses. This is achieved by controlling
both the composition of the sheet and its crystallographic texture.
An ideal alloy is grain orientated silicon sheet (Goss texture) Fe-3wt.% Si iron since the addition of Si
modifies the equilibrium diagram so that there are no phase transformations on heating, as well as:
- lowering the anisotropy constant K (a measure of the magnetic anisotropy);
- increasing the electrical resistivity which reduces eddy current losses;
- enabling one of two optimal textures (see Fig. 5.13)
•
cube on face or {100}<100> texture in which a {100} plane lies in the sheet plane, or
•
cube on edge {011}<100> Goss texture in which a {011} plane lies in the sheet plane which
can increase the magnetic flux density by up to 30% relative to a steel without this texture.
The silicon modifies the Fe-C phase diagram to that shown in Fig. 5.14. This stabilises the ferrite α
phase and suppresses the transformation to austenite γ even at the high processing temperatures.
1 http://aluminium.matter.org.uk/content/html/eng/default.asp?catid=84&pageid=-1941055071
rob.wallach@msmcam.ac.uk
Page 35 of 61
2014-15
Materials Science Part II
University of Cambridge
C9 Alloys
Lecture 5: Anisotropy
Figure 5.13. Cube-on-edge texture and the cube texture in grain oriented silicon steel.1
This avoids phase changes during processing which otherwise result in too small grains in the final
sheet and poorer soft magnetic properties; the final optimal soft magnetic grain size is ~ 5 mm.
A disadvantage of Si additions is that Si raises the ductile-brittle transition to room temperature for
alloys with 4 wt.% Si as is shown in Fig. 5.14. To maintain the single ferrite α phase field at lower Si
contents, e.g. between 2 - 3 % Si, the C content also needs to be low. Hence a typical composition
(in wt.%) is:
3.2 %Si, 0.03 %C, 0.08%Mn, 0.02%S.
Figure 5.14. Effect of Si on the equilibrium diagram of Fe-C alloys and
the effect of Si on the ductile brittle transition temperature.
The magnetically favoured Goss texture is produced by secondary recrystallization during high
temperature anneals in controlled atmospheres of hydrogen.
Texture formation is aided by fine particles of MnS which inhibit normal grain growth; the number is
subsequently reduced to avoid domain wall pinning in usage.
A typical manufacturing sequence will be of the form:
- hot roll at 1300°C to 2 mm and then remove oxide;
- cold roll to 0.2 mm in two steps with intermediate softening anneal at 800 - 1000°C;
- decarburise at 800°C in moist H2 which also allows recrystallisation during which the presence of
MnS particles helps to stop excessive grain growth;
- anneal in dry H2 at 1100 - 1200°C for several days to form the Goss texture by grain growth.
MnS, having fulfilled its role in the earlier recrystallisation step, also is reduced and this avoids
domain wall pinning in service; the resulting Mn goes into solution;
- shaped by cutting or punching followed by a stress relieving anneal at 800°C in dry N2.
Sheets may be coated with MgO before the anneal in dry H2 as this then produces a surface coating
of magnesium silicate which keeps the sheet in tension and so minimises stress magnetostrictive
losses.
1 http://softmagneticalloy.com/soft_magnetic_materials.htm
rob.wallach@msmcam.ac.uk
Page 36 of 61
2014-15
Materials Science Part II
University of Cambridge
C9 Alloys
Lecture 6: Steels
6. Steels and high-temperature materials
6.1 Iron-base alloys: overview
Iron
Steels
Stainless steels
Cast irons
---- plain carbon
---- ferritic α
---- low alloy & creep resistant
---- austenitic γ
--- grey -- flake or
-- spherodised
graphite (SG)
---- IF interstitial free
---- martensitic
--- white
---- thermo mechanical
---- duplex (α/γ)
-- whiteheart
-- blackheart
---- controlled transformn
-- HSLA
-- bainitic
--- others -- austenitic
-- dual phase
-- martensitic
-- TRIP
-- silal (6-8% Si)
-- TWIP
---- ultra high strength
-- secondary hardening
-- maraging
6.2 Alloying in steels - a brief guide (nearly all elements result in hardenability increasing)
Element
Influence on
ferrite
austenite
Effects on carbides
formation
tempering
Principal function
C
limited
solubility
stabilises γ
essential
coarsen
Mn
strengthens
markedly
A3 ↓ γ stabilised
less than Cr
little effect
Si
strengthens
stablises α
restricts
forming
Ni
toughens &
strengthens
stabilises γ
Cr
strengthens
stablises α
forms
carbides
resists
softening
- corro /oxid resistance ↑
- hardenability
ndry
- higher T usage (2
harden)
Mo
stablises α
creep props↑
promotes
carbides
[> than Cr]
resists
softening
- reduces γ coarsening
- minimise temper embrittle
ndry
- higher T usage (2
harden)
V, Nb
Ti
slight solid
solution
very strong
resists
softening
- reduce γ coarsening by
ppte pinning of gbs
rob.wallach@msmcam.ac.uk
carbide formers
- reduce S embrittlement
- cheap way ↑ hardenability
- deoxidiser
- oxidation resistance ↑
- use in transformer steels
- γ stabiliser (stainless steels)
- graphite former
- improve ferrite properties
Page 37 of 61
sion
n
2014-15
Materials Science Part II
University of Cambridge
C9 Alloys
Lecture 6: Steels
6.3 Range of available steels.
The wide range of commercially available ferrous alloys is summarised in section 6.1. They utilise the
majority of the different strengthening mechanisms summarised in Section 1.3 and take advantage of
the possible phase changes (both equilibrium and non-equilibrium) that exist in Fe-X systems, where
X represents one or more element.
Very different mechanical processing approaches and heat-treatments are used, and are tailored to
optimise the various strengthening mechanisms for individual compositions and subsequent usage.
The ultimate tensile strengths (TS) of steels range from 10 MPa for pure iron to 13 GPa for whiskers.
Material
TS (MPa)
Iron whiskers
Scifer
Ausformed (hardened) steel
Martensitic steel
Bainitic steel
Pearlitic steel
Cold-worked C steel
Low C steel
Pure, single-crystal iron
13000
5500
2930
2070
1380
1200
690
340
10
Scifer is a wire made by drawing 10 mm diameter rods of a dual-phase steel with a microstructure of
martensite and ferrite into fine wires of ~ 10 µm diameter, a very high true strain in excess of 9. The
very high strength arises from the very fine grain size and a dislocation cell size of ~ 10–15 nm.
While commercial Scifer wires have exceptionally high strength, components for many applications
require much larger dimensions and also have to be simple to fabricate. To achieve this:1
- the material must not rely on perfection to achieve its properties: strength can be generated by
incorporating the large number density of defects such as grain boundaries and dislocations, but
the defects must not be introduced by deformation if the shape of the material is not to be limited;
- defects can be introduced by phase transformations, but to disperse them on a sufficiently fine
scale requires the phase change to occur at large undercoolings (large free energy changes);
- a strong material must be tough in order to fail in a safe manner if it should do so;
- recalescence (heat evolved during a phase transformation) limits the undercooling that can be
achieved: therefore, the product phase must be such that it has a small latent heat of formation
and grows at a rate that allows the ready dissipation of heat.
These design criteria are helpful to bear in mind when considering the range of steels shown below.
IF
Interstitial free
(Section 6.3 3)
HS
High strength
HSLA High strength low alloy
(Section 6.4 a)
DP-CP Dual phase-complex phase
(Section 6.4 c)
HSS
High speed steel
ation
TRIP Transform
induced plasticity
(Section 6.4 d)
TWIP Twinning induced plasticity
(Section 6.4 e)
MART Martensitic
Figure 6.1. Strength and elongation for major classes of steel (categories as above)1
1 Bhadeshia H.K.D.H., “Bulk nanocrystalline steel”, Ironmaking and Steelmaking, 32 (5), pp 405- 410, 2005.
rob.wallach@msmcam.ac.uk
Page 38 of 61
2014-15
Materials Science Part II
University of Cambridge
C9 Alloys
Lecture 6: Steels
6.4 Steels – major categories
6.4.1 Plain carbon (C - Mn)
(a) Very low C C< 0.08%, Mn up to 0.4%
High ductility and toughness.
