Abstract
A study was commissioned to examine the wear characteristics of six high performance
engineering plastics currently used in tribological applications. The test consisted of a thrust
bearing 3 x 3 grid High-PV wear test of each material, which was then extended through
increased PV conditions until material failure for the survivors of the 3 x 3 grid.
The study produced valuable data on the comparative performance of these materials. The two
most outstanding materials were VICTREX TL-60 and Vespel SP-21. VICTREX TL-60 demonstrated
superior performance with its consistently low wear factor, low coefficient of friction and cool
counter-surface temperatures. It was the only material to survive loads at or above 2500 psi (at
speeds greater than or equal to 50 ft/min) and in fact performed at 4000 psi at a PV of 200,000
psi-ft/min. Vespel SP-21 was also remarkable in its demonstration of achieving the absolute
highest PV of 400,000 psi-ft/min at a surface velocity of 800 fpm.
1. Overview
Many industrial processes and a number of consumer products require high performance
materials in tribological applications. Traditionally this has been the realm of lubricated metals,
but with increasing demand for performance and light weight, advancements in material science
have brought engineering plastics to the forefront. These engineering plastics bring the benefit of
light weight, long part life-cycles, low failures rates, and enable some applications heretofore not
possible with metal. The material space in play requires wear resistance, high strength, high
temperature resistance, and low creep in a non-lubricated environment. This study reviews six
materials competing for this space and shows that engineering plastics exist which are up to the
task.
ASTM defines wear as damage to a solid surface, generally involving progressive loss of material,
due to relative motion between that surface and a contacting substance or substances. Wear
resistance is a paramount characteristic of a material’s suitability for tribological service.
Technicians measure wear rate by standardized methods and commonly report wear rate as a
measured change in thickness or weight loss per unit of time or distance. Since wear rate is a
function of the subject material and the mating surface material, their hardness, their surface
finish, the ability for the material and the system to dissipate heat, surface velocity, load,
temperature, other environmental conditions, and exposure time, it is necessary to standardize
test methodology and conditions to enable comparison. This study attempts to do so by
measuring equilibrium wear rate, coefficient of friction and counter-surface temperature of
engineering plastics in a thrust washer test configuration against a smooth mild steel surface in
air. The objective is to determine which are best suited for the most demanding tribological
applications with steel. Ideal materials will exhibit a low coefficient of friction and low wear rate
while generating little frictional heat.
2. Experiment Design
2.1 Description of Test
The test employed was a High-PV thrust bearing wear test based on ASTM D-3702 utilizing
engineering plastic thrust washer test specimens rotated against a steel counter-surface. The
test specimens measured 1.0625” in diameter with 0.20 square inches of contact area. The
counter-surface was annealed AISI 1018 carbon steel machined to a 16 +/- 2 micro-inch (AA)
surface finish. The test specimens were rotated under load without lubrication against the
stationary steel counter-surface in air at ambient conditions for fixed time periods. Wear rate,
coefficient of friction, and counter-surface temperature were measured as the material
progressed through a matrix of pressures and velocities.
The product of the pressure (P) on the contact area in psi and surface velocity (V) in ft/min (fpm)
provides a PV value at which wear rate, counter-surface temperature, and the coefficient of
friction were measured. A wear factor (K) based on measured wear is calculated and reported.
K = (the measured change in specimen thickness in inches)/(PVT) where (T) is duration in hours.
The coefficient of friction (f), a dimensionless unit, is calculated as f = 2Fa/(PAd), where F = the
frictional Force in lbs, a = frictional arm length in inches, A = thrust washer specimen contact Area
in square-inches, and d = mean annular diameter of thrust washer specimen in inches. Countersurface
temperatures were measured by a fixed thermocouple to the side edge of the countersurface
and reported in degrees Fahrenheit. Temperature readings are periodic and therefore
may not represent the maximum temperature encountered.
