DESIGN, CONSTRUCTION, AND TESTING OF
A RAM EXTRUDER FOR HIGH VISCOSITY POLYMERS
BY
S. LUBISCHER
JOSEPH
SUBMITTED IN PARTIAL FULLFILLMENT OF
THE REQUIREMENTS FOR THE
DEGREE OF BACHELOR OF SCIENCE IN
MECHANICAL ENGINEERING
AND THE DEGREE OF
MASTER OF SCIENCE
IN MECHANICAL ENGINEERING
AT THE
MASSACHUSETTS INSTITUTE
OF TECHNOLOGY
f
JANUARY, 1973
SIGNATURE OF AUTHOR ,,,
./PARTMENT OF MECHANICAL ENGINEERING
OCTOBER 6, 1972
CERTIFIED BY
II
I
I I
I
I
I
t 1
l
Ir
l
I
I
I
r
II
I
I
I I
I
I
I
I
I I
I
I
I
I
I
I
I I
I
I
.HSIS SUPERVISOR
ACCEPTED BY
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
II
I
I
I I
I
II
I
II
I
I
CHAIRMAN, DEPARTMENTAL COMMITTEE
ON GRADUATE STUDENTS
Archives
FEB 141973
-2TABLE
OF CONTENTS
CHAPTER
PAGE
Title Page .........
....
.....
Table of Contents ..
List of Figures
....
· · · · · · · · ·...........
4
Abstract ...........
Acknowledgements
I
II
I II
IV
Introduction
· · · · · · · · ·...........
··
5
...
.......
... ........
7
...........
Literature Review ..
10
an d Flo w of Granu lar Media
II.1
Compression
II.2
Polymer Properties
II.3
Flow Behavior
of Po lymers
...
Experimental Work ..
. 10
...........
12
...........
19
..........
22
ults
Id Flo
III.1
Basic Concept of t he Ram Ext ruder ..... 22
III.2
Design
III.3
. .... ....
...... ties·
...........
22
Instrumentation ..
...........
28
III.4
Tests
and Res
...........
30
III.5
Limitations .
...........
32
...........
36
Analysis.
IV.1
Inapplicability
IV.2
The High Elastic St ate and St ick-Slip
of Viscous
Fl ow Analysis
Flow...............
IV,3
...........
36
38
The Theory of the H igh Elasti c State
Applied
to the Ram Extruder ............
IV.4
Heating
Time
IV.5
Powder
IV.6
Maximum Output Conditions .............. 48
...........................
Compression
and Air Expulsion
42
45
... 48
-3CHAPTER
V
P AGE
Conclusions
..................................
51
52
Appendix
A:
Detail Drawings .................
Appendix
B:
Equipment List .................. 62
Appendix
C:
Test Conditions, Results, and
Calculations
Appendix
D:
References
64
....................
68
Density Measurements ............
..................................
71
-4LIST OF FIGURES
FIGURE
PAGE
1.
Schematic of Screwless Injection Machine
2.
Viscosity
Data for UHMWPE AC 260-100
3.
Schematic
Conce ption
4.
Basic
5.
Top View of App aratus
Design of the
o f the
Ex
truder
14
.. .
*o
o
.
om
eeeoe
View of Apparat us
Bottom
7.
Transducer
8.
Velocity Traces
9.
Velocity Traces
Velocity Traces ress
ure
. .
11.
Flow
Rate
. .
Layo ut ....
10.
vs.
....
23
Ram Extruder
*
·..
US rop
..
eoooe
oemm
moeoo
mo·
mooeo
mo·
me·m·
moom
·
· ·
.
mo·
*
6.
....... 9
co
mo·
.
......
24
......
25
......
25
......
29
......
33
......
34
......
35
P ressure Drop for
Polybutadiene
12.
Calculation
of Tube Entrance
Temperature
Gradients ........................
41
45
.
-5-
ABSTRACT
DESIGN, CONSTRUCTION, AND TESTING OF
A RAM EXTRUDER FOR HIGH VISCOSITY POLYMERS
by
Joseph S. Lubischer
Submitted
to the Department
Engineering
fullfillment
of Mechanical
on October 6, 1972 in partial
for the requirements
for the
Degree of Bachelor of Science in Mechanical
Engineering
and the Degree of Master of
Science in Mechanical Engineering.
A ram extruder
for the direct
extrusion
of high viscosity
polymers from powder has been successfully designed and tested.
The extrusions,
in terms
of density,
surface
finish,
dimensional accuracy, were of very high quality.
heated by radial heat conduction,
stick-slip phenomena.
by theoretical
polymers.
and
The flow,
was found to exhibit
a
This has been satisfactorily explained
considerations
of the high elastic
state of
The mathematical technique and experimental data
needed for determining the maximum output of the extruder
while retaining a high quality extrudate is explained.
Thesis
Title:
Supervisor:
Nam P. Suh
Associate Professor of Mechanical Engineering
-6ACKNOWLEDGEMENTS
Any thesis
experience.
is first and foremost
As such,
suggestion
and insights
writer from many different sources.
acknowledge
all, but I would
to Nam Suh, my thesis
and encouragements.
Whitemore,
cone to the
It is impossible to
like to thank a few.
supervisor,
First,
for his help, criticisms,
To Jack Leach,
and Ralph Bowley,
an educational
Fred Anderson,
I express my thanks
Ralph
for their
assistance and advice in constructing the experimental
apparatus.
To my fellow students, I thank you for many
thought provoking discussions.
the Shell
Companies
Foundation
Thanks must also go to
and the Department
of
Mechanical Engineering for financial support and to
Mr. Allan Margolies, Allied Chemical Co., for providing
the polyethylene
used.
-7-
I,
INTRODUCTION
Among the many new polymers
are polyethylenes
introduced
in recent years
with high molecular weights.
Especially
interesting are those with weight average molecular weights
in excess of 1,750,000
u ltra-high
Termed
molecular weight
polyethylene (UHMWPE) [ 1], these polymers are seeing increasing usage because
particular,
of t:h eir unique physical properties.
UHMWPE wit:h a molecul ar weight
of 8 million,
like other polyethyler le s, is very tough and chemically
has a coeffieient
of fFr iction
flouroethylene (teflorl)
cessable
[2].
and a very high abrasion
c)n g
However, these
polymers
chained
, they simply
cessing mod es known
to be
inert,
polytetra-
are not easily proof UHMWPE
in conventional screw extruders.
"jam up" in the extruder.
in
which
resistance.*
Because of the very high viscosity
melts, they cannot be extruded
In practice
to that of
cornparable
In
use
The pro-
are counter-rotating twin-screw
extruders, compression molding, and a "screwless injection
machine".
Twin screw extruders are undesirable because the
large shearing
forces created
to the point where molecular
break up the molecular
weight,
ties, are significantly altered.
traditional
*
drawbacks
and thus physical
proper-
Compression molding has the
of low production
This UHMWPE has been kindly provided
is sold
chains
rates and of limitation
by Allied
under the designation AC 260-100.
Chemical
and
-8-
to single parts.
A "screwless
injection
machine"
has been
introduced by Borg-Warner Machines which utilizes mating conical
shear surfaces to heat the plastic, a powder metering screw,
and two stuffer plungers
respectively
for transporting
(Figure 1).
ment in processing
This machine
high viscosity
the powder
and melt,
is a significant
polymers.
However,
improve-
it is
still complicated, expensive and, more importantly, utilizes
shear work to melt the polymer.
The aim of this research
work has been the design,
struction,
and testing of a ram extruder
polymers.
In contrast
it is simple,
for high viscosity
to all of the above processing
inexpensive,
and introduces
con-
methods,
an absolutely
minimal amount of shear work, thereby minimizing physical
deterioration
of the polymer.
-9-
7-
'-
FE
S
Uw
\ 4AUzLUS (YARMerSC-)
\
OA/C
\ RTOTOR
SSAL
rmLOW v,4 tvr
FIGURE
1.
Schematic of Screwless Injection Machine
made by Borg-Warner Machines.
-10II,
LITERATURE REVIEW
This chapter reviews
to the design
and analysis
previous
work which has been relevant
of the ram extruder.
Not surprising-
ly, no work has been found dealing with the principles
extruder operation.
of ram
The areas searched were powder flow, powder
compression, polymer properties, and polymer flow.
II,1
COMPRESSION AND FLOW OF GRANULAR MEDIA
The only work found in the area of powder
the free flow of powdered
and flow along moving
views and analyzes
of powders.
and granular
surfaces
shear of granular
media
and Poltorak
along a surface
They found that the resistance
depended
considerably
also the rate of increase
covered a two-phase
force.
Adler
hoppers
[3] re-
the gravity
flow
force.
[4] investigated
at low normal
to shear (coefficient
of the shear stress.
of external
and
They also dis-
The first phase was a
by an increase
in the shearing
phase was uniform movement
One of the materials
the
stresses.
on the normal pressure
process of flow.
flow accompanied
The second
shearing
(such as blades).
through
Little work has been done concerning pressure
Platonov
stick-slip
materials
the factors which affect
flow of powders.
friction)
flow dealt with
under a constant
used was granular
poly-
styrene.
