Prediction of sliding speed and normal force effects on friction and wear rate evolution in a dry oscillating-fretting PTFE/Ti-6Al-4V contact

Dry fretting wear tests were performed on a semicrystalline polytetrafluoroethylene PTFE cylinder rubbing on a flat Ti-6Al-4V surface at room temperature. Various sliding speeds and normal loads were imposed for test durations from 5×104 to 106 cycles. The purpose of the investigation was to evaluate how sliding speed and normal load influence friction and wear rate evolution for gross slip fretting-reciprocating sliding conditions. Analysis confirmed an initial transient regime related to the progressive formation of a PTFE transfer film on the titanium counterface, followed by a linear steady-state wear regime. Focusing on the steady-state wear regime, a rising staircase evolution versus sliding speed was observed for both friction and Archard's wear rate. This continuous “sigmoid” wear rate evolution was related to friction heating, inducing transition from micro to macro surface-ploughing wear damage. In addition, a discontinuous cracking wear component was observed from a threshold sliding speed (vth=12 mm/s), due to friction-induced thermal activation of the crystalline phase transition then the principal transition (associated with the glass transition) of the PTFE polymer. The analysis also showed that an increase in normal load induced a parallel reduction in the coefficient of friction and parallel increase in wear rate. These parallel evolutions suggest no direct interaction between sliding speed and normal force. Hence the normal force effect can be analyzed using a (P/Pref)n weight function, where the n exponent quantifies the relative effect of normal load versus friction and PTFE wear rate. These results, which are discussed regarding former research works of Tanaka, are correlated with PTFE thermo-mechanical response, friction thermal considerations and competition between several wear processes. Finally, global wear rate could be formulated, including sigmoid (ploughing wear), bell shape (cracking wear) and weighted (normal force effect) functions. Good correlation between experiments and predictions confirmed the stability of this formulation.