Easily deep drawn (e.g. cans) and formable (car bodies, domestic appliance bodies).
(b) Low C
C< 0.2%, Mn up to 1.5%, Si 0.15-0.35%
Mn present for solid-solution strengthening and also to tie up residual S (as above also).
Structural steel e.g. rolled products - plates or sections (e.g. I-beams).
(c) Medium C
C ~0.2-0.5%, Mn 1-2%
Can be quenched and tempered but low hardenability (better to use low alloy steels).
Forgings such as axles, gears, crank shafts.
(d) High C
C> 0.4%
Greater strength but at expense of ductility and toughness.
~ 0.4 -0.6% rails, railway wheels.
~ 0.8%
springs (e.g with Si 1.5%, Mn 1%) and drawn wire (aligned pearlite)
6.4.2 Interstitial free IF steels2
Controlled low levels of carbon and nitrogen give very high levels of formability – used to fabricate car
body panels.
Interstitial concentrations are <0.003 wt.% C and <0.004 wt.% N such that there is no C or N in solid
solution and this is achieved principally by:
- control of the melt chemistry during steelmaking;
- Ti, V and/or Nb additions to form precipitates, control of which during hot mill processing ensures
grain size of hot rolled sheet is as small as possible. Precipitates include Ti2CS (carbosulphide),
TiC, TiN, NbC, NbN (and V equivalents).
To optimise formablitiy (including deep drawability):
- Control (restrict) strain rate;
- ensure small grain size in the final cold rolled and annealed product;
- control crystallographic texture and so penultimate rolling of austenite is at temperatures just above
A3 to avoid rolling of ferrite at temperatures where it would recrystallise;
- cold roll ferrite and subsequently anneal to enhance drawability.
Solid solution strengthening is from Mn, P and Si. The absence of interstitials means deformation is
continous, i.e. no yield drop, and no Lüders bands which are detrimental to surface appearance.
1 www.keytometals.com/page.aspx?ID=CheckArticle&site=kts&NM=207
2 Hoile S., Materials Science and Technology, 16 (10), pp 1079-1093, 2000,
rob.wallach@msmcam.ac.uk
Page 39 of 61
2014-15
Materials Science Part II
University of Cambridge
C9 Alloys
Lecture 6: Steels
6.4.3 Thermo-mechanical processed steels
(a) High strength low alloy (HSLA) or microalloyed e.g. 0.1% C, 0.3% Si, 1.4-2.0% Mn, 0.05% Nb
Microalloyed with small amounts of carbide and/or nitride from elements such as Nb, V.
Precipitates (form in γ at 1450°C) help stabilise fine grain size, plus precipit hardening.
n
Additional strength from texture - recrystallisation & deformation by "controlled rolling"
Plate
thickness
Mean
plate temperature
Texture strengthening
Solution of precipitates (NbC, VC)
initial
rolling
Temperature
565
Rapid recrystallisation of γ
delay
535
Yield stress
505
MPa
Partial recrystallisation of γ
final
rolling
Precipitation hardening
445
A3
No recrystallisation of γ
Calculated base strength
800
A1
Time
Work hardening
475
750
700
650
600
Finish rolling temperature oC
Figure 6.2. Fabrication route for high-strength low-alloy HSLA steels and
effect of finish rolling temperature on the final yield stress
(b) Bainitic steels compositions similar to HSLA plus 0.002% B and 0.5% Mo
A wide range of bainitic steels now are available.1 Their strengths can be attributed to their small
ferrite grain size (bainite laths), uniform fine carbide dispersions, high dislocation density, and
solid-solution hardening.
These include high strength and toughness carbide-free bainites, whose microstructures comprise
1
fine plates of bainitic ferrite separated by C-enriched austenite. The advantages include :
- absence of cementite which can initiate fracture so the bainites are more resistant to cleavage
failure and void formation;
- the bainitic ferrite is almost C free, which would otherwise strenghen and hence embrittle it;
- the bainite ferrite plates, and so the effective grain size, are < 1µm thick – this fine grain size
accounts for the high strength and toughness;
- the ductile austenite films separating the ferrite plates have a crack blunting effect and
additionally can transform locally to martensite as a consequence of the stress field associated
with an advancing crack – this is the basis of TRIP steels and also partially stablised zirconia;
- stress corrosion resistance is higher as hydrogen diffusion is slower in austenite than ferrite.
These carbide-free bainites do not require expensive alloying. Sufficient Si is needed to suppress
carbide formation; a typical composition is 0.3 C, 1.5 Si, 1.5 Cr, 0.25 Mo, 3.5 Ni, balance Fe.
Figure 6.3. Mechanical properties of carbide-free
bainites versus quenched and tempered low-alloy
martensitic steels (QT) and marageing steels.
The two bolder points are the latest bainitic steels and
are some x30 cheaper than the equivalent marageing
1
steels.
1 Bhadeshia H.K.D.H., “Bainite in Steels”, 2nd ed., Institute of Materials. 2001. Chapter 13 “Modern bainitic steels”
Also download from
www.msm.cam.ac.uk/phase-trans/newbainite.html
rob.wallach@msmcam.ac.uk
Page 40 of 61
2014-15
Materials Science Part II
University of Cambridge
(c) Dual-phase steels
C9 Alloys
Lecture 6: Steels
0.1-0.2% C, 0.6% Si, 1.5% Mn, 0.05% V
Dual-phase steels have a soft ferrite matrix containing islands of martensite as the secondary
phase, the volume fraction (20 - 50%) of which determines the tensile strength.
Produced by holding / deforming at temperatures in the A1 - A3 region followed by quenching.
The properties of dual-phase steels include:
-
relatively low yield strength
-
low yield to tensile strength ratio (yield strength / tensile strength = 0.5)
-
high initial work-hardening rate
-
good and uniform ductility
-
strain rate sensitivity (“unusually” the faster it is deformed the more energy is absorbed)
-
good fatigue resistance
Due to the strain distribution in the two-phase alloy, there is no yield drop on plastically deforming,
i.e. no Lüders bands which are detrimental to surface appearance.
Dual-phase steels are used for body panels, wheels, and bumpers.
(d) TRIP steels
0.2%C, 2%Mn, 2%Si.
TRIP steels have excellent formability due to their high work hardening rates which can continue to
higher strains (up to 35%) than for dual-phase steels allowing more stretch forming. Their retained
austenite also provides energy absorption by its transformation to martensite in crashes when used for
cars.
TRIP steels have a matrix of ferrite α with a significant volume fraction (> 5%) of retained austenite γ
and small amounts of martensite and bainite. The Si and C contents are higher than in dual-phase
steels to assist stabilising the austenite to below room temperature.
During their heat-treatment, an isothermal hold at an intermediate temperature is necessary in order to
form the bainite. The Si (together with Al) helps to suppress carbide precipitation as the bainite forms
and this also is helped by the austenite which retains C in solution.
During subsequent deformation, the strain at which the retained austenite starts to transform to
martensite is controlled by adjusting the C content:
-
at low C contents, retained austenite begins to transform almost immediately to form martensite
and this increases both the work hardening rate and formability;
-
at higher C contents, retained austenite is more stable and transforms to martensite only at
strain levels beyond those produced during forming. Hence the retained austenite is part of the
microstructure of the final steel. It is beneficial as it then transforms to martensite if later
deformed, which can be beneficial as this absorbs energy, e.g. in a car crash.
Austenitic manganese steel 1.2% C and 12% Mn.
Hadfield manganese steel can be regarded as an historic variant of TRIP steel and are still widely
used today, generally for as-cast large tools. They are capable of withstanding extensive deformation
and wear and so are used in applications such as mining, quarrying, material handling, and coal
grinding mill rings used in power plants, as well as in earth moving machinery such as crushers,
shovels, and bulldozers.
Austenitic manganese steels are fully austenitic, obtained by solution heat-treated at temperatures
above the A3 to ensure complete homogenisation and then water quenched. During initial use, their
surface layers are plastically deformed them to a hard martensite which then gives excellent wear
resistance in service while the bulk remains austenitic with good strength, toughness and ductility.