The compression of powdered materials has received signifi-
-11-
powders,
and Holland
Schwartz
cant attention.
[5], working with
show that the stress state for initial yielding
iron
is
given by the Mohr-Coulomb yield criterion
y
=
C +
tan~
(1)
where y is shear stress, a is normal stress, C is the cohesion
at zero normal stress,
is the angle of internal
and
friction.
decreases as densification proceeds.
However, C increases and
As a result, the failure envelope
curves
down, but can still
be reasonably represented by the classical Mohr-Coulomb yield
criterion.
Lee and Schwartz [6] applied the modified Mohr-
Coulomb criteria in an analysis of powder rolling.
papers
Several
[7,8, and 9] have studied the failure of solid polymers
pressure.
under hydrostatic
The data taken shows reasonable
agreement with the Mohr-Coulomb yield criterion.
Kawakita
and Ludde
[10] approach
the problem
compression via pressure-volume considerations.
of powder
They give
pressure-volume equations for piston, tapping, and vibrating
compaction.
For piston compression, the following expression
is given.
C
- V
_
Uo
o - V
Vo
abP
ab
1 + bP
(2)
-12-
where
C is the degree
of volume
V o is the initial
reduction,
apparent volume, V is the volume under pressure P,and a and b
are empirical constants.
empirical,
However, this equat ion is entirely
has not been related
the constants
to the state of stress, and
have not been related to any ph ysical
of the material.
Thus the usefulness
properties
of this expression
is
extremely limited
In summary, previous work indicates that the deformation
of powder compaction can be idealized through the use of the
Mohr-Coulomb yield criterion.
11 ,2
POLYMER PROPERTIES
Knowledge
of the physical
a great deal to be desired.
properties
of polymers
Because of the
leaves
wide number
of
polymers, co-polymers, and polymer mixes, data must be collected from individual manufacturers.
In addition, variations of
many properties with temperature, state of stress, and stress
rate are not well understood,
let alone readily
available.
Also the inherent non-linearity of such properties as viscosity
has made the synthesis
of useful constituitive
relations
very
difficult.
As a result, almost all design and analysis related
to polymers
and polymer flows begins with very limiting
assumptions regarding the nature of the properties.
VISCOSITY
In general,
the viscosity
the temperature-shear-shear
rate variation
of a polymer melt has been very difficult
to
of
-13characterize,
Recently,
even though a great deal of work has been done.
a computer
simulation
and 13] has utilized
several
of flow in extruders
power law models
[11, 12,
to give a single
viscosity curve.
Viscosity relations which have been proposed
for non-Newtonian
fluids are given in Table 2-4, reference
The most common,
power
law model
primarily
given
of its "usability",
because
[14].
is the
below:
n: fl
n =
(2-3)
nl1l
I
no
A common modification is to include temperature dependance
based on the Arrenh ius equati on:
n
The major
= Aexp(-AE/RT)II
problem w ith this model
empirically determined Newtonian
rates.
is shown
The actual viscosity
in Figure
also plotted.
2.
(2-4)
n - 1
is that it does yield
beha vior at high and low shear
data for the UHMWPE, AC 260-100,
For compar ison , the power law model is
This model,
using equa tion (2-4), is
n = (1.53-107) exp(-22.8/T) IjI-,7
The value of AE
the
was taken from McKelvey
[14].
(2-5)
-14,Inq
_
VIZCOSITYDA T AR
AC
Za-I
CI.
rt
DO
UHMWPF
viL.
co -eR
3o0 °C.
#PWov!ID)D BV
* '/%
ALLIED CHMetICA1tL C.,
/0'
Q
/O
1V7
ri
%J
r
WER,LQ
01T
rces
,
-
2
s
.f
SU4qr
ICr6'
To
(se )
FIGURE 2. Viscosity Data and Power Law Model
for UHMWPE AC 260-100,
-15SPECIFIC HEAT
A common
the
literature
value of specific
[15 and
16]
cp for polyethylene
given in
is
CPE = 0.55 Btu/lbm
The data in reference
heat for polyethylene
°F
(T=250C)
(2-6)
[16] shows the temperature
up to 320°K.
Forsythe
dependance
of
[17] gives the
temperature effect as
Cp = cpo(l + bT)
However,
(2-7)
no data has been found to support this equation
range of the polymer melt.
temperature
done to the author's
in the
Also, no work has been
knowledge on the relationship between
molecular weight and specific
heat.
manufacturer
heat of AC 260-100.
for the specific
No data is given by the
DENS ITY
The specific
gravity
of AC260-100
obtained by A.S.T.M. D792-62T.
PE
Naturally,
.g.PEPH2
=
is gi ven as .930,
The density is thus
93 .9975 = .928 gr/cc
due to compressibility
effects,
(2-8)
the density has a
strong temperature and pressure dependance. Though this phenomena
has been studied
, data is not generally
generated
by the
-16-
manufacturers.
Forsythe [17] expressed the density change as a
function of pressure and temperature:
p
[=
aTPdT+
pTdP
(2-9)
Substituting the definitions of thermal expansion and volume
compressibility he obtained
dp = -pc*sdT + pdP
Values of the constants
were obtained
equation
and
(2-10)
, not being generally
from manipulation
available,
of the Spencer-Gilmore
of state as will be noted below.
THERMAL CONDUCTIVITY
The thermal
temperature,
conductivity
has not been well established.
of heat transfer
in polymers
the thermal conductivity
not available
temperature
data, and its variation
except within
dependances
Nor has the mode
been well explained.
of high molecular
with
weight
an order of magnitude.
Values for
polymers
are
The
found [18] are in such disagreement
as
to be meaningless.
Data for thermal
[16] show a decrease
range -100°C to 50 0 C.
Lohe [19], and Hennig
conductivity
of polyethylene
in k with increasing
However,
temperature
in the
work by Fuller and Fricke
[20] in the temperature
300°C is in serious disagreement.
in reference
[18],
range 100°C to
As a result, no conclusions
-17can be made concerning
the temperature
dependance
of the thermal
conductivity of polymer melts.
Fuller and Fricke also noted that thermal
conductivity
decreased with increasing complexity of the molecular chain.
Hanson
and Ho [21] propose a theory, and present limited
to show that thermal conductivity
weight.
with molecular
They also give data that shows thermal
decreasing
reached,
increases
with temperature
at which
increasing very slowly.
for drawing
conductivity
until the melt temperature
point the thermal
data,
conductivity
is
starts
In general, however, data is insufficient
any hard conclusions.
CONSTITUITIVE RELATIONS
The author has found only one attempt
constituitive
relation.
In 1949, Spencer
to develop
a basic
and Gilmore
[22 and
23] propsed equation (2-11), the Spencer-Gilmore equation of
(P + r)
state, where
(V - b) = RgT/Mw
P is the external
pressure,
(2-11)
V is volume,
Rg is the
gas constant,
T is absolute
mer weight,
is internal pressure or cohesive energy density,
and b is volume
at absolute
to have been arrived
temperature,
zero.
Mw is the molecular
Equation
at from considerations
(2-11) appears
of the ideal gas
law
PV
= RT
(2-12)
and Van der Waale's
equation
(P + a/v
a
)(v - b)
(2-13)
= BT
cation of eq uation of (2-12) for higher pressures.
modifi
Though
2
-18of state,
1
imited, this equation can be used to derive expressions
for the coefficients of therm al expansion and volume compress-
ibility and the dens ity.
is strai
The derivation
+)
of the two coefficients
ghtforward, involving only substitutions of their
definitions , giving
1
(P
1
+
Rwb
(P
+
(2-14)
)4
and
=
An expression
equation
(2-15)
g
for density
can also be found ( by combining
(10), (14), and (15) and integrating).
the resulting
use.
+R
expression
is much too complicated
It should furthermore
equation
of state
However,
to be of any
be noted that the Spencer-Gilmore
has never been tested for polymers
ultra-high molecular weights.
with
-19-
II,3
FLOW BEHAVIOR OF POLYMERS
of this work, the author
At the beginning
on the flow characteristics
papers
work showed
later experimental
of polymer melts.
many
However,
that the flow in the ram extruder
exhibited a stick-slip phenomena.
Since the work on viscous
only a brief review
flow is not applicable,
reviewed
of major work will
be given.
In 1951, Brinkman
problem
[24] solved the steady
for a Newtonian
state flow
fluid with viscous heat generation.
The governing equation
=k
MrDTJ
_.
r2(aPJ2
arar
(2-16)
4+ [z
was solved by evaluating an eigenvalue series for boundary
conditions
(1) where
the tube walls are at To and the entering
fluid is at To and (2) where the walls are insulated.
results showed
very significant
viscous heat generation.
temperature
The
rises due to the
Bird [25] extended Brinkman's work
to the case of non-Newtonian
flow characterized
by a power law
viscosity.