The test battery consisted of an Initial Test of six high performance wear-grade engineering
plastics and an Extended Test of those materials which survived the Initial 3 x 3 grid. All tests
were conducted by Lewis Research, Inc. on a LRI-1a Thrust Bearing Tribometer.
The tests were conducted using three washer specimens (a, b, and c in Table 1 below) for each
thermoplastic material. These washers were first put through a break-in test interval (designated
by subscript 1 to the specimen ID in Table 1 below) at the lowest PV and middle velocity
condition (PV = 50,000 psi-ft/min; Velocity = 200 fpm) for 8 hours. The specimens were then
advanced through the PV grid as indicated by sequential subscripts to each specimen “a”, “b” and
“c”. By example, after the break-in cycle specimen “a” was next tested at 50,000; then 75,000
and then 100,000 PV at a velocity of 50 fpm (respectively a2, a3, and a4); broken-in specimen “b”
was tested at 50,000; 75,000 and 100,000 PV at a velocity of 200 fpm (b2, b3 and b4); and
broken-in specimen “c” was tested at 50,000; 75,000 and 100,000 PV at a velocity of 800 fpm (c2,
c3 and c4). The duration of each test interval was 8 hours. This test grid is repeated in the same
manner for each of the six thermoplastic materials.
Three thermoplastic materials survived the Initial 3 x 3 grid and were pushed to higher PV’s in
increments of 25,000 psi-ft/min until failure in an Extended High-PV Wear Test. These tests were
also run at 50, 200 and 800 fpm surface velocities while increasing the pressure to achieve higher
PVs (subscripts 5 through 16 in Table 1). Results are presented below.
2.2 Materials Tested and Material Preparation
Six high performance engineering plastic materials were tested. Five were aromatic
thermoplastics and one was an aromatic thermoset.
Two sources of VICTREX TL-60 parts were tested. This material is designed to be injection molded
and used without further treatment or machining; however, since it is available in stock shape and
molded part form, both machined parts (from stock shape) and injection molded parts were tested.
Two approved sources for VICTREX T-Series products supplied specimens. Parkway Products, Inc.
molded the injection molded VICTREX TL-60 thrust bearing specimens. These specimens were
tested as molded after boring the center axis and drive holes. Piper Plastics, Inc. molded the TL-
60 stock shapes and machined the respective machined VICTREX TL-60 thrust bearing specimens.
Three wear grades of Solvay Torlon Polyamide-imide (4203L, 4275 and 4435) were injection
molded into part form and post cured by Parkway Products in accordance with Solvay’s
processing instruction. After post-curing, these specimens were bored with center axis and drive
holes.
The sole aromatic thermoset wear grade material tested was DuPont’s Vespel Polyimide SP-21;
molded by DuPont. Thrust washers were machined from molded stock shape.
And finally, VICTREX PEEK wear grade material 450FC30 was tested. Parkway Products injection
molded these parts and annealed them according to Victrex’s instruction. After annealing, these
parts were bored with center axis and drive holes.
Reference Table 2 for further details on the materials.
Product notes:
VICTREX® TL-60 made with Celazole® Polybenzimidazole (PBI) is Victrex’s injection moldable, high performance, wear-grade T-Series compound.
Vespel® SP-21 is DuPont’s 15% graphite filled Polyimide (PI) with enhanced wear resistance.
Torlon® 4203L is Solvay’s un-reinforced all purpose-grade Polyamide-imide (PAI) compound made with 3% TiO2 and 0.5% fluoropolymer.
Torlon® 4275 is Solvay’s low-friction and low-wear Polyamide-imide compound with 20% graphite and 3% fluoropolymer.
Torlon® 4435 is Solvay’s exceptionally low wear performance grade Polyamide-imide (PAI)
compound made for non-lubricated applications.
VICTREX® 450FC30 is Victrex’s lubrication grade PEEK, with standard viscosity PEEK
compound and 30% carbon/PTFE for injection molding and extrusion.