The basis of the surface transformation is a form of TRIP, i.e. a transformation that occurs due to the
additional energy available from the imposed deformation.
rob.wallach@msmcam.ac.uk
Page 41 of 61
2014-15
Materials Science Part II
University of Cambridge
C9 Alloys
Lecture 6: Steels
start (1% transformation)
finish (99% transformation)
solvus
ΔG
ferrite α
martensite α ′
Md
Ms
Mf
gamma γ
Mf
Ms
Md
ln (time)
temperature
Figure 6.4. Martensite formation at a temperature below the martensite deformation temperature Md
but above the normal martensite start temperature Md.
Ms Martensite starts to form when free energy difference between metastable austenite and
martensite is sufficient to allow transformation (some energy needed to overcome strain and
surface energies).
Mf
Martensite transformation complete: the amount (%) of martensite when Ms < T < Mf depends
only on T, not time.
Md If the temperature is slightly above Ms, the transformation can be initiated by mechanical
deformation as this provides an additional contribution (strain energy).to the driving force for
the transformation which is sufficient to allow martensite to form.
(e) Twinning induced plasticity steels (TWIP)1 15 – 30% Mn, 3% Si, 3% Al, balance Fe.
The high Mn contents ensure the steel is fully austenitic at room temperature. When plastically
worked, extensive twinning occurs which results in a high work-hardening rate (i.e. a high value of the
work hardening exponent n, see Section 3.4) as the microstructure becomes finer and finer and the
resultant twin boundaries act like grain boundaries to strengthen the steel.
TWIP steels have both extremely high strength (> 1GPa) and high formability (elongations of 50%).
The n value increases to a value of 0.4 at an engineering strain of ~ 30% and then remains constant
until the total elongation reaches 50%. TWIP steels can show elongations of up to 90%
The ductility of TRIP steels is ~ 35% as a consquence of the collective shear of the austenite to
martensite but this then limits further ductility in that the martensite has a rigid bct structure.
The higher Mn, Si and Al in the TWIP steels provides additional ductility since two martensitic
transformations can now occur – first the austenite twins to an hexagonal martensite, and this
subsequently transforms to the bct martensite. The twinning is the cause of the additional ductility.
1 www.keytometals.com/page.aspx?ID=CheckArticle&site=kts&NM=207
rob.wallach@msmcam.ac.uk
Page 42 of 61
2014-15
Materials Science Part II
University of Cambridge
C9 Alloys
Lecture 6: Steels
6.4.4 Ultra-high strength (σ γ > 1400 MPa)
(a) Secondary hardening - tool steels (0.4% C, 5% Cr, 1.5% Mo, 1.0% V)
Alloyed heavily so good hardenability (air hardening) readily achieved.
The strong carbide formers result in alloy carbides on tempering martensite. The alloy
carbides are more stable than Fe3C which coarsens at higher temperatures (600-700°C).
Figure 6.5. Strength of tempered martensite as a function
of tempering temperature for both a plain C steel and an
alloy steel containing strong carbide formers.
(b) Maraging steels (0.02% C max., 18% Ni, 5% Mo, 7.5% Co, 0.4% Ti)
High alloy content allows martensite to form on air-cooling without the need for a high quench
rate and so residual stresses are minimised.
The low C content means the martensite is relatively ductile and can be formed, and with a low
work-hardening rate there is a low extent of uniform ductility.
Strength is then achieved by precipitation of Ni3Mo (and Ni3Ti) on ageing at 700-800°C.
1
6.4.5 Ferritic creep resistant steels e.g. 2¼% Cr, 1% Mo, 0.25% V and 0.1% C.
Ferritic steels based on Cr-Mo-V have been developed over many decades for use in power
generation applications where creep resistance is crucial.
Typically quench and temper to give an alloy carbide precipitate distribution of the form M23C6, NbC,
VN. These are stable at service temperatures at which these steels are used – so many variants.
Steels with lower alloy contents typically may have allotriomorphic ferrite/pearlite microstructures.
The higher alloy steels have greater hardenability (hence give a more uniform microstructure with
lower quenching strains) than plain C steels.
Typical compositions:
0.1-0.5% C, ~1.5% Mn with up to 3% Cr, 1.5% Mo, 0.5% V and/or 5% Ni
Higher chromium steels, up to ~10 % Cr have been developed to meet the increased demands of
operating the turbines at higher temperatures to improve their efficiency.
Usage:
gear, machine, car parts
easily surface hardened
power plant turbines, boilers and heat exchangers.
C 0.1% max., high Ni 5-9% cryogenic applications as good toughness at low temperatures.
rob.wallach@msmcam.ac.uk
Page 43 of 61
2014-15
Materials Science Part II
University of Cambridge
7.
C9 Alloys
Lecture 6: Steels
Temper embrittlement
Temper embrittlement can occur particularly in some quenched and tempered steels steels and is a
marked reduction in impact toughness (generally shown by an increase in the ductile-brittle transition
temperature) and fatigue properties after heat-treatment; ambient tensile strength is not affected.
There are two principal forms of temper embrittlement characterised by temperature:
(a) tempered martensite or 350°C embrittlement or “blue
brittleness” is irreversible and occurs in the range of
approximately 250–400°C;
(b) reversible temper embrittlement may occur when steels
are slowly heated or cooled through a temperature range
of 450–650°C.
Figure 6.6. Temper embrittlement: temperature dependence.1
(a) Tempered martensite embrittlement arises from the nature of the cementite precipitation on prioraustenite grain boundaries or interlath boundaries.
The prior austenite grain boundaries remain during martensite formation and so are where
carbides will heterogeneously nucleate preferentially during tempering. As the carbides grow and
coarsen [including metastable Fe2.4C], cracking and/or void nucleation occurs which leads to a
reduction in toughness.
(b) The higher temperature embrittlement is associated with just segregation of the impurity elements
(P, Sn, As, Sb) to the prior austenite grain boundaries. This leads to decohesion at the
boundaries, and again lower toughness subsequently. Since it is segregation, this embrittlement
is reversible by reheating to a temperature at which the impurity elements go back into solution
just below the A1 and then cooling again but at a faster rate so there is insufficient time for
diffusion to the grain boundaries.
The extent of embrittlement is affected by the prior austenite grain size and the hardness of the steel,
a fine-grained and softer steel will be less susceptible.
Alloying elements such as Cr, Mn and Ni tend to promote temper embrittlement and so it can be
prevalent in the ferritic creep resistant steels (section 6.4.5).
Small additions of Mo are beneficial in reducing embrittlement due to P.
1 http://steel.keytometals.com/articles/art102.htm
rob.wallach@msmcam.ac.uk
Page 44 of 61
2014-15
Materials Science Part II
University of Cambridge
8.
C9 Alloys
Lecture 6: Steels
Stainless Steels
Schaeffler diagram
(a) Ferritic (15-30% Cr, 0.1% C)
Good ductility (so formable) and stress corrosion resistance, cheaper than austenitic stainless
steels but high-temperature mechanical properties poorer. Toughness may be problematic at low
temperatures and in heavy sections.
(b Martensitic (12-17% Cr, 2% Ni max, 0.1-1.0% C)
High hardenability results in martensite on air-cooling from γ phase.
Often tempered to optimise strength and toughness.
(c) Austenitic (18-25% Cr, 8-20% Ni, <1% C)
Austenitic (fcc) at all temperatures with good corrosion resistance and ductility over wide
temperature range, depending on precise composition.
No ductile-brittle transition so used for cryogenic applications.
(d) Duplex (22% Cr, 5% Ni, 3% Mo, 0.1-0.2% N, 0.03% C)
Formed in α/γ region and controlled rolled to give very fine grain size (can be superplastic).
Similar corrosion resistance to austenitic alloys but better stress corrosion cracking resistance.
Higher tensile and yield strengths, but poorer toughness, than austenitic stainless steels.
(e) Controlled transformation (14-17% Cr, 3-5% Ni, 3% Mn, 0.1-0.3% C)
Precipitation hardened stainless steels with metastable γ at room temperature so that martensite
forms during fabrication or service deformation (e.g. TRIP).
As martensite forms during deformation, work-hardening rate increases so uniform ductility occurs
before plastic instability.