Toor [26] investigated
the effects
of compressibility
the temperature profiles of power law fluids.
energy equation
=
=k
r
where
on
The governing
used was
rT
r r
+ n+2
n
2Te + n
n is the power law index,
2
rn
T
is the coefficient
(2-17)
of thermal
-20expansion,
and W is the average
across the tube.
Toor
[33] continues
the inlet region in order
to reach
steady
by the addition
entering
of the enthalpy
flow.
the distance
He solves equation
state flow.
dissipation
this work by examining
to determine
a tube with constant
the entering
volume rate of energy
transport
necessary
(2-17), amended
term, for a fluid
wall temperature
equal to that of
He shows that that the center
temperature
initially decreases due to expansion, but then increases with
axial distance
due to viscous
The Reynolds
heat generation.
number is a measure of distance
required
to reach a steady
velocity profile and the Peclet number is a similar measure of
the development of the temperature profile.
equation
for the distance
less group analysis.
profile
to reach steady
required
files and notes basic theoretical
agreement
For high viscosity
is pointed
The Prandtl number
relative
rates of development.
Lyche and Bird [27] obtained
state pro-
with the dimension-
fluids,
the velocity
than the temperature
develops much more rapidly
profile.
Toor develops an
out as a measure
a semi-analytical
of the
solution
for the Graetz-Nusselt problem extended to non-Newtonian flow.
Reduced temperature profiles for heating and cooling for
different
power
law indexes
are given.
out the "urgent need for experimental
tivities
years
and heat capacities
A final comment
data on thermal
of non-Newtonian
Toor [28] presented
Newtonian
a heat transfer
analysis
fluids with viscous heat generation
conduct-
fluids."
later this comment is still very applicable.
pointed
Fifteen
In 1958,
of power law nonin a pipe.
Gee
-21-
the basic equations governing
and Lyon [29] derive
here and give a computer
under discussion
the flow
solution thereof for
a particular material.
Christiansen
Newtonian
be important
a transient
generation.
[30,31, and 32] have worked
but neglected
flow problems
heat generation
Forsythe
et.al.
analysis
and 33].
due to viscous
the effects
which have been shown to
and compressibility
[24,25,26,
Gruntfest
et.al.
[34] present
of a power law fluid with viscous
heat
The work was based partly on that of Kearsley
[17 and 36] developed
state flow of a temperature
a numerical
dependent
solution
improvement
over the methods
for the cooling of polymers
well with experiments,
solution--application
a
for steady,
The results
of Bird [25] and Toor
[26]
These results, though agreeing
are limited by the nature
to
[35].
power law fluid with
viscous heat generation and compressibility effects.
showed
on non-
particular material.
of a numerical
-22-
III,
III.1
EXPERIMENTAL WORK
BASIC CONCEPT OF THE RAM EXTRUDER
The ram extruder
of high viscosity
was noted
polymers.
for continuous
The usefulness
in the introduction.
illustrated
stroke, uncovers
(5) through which
powder
A reciprocating
the wall friction
of the powder
pressure
in section
as A.
where
III,2
the polymer
(2) falls into
On the forward
After a few strokes,
is sufficient
to cause a
B every time the ram pushes
forward, thereby compacting the powder.
powder
ram
a port (3) in the feed
stroke, the ram pushes the powder forward.
of compacted
of such a machine
in the hopper
that section of the chamber denoted
build-up
processing
The basic concept may be
through the use of Figure 3.
(1), on every return
block
was designed
This continuous slug
is then forced through
a heated
tube (4)
is melted.
DESIGN
Figure 4 shows the basic extruder
contains
detail
of standard
drawings
equipment
design.
Appendix
of the parts and Appendix
used.
A
B is a list
Figures 5 and 6 are photographs
of the experimental apparatus, including instrumentation.
The extrusion
a
inch ID.
tube was 6 inches long with a 1 inch OD and
Four 125 watt band heaters were used.
Two
iron-constantan thermocouples with off-on controllers monitored
-23-
V)
'I,
5.-
w
-0
S-
x
w
E
co
-1
C)
0
c
0
c-,
C
ua
uL
c
4-)
E
3
0
a
P-
UC,
E
0)
Cr4
c-l
C;
t
LU
K~-
-24-
~.
=
9;
t
%
r-J
.J
x
CS
1
t
,
o
L
a
C)
k
cv
Lz
C.)
-_ c
H
4--
CC
7YM'~--II
FIGURE 5. Top View of Apparatus
FIGURE 6.
Bottom
View of Apparatus
-26the barrel temperature.
Thermocouples
A and B were mounted
11 inches from the exit and entrance
ends of the tube, respec-
tively.
probe was
The tip of the thermocouple
3 inch from
the inside wall.
The thermal insulator between the feed block and the
extrusion
tube was included
for two reasons.
First, it was
desired to maintain as sharp a temperature gradient as
possible for later theoretical considerations.
melted
polymer near the piston
cylinder.
rise in this part.
to jam in the
and materials
any
(During start up it was always
to turn on the water before
Several designs
tested.
causes the piston
The feed block was water cooled to prevent
temperature
necessary
Second,
powering
for the thermal
up the heaters.)
insulator
were
These included teflon, steel, machinable glass,
water cooled epoxy, and water cooled steel sleeve with a teflon
insert.
Ceramic materials were judged to be unsuitable.
of the materials
were not acceptable
because
Most
of low mechanical
strength, high thermal conductivity, surface roughness, or
brittleness.
steel-teflon
strength
The only successful design was the water cooled
composite.
teflon tended
(See Appendix
A.)
However,
the low
to flow out between the steel sleeve
the heated extrusion tube.
filled this space without
and
Fortunately, the polyethylene
apparent
effect
on the flow.
The
design was thus suitable for testing purposes though in practice
a more durable design would be necessary.
Some problems
were encountered
through the port in the feed block,
in the free flow of powder
particularly
after the
-27-
powder had absorbed humidity.
by simply enlargening
All dificulties were eliminated
the port.
Significant problems resulted from the dry sliding friction
between piston and cylinder.
mild steel.
Initially, both parts were made of
However, after a break-in period piston seizures
were repeated
and rapid.
Replacement
with a brass piston
and
additional securement of the bushing prevented further problems.
Periodic inspection of this arrangement showed signs of wear
on the piston which
The problem
eliminate
indicated
that the design was defficient.
is not overwhelming.
non-axial
Use of dual bushings
forces and use of hardened
to
steel or low
friction materials (such as teflon impregnated aluminum) should
eliminate difficulties in the future.
The air motor was run off an 80 psig line through
lator.
Valving on the air motor is controlled
switches accuated by the lever arm.
a regu-
by two micro-
By adjusting the switch
positions, the stroke distance can be controlled.
Shaft speed
is influenced by supply pressure, adjusting screws which restrict
the flow passages,
and load.
As a result,
piston speed was quite difficult.
a much
less sensitive
speed
fine adjustment
of the
Future apparatus should employ
control.
The lever arm was designed
stroke reduction factor of 3.
to give a force magnification
and
-28-
111,3
INSTRUMENTATION
Among the variables to be recorded were piston velocity,
extrudate velocity, piston displacement, and piston force.
The velocities were measured with a self-generating,
magnetic core type transducers (ISA#1-313 [37]).
of transducer
is suitable
of the measured
specifics
parameter
are given
apparatus,
for uses where essentially
can be tolerated.
in Appendix
it was necessary
B.
straight
coupling was thus employed.
arms.
were encountered
The arms were strengthened
was inserted
The instrument
to mount the transducers
piston and extrudate.
Initial problems
in the transducer
no loading
Due to the set-up of the
with the moving
shaft,
This type
in parallel
A bent arm, rather than
(Figures 5 and 7.)
due to the spring
effect of these
at the bend and a teflon
to prevent
lateral motion.
guide
This
eliminated all problems except that of piston velocity overshoot at the beginning of the return stroke.
Since the unloaded
piston velocity
was obvious
is a constant,
could simply be ignored.
this overshoot
and
The transducer outputs were fed into
an oscilloscope with camera attachment.
60 cycle pickup was
12 mp
for extrusion velocity.
for piston velocity and 4 mp
An LVDT (ISA#1-034)
ment.
was used to measure
Output was fed into a Sanborn
recorder
piston displaceand calibrated
with
depth micrometer measurements of piston position.
Piston
force was to be calculated
from measurements
differential pressure across the air motor piston.
taps were located in the cylinder
of the
Pressure
walls at each end of the
-29-
s,
x
-n
I-
'-
It"
.
ik,
0
%I
Co
-J
LLr!
VIl
WC
It,
c) Q
'hS,
-4j
1U
5-
V
-0
FiS i4
4.
,v
Li
CD
i~
t7
A
cr
-30motor and connected via
pressure
transducer.
"Poly-Flo" tubing to a differential
(See Figure 6.)
Output was fed to the
Sanborn recorder and calibrated by comparison to a calibrated
Bourdon
tube gage at 20 psid.
The transducer
was mounted
separately from all other apparatus to prevent vibration
pick-up.
111,4 TESTS
AND RESULTS
A series of test were run to determine
characteristics of the ram extruder.