3. Results and Discussion
3.1 Initial 3 x 3 High-PV Wear Test Results
To enable the reader to obtain a more complete understanding of the test results, the wear test
data is included in the attached Tables 3, 4 and 5; and Figures 1-9. The reader will find it
beneficial to read this discussion in conjunction with these tables and figures.
No wear performance data could be generated for Torlon 4203L. All three specimens melted and
were completely destroyed within 2.5 hours of starting the break-in interval at PV = 50,000 psift/
min; Velocity = 200 fpm.
VICTREX 450FC30 survived just over half of the Initial 3 x 3 grid. Wear factors, coefficient of friction
and running temperatures were amongst the highest. (Lower is better for all three measures.)
One specimen melted during the PV = 100,000 / 200 fpm condition. Another melted at the end of
the PV = 100,000 / 50 fpm condition. The third specimen, plus two additional specimens melted
very early in the PV = 50,000 / 800 fpm condition; therefore, no data could be generated at 800
fpm for this material.
Torlon 4275 demonstrated measurable improvement relative to the first two materials. It
performed well at the lower PV level (50,000 psi-ft/min/) under the conditions of low surface
velocities 50 and 200 fpm. Nevertheless, its wear rate was twice that of the best material. It
melted or showed signs of melt at 100,000 PV for all three velocities, performing best at low
speed (50 fpm). Its coefficient of friction was notably high, and thereby exhibiting high countersurface
temperatures. Due to melting, it could not continue to the extended high-PV test.
Torlon 4435 survived the entire Initial 3 x 3 grid. Its frictional coefficient was very similar to
Vespel SP-21, or perhaps slightly lower. At 50 and 200 fpm surface velocities its wear factor was
similar to SP-21, but was nearly twice SP-21 at 800 fpm, and its wear factor was nearly twice VICTREX
TL-60 under almost any condition. At the 800 fpm condition, its counter-surface temperatures were
50 to 90°F hotter than VICTREX TL-60.
Vespel SP-21 had difficulty with tests at high load and low speed, but performed well at moderate
and high speeds despite increasing temperature due to friction. One of the Vespel SP-21
specimens broke during the break-in cycle under the 50,000 PV / 200 fpm condition (250 psi), so
a substitute specimen was broken in and introduced to fill this space in the grid. A second
specimen broke during testing under the 100,000 PV / 50 fpm condition (2000 psi). The other two
specimens completed the 3 x 3 test grid. Wear factors for SP-21 in this 3 x 3 grid were about
twice that of the VICTREX TL-60 at the lower velocity and measurably higher at the middle and
upper velocity conditions. The Vespel SP-21 frictional coefficient was notably higher than VICTREX
TL-60; therefore, counter-surface temperatures were also higher – by about 50°F.
VICTREX TL-60 exhibited the lowest wear factors, the lowest coefficients of friction and the coolest
counter-surface temperatures of the group. At the highest surface velocity (800 fpm) the
differences in counter-surface temperatures were dramatic. VICTREX TL-60 ran about 50°F cooler
than the next lowest material – Vespel SP-21, and about 200F cooler than the highest – Torlon
4275.
Overall, machined VICTREX TL-60 parts from stock shapes produced almost identical results to the injection molded TL-60 parts. At several conditions, the numbers were slightly better for the
machined parts, but within the variability seen at the replicate point (+/- 20%) the results were the
same. A difference was observed, however, at 800 fpm, where the counter-surface temperatures
for machined TL-60 parts ran about 30°F cooler than those coupled with their injection molded
cousins. Also, the injection molded TL-60 parts showed some slight mushrooming beginning at
1500 psi loading in the 50 fpm tests; whereas, the machined TL-60 parts from stock shapes did
not. The slight mushrooming of the iInjection molded TL-60 parts did not change through 3500 psi.
Scanning Electron Microscope images of untested injection molded TL-60 parts showed filling
fibers horizontal to the thrust surface, supporting a view that filler orientation in injection molded
parts could allow mushrooming. Parts with more randomly oriented fibers would benefit from
added structural reinforcement.