(f)
ODS ferritic (14% Cr, 2% Al, 1% Si, 0.3%Ta, 1% Y2O3)
These alloys are being of interest for creep resistant applications, e.g. interconnects for oxidefuel cells and in nuclear plants.
rob.wallach@msmcam.ac.uk
Page 45 of 61
2014-15
Materials Science Part II
University of Cambridge
C9 Alloys
Lecture 7-8: Copper and nickel alloys
7. Copper alloys
Pure copper
Good electrical and thermal conductivities, reasonable corrosion resistance ⇒ electrical applications.
Brasses
α
<36 wt% Zn
solid-solution strengthening
5 wt%
15
30
34
α/β
>36 wt% Zn
cap brass
gilding brass
cartridge brass
yellow brass
plumbing, jewellery
very ductile, deep drawing
major α brass, cheapest, workable, screws, tubes
two-phase alloy, stronger than α but need to hot-work
40 wt%
40 + 1-3 Pb
Muntz metal
leaded brass
extruded pins on electrical plugs
free machining
free machining brasses contain upto 1.5 wt.% Pb to aid cutting without long metal swaths
high tensile
Al, Fe, Mn, Sn, Ni additions
mainly cast, e.g. admiralty, but some can be wrought
manganese “bronze” α/β + ~ 2 % Mn
Bronzes
Sn
<10 wt%
9 Sn
19 Sn
leaded bronze
α phase solid-soln strength, corrosion resistant, ductile e.g. 3.5 Sn 1.5 Zn coins
“gun metal” – workable
casting alloy
< 4% Pb
cast alloy, free machining
~30% Pb
powder formed, sliding “self lubricated” bearings – low friction
phosphor bronze 5 Sn 0.2 P
high modulus, corrosion resistant
5-13 Sn 0.3-1.0 P
⇒ non –magnetic springs
bearing alloys containing Cu3P
Bronzes: non Sn
aluminium bronze 4 - 7 Al
7 - 9 Al
high strength wrought alloy, can be hot or cold worked
casting alloy with α/γ2 eutectoid
Nickel alloys
[see nickel sheet] cupro nickels,
monels,
nickel silver
Copper-beryllium (only precipitation hardened copper alloy)
2% Be
precipitation hardened (solution heat treat ~850ºC, age at 450ºC) ⇒ non–magnetic springs,
non–sparking tools
α brass [micrograph 430]
rob.wallach@msmcam.ac.uk
α/β brass: air cooled [442]
α/β brass: annealed [446]
Page 46 of 61
leaded gun metal bronze
Cu 85, Sn 5, Zn 5, Pb 5
[micrograph 513]
2014-15
Materials Science Part II
University of Cambridge
rob.wallach@msmcam.ac.uk
C9 Alloys
Page 47 of 61
Lecture 7-8: Copper and nickel alloys
2014-15
Materials Science Part II
University of Cambridge
C9 Alloys
Lecture 7-8: Copper and nickel alloys
8 . Nickel alloys
8.1 Range of nickel alloys
pure nickel
electrodeposit, electroformed, sintered powder (catalysts, filters, fuel cells)
ferrous alloys
γ stabiliser, α strengthening, toughness ↑
stainless steels
γ stabiliser
low alloy steels
γ stabiliser
ultrahigh strength
marageing steels
cast irons
Ni-hard (5 Ni, 2 Cr)
non-ferrous alloys
solid-solution strengthening ↑
> 50% Cu
cupro nickels
corrosion resistant
< 50% Cu
monels
improved properties
Ni Cu Zn
c.f. brasses but stronger
decorative, springs, zips
high-temperature alloys
corrosion/oxidation resistant
Ni Cr
Nichrome 80
oxidation resistant
up to 1000°C
Ni Fe Mo
Hastalloys
Ni Fe Cr
Inconels
hot acid resistant
Ni Cr Al Ti
“superalloys”
ODS
C tolerant
Nimonics
use at <1000°C
aerospace alloys
magnetic alloys
Ni Fe
40 – 80% Ni
soft
Al Ni Co
Al Ni Fe
hard
thermal expansion coefficient (α )
low α
Invar (60% Ni)
high α
72 Mn, 18 Cu, 10 Ni
shape memory alloys Nitinol
superelastic
rob.wallach@msmcam.ac.uk
constant modulus
Fe-Ni-Cr and Fe-Ni-Co
50% Ti - 50% Ni
shape memory
Page 48 of 61
2014-15
Materials Science Part II
University of Cambridge
C9 Alloys
Lecture 7-8: Copper and nickel alloys
8.2 Materials for gas turbines
8.2.1 Background
Requirements for materials in the hot regions of gas turbines of aircraft engines include:
- strength, toughness and ductility at both ambient and elevated temperatures
- stiffness
- creep and fatigue resistance
- oxidation resistance
- light weight (low density)
- reliability and reproducible
- easy fabrication at acceptable cost - even for high-technology products such as turbine blades, the
cost of the main constituent metal can be significant.
Additional factors to consider include:
- polymorphism since blades are cycled through wide temperature ranges and so repeated
changes of phase (e.g. bcc ↔ fcc in iron) would cause problems (e.g. refinement of grain size).
- intrinsic diffusivity since atomic diffusivities at a given homologous temperature T/Tm vary with
crystal structure. Relative diffusivities at Tm are fcc < bcc for many metals and so fcc alloys are
preferred.
- avoidance of high density to minimise weights of aeroengine components. Use of high-density
materials in rotating parts also results in higher centrifugal forces and so shorter creep lives and
greater chance of fatigue failure.
Initially consider melting points and ductility. High melting point alloys based on
- titanium: problem of strong affinity for and solubility of oxygen limits alloy use to T < ~600°C;
- cobalt: limited range of strengthening mechanisms compared to nickel alloys, also two crystal
structures (hcp and fcc) with transition temperature of 450 °C;
- steels: used initially for turbines but replaced by nickel alloys;
- nickel alloys: more options for high temperature strength and good oxidation resistance.
Historically: the range of alloys available to Whittle in 1937 for his gas turbine was relatively restricted:
- austenitic steel ("Stayblade"), originally developed for steam turbines, used for blades and discs
0.22 C
20.0 Cr
8.5 Ni
1.2 Ti (weight %)
- problems with yielding in service led to replacement by the more complex alloy G18B
0.4 C
13.0 Cr
13.0 Ni
10.0 Co
1.8 Mo
2.5 W
3 .0 Nb
- used for discs in Rolls-Royce Derwent engine (powered Meteor fighter aircraft in early 1940s).
Developments to find alloys stable at higher temperatures resulted in the awareness that:
- face centred cubic (fcc) alloys were more stable than body centred cubic (bcc) alloys;
- larger amounts of solid-solution alloying additions could be accommodated in fcc alloys;
- precipitation or age-hardening (initially interest focused on carbides) was readily achieved.
rob.wallach@msmcam.ac.uk
Page 49 of 61
2014-15
Materials Science Part II
University of Cambridge
C9 Alloys
Lecture 7-8: Copper and nickel alloys
Next generation of steels were strengthened by precipitation-hardening:
- precipitates based on Ni, Ti and Al such as ordered Ni3(TiAl) gamma prime γ'
- improved relatively low ratio of proof strength to ultimate tensile strength
- allowed the steels to be used at higher temperatures up to ~550°C
- further increases in age-hardening alloying elements led to decreases in ductility at both ambient and
elevated temperatures.
Substitution of nickel for iron led to the development of super alloys based on nickel matrix:
- nickel-base superalloys now used extensively for turbine blades enabling gas temperatures to
increase from 850 to 1500°C (i.e. above melting point of blade alloy), as well as for turbine disks;
- Ti 6Al 4V alloys (high strength to weight ratio) have replaced steels for fan and compressor blades,
and discs, with additional option of diffusion bonding and superplastic forming of fan blades.