The factors varied were
the stroke length, the barrel temperature,
cycling rate.
the operating
and the speed or
Stroke length was adjusted with the extruder
operating by moving the valve controlling micro-switches.
The travel was read directly
from the LVDT output.
The extruder
tube temperature was controlled by setting the temperature
controllers.
to adjust,
The speed of operation, as noted above, was difficult
This was accentuated
by the fact that if the machine
began produced a partially melted extrudate, the process usually,
even when slowed
down, went to a new steady state situation
only a thin outer layer of the extrusion
was melted.
where
The
author attempted to set the air motor speed adjusting screws and
the regulator pressure to yield a maximum output condition for
a given stroke and temperature.
doing so is discussed
Measurement
of the success
below.
Variation of differential pressure and piston position
were continuously recorded versus time.
Pictures of piston
in
-31and extrudate velocities were taken during the test period of
each specimen.
return
The specimens were marked, during the piston
stroke, with a knife.
A notation
was made on the
pressure
trace to indicate
marked.
Appendix C contains the specimen notation, tempera-
ture conditions,
at which cycle the specimen
and readings
recordings and the specimens.
taken directly
was
from the output
These later include time,
cycles, stroke, pressure, and specimen length, diameter, and
weight.
Calculations included are frequency, mass flow,
average extrusion velocity, and residence time.
Typical velocity traces under different conditions are
shown
in Figures 8,9, and 10.
The upper trace is piston
velocity, the lower is extrudate velocity.
Measurements
were made.
Also,
of the density of samples
The method
and results
are included
in Appendix
at the end of tests for a given stroke
melted polymer was pulled out of the tube.
of the specimen,
slicing
from each specimen
at the tube entrance,
the extrudate
axially,
the
D.
the partially
Only about
inch
did not pull out.
By
axial point at which
the
polymer first becomes completely melted was found.
Presumably,
in the ideal situation the polymer would first become completely
melted
at the tube exit.
was judged to be about
The specimens
On this basis, the speed of operation
30-40% below the maximum
themselves
were of very high quality.
Dimensional tolerance was within .001 inch.
was excellent.
extrudate
Surface quality
A few small bubbles were noticed
for a short
speed.
(5/16) stroke.
in the hot
With the longer strokes,
-32no bubbles were noticed at all.
Density measurements, however,
showed no dependance
A large amount of iron oxide
on stroke.
did appear on the inside wall of the steel extrusion
However,
III,5
none of the specimens
in any way.
LIMITATIONS
There are two primary
work.
were contaminated
tube.
limitations
in this experimental
First, difficulty in controlling the speed of operation,
and therefore
the residence
time of the polymer
in the heated
section, prevented the extruder from attaining the maximum
output steaty state condition.
heaters were insufficient
above 450°F.
As a result,
Second, the four 125 watt
to keep the barrel temperature
higher
for short times, the polyethylene
not studied.
temperature
ranges
will not decompose)
much
(at which,
were
FIGURE 3a.
Test No. 203
Vertical Scale: .667 in/sec/cm
Sweep Speed: .5 sec/cm
TemoeratureThermocouple A: 400F
Thermocouple B: 400°F
Stroke: .311 in
FIGURE
8b.
Test No. 221
Vertical Scale: .667 in/sec/cm
.5 sec/cm
Sweep Seed:
TemperatureThermocouple A: 450F
Thermocouple B:
Stroke: .315 in
50°F
FIGURE Sc.
Test No. 242
Vertical Scale: 1.33 in/sec/cm
SweeD Speed: .2 sec/cm
TemperatureThermocouple A: 475°F
Thermocouple B: 460°F
Stroke: .320 in
FIGURE
?a.
Test No. 303
Vertical Scale: .667 in/sec/cm
Swee Speed: .5 sec/cm
Te moeratureeThermocouple A: 400'
Thermocourle
: 400 F
Stroke: .488 in
FIGURE
9b.
Test No. 321
Vertical Scale: .667 in/sec/cm
Sweep Seed:
.5 sec/cm
TemperatureThermocouDle A: 450'F
Thermocouple B: 450°F
Stroke: .438 in
FIGURE 9c.
Test No. 343
Vertical Scale: .667 in/sec/cm
Sweep Speed: .2 sec/cm
TemperatureThermocouple A: 475°F
Thermocouole B: 460'F
Stroke: .488 in
FIGURE 10a.
Test No. 402
Vertical Scale: .667 in/sec/cm
Sweep Speed: .5 sec/cm
TemperatureThermocouple A: 400°F
Thermocouple B: 400°F
Stroke: .582 in
FIGURE
lOb.
Test No. 422
Vertical Scale: .667 in/sec/cm
Sweep Speed: .5 sec/cm
TemperatureThermocouple A: 450°F
Thermocouple B: 450°F
Stroke: .582 in
FIGURE
10c.
Test No. 443
Vertical Scale: .667 in/sec/cm
Sweep Speed: .5 sec/cm
TemperatureThermocouple A: 475°F
Thermocouple B: 460°F
Stroke: .574 in
-36-
IV,
IV,1
INAPPLICABILITY
ANALYSIS
OF VISCOUS FLOW ANALYSIS
It was initially
supposed
that the flow of polymer
out the
tube would behave as standard non-Newtonian fluid flow.
There-
fore, the governing equations were developed [38] assuming
axi-symmetrical conditions , negligible axial heat conduction,
negligible radial velocity and steady state.
the effects of compressibility
and thermal expansion
developed following Toor [26].
momentum,
and energy
az{pz
Davz
1
z az
respectively,
The equations of continuity,
then reduced
to,
r
(4-1)
a
=
where
aP
rz
r
r rT
Trz
(4-2)
az
T.S.
=
n
v
7 +
An order of magnitude
analysis
under the approximate
test conditi ons.
as
avZ
rz ar
(4-3)
n is the viscosity,
coefficient of thermal expansion, and
was rewritten
were
0
=
WpcvT
Expressions for
e
is the
is the compressibility.
was then performed
for UHMWPE
The continuity
equation
-37-
3az
following
Forsythe
negligible
V
[17].
aZ
z
(4-4)
showed
The analysis
in both the energy and momentum
that the temperature
were negligible.
v7- was
that
equations
changes due to compression
and hence
and expansion
The resulting momentum and energy equations
reduced to:
apa
1 ar
PCVVZH
r rz
1
= i
kr.
3v~z'z r
(4-5)
z
+
Trz
a
r
r
However, over the temperature range (300-450°F) for which
the extruder was tested, the flow conditions were not describable
by equations
(4-5) and (4-6).
and above the flow exhibited
At test temperatures
a definite
This can be seen by comparing
(upper and lower traces,
In the range
rod was clean and showed
assumed
respectively)
of the extrusion
on the interior
wall.
tube showed
to be viscous
polyethylene
it must be
to the tube walls.
to the wall and application
to this problem
flow.
a heavy iron oxide
Since the extruded
that the UHMWPE was not adhereing
a viscous flow analysis
velocities
in Figures 8, 9, and 10.
no signs of contamination,
The flow was thus not sticking
phenomena.
the piston and extrusion
of 300-325 0 F the flow appeared
But inspection
coating
stidk-slip
of 400°
is impossible.
of
-38THE HIGH ELASTIC STATE AND STICK-SLIP FLOW
IV.2
of this section is to explain
The purpose
and reasons for
the stick-slip
flow observed
of stick-slip
The phenomena
attention in the literature.
the nature of
in the experiments.
almost no
flow has received
However, a very recent paper by
et. al. [39] sheds a great deal of light on the
Vinogradov,
In particular, it is shown that under certain
subject.
conditions
in the high elastic
a polymer
state can exhibit
stick-slip flow.
Linear uncured polymers, with molecular weight above a
temperature,
can exist
The high elastic state
of the polymer
The critical
state is given
viscosity
nature of deformations.
irreversible.
for transition
to the high
as
2M
2Me
M is the molecular
corresponding
states.
by the rubbery nature
are of course
weight
molecular
M
weight
is characterized
deformations
Mc -
where
in the fluid and high elastic
and by the reversible
In the fluid state,
elastic
above the glass transition
value and at temperatures
critical
>
10
weight,
(4-7)
Mc is the critical
to the change
power dependance
in the nature
on molecular
weight,
molecular
of the
and Me is the
molecular weight of a chain section between nodes (entanglements).
The values of Mc and Me for polyethylene
about 2,000 and 4,000
respectively.
are extremely
low,
Thus flow instability
has
-39has been widely
noted for
For AC260-100,
polyethylene.
we find
that since
=c
4,000 >> 10
(4-8)
the necessary
this UHMWPE easily meets
criterion for possible
transition to the high elastic state.
Transi tion of a polymer
system
high elasti c state is dependent
shear strai n rate.
from the flui d state to the
on the shear stre ss and the
The critical
transition to occur are denoted
values of these parameters
Tc
for
and Ic. Vinogr adov and
co-workers state that the critical shear stress, Tc, does not
increase wi th molecular
weight
and is, in fact, i ndependent
the molecul ar weight.
