Reciprocating sliding contact, Thermal contact conductance, Heat flow distribution coefficient, Surface contact temperature, BP neural network, Engineering (General). Civil engineering (General), TA1-2040

The frictional heat generated leads to elevated surface temperatures, which markedly influence the reliability and service life of friction pairs. Considering the dynamic changes in the thermal contact conductance and friction coefficient, the heat transfer equations which consist of heat exchange due to temperature difference and frictional heat generation at the sliding interface have been established. A finite element heat transfer model is constructed between two rough reciprocating sliding surfaces. The influences of the reciprocating motion and the interface thermal contact conductance on the heat flow distribution coefficient and surface contact temperature are solved by numerical simulations respectively. Furthermore, a prediction model is developed based on the BP neural networks. The results indicate that the heat flow distribution coefficient and surface contact temperature increase with rising motion frequency or interface thermal contact conductance and eventually reach a steady state. Moreover, for