Figure 8.1 Increase in turbine entry temperature (TET), by more than 700 K,
as seen in Rolls-Royce civil aero engines.
rob.wallach@msmcam.ac.uk
Page 50 of 61
2014-15
Materials Science Part II
University of Cambridge
C9 Alloys
Lecture 7-8: Copper and nickel alloys
8.2.2 Turbojet efficiencies
Engine efficiencies can be defined through the Brayton cycle, the constant pressure gas heating cycle
used by gas turbines. There are four stages:
stage 1: adiabatic compression of fresh air on entering the engine (S = constant)
stage 2: heating in the combustion chamber (P = constant)
stage 3: adiabatic expansion, with the turbine extracting work (S = constant)
stage 4: dissipation of hot gases (P = constant).
Figure 8.2 Thermodynamic Brayton cycle for gas turbines1
[ P, V, T and S are pressure, volume, temperature and entropy, respectively, and q represents heat in or out.]
Modern aircraft engines are turbofans in which some of the energy produced is used to power the fan
blades which can move a larger mass flow and so provide an increase in thrust and efficiency and a
similar temperature versus entropy diagram is shown in Fig 8.3.
Figure 8.3 Thermodynamic cycle for turbofans2
Energy changes are denoted by TdS on the T versus S curves. For both cycles (as well as for the
idealised Carnot cycle where there is no heat exchange with the surroundings), the area enclosed by
the lines on the T versus S curve is a measure of the amount of work energy W. The efficiencies are
given by expressions of the form:
Brayton efficiency
T − TC
η= H
TH
⎛P ⎞
T − TC
= 1 − ⎜ C⎟
Turbojet efficiency η = H
TH
⎝ PH ⎠
(γ −1)/γ
where γ is the heat capacity ratio, the ratios of the specific heats at constant pressure and constant
volume = Cp /Cv
Hence the higher the operating temperature, the more efficient is the engine and CO2 emissions also
are reduced. This provides the rationale for increasing the operating temperatures of gas turbine and
the need for materials which can operate at the higher temperatures.
1 http://en.wikipedia.org/wiki/Brayton_cycle
2 http://en.wikibooks.org/wiki/Jet_Propulsion/Thermodynamic_Cycles
rob.wallach@msmcam.ac.uk
Page 51 of 61
2014-15
Materials Science Part II
University of Cambridge
C9 Alloys
Lecture 7-8: Copper and nickel alloys
8.2.3 Development of nickel alloys for gas turbine applications
The use of nickel alloys for gas turbines evolved progressively in order to exploit the various
mechanisms strengthening mechanisms summarised in Section 1.3.
Ni matrix.
Nickel has a high melting point and is face centred cubic, and so has the potential for
good tensile and creep properties at both room and high temperatures (compared with
bcc alloys in which, for instance, diffusion rates will tend to be higher).
Ni-Cr alloys.
Nickel alloys alloyed with sufficient chromium have excellent:
- oxidation resistance and
- chromium also provides solid-solution strengthening.
Nichrome (80%Ni - 20%Cr) is still used extensively in heating elements and is resistant
up to approximately 1000°C. Alloys of this composition were originally used in a forged
condition in early turbines.
Ni-Cr with Ti and Al. Work on steels led to awareness that improved strength achieved by:
- gamma prime γ' ordered precipitates which have a nominal composition Ni3(Al)
- up to 65% of the Al can be replaced by Ti and so γ' is normally written as Ni3(TiAl)
- γ' precipitates are both
ordered:
providing additional strengthening
coherent:
minimising coarsening at elevated temperatures.
- adjustment of the Ti-Al ratio also affects the coherency with the nickel matrix
- two competing requirements:
- increased coherency for temperatures > 0.6 Tm in order to minimise coarsening;
- decreased coherency for lower temperatures to maximise the local strains and so
improve strength.
Optimisation of γ' leading to Mo, W and Al additions.
- reductions in the Cr content lead to higher volume fractions of γ'
- but also lead to reduced solid-solution hardening and oxidation resistance
- to maintain
strength: other solid-solution alloying additions added (e.g. Mo and W)
oxidation resistance: Al (4%) increased to form Al2O3 surface oxides.
Additions of Co and C led to higher creep resistance.
Co raises the γ' solvus (temperature at which precipitates go into solution in the nickel).
In addition, Co is hcp and can diffuse to and stabilise the stacking faults which exist
between partial dislocations. This Suzuki locking improves creep resistance by making
cross-slip of dislocations more difficult. The same phenomenon occurs in α brasses by
Zn segregation to the annealing twins during recrystallisation.
C forms carbides e.g.MC, M23C6 and M6C where M is a metal such as Ti, Nb, Cr, W, Mo.
These are very stable and form at high temperatures both within grains and at grain
boundaries. Boundaries then are pinned so creep resistance is improved.
Since the C is added principally to pin grain boundaries, it is not added to blade alloys but
is used in disc components.
Precise composition depends on the component as this determines both
the required properties and the fabrication practice.
rob.wallach@msmcam.ac.uk
Page 52 of 61
2014-15
Materials Science Part II
University of Cambridge
C9 Alloys
Lecture 7-8: Copper and nickel alloys
8.2.4 Grain size control and optimisation for different components
- key factor in optimising high temperature mechanical properties
- as grain size increases, creep properties enhanced considerably
strength, toughness and fatigue behaviour deteriorate
- gas turbine blades need maximum creep resistance (large grain size)
blades are relatively lightly stressed but operate at highest temperatures
- discs onto which blades attached operate at lower temperatures at high stresses. Hence
they need high strength and excellent fatigue resistance (small grain size).
8.2.5 Fabrication practice.
Vacuum casting of alloys plays a crucial role in the optimisation of properties:
- vacuum induction melting allowed the use of more reactive element additions but resulted in
extensive macro and micro segregation;
- vacuum arc remelting (now of various complexities) reduced solidification difficulties and also
provided an additional benefit in the cleanliness of the final alloy with a dramatic increase in
ductility (reduction of inclusions which act as stress raisers).
Investment casting of turbine blades allows
- economical fabrication of near finished (high dimensional control e.g. ± 0.06 mm) components
- inclusion of complex internal passages (using ceramic core technology)
- properties (creep particularly) of resulting blades optimised by using:
directional solidification (aligned grains)
single crystal growth.
Ceramic cores (based on silica or alumina) or drawn silica rods are included to form complex internal
patterns in turbine blades to improve gas cooling in service and hence increase operating
temperatures. The silica rods form fine channels (diameter ~ 0.4 mm, lengths up to 400 mm); the
silica is leached out of the final casting using pressurised KOH at elevated temperatures.
Optimisation of the castings is achieved by control of casting geometry, mould temperature, casting
temperature and cooling rate since these will govern grain orientations, carbide and γ' distributions.
The turbine blades, even single crystals, have to be homogenised after casting due to the inevitable
micro segregation that occurs during solidification.
Directional solidification of turbine blades is used to obtain <100> grain orientations as
solidification in these directions is preferred. Turbine blades with columnar grains with this orientation
extending along their length were favoured and these now are replaced by single crystal blades made
in a similar manner, with the addition of the “pig tail”. The pig-tail is a spiral crystal selection passage
located between the starter block and the blade cavity. As the aligned grains grow through this spiral
passage, their number is reduced until one remains and this grows into the turbine cavity to form the
blade. As this approach does not control grain orientations normal to the <100> growth direction,
seed crystals can be used in the starter block to optimise all orientations.
8.2.6 Blade coatings
- competing requirements of high temperature strength and excellent oxidation resistance
- wide range of coatings used to improve oxidation (and corrosion) resistance
- include vapour deposition of relatively thick (~ 1 µm) overlay coatings
- composition: mixed alloy comprising Cr Co Al Y
rob.wallach@msmcam.ac.uk
Page 53 of 61
2014-15
Materials Science Part II
University of Cambridge
C9 Alloys
Lecture 7-8: Copper and nickel alloys
8.2.7 Wrought alloys for discs
•
Requirements:
- high tensile strength so fine grain size & optimised precipitate size/spacing
- good ductility (relaxation at stress concentrators)
- high fracture toughness
- fatigue resistance
- oxidation resistance
•
Factors to consider - operating temperatures ~ 600-700°C are lower than turbine blades so
creep is less of an issue
- design based on tensile properties and not creep resistance
- toughness very sensitive to:
defects (need non-destructive NDT evaluation and quality control)
inclusions (minimise by use of clean melts)
•
Fabrication carried out by:
- forging at high temperatures, i.e. at temperatures where γ' is in solution (above γ' solvus),
and so alloy is easily deformed and ductile;
- solution heat treating to dissolve precipitates (especially γ' ) prior to controlled precipitation
and the precise temperatures (as high as 1200°C) and times are alloy dependent;
- precipitation hardening in a multiple ageing sequence to optimise size, spacing,
coherency.