It is also suggested
relatively insensitive
to temperature.
rate,
c,
is however
dependent
that
Tc
of
should be
The criti cal shear strain
on molecular weight.
The
expression
(4-9)
ic= Ma
is given.
An effect
of this relationship
molecular weight distribution
is that a narrow
(MWD) implies
the fluid state to the high elastic
state.
a sharp change from
Since the initial
viscosity dependance on molecular weight is
no
=o M(
(4-10)
-40-
(where a is the power dependance of initial viscosity on the
molecular weight), it is also found that the critical shear
strain rate is inversely proportional to viscosity and is
therefore proportional to temperature, equation (4-11).
-1
Yc =
Transition
no
T
(4-11)
to the high elastic
in the nature of deformation.
state also causes a change
The primary
changes
are non-
adhesion to solid surfaces and extrudate distortion. The latter
has been referred
to in [39] as elastic
turbulence.
In the
literature, extrudate distortion has been generally referred
to as flow instability
(such as the so-called
melt-fracture).
The experimental work by [39] utilized a constant pressure
drop apparatus
rectangular
to push a narrow MWD polybutadiene
duct.
The flow rate was measured
the flow and extrudate
a flow rate--pressure
here as Figure
region
11.
distortion
stick-slip
pressure
to continuous
Branch
the beginning
is confined
slippage
as the
III is characterized
by very
of small scale distortions
extrudate during viscous flow.
however,
I is the
The orthogonal arrows on the
strong extrudate distortions.
curve indicate
and is reproduced
Branch II is the spurt region, with
flow progressing
drop increases.
of
From the data,
The branch of that curve denoted
of viscous flow.
a
and the nature
was observed.
drop curve was obtained
through
in the
Deformation in branch II,
to the rounding
off of the corners
of the
-41-
9dA"
1-
IT
0o
I
-f
|
-
--
6
FIGURE
11.
6.'
/o
,
h, L
hA_
Flow rate dependance
drop for polybutadiene
on pressure
[391.
-42rectangular
extrusion.
of deformation
from branch
l6, for transition
T 6 and
II to branch
to define a second set of critical
convient
IV,3
going
Due to the major change in the nature
III, it is
parameters,
between the two regions.
THE THEORY OF THE HIGH ELASTIC STATE APPLIED
TO THE RAM EXTRUDER
In this section,
explained
the nature of the high elastic
by [39] will be compared
experimental work.
to the results
state as
from this
Similar to the experimental work of [39],
this work uses a device which applies a relatively constant
force to the polymer throughout the extrusion stroke. Also
in this section,
the possible
factor on the pressure
slippage
effects
gradient
of the wall-polymer
friction
and the iron oxide coating
on
at the wall will be considered.
Two situations from this experimental work are noted.
First, at temperatures
slip.
of 4000 F and up, the flow was stick-
Second,at temperatures of 300-3250 F both continuous
slippage
at the exit and some surface
ruptures
were observed.
There are two possible explanations which account for these
observations.
The first explanation
takes into account
only
the nature of the high elastic state, while the second includes
considerations
of the polymer-metal
the pressure gradient.
coefficient
of friction
and
-43EXPLANATION ONE
In both cases
of critical
(high and low temperature),
parameters
are obviously passed.
flow is observed
for transition
to the high elastic
and the polymer
is therefore
This suggests
that the
distortion
is in the area where
II of
slippage
and surface
are noted.
in region
continuous
At the lower temperature,
II and III meet.
regions
These observations may be explained as follows.
Tc
state
At the higher temperature, stick-slip
Figure 11.
polymer
the first set
Since
is not temperature dependent, we might well assume that
(the critical
independent
shear
stress for ruptured
of temperature.
extrudate)
And since pressures
is also
for both cases
are roughly equal, we can assume that the shear stresses
not very different.
(the critical
From section
shear
Thus the important
IV.2, equation
shear strain rate,
parameter
strain rate for a ruptured
c,
(4-ll),we
Tc
is
are
'
extrudate).
note that the critical
is temperature dependent.
this is also true for f'. Then since piston
Let us assume
velocities
were
roughly the same, the shear strain rates induced would not be
very different.
Therefore, raising the barrel temperature
could have raised the second
critical
such that the polymer was in region
serious distortion.
shear strain
II--spurts
rate, ¢{c,
without
-44Two
EXPLANATION
In this explanation,
the exception
assumed
the reasoning
that the strong
above is taken
observed phenomena.
above is accepted
temperature
to be insufficient
dependance
of
in explaning
with
c
the
Now the coefficient of friction between
metals and polymers decreases with temperature after the
polymer melting temperature is exceeded.
Then for a given
apllied pressure, the pressure gradient will become steeper
as the temperature decreases.
For low temperature, therefore,
the state of stress would be very high near the tube entrance
and low elsewhere.
As the temperature is increased, the
pressure gradient and the maximum stress decrease and a larger
portion of the polymer is subjected to higher stresses.
Observations indicate that for temperatures around 4000 F,
the state of stress
the polymer
is sufficient
in region
smaller portion
state which
II.
For temperatures
of the polymer
could account
to put a large part of
around 300°F,
would be in a much higher
for the distortions
observed.
a
stress
The
continuous slippage noted at this temperature can be simply
attributed
to a combination
polyethylene
of the non-adhesiveness
and the iron oxide acting
as a dry lubricant.
should be pointed out that the non-adhesiveness
may become pronounced
well below
Tc
and
However,
drawing a conclusion.
the data is really
It
of the polymer
c due to the MWD.
The author feels that the second explanation
reasonable.
of the
is more
insufficient
for
-45-
IV,4
HEATING TIME
Of utmost
in the design of this extruder
importance
the time necessary
as the residence
to melt the polymer.
time (of an infinitely
This is referred
is
to
thin disk in the heated
tube).
Comments regarding the nature of gradients in the extruder
tube are in order here [40].
into the polymer
First, using a maximum
(1.6 Btu/hr) the radial
negligible.
was found to be completely
at the tube entrance
temperature
Second,
heat flux
gradient
the gradient
was found to be significant,
as follows.
Temperature conditions and physical dimensions are shown in
Figure 12.
extruder
If the radial
tube interface
at the thermal
gradients
are small,
98% of the heat flux in
region A will go through the steel sleeve.
We can now equate
j
(
7
.
1
I
70°
12.
/7 .
I
I
I
FIGURE
insulator--
7
Calculation
'C000
of Tube Entrance
Temperature Gradients
-46for heat flux in the two regions provided
expressions
an equivalent
with the area of the extruder
T is then
Averaging
area for region B.
given
the interface
area
tube, we obtain AB = .56 sq. in.
by
(4-12)
T
T = 2150 F.
which yields
Taking
sleeve is cooled by assuming
the fact that the
into account
= 2- AxA, we find T = 1600 F.
Ax
Thus it seems that "shortening"
10% would
we select
of the heated
section
by about
be reasonable.
As noted
too slow.
in Chapter
Thus another
were run 30-40%
III, the specimens
correction
factor,
of about 35%, is
necessary to obtain the effective residence time:
tcorrected
We can now treat the problem
temperature
ature.
subjected
(4-13)
tmeasured (.9)(.65)
to a sudden
as a cylinder
increase
[41] have solved
Carslaw and Jaeger
at constant
in the wall
temper-
this problem.
solution for average and center temperature is plotted with
V versus S as the coordinate
axes where
The
-47v
T
V
Tw
S
=
- Ti
(4-14)
Ti
(4-15)
at
r2
to be either
T is taken
Ttr
t is the corrected
diffusivity,
a is the thermal
or Tave
residence
time, and r is the
cylinder radius.
Residence
times
for the tests at 4000 F and 450°F were
At 4000 F the average
averaged.
and at 4500 F it was 63.8 sec.
effective
residence
approxima t , lI
is
1
.23)
that
a
fl
diffusivity
01 ft 2 /hr.
·
(S at 450 °
a nd
equation
(4-13), the
of polyethylene
Calculating
is .96).
a t tem perature
is
S, we find (S at 400 °
These values of S indicate
distribution
exists
in the polymer
[41].
fla t temperature distribution could be explained
A
in several
Applying
time was 82.0 sec
times are found to be 48.0 sec and
The thermal
37.3 sec.
residence
w ays
dependent.
First, the melting
Thu s the melting
temperature,
temperature
for a given residence
time could b e v ery close to the wall temperature.
value of Tm, at this time,
300°F and be 1
ow
Tm, is time
The actual
can only be stated as being above
the wall temperature
non-adhesion of the polyethylene
(4000 or450°).
and the presence
Second,
the
of the oxide
layer indicate the possible existence of a significant contact
resistance.
thermal
of 2.
Third, the thermal conductivity, and thus the
diffusivity,
In all
could be in error by as much as a factor
probability,
all
three
explanations
probably
-48contribute to obtaining the calculated results.
Modification
of the experimental apparatus and further testing is definitely
necessary.
However,
it is still believed
temperature
increase
is caused
and therefore
that the polymer
soley by radial heat conduction
of the problem made above is
the idealization
valid.