3.2 Extended High-PV Wear Test Results
To enable the reader to obtain a more complete understanding of the test results, the wear test
data is included in the attached Tables 3, 4 and 5; and Figures 10-18. The reader will find it
beneficial to read this discussion in conjunction with these tables and figures.
The specimens which were continued to the Extended High-PV grid included: VICTREX TL-60,
Torlon 4435, and Vespel SP-21. Note that the machined TL-60 parts could have undergone the
xtended test, but in the interest of efficiency, and because the injection molded parts represented
the most conservative and critical review of the material, only the iInjection molded TL-60 parts
were carried forward. The Torlon 4203L, Torlon 4275 and VICTREX 450FC30 did not survive the
initial 3 x 3 grid and therefore, could not be extended.
All three Torlon 4435 specimens failed at the first interval beyond the initial 3 x 3 grid at 125,000
PV at all three velocities by melting within the first hour VICTREX TL-60 continued to do well up to
200,000 PV at 50 fpm, 225,000 PV at 200 fpm, and 175,000 PV at 800 fpm. Without exception, the
VICTREX TL-60 exhibited the lowest wear factors, the lowest frictional coefficients and the lowest
counter-surface temperatures at a given PV throughout the extended high-PV test, until failure.
At the 125,000 PV / 50 fpm condition (2500 psi), only VICTREX TL-60 survived; all other materials
had either broken or melted by this point. VICTREX TL-60 not only survived this condition, but went
on to 200,000 PV / 50 fpm, a 4000 psi load.
Vespel SP-21 achieved the absolute highest PV in the group. It did very well at high speeds and
in fact achieved 400,000 PV at 800 fpm and 350,000 PV at 200 fpm. Above 250,000 PV, the 800
fpm specimen needed help to get started, however. That is, when the test first started the friction
was too high for the ¾ hp motor to turn the specimen so the load was partially lifted off the arm
until the counter-surface heated enough to drop the friction. This took between 25 seconds and 2
minutes to level out, after which it was stable. This is not unexpected for Vespel SP-21 as its
static coefficient of friction at ambient temperature is about 0.3, but drops substantially when the
surface temperature reaches 300°F.
The high velocity (800 fpm) Vespel SP-21 specimen was stopped when the torque within a test
became too great for the test machine to maintain the proper speed. The final test condition
(400,000 PV/ 800 fpm) duration and wear factor are approximate as the torque became too high
during the test and stopped the machine. This duration of this test was limited to about 3 hours.
While Vespel SP-21 demonstrated excellent extension of PV range, it consistently demonstrated
a high coefficient of friction, high counter-surface temperature, and a high wear rate relative to VICTREX
TL-60. At any given PV the counter-surface temperatures were higher by about 50F. In
the extreme, the counter-surface temperature in the Vespel SP-21 test reached 638F. It is
impressive from the standpoint that it ran without melting, but more notable for that fact that the
high temperature was induced by its high coefficient of friction. Vespel SP-21 also did not
tolerate the high-load/ low-speed conditions as well as VICTREX TL-60 since it is inherently lower
strength. Compare (ambient) flexural strength – 16 ksi for machined SP-21 vs. 23 ksi for injection molded TL-60. This was exhibited by three Vespel SP-21 washer fractures – two in the
Initial 3 x 3 grid (250 and 2000 psi) and the third in the extended grid (at 1875 psi). Had the
Vespel SP-21 been Direct Formed, Flexural Strength would be lower still at 12 ksi.
4. Conclusions
Injection molded Torlon 4203L melted due to frictional heat created at PV of 50,000 psi-ft/min;
200 fpm velocity during the break-in cycle, making it ineligible for comparison in this High-PV test.
And while injection molded VICTREX 450FC30 did survive several conditions of the initial 3 x 3 PV
grid, its wear factor and counter-surface temperatures were quite high.
Injection molded Torlon 4275 performed reasonably at low velocity (50 fpm) for the initial 3
PV’s: 50,000; 75,000 and 100,000; however, it generated a significant amount of heat and the
wear rate was high in comparison with the rest of the field.
Injection molded Torlon 4435 reached a PV of 100,000 psi-ft/min (for all three velocities) –
which is high for most engineering plastics; however, its wear rate was measurably higher than
the best. At PV’s of 50,000 and 75,000 with surface velocities of 50 and 200 fpm its wear factor
was about equal to Vespel SP-21. Its wear rate was very high at 800 fpm. Beyond 100,000 PV
(at all three velocities) it melted.
Interestingly, all three Torlon materials and VICTREX 450FC30 failed by melting, yet the highest
counter-surface temperatures recorded prior to melt and/or destruction were well below the test
specimen’s respective melt points. Therefore it should be noted that the counter-surface
termperature recorded may not reflect the peak temperature of the specimen at the time of failure.
Also, the samples are under considerable load, so damage can be expected to occur much
nearer to the glass transition temperature. So, while these materials failed relatively early in this
test as compared with SP-21 and TL-60, they are known wear grade materials and their
widespread use in lower PV situations indicates they provide value in less demanding applications.
Machined Vespel SP-21 was capable of the highest operating PV’s of the group; reaching a
PV of 400,000 psi-ft/min. The final condition, 400,000 PV / 800 fpm had to be stopped, however,
as the torque became too high to maintain the speed. SP-21’s wear rate and counter-surface
temperatures were not all that low, however. In fact, Vespel SP-21’s counter-surface temperature
typically ran 50°F hotter than VICTREX TL-60 and wear rates were frequently 50-100% higher than
TL-60 at any given condition. Vespel SP-21 was also prone to failure under load as exhibited by
washer fractures at 250, 1875 and 2000 psi. It did best with low loads and was the only survivor
beyond a PV of 225,000 psi-ft/min.
VICTREX TL-60 is an exceptional wear grade material reaching a PV of 225,000 psi-ft/min at 200
fpm. It exhibited the lowest wear factors, the lowest coefficients of friction and the lowest countersurface
temperatures of the group. At the highest surface velocity (800 fpm) the differences in
counter-surface temperature were dramatic; as it ran about 50°F cooler than the next lowest
material – Vespel SP-21, and about 200°F cooler than the highest – Torlon 4275. The value of
low operating temperature should not be discounted, since the temperature of a bearing surface
frequently determines the PV limit as long as mechanical strength is sufficient. And since
engineering plastics lose strength with increasing temperature, it frequently comes back to this
point.
Choice of a bearing material requires a match between the conditions of the application and
the attributes of subject materials. And as the empirical evidence within this study shows, it is not
sufficient to look only at PV. It is important to consider wear rate, operating temperature and
strength requirements. And when looking at PV, both the velocity and pressure components
should be considered because some materials are better at high speed/ low load and vice versa.
Opening one’s mind to consider all the repercussions will result in the best material choice.
Acknowledgements
Lewis Research, Inc., Lewes, Delaware
Parkway Products, Inc., Atlanta, Georgia
Piper Plastics, Inc., Chandler, Arizona
Polymics, Ltd., State College, Pennsylvania
Victrex plc, Thornton Cleveleys, Lancashire, United Kingdom
References
ASTM D 3702 -94 (1999) Standard Test method for Wear Rate and Coefficient of Friction of
materials in Self-Lubricated Rubbing Contact Using a Thrust Washer Testing Machine.
DuPont Company, Summary of Typical Properties Standard SP Polyimide Resins, www.dupont.com
DuPont Company, Using Vespel Bearings – Design and Technical Data, www.dupont.com
About the author
Michael Gruender is VP of Marketing and Sales for PBI Performance Products, Inc. He holds a
BS Chemical Engineering from the University of Missouri in Columbia, MO, and a MBA in
Business Administration from Winthrop College in Rock Hill, SC. He is employed by PBI
Performance, Products, Inc., a producer and innovator of high performance materials based on
polybenzimidazole.