8.2.8 Improvements in creep properties
The mechanisms for creep deformation are summarised on deformation-mechanism maps and two
such maps based on nickel alloys are shown below to show the improvements in potential creep life.
Map for superalloy1 with 10 mm grain size
similar to single crystal turbine blade
Map for pure Ni with 100 µm grain size
-1
The contours on the maps indicate creep rates (strain rate in units of s ). Alloying and the absence of
grain boundaries dramatically reduce the creep rate for the operating conditions under which turbine
blades in an aircraft engine are likely to be used (indicated by the rectangular box in the figures).
Diffusion, and hence creep rates would be further reduced in single crystal turbine blades by the
elimination of grain boundaries.
1 Composition of MAR–M200 in weight % is 5 Al
2 Ti
rob.wallach@msmcam.ac.uk
Page 54 of 61
9 Cr
12 W
10 Co
1 Nb
0.15 C
balance Ni
2014-15
Materials Science Part II
University of Cambridge
C9 Alloys
Lecture 9: Non-destructive testing
9. Non-destructive testing
9.1 Introduction
An increasing number of techniques can be used to examine, without causing damage, the integrity
of an individual component or an assembly, e.g. a completed printed circuit board on which there are
various components. It is carried out both to ensure that standards are being adhered to as well as
for the detection defects within components, either to monitor or to prevent jeopardising service life.
Common NDT methods, apart from visual, for examining individual metallic components include:
liquid dye penetrant
magnetic particle
eddy current
radiography (X-ray & neutron)
ultrasonics
acoustic emission.
When selecting a technique, factors to consider include:
defect location
type of defect
orientation of defect
ease and speed of testing
information required.
possible size of defect
number of componets to test
9.2 Dye penetrant – suface defects only
Basis:
Non-viscous dye (or fluorescent fluid) applied to penetrate cracks and then excess removed.
Observe cracks using a developer (or fluorescent light) which "pulls out" penetrant from defects.
Figure 9.1 Dye penetrant. 1. Clean surface. 2. Apply
dye penetrant and leave to penetrate any cracks.
3. Remove excess penetrant. 4. Apply developer into which dye from cracks bleeds out. 1.
Advantages:
cheap – minimal investment
sensitive to fine cracks
defect location clear
simple and non-skilled interpretation
copes with irregular and complex shapes
large surface areas rapidly inspected
Disadvantages:
surface defects only
need relatively smooth surface
non-quantitative
surface cleanliness critical as contaminants mask defects
slow (10-15 minutes)
post cleaning necessary to remove chemicals
[Excellent web site: www.ndt-ed.org/EducationResources/CommunityCollege/communitycollege.htm]
1 http://brennanndt.co.uk/#/dye-penetrant-inspection/4547126320
rob.wallach@msmcam.ac.uk
Page 55 of 61
2014-15
Materials Science Part II
University of Cambridge
C9 Alloys
Lecture 9: Non-destructive testing
9.3 Magnetic particle – suface and near-surface defects only
Basis:
Induce a magnetic field across a ferromagnetic sample and sprinkle on magnetic particles to show
the location of defects.
A magnetic field is induced in a ferromagnetic specimen either locally or overall, e.g. using a
permanent magnet or electromagnet. The induced magnetic flux remains concentrated below the
surface of the specimen if there are no defects, but a defect, e.g. cracks or voids, cannot support as
much flux. Hence the flux lines are distorted locally and some of the magnetic flux is “forced” outside
the specimen, i.e. flux “leaks” from the surface. This is detected using fine magnetic particles which
are applied to the surface and which concentrate in regions where defects are located.
Only surface and near surface flaws can be detected.
The induced magnetic flux lines are directional and depend on how they are introduced. Accordlingly,
the orientation of the defect is relevant and it is ideal to magnetise in two perpendicular directions, as
shown in the figure below, to ensure defects, and thin cracks especially in different orientations are
detected.
(a) Disruption of flux lines around defect
Magnetic
flux lines
(b) magnetic flux along bar
(c) circular magnetic flux
Figure 9.2. Magnetic particle inspection.
Adapted from1
The disruption of magnetic flux lines where there is a defect is shown in (a). Hence it is ideal to magnetise in two
perpendicular directions, as shown in (b) and (c), to ensure defects in different orientations are detected.
Advantages:
cheap and easy to use
low equipment costs
defect location clear
not affected by surface paint films <0.1 mm
surface preparation less critical than for dye penetrant
large surface areas rapidly inspected
Disadvantages:
ferromagnetic materials only
defect alignment is critical
non-quantitative
need to demagnetise after inspection
surface & near-surface defects only
non-magnetic coatings adversely affect sensitivity
1 www.milinc.com/services/nondestructive-testing-ndt
rob.wallach@msmcam.ac.uk
Page 56 of 61
2014-15
Materials Science Part II
University of Cambridge
C9 Alloys
Lecture 9: Non-destructive testing
9.4 Eddy current – suface and near-surface defects only
Basis:
Induce eddy currents near surface of component – detect defect by change in local current flux.1
Figure 9.3. Eddy current inspection. 2
When a current flows through a wire coil, designed as a probe, it generates a magnetic field. The
probe is then brought near to a metal specimen and so an eddy current is induced in the metal in its
near-surface region, the depth of which is material and test frequency dependent.
The eddy current in the metal metal generates its own magnetic field, and this interacts with the probe
coil through mutual inductance. Any change in the metal at or near its surface, e.g. defects,
thickness, conductiviity, alters the nature and amplitude of the local induced eddy current and the
resulting magnetic field. This feeds back to change the impedance amplitude and phase angle within
the probe coil alters and so allows identification of the changes in the metal; it can be quantitative.
The induced current density is highest very near the surface of the metal and then decreases. The
depth of penetration δ is usually defined as the depth at which the eddy current intensity I is 37% of
its surface value is Io when using a signal frequency f. This can be calculated from:
(
I = Io exp −δ
π f µσ
)
δ =
i.e.
(
π f µσ
)
−1
for 37% loss of signal
The signal intensity, test sensitivity, resolution, and penetration are affected by the material’s
magnetic permeability µ and electrical conductivity σ, and also by the nature and geometry of the
probe coil. As an example, a signal depth of ~1 mm can be achieved in a mild steel when using a
simple probe and a signal frequency of 500 Hz.
Higher signal frequencies increase near-surface resolution but decrease the depth of penetration.
The probe coil type, and size are selected to optimise test depending on the material being tested
and the purpose of the test.
The magnitude of the signal change can be calibrated using known standards and then this allows
the sizes of defects or coating thicknesses to be determined. The technique also can be used for
high speed applications such as wire drawing or inspecting the surfaces of plates, as the responses
to changes are very rapid. In such applications, differential probe coils are used, i.e. the coils are
wound in opposite directions and so give a null signal except when a discontinuity, such as a crack,
occurs in the metal being examined.
Advantages:
easy to use and automate
-1
high speed usage (up to 50 m s )
quantitative for some applications
non-ferrous and ferrous materials can be tested
no physical contact with the specimen
Disadvantages:
limited depth of inspection
orientation of defects critical
surface & near-surface defects only
1 www.eurondt.com/EDDY%20CURRENT.html
2 www.milinc.com/services/nondestructive-testing-ndt
rob.wallach@msmcam.ac.uk
Page 57 of 61
2014-15
Materials Science Part II
University of Cambridge
C9 Alloys
Lecture 9: Non-destructive testing
9.5 X-ray radiography
Basis:
Measure change in X-ray intensity (due to variations in absorption coefficients for different materials)
Figure 9.4. X-ray radiography. 1
The extent dI by which a material absorbs X-rays depends on its (mass) absorption coefficient µ.