IV.5
POWDER COMPRESSION AND AIR EXPULSION
During all the tests,
no problem was ever encountered
to inadequate powder compression or entrapped air.
due
As
mentioned in the literature review, the powder compression can
be analized
by use of the Mohr-Coulomb
cation of this criterion
shows that the pressure
The expulsion
follows.
criterion.
Appli-
in a cylinder
drop is exponential.
is explained
as
Initially the powder is compressed to near its
(Preliminary data indicate that the com-
results in an initial
density.)
diffuses
compaction
of air from the powder
theoretical density.
paction
to powder
yield
As the temperature
density of 90-95%
increase
back through the powder
escapes past the piston.
of the theoretical
in the powder,
to the tube entrance
the air
and
This temperature activated diffusion
process completely degases the polymer.
IV,6 MAXIMUM
OUTPUT CONDITIONS
The residence
proportional
time necessary
to the radius squared
for melting
of the powder
and is also related
is
to the
-49values of the melting
the solution
and the wall temperature
temperature
to the heating
problem of an infinite
via
cylinder.
If the melting temperature--time and degradation temperature-graphical
time relationships
are found, then an iterative
solution
may be used to find the minimum
time.
technique
The only other necessary
consideration
that the shear stress and shear strain
polymer
Figure
into range of extrudate
1l.
background
tm = f(Tm,Tda)
=-
where tm is the minimum
extrusion
velocity,
temperature,
thermal
polymers.
for this is simply
length, Tm is the melting
temperature,
rY is inversely
c
even a low extrusion
rate exceeding
III in
time, v is the average
Td is the degradation
to viscosity,
strain
residence
Now, since
rate do not move the
(4-16)
L is the barrel
diffusivity.
is to make sure
distortion--branch
The theoretical
residence
velocity
and a is the
proportional
will cause of shear
Y' for the case of high viscosity
c
Therefore the shear stress will be the primary factor
controlling
whether
the polymer
flow or in the region
is primarily
affected
barrel length.
is in the region
of extrudate
distortion.
by the pressure
Increasing
would
raise the pressure
would
increase.
of stick-slip
The shear stress
drop which depends
the length of the extrusion
drop and therefore
on the
tube
the shear stress
For a given barrel temperature,
the output
mass flow is limited by that length of the extrusion
tube
-50which
causes a pressure
shear stress.
establish
maximum
drop corresponding
to the critical
Much more experimental work will be necessary to
all of the functions
flow rate.
needed
to determine
the
-51-
V.
The ram extruder
extrusion
has been shown to be feasible
of high viscosity
is expected
extruded.
CONCLUSIONS
UHMWPE
in cylindrical
that other high viscosity
polymers
for the
form.
It
can also be
The nature of the flow has been found to be stick-
slip due to the fact that the polyethylene
is in the high
elastic state.
Heating of the polymer
tion.
is by means of radial
Thus the time required
tional to the square
to heat the polymer
of the radius.
mass flow is greatest
As a result,
heat conducis proporthe extruder
for thin extrusions.
The piston driving source should have separate velocity
controls for extrusion and return strokes and a constant force
should be applied by the piston for maximum output.
stroke speed is limited by the time required
The return
for the powder
to fall through the port.
Maximum output may be calculated, if the melting temperature--time and degradation temperature--time relations are
known, through
circular
use of the solution
cylinder.
the polymer
appropriately
Extrudate
to heating
quality
can be kept high (i.e.
stays in the region of stick-slip
limiting
the barrel
of an infinite
length.
flow) by
-52-
APPENDIX A
DETAIL DRAWINGS
COMPONENT
PAGE
Extrusion Tube ......................... 53
Piston
.................................
54
Air Motor Adaptor ...................... 54
Lever .................................. 55
Connecting
Rod .........................
56
Spacer ................................. 56
Bushing ................................ 56
Lever Mount ............................ 57
Extruder Tube Mount .................... 57
Pins
...................................
Feed Block
.............................
58
59
Thermal Insulator ...................... 60
Velocity Transducer Mount .............. 61
Displacement Transducer Mount .......... 61
-53-
I
I
.o
-.
1
N
QcIIt
-(
? 4i
.
t
A
II
o1wo '.
I~~~I
t l~~~~~~al W
k
_
__
41
LI)
a
-
~~~k~~*
z
ji;TC
IiU
I
C
I
1
I
XN
!
0
(1
Ii
--
i.k
-
~
_
_
Nt
4c1
.....
©_[_T
-eI
I
3.\-
'V
I
I
Ra
i
!
I
i
7i
T
.
K 41
z - 'o
t
-54_.,
.
I
.
.
.
..
I..
,
%,.
!s
Wf
24ky--
--
-1~~~
------
. 6215
"1.--
-
i
U-
.... -.--v-,,,
I Il -
, IT
V-o ur...
Cs ,lI(/V/
IvC
,.-
1P9ppleaAyRC4rR
CAC
1A4T;
RA=5
rTOZ-c
RACs ;
Ft ,q T /A/S -:
Z)c c t A
II It
1
L
6
.
ol
--.of ---
I
__
..I
:
7r
. I-
TrH? 4V
PIs T/
II
) /-/
1
/---
i
i
.
-
_ 1 9/3 D,,.
I-.
-,
II;.
i
A)P P-11-
--.
0,
l
,
s/
1
I i '.-~~~~~~~~~~~~i
_---1
L,
l ,~I!:
I
i
Ji
//MorR
AlreR
MN4 (f)
-- _
rTOt RAA6/ce;:
-- - -/ !
-
-.----- _
RE'ACYOMS2±
k.,L
iGFCIIS
L. oo0
-55-
V, RADA4S
,'y
MxiAttM
_... .007
--F o,-
I
i
II i
:
----
II
1
--I
0
2DR4L,
e 4
2
/ cekc
d
I;
to
i
I
I
I
0
l
i
i
"
I,---X Z)RLL,
/-)
e
I L_
I
,
-
'
f
-
r-i
l.__
/
I f
-_.
1
.ooZ
. FvL#7-rC
e-k
<
.2.O
(X)
T L&rAAE
C S : F R4 CT/(AS
,C}
Ce
IAIt cE.
A- P
-56_
z
-
X
s
_ At
'
'I_
j
ic
:11 -
-N
I
*-
.
r,
4
0~~~~~~~~~~~~~~I
,~~~~~~4
k Q$
0
I
,
_
T9
1.
td
A
i
6
,
'$- ''
~
_
Li
y
. t
,-
t
IZ
XrX
-
-4 2SZ'
0
o
_~~~~~~~~~V
_
'W
tY Pi,_
s<
0~~
*~~)~~~~~~
s
I
T.
I
R"
|
V
k 4
34
d , -s t e 4 b~~~~~R~
C
1, G,
V,
W
4
-57-
_.-.
-----
--
I-
"·n
-·-rr IU r.rr·--·l
I~ ~
...
it
--IA
F
1V-
u
I
I
Cn*
i--z~I
-II
0
iIi
~
I
t~~3
·CIq1
.--
_l
L
*
i
14j
t. ,
1.
1\1
z!
-\I ~
.
c7
)
;j
-- %
4i
,
1;
%A Q
b,-Z;
J
-58__
m
IV
I
t
mo
t
Na
]t
IU
kv
rTQ
x
00
aLI~~
i!
el~~~~e
i,
/
.<..
.
.X
1
4J
:I1 to
..
,
Q
k
$
'k1
0. Z'
..-...
o0'-c
rr
c'r
,goal0-f-
Os'-e '
4
C
c'
c
.,
L
'
0
' z
1.
.,
-90o·
_qoo
I
o.
r3
e
l$
Ira
-
--.-
0
0Sw
1'q~
3t
P Z'
vi
.
-59-
I +
N
-
00'/
-
-
-\
:t G, a
4i
_
LI
;Sl
1-.
\-
R- t
Cl
k
.I
.,
C4
-:t
I
V
L.L
2
'1
I
t
-1
-
-----
'N
/
_
I_
-it
__'
_
_
_
-4
L)
I
ti
;'
_
1,k
I
/
I
f~
1
11
t'
4
-.
I-
_
-Z
k
\6
Q
C-
J.
I
til
Q
:
__
---,
-
IA
,
-
i
*-
j.
.
1th' -J
I.
k11
k- t
c
..
:?u
k
0
l'
a]
icq
i-,:
%.7-
,; J
----I
IA
i. W-
.
0-
I
.
I,
;!r----
I
I1
0
I
I
I
I
-T-
-- 4-
,L
i
I
/
LI~
--
?P
'z
4i
-
I
I
V
i:
I
Ii,
- -
'4
1- r.
. ___
. ___
.--r
- -__
I
11
I
0 ..
