Experimental extrapolation model for friction and wear of polymers on different testing scales

Abstract

The friction and wear behaviour for polyoxymethylene homopolymers (POM-H) and polyethylene terephthalate with teflon additives (PET/PTFE) is compared on small-scale cylinder-on-plate tests (50–200 N normal loads) and large-scale flat-on-flat tests (190–3880 kN normal loads). A common parameter to characterise tribological data is the contact pressure×sliding velocity ( pv-value), but its use seems restricted to a single testing scale. Four experimental models are presented to extrapolate tribological data from one to another testing scale, based on (i) one single mechanical parameter (normal load or contact pressure), (ii) two mechanical parameters (normal load and sliding velocity), (iii) the contact pressure–sliding velocity model ( pv-temperature limit), (iv) macroscopic geometry model. The latter model is most extensive, considering the influences of thermal effects (frictional heat generation and dissipation), sample geometry (geometry factor G) and visco-elastic contact (critical contact pressure p 0). For unfilled polymers, the introduction of macroscopic scaling factors allows for the extrapolation of coefficients of friction obtained on different testing scales. Specific or volumetric wear rates cannot be extrapolated because they strongly depend on the sample geometry, while linear wear rates are in better agreement when considering the transitions between mild wear, softening and melting. For internally lubricated polymers, extrapolation is more difficult. The differences depending on the testing scale are attributed to contact stress concentrations near the sample borders and limited wear debris mobility within large contact areas, promoting a homogeneous film formation onto the polymer surface.