Io
t
dI
= − µ dt
I
(
I = Io exp − µ t
x
Ix
I
)
⎛ ⎛ µ⎞
⎞
= Io exp ⎜ − ⎜ ⎟ ρ t ⎟
⎝ ⎝ ρ⎠
⎠
where (µ/ρ) is the mass absorption coefficient and ρ is density.
If there is a region of thickness x within the sample with a different absorption coefficient µx, the final
intensity for X-rays having passed through that region is given by:
(
Ix = Io exp − µ(t − x) − µ x x
)
Ix − I
≥ ~ 2%
I
and hence cracks in metals often are difficult to detect. This is especially problematic if fatigue is an
issue as well as for fracture toughness assessments.
To detect a difference, the contrast difference on the film (or recording medium)
X-ray radiography can be used to inspect almost any material for surface and internal defects. The
location of the defects will be shown but not their depths within the sample; the overall thickness of
the sample can be measured.
Advantages:
good for larger defects
inspect virtually all materials
minimal sample preparation
defect identification from “image” is straightforward
surface and subsurface defects detected
inspect complex shapes without disassembly
Disadvantages:
“thin” defects (cracks) not detected
radiation hazard potentially
equipment cost relatively high
sensitivity decreases with test piece thickness
access to both sides of structure generally needed
extensive operator training and skill required
Neutron radiography
Neutron and X-ray radiography are complementary techniques as different information is provided
due to the different nature of signal attenuation. X-rays are scattered and absorbed by electrons, and
so higher atomic number atoms interact more strongly. In contrast, neutrons interact with atomic
nuclei. Due to the complexity in generating a neutron signal, the technique is primarily used for
research, especially neutron diffraction for accurate stress analysis.
1 www.ndt-ed.org/GeneralResources/MethodSummary/MethodSummary.htm
rob.wallach@msmcam.ac.uk
Page 58 of 61
2014-15
Materials Science Part II
University of Cambridge
C9 Alloys
Lecture 9: Non-destructive testing
9.6 Ultrasonics
Basis:
Observe ultrasonic waves reflected from interfaces within a sample.
Figure 9.5. Ultrasonic testing.1
A piezoelectric crystal transducer is used to both generate and receive back ultrasonic sound waves
which are transmitted through a specimen. Reflection of part of the sound waves occurs from the
interface of any discontuity encountered (including the front and back surfaces of the specimen).
The reflected wave signals are displayed electronically in real time as a graph of reflected signal
intensity versus time elapsed from signal generation to when the reflection is processed. The
reflected signal travel time is directly proportional to the distance travelled by the sound wave in the
specimen. This provides information about the location, size, orientation of any defect encountered.
The velocity u of a sound wave is related to the density ρ of the material by the acoustic impedance Z
Z=ρu
and ultrasonic signals are a consequence of the reflection at a boundary betwen two materials with
different acoustic impedances. Hence the thickness of a “defect” is unimportant.
Ultrasonic testing has excellent penetration and so can be used on thick samples and also even “hairline” cracks give strong reflections. The signal from the transducer has to be coupled into the
specimen and so relatively good surface finishes are required – moreover, a front surface signal will
always be seen so that surface defects cannot be detected.
Where a good surface finish is not possible, e.g. on many fusion welds, a shear wave transducer can
be used as this does not have to be placed directly above the possible location of any defect as is
required for a longitudingal wave transducer (such as is shown in Fig. 9.5 above).
Relating a signal to the nature of a defect is extremely difficult in that reflections from boundaries are
similar regardless of the nature of the defect. A comparision between ultrasonics and X-ray
radiography is shown in Fig. 9.6.
a) incomplete penetration
b) slag inclusions (linear)
c) porosity cluster
d) crack parallel to sidewall
Figure 9.6. Ultrasonic versus X-radiographt for defect detection in welds
1 www.virtualengg.com/ultrasonic.html
rob.wallach@msmcam.ac.uk
Page 59 of 61
2014-15
Materials Science Part II
University of Cambridge
C9 Alloys
Advantages:
easy to automate
can inspect virtually all materials
single-sided access only needed
“hair line” cracks detected
Lecture 9: Non-destructive testing
defect size and location, including depth, recorded
depth of penetration superior to other NDT methods
instantaneous and permanent record
Disadvantages:
skilled interpretation needed
thin samples difficult to inspect
detects only sub-surface defects
surface finish can attenuate signals – need good finishes
Expression for ratio of intensities of reflected to incident waves in an ultrasonic test.
1. When a sound wave travels through a medium, the pressure P that it sets up locally is related to the
speed ú at which it is travelling by:
P =úZ
where Z is the acoustic impedance of the material, equals to the product of density ρ and speed of
sound cL in the material.
Z = ρ cL
2. Consider conditions at an interface during ultrasonic testing.
At the interface, must have
(a) no net pressure: continuity of strain hence stress
(b) no displacement u, hence nonet velocity ú
3. Consider 2(a) and treat Pi and Pr as forces (by considering unit area on which they act)
Pi
Pt
Pr
hence Pt = Pi + Pr
[equation X]
4. Consider 2(b), with no net displacement in x-direction, hence ut = ui + (–ur)
and
út=úi –úr
Pt
P P
= i − r
Z2
Z1 Z1
Since P = ú Z, then can substitute for ú, and so
5. Combining [X] and [Y] gives
Pr
Z − Z1
= 2
Pi
Z2 + Z1
6. Intensities are given by
Ir
Ii
rob.wallach@msmcam.ac.uk
⎡P ⎤
= ⎢ r⎥
⎣ Pi ⎦
2
and
Pt
2Z2
=
Pi
Z2 + Z1
⎡ Z − Z1 ⎤
= ⎢ 2
⎥
⎣ Z2 + Z1 ⎦
Page 60 of 61
[equation Y]
2
2014-15
Materials Science Part II
University of Cambridge
C9 Alloys
Lecture 9: Non-destructive testing
9.7 Acoustic emission
Basis:
Constant monitoring of “sound” waves to detect sudden change in stress in a material.
Figure 9.7 Acoustic emission1
If there is a sudden change in stress within a material as a consequence of a change in pressure,
temperature, load or defect propagation, transient elastic waves are produced. Sensors permanently
installed on the surface can detect these waves, which are similar to those used in ultrasonic NDT
although much reduced in intensity.
Acoustic emission sensors can be used on the surfaces of structures such as pressure vessels to
monitor the initiation and growth of cracks, slip and dislocation movements, twinning, and phase
transformations.
Note two key differences from the other NDT techniques:
no energy required to operate acoustic emission; it is a passive device waiting to record any
energy change within a component;
- it can monitors any changes occurring in a component in live time rather than measuring the
extent of any change after the change has occurred, providing that the changes are of sufficient
magnitude to be detected.
-
The technique is not quantitative. Also, given the relatively low signal intensity for many situations,
signal discrimination and noise reduction are required to ensure the signal is valid.
Other similar approaches are now being used and/or are under developed. These include:
-
piezoelectric paint2 – as being tested to measure vibrations on the Millennium bridge in
Newcastle as conventional sensors, accelerometers and strain gauges, are not appropriate. The
paint is spray painted to ~ 50 µm thick and is based on lead zirconium titanate PZT.
[Accelerometers are too expense and large numbers would be required while strain gauges,
although cheaper, are difficult and time consuming to install.]
-
carbon nanotubes mixed into paints3 – changes in electrical resistance can be monitored by
passing an electrical current through the paint. Cracking of the substrate material changes the
nanotubes’ electrical resistance and the information is transmitted wirelessly and then recorded.
This allows early detection of crack propagation and hence any cracks can be repaired.
1 www.ndt-ed.org/EducationResources/CommunityCollege/Other%20Methods/AE/AE_Intro.htm
2 www.ncl.ac.uk/press.office/press.release/item/989496353
3 www.strath.ac.uk/civeng/research/students/phd/davidmcgahon
rob.wallach@msmcam.ac.uk
Page 61 of 61
2014-15
Materials Science Part II
University of Cambridge
C9 Alloys
Question sheet
C9 Alloys Question sheet
[Additional questions are on the Examples Class sheet]
1. Explain:
(a) briefly by giving examples only and not listing all the classifications used, the basis of the
International Alloy Designation System for wrought aluminium alloys;
(b) the advantages of two-stage heat treatments for precipitation hardened aluminium alloys;
(c) why two alloys, designated 2014 in the T4 and T6 condition are used in the Airbus 320 and
why their respective yield strengths would be in the region of 275 MPa and 410 MPa;
(d) the absence of martensitic transformations in aluminium alloys;
(e) why precipitation hardened aluminium alloys may be clad with thin layers of pure aluminium or
an aluminium-zinc alloy.