0
-
/
CO
-"
I,,
- v
-S'Zff
I
-#
F" .. Ll
I
I
1
__
I)
NZ,
I
t
I
527/
t
W
-60-
14.-
t I'
-f.w
19 v
h B
77
,- ,
- uI '.7
-
i, i
U)
7- -
"Aa
.42
'It
(i3
1
t-%< >
·
Ni
-
J
Z
,it
,\ pj
Iz10
Zli
%
6
Q~z
Lk
t
< -d-
/
X4
7
11
- 11
.-
I
s
£~~
i
-61-
i
I
I
I
1
,?.7 UWA- .V's
b-pro
le/1.
Y - Zo urV
( )
Va
A'C
RA /P DL1C
T
f A1 T: J/ 8 Rh. P
l.E<-
No10v-vr
R
r7r
()
Fk/,eqC-r I's 5
Y*JZ
P,6 C.IM4 L. i , /
---,A6
r--
... -IL.
j
\
~....,.~.,......~
_}-
-Lt-
,
-1=1I
1=D
I
.z
L - __
I_____
IWL
.1
I
0
f
·
__
_
Q
Q
O
O
1#4
,
n1lll
·
\
. , \
- --
--T
-
...
4
/vz.1
Lj
____··___·_
___P
t
.- I,
;Z .....
B
1
W
I
I I
L_.L
L-L
iV
F 1
I-.
-
-- ,
- - 1
O
O0
0
'"Nt9
ii
__.
't
©
I
3
iI
/o -32£A/. ()
( 3 /e J'-Pr#)
DisPlAeemEcA) r
©)
0
i
- ..
----
TZAD VLC-C-7R
/t OUA~r
MA r:
.
tA1AKt.
/114 x rlf) ()
JO0LC-Rq i CCS'.
FRALTb~lt
Y*
D6~'Cr#t4v.S± Y,
o
FSCJA4,1S.S
-62-
APPENDIX
B
EQUIPMENT LIST
UHMWPE
AC 260-100, Allied Chemical Co.
Weight Average Molecular Weight
of 8,000,000
Air Motor
Piston type, Bellows-Valvair Inc.
Bore: 3- in.
Stroke:
Force:
2 in.
up
Heaters
1500
Rating:
at 150 psig.
wide, variable range.
Band type, Watlow
Model:
lbs
electrically initiated.
Valving:
Speed:
to
Inc.
11344G
125 watts at 120 volts.
Dimensions:
1 in.
Thermocouples
Iron-constantan.
Temperature Controllers
Off-on
control
ID,
1 in. width.
type.
West Model JP and
Barber-Colman
Pressure Transducer
Model 297C.
Differential type, Fredric Flader Inc.
Model:
PSB
Range:
0-160 psi.
-63-
Displacement Transducer
LVDT, Hewlett-Packard
Model:
Stroke:
LinearSYN 585DT-1000.
1 in.
Non-linearity:
Velocity Transducers
Inc.
less than 1%.
Magnetic core, Hewlett-Packard
Model:
LVSYN 7LV6-NB6.
Stroke:
Inc.
6 in.
Non-linearity:
Sensitivity:
less than 1%.
150 mv/in/se c within
5% over full stroke.
Recorder
Sanborn
321 Dual Channel
Carrier
Ampl i fi er-Recorder.
Oscilloscope
Tektronix Type 564 Storage Scope
with Camera C-27.
-64APPENDIX
C
TEST CONDITIONS, RESULTS, AND CALCULATIONS
Specimen Nomenclature:
Test specimens
first digit
were
labeled with three digit numbers.
(2,3, or 4) indicates
the second digit
(0,2, or 4) indicates
last number specifies
Test conditions,
through C-4.
melted.
the nominal stroke
results
Specimens
a particular
temperature.
specimen
and measurements
The
and
The
in a test set.
are on pages C-2
202, 342, and 402 were
incompletely
-65Test#
Stroke
Thermocouple
Temperature( 0 F)
(in.)
B
201
202
203
221
222
223
241
242
Number of
Total
Time of
Cycles
Test
Average
Piston
Frequency
(sec)
A
I
.
Total
__
(Hz)
|
.311
400
400
87
.311
400
400
114
118.8
.78
.96
. 311
400
400
113
130.3
.87
.315
.304
450
450
84
89.7
.94
450
450
64
59.8
1.07
.304
450
450
144
110.3
1.30
.312
475
475
130
79
1.65
.320
460
475
119
69.6
1.71
.480
.484
400
400
51
105.5
.48
400
400
56
400
400
106
176.7
450
450
59
86.8
322
.488
.488
.492
450
450
65
85
323
.492
450
450
63
98.7
.765
.64
341
.492
475
475
82
89
.92
342
.496
460
475
81
65
1.25
.488
460
475
78
68
1.15
-
-
301
302
303
321
343
-
111
.57
99
.60
.68
|-
-
401
.582
400
400
37
89.8
.41
402
.582
400
400
35
96.4
.36
421
450
450
43
74
422
.582
.582
450
450
41
79.2
.58
.52
441
.582
475
475
62
84.3
.74
442
.578
.574
460
480
54
55.4
.975
460
475
46
54.6
.84
443
-66Test #
Specimen Dimensions
Average Air
Motor Pressure
(psid)
Extrusion Ret urn
-
S
Length
(+.03in )
Weight(grs)
(+.5mg)
Diam.
(+.001 i n .)
a
201
5
1
7.25
.235
4.752
202
4
1
8.88
.234
203
4.5
1
9.42
.234
221
5.1
1 .2
6.07
.237
5.822
6.183
4.0485
222
5.25
1 .1
4.80
.236
3.203
.236
7.331
223
5.5
1 .1
A
,"I,
I I. UU
241
7.75
2
10.57
.237
7.072
242
7.25
2
9.70
.237
6.5345
301
5.375
1
7.25
.235
4.782
302
3.5
1
8.17
.235
5.370
303
4.5
1
15.15
.235
9.984
321
5.5
1
.2
8.13
.236
5.432
322
4.75
1 .1
9.20
.236
323
6.25
1
.2
8.60
.237
6.098
5.744
341
6.5
1 .5
11.02
.236
342
7.5
2
10.48
.238
343
-
7.5
1-
9.98
2
·
S
·----
·-
·
·
7.368
7.078
6.739
.238
·---
-
401
4
1
6.00
.234
3.921
402
4
1
7.12
.234
4.658
421
3.75
1.2
6.95
.235
4.581
422
4.5
1.3
6.34
.236
4.194
441
5.5
1.5
9.61
.236
442
5.5
1.75
10.08
.236
6.368
6.7215
443
6
1.7
6.83
.238
4.553
-67Test#
Average
Mass Flow
( lbm/hr)
Residence
Time
Average
Extrusion
Velocity
Average Pressure
At Piston Face
(sec)
(psi)
(ft/hr)
_ _I
:
201
.0685
19.6
91.9
3160
202
22.4
80.4
2530
203
.0801
.0775
21.7
83
2840
221
.0736
20.3
88.7
3220
222
.0875
24.1
74,7
3320
223
.1085
29.9
60.2
3470
241
.146
44.9
4950
242
.153
40.1
41.8
43
4580
I
301
.0740
20.6
87.4
3390
302
.0886
24.8
72.6
2210
303
.0924
25.7
70
2840
321
.102
28.1
64
3470
322
.117
32.5
55.3
3000
323
.095
26.1
69
3940
341
.135
37.2
48.3
4100
342
.178
48.3
37.3
4730
343
.162
44.0
40.9
4730
-
i
-
.r
F
__
-
20.1
22.2
89.5
2530
81.1
2530
63.8
2370
.0865
28.2
24.0
75
2840
441
.1235
34.2
52.6
3470
442
.198
54.6
32.9
3470
443
.136
37.5
48
3790
401
.0714
402
.0790
421
.101
422
-68-
APPENDIX D
DENSITY MEASUREMENTS
Two 1.25 inch (nominal)
The samples were taken
samples were cut from each specimen.
approximately
from the ends of the specimens.
one-third
The density was then found by
the hydrostatic weighing technique.*
weighed
in a liquid, then in air.
one may then calculate
s
mb + ms
The specimen is first
Knowing
the specimen
Ppm'm
of the way in
the liquid density,
density.
The equation
is
(D-1)
mnet
where Ps = density of specimen,
= density of liquid,
m
= mass of specimen,
mb = mass of basket
(specimen
mnet = mass of specimen
The apparatus
holder)
and basket
has been used to measure
of 10- 5 gr/cc.
The author,
in liquid, and
in liquid.
densities
to an accuracy
due to the large number
of specimens,
attempted to achieve a final accuracy of 10- 4 gr/cc.
mass was corrected
*Apparatus
for the bouyancy
at the Materials
assistance
description
Science
effect
of air using
Center, MIT, was used.
of Dr. J.A. Kalnajs is gratefully
of the apparatus
Specimen
see A.Smakula
acknowledged.
The
For a
and V.Sils,"Precision
Density Determination of Large Single Crystals by Hydrostatic
Weighing",Phys.Rev.,99,#6(1955).
-69m
ms
[1 + P a i r _ Pair
~s S
where
Pw is the density
balance
of the weights
(+.02 mg), with stainless
The bouyancy
these measurements.
static weighings
therefore
densities
used.
steel weights,
effect
H16
was used for
of air in the hydroand was
Ethanol was used for the liquid and a Wang
for calculations.
