Engineering plastics are widely used in many tribological applications due to
their inherent advantages such as reduced weight, ease of manufacturing,
improved chemical compatibility, and damping characteristics. However, the
process of selecting an appropriate polymeric material system for a specific
application involves significant experimentation. Although, standardized methods
of evaluating tribological performance of engineered plastics exist, their
ability to be indicative of part performance in the end application is rather
poor, primarily because of (a) the typically low pressure-velocity combinations
used in these tests, (b) their inability to properly differentiate between
various failure modes (wear vs creep vs melting of the plastic in case of
thermoplastic polymers) and (c) the large number of variables that exist in a
tribological system including pressure, relative speeds, thermomechanical
properties of mating components, surface roughness/hardness parameters, type of
lubricant and most importantly, part geometry and large variability of thermal
management in the application. In the present work, a four hour test to evaluate
dry and lubricated wear of high performance plastics is developed. This allows
for multiple repeats to understand variation within samples. A modified
multi-grooved test specimen is used to promote two-body wear by removal of wear
debris through the grooves. For the four hour lubricated test, the
pressure-velocity conditions are chosen to ensure that the test is in mixed
lubrication regime. The experimental set-up is described and tribological data
for a range of engineering plastics are summarized. A case study is presented
where a thrust washer molded using a thermoplastic material that was developed
and rated using the proposed methodology, is compared to a thrust washer molded
from a commercially available material, using test parameters as in the end
application. Results show that the proposed test methodology is suitable for
rating tribological performance of materials and correlates well to component
performance in the end application.