2. Compare and contrast the strengthening mechanisms exploited in titanium alloys and in steels
(consider presenting your answer succinctly in tabular form). Illustrate your answer with specific
examples.
Explain why a titanium plate might show a higher yield strength when tested in tension than when
tested in compression.
3 (a) Aluminium is commonly produced by the Hall-Heroult process. Summarise the basis of the
process.
Using values from the Ellingham diagram (see your Pt. IB Data Book or available on the web),
estimate the open circuit voltage which would be needed to produce aluminium from alumina
using carbon. Explain why a voltage drop of 4-5 volts occurs in practice.
Estimate the energy [kWh] necessary to produce 1000 kg of aluminium. Comment on your
answer.
(b) Magnesium can be produced by the Pidgeon process in which dolomite ore is reduced by
ferrosilicon. Assuming values of the relevant activities in the dolomite, ferrosilicon and calcium
silicate for the MgO, Si and SiO2 respectively, calculate the partial pressure of the Mg vapour in
the process.
(c) Explain the Kroll process for the production of titanium and why the FFC process offers
improved energy efficiencies.
4. The properties of bulk materials are readily isotropic. Discuss the incidence of anisotropy in single
and polycrystalline materials, including how anisotropy may be an advantage in some applications
and a disadvantage in others.
Explain briefly the significance of anisotropy in the following:
(a) a deep-drawn drink can;
(b) a silicon-steel transformer core sheet;
(c) high-strength low-alloy steel plate;
(d) a quartz crystal piezoelectric oscillator.
5. Distinguish between “350°C embrittlement” and “reversible temper embrittlement” in quenched and
tempered steels, describing the micro-mechanism of failure in each case.
How might the steel-making practice and /or subsequent heat-treatments be modified to alleviate
such embrittlement problems?
[… continued overleaf]
rob.wallach@msm.cam.ac.uk
Page 1 of 2
2014-15
Materials Science Part II
University of Cambridge
C9 Alloys
Question sheet
6. Discuss the following:
(a) the possibility of improving the properties of materials by control of grain size and shape;
(b) the major alloying additions, and their rôles, to mild steel as a construction material;
(c) development of different categories of steels to improve formability;
(d) the requirements for high-temperature nickel-base superalloys and how their strengths are
optimised.
7. Explain the main distinguishing features in the transformations from austenite to upper and to
lower bainites.
What is the basis of the very high strength and toughness in some bainites, and how are these
properties obtained?
8. Many steels are highly alloyed, particularly with chromium. Explain why a minimum of ~ 12wt% Cr
would usually be added to a steel if good corrosion resistance is required.
Distinguish between ferritic and austenitic stainless steels, and summarise typical applications for
the two types of steel.
9. Briefly describe the basis of eddy current non-destructive testing (NDT). What are its advantages
and disadvantages compared to other NDT approaches?
[Calculation question on NDT included on Examples Class sheet.]
10. For what applications would the following materials be particularly suitable? What fabricated form
are they likely to be in, and what thermal or mechanical treatments (if any) might have been used
to optimise their properties? All compositions are weight percentages.
(a) 0.6% C, 4.0% Cr, 1.0% V, 18% W, balance Fe.
(b) 11.5% Si, balance Al.
(c) 0.3%C, 0.2%Mn, 13% Cr, balance Fe.
(d) 4.4% Cu, 1.0% Mg, 0.75% Mn, 0.4% Si, balance Al.
(e) 0.25%Si, 0.5% Mo, 5% Zr, 6% Al, balance Ti.
(f)
3.2% Si, 0.03% C, 0.1% Mn, 0.01% S, balance Fe.
(g) 0.15% C, 1.4% Mn, 0.04% Nb, 0.05%V, 0.25% Mo, balance Fe.
(h) 4.5% Mg, 0.15% Cr, 0.5% Mn, balance Al.
(i)
0.08% C, 0.5% Mn, 9% Ni, 18% Cr, 0.4% Ti, balance Fe
(j)
6% Al, 4% V, balance Ti.
(k) 3.6% C, 2.4% Si, 1.2% Mn, 0.05% Mg, 0.015% S, 0.07% P, balance Fe.
rob.wallach@msm.cam.ac.uk
Page 2 of 2
2014-15
Materials Science Part II
University of Cambridge
C9 Alloys
Examples Class questions
1. (a) Which properties result in titanium alloys being the preferred choice to aluminium alloys and
steels for certain applications? Give two specific applications.
In what respects are titanium alloys less advantageous?
(b) Titanium alloys are extremely versatile as many different microstructures can be obtained
through alloying and heat treatment, in a similar manner to steels. Describe which titanium
alloy would be most appropriate for each of the following applications, commenting on the role
of the alloying additions, and which thermo-mechanical treatment would you recommend:
(i)
moderate strength, very high resistance to corrosion and ability to store chemicals under
cryogenic conditions;
(ii)
very high strength and toughness combined with possibility of superplastic forming;
(iii) high temperature oxidation resistance?
2. What are the material requirements for an alloy to be superplastic? In what different ways can a
fine grain size be achieved in metallic alloys?
3. What do the A1 and A3 lines represent on an iron-carbon phase diagram?
Why are both dual-phase and TRIP steels partially fabricated in the region bounded by the A1 and
A3 lines? In what way do the final microstructures differ?
Describe the basis of the strengthening and toughening mechanism in TRIP steels, including both
sketches of free energy versus temperature and TTT curves in your answer.
4.(a) Discuss critically the statement that ultrasonic testing is more useful than x-ray radiography in
non-destructive testing.
(b) Derive expressions for the ratios of the intensities of:
(i) transmitted to incident x-rays for x-rays passing through a thickness t of material;
(ii) reflected to incident waves in an ultrasonic test.
(c) A welded aluminium alloy block, 2 mm thick, contains sporadic isolated oxide inclusions each
of thickness ~40 µm. By means of calculation, compare the relative merits of ultrasonic and Xray radiography in detecting such inclusions.
-1
-1
[Data. Speeds of sound in Al alloy and in oxide inclusions are 6350 m s and 8700 m s
-3
-3
respectively, densities of aluminium and oxide are 2700 kg m and 4020 kg m respectively,
2
-1
mass absorption coefficients of aluminium and oxide are 0.0063 and 0.0054 m kg
respectively for the particular X-ray energy used.]
5. Past Tripos question 2007 (18 on paper 2)
Time-temperature-transformation (TTT) diagrams are shown in Figures 1 and 2. One of these
corresponds to a 0.4 C, 0.2 Mo steel and the other to a 0.3 C, 2.0 Mo steel.
(a) Discuss the general form of these diagrams, explaining the terms, A3, A1, M5, M50 and M90.
(b) Explain the reasons for the curved solid and the curved dashed lines in the diagram and
confirm which diagram is the TTT diagram for the 0.4 C, 0.2 Mo steel and which is the TTT
diagram for the 0.3 C, 2.0 Mo steel.
(c) Explain which of these two steels would have greater hardenability.
(d) Draw possible cooling curves on Figure 2 which would produce:
(i) a fully pearlitic structure;
(ii) a fully bainitic structure.
(e) For a fully bainitic microstructure describe, using sketched micrographs, the microstructural
features present in both upper and lower bainites and explain the kinetic differences accounting
for each phase.
rob.wallach@msm.cam.ac.uk
Page 1 of 2
2014-15
Materials Science Part II
University of Cambridge
(f)
C9 Alloys
Examples Class questions
Draw on Figure 1 cooling curves which might indicate the cooling of the outer and inner regions
of a large ingot of steel. Explain why this might cause subsequent surface cracking problems.
(g) How might such problems be overcome using martempering?
rob.wallach@msm.cam.ac.uk
Page 2 of 2
2014-15