All masses
in grams per cubic centimeter.
calculations
A Mettler
was less than the error involved
ignored.
calculator
(D-2)
W
P
are in grams; all
Constants
are:
= .78505
Pair = .001197 (for test conditions)
Pw = 7.77
mb = 2.17920
The results are tabulated on the following page.
for all
-70Test
#
m
s
mnet
+.05mg
ms)corrected
+.02mq-
+.02mg
I
-_
Room Temp.(°C)/
PS
+. 0001
(mmHg)
-11
I
g -
-
201
1.57832
1. 5 8011
2.41870
.9253
202
1.56935
1. 57108
2.41748
.9254
203
11.62245
1.55463
1.624235
2.42667
.9262
1.55634
2.41635
.9262
221
1.64971
2.43100
.9265
223
-1.64790
1.67755
1.679395
2.43498
.9261
241
1.68686
1.68872
.9264
242
1.63830
1.64010
2.43681
2.42962
~~-
222
I
301
.
1.62172
Humidity(%)/
Barametric Pressure
--
22/60/771
22/61/768
22
.
_
_
_ _ _
1.62350
2.42620
.9259
ll
t1.66978
1. 67162
2.43366
.9260
303
1.65418
1. 65600
2.43119
.9260
321
1.62280
1.624585
2.42672
.9262
322
1.65149
1.65331
2.43097
.9261
323
1.64294
1.64475
2.42951
.9260
341
1.65940
1.661225
2.43229
.9262
342
1.73451
2.44351
.9260
343
1.67297
1.73642
1.67481
2. 43436
.9262
__I·_
·-----D-·(-·-·-··UI·-·
.9256
.
_I
1.56039
1.63285
1.56211
1. 63465
2.41644
2.42736
1.62900
2.42702
.9259
422
1.62721
1.63533
1. 63713
2.42802
.9258
441
1.74651
1.74843
2.44507
.9258
401
402
421
..
X
.
-I
3
1
223/601/766111
3
1
22-8/60-/766
1
22-1/60/765
1
1
22 -/6 0/764 2
.
_
...
.-
.9256
442
I 1.78671
1.788675
2.45109
.9258
443
i1.67472
1.67656
2.43434
.9260
(Theoretical density
1 6
1/76 7
.9265
302
.
-
is .928 gr/cc.)
1
2 2-/60/764
-
-
-71-
REFERENCES
1.
Margolies, A.F.,"Effect of Molecular Weight on Properties
of HDPE",
2.
Craig,
SPE J.,27(1971)
R.F. ,Barnes, E.N., and Rakes, J.L.,"Extrusion
of
Extra High Molecular Weight High Density Polyethylene",
SPE J.,24(l1968)
3.
Adler,A.,"Flow
Properties
of Metal Powders",Intl.J.
Powd er
Metallurgy,5(1969)
4.
Platonov,P.
and Poltorak,V.,"Investigation
of Shear of a
Granular Material along a Bordering Surface",
Powder Tech.,3(1969/70)
5.
Schwartz,E.G.
Criterion
and Holland ,A.R.,"Determination of Yield
for Iron Powder Undergoing Compaction",
-1ntT.J. Powder Metallurgy ,5(1969)
6.
Lee,R.S. and Schwartz,E.G.,"AnAnalysis
of Roll Pressure
Distribution in Powder Rolling",Intl. J. Powder Metallurgy,
3(1967)
7.
Bowden,P.B. and Jukes,J.A.,"The Plastic Yield Behavior of
Polymethylmethacrylate" ,J.Mat.Sci.,3(1968)
8.
Rabinowitz,S.,Ward,I
.M., and Parry,J.S.C.,"The
Effect of
Hydrostatic Pressure on the Shear Yield Behavior
Polymers",J. Mat. Sci. ,5,#1(1970)
9.
of
Mears,D.R.,Pae,K.D., and Sauer,J.A.."Effects of Hydrostatic
Pressure on the Mechanical Behavior of Polyethylene and
Polypropylene",J.App.Phy.,40,#11(1969)
-7210.
Kawakita,K.
and Ludde,K.,"Some Considerations on Powder
Compression Equations",Powder Tech.,4(1970/71)
11.
and Klein,I.,Engineering
Tadmor,Z.
Principles
of
Plasticating Extrusion, Chapter 8,New York, Van Nostrand
Reinhold Co.(1970)
12.
Klein,I.
and Marshall,D.I.,eds.,Computer
Programs
for
Plastics Engineers,New York,Reinhold Publishing Corp.(1968)
13.
-
-
s
Klein,I.
and Tadmor,Z. ,"The Simulation of the Plasticating
Screw Extrusion Process with a Computer Programmed
Theoretical Model",Poly.Sci.Eng.,9,#l(1969)
14.
McKelvey,J.M.,Polymer
IProcessing,New
York,John Wiley and
Sons,Inc.(1962)
Encycl opedia,New York,McGraw-Hill(1971)
15.
Modern Plastics
16.
Brandrup,J. and Immergut,E.H.,eds.,Polymer Handbook,
New York,Interscience Publishers(1965)
17.
Forsythe,T.H.,Ph.D. Thesis,Virginia Polytechnic Inst.,
Blacksburg(1967)
18.
Fuller,T.R.
and Fricke,A.L.,"Thermal
Conductivity
of
Polymer Melts",J.Appl.Poly.Sci.,15(1 971)
P., Kolloid - Z.Z. Poly.,
19.
Lohe,
20.
Hennig,
203 (1965)
J., Knappe, W., and Lohe, P. , Kolloid - Z.Z.Poly.,
189 (1963)
21.
Hanson,
D., and Ho, C.C., J. Poly Sci, A3 (1965)
-7322.
R.S. and Gilmore, G.D., "Equation of State for
Spencer,
Polystyrene", J. Appl. Phy., 20 (1949)
23.
24.
Spencer,
R.S. and Gilmore,
G.D., "Equation of State for
High Polymers",
J. Appl. Phy., 21 (1950)
Brink-nan,
"Heat Effects in Capillary
-Appl. Sci . Res., A2
. 'l .m
Flow I",
1951)
25.
Bird, R.B .,
'Viscous Heat Effects in Extrusion
Plastics" , SPEJ. (1955)
26.
Toor, HL ., "The Energy Equation for Viscous Flow",
Ind.
Eng. Chem., 48 #5 (1956)
~~e~T-_
of Molten
27., Lyche, B.C. and Bird, R.B., "The Graetz-Nusselt Problem
for a Power Law Non-Newtonian Fluid", Chem. Eng. Sci.,
-
6 (1956)
28,
Toor, H.L., "Heat Transfer in Forced Convection with
Internal Heat Generation", A.I.Ch. E.J., 4, #3 (1958)
29.
Gee, R.E. and Lyon, J.B., "Non-Isothermal Flow of Viscous
Non-Newtonian Fluids",
Ind.
Eng. Chem., 49, #6 91957)
30.
Hanks, R.W. and Christiansen, E.B., "The Laminar NonIsothermal Flow of Non-Newtonian Fluids", A.I.Ch.E.J.,
15, #4 (1969)
31.
Christiansen, E.B. and Craig, S.E., "Heat Transfer to
Pseudoplastic Fluids in Laminar Flow", A.I.Ch.E.J., 8,
#2 (1962)
32.
Christiansen, E.B. and Jensen, G.E., "Laminar, Nonisothermal
Flow of Fluids in Tubes of Circular
15, #4
1969)
Cross Section",
A.I.Ch.E.J.,
-74..
:
33
tEA., t'Ie.at
Too,
Generation
and
Liquid",
of a Viscous. Compressible
in
Conduction
the Flow
Trans. Soc. Rheol.,
r C1957)
34.
Gruntfest,
tures
I.J., Young,
Generated
J.P., and Johnson,
by the Flow of Liquids
N.L. , "Tempera-
in Pipes ," J. Appl.
Phy., 35, #1 (1964)
35.
Kearsley,
36.
Forsyth,
E.A., Trans.
Soc. Rheol.,
T.H. and Murphy,
6 (1962)
N.F., "Temperature Profiles
of Molten Flowing Polymers in a Heat Exchanger", A.I.Ch.E.J.,
15,
#5
(1969)
37.
Minnar, E.J., ed., ISA Transducer Compendium, New York,
Plenum Press (1963)
38.
Yuan,S.W.,Foundations of Fluid Mechanics,Englewood Cliffs,
N.J..,Prentice-Hall (1967)
39.
Vinogradov,G.V.,Insarova,N. I.,Boiko,B.B.,
and
Borisenkova ,E.K.,"Critical Regimes of Shear i n
Polymers",Poly Eng Sci .,12, #5(1972)
Linear
40.
Holman,J.P.,Heat Transfer,2nded.,New York, McGraw-Hill(1968)
41.
Carslaw,H.S.
and Jaeger,J.C.,Conduction
Oxford,University Press(1947)
of Heat in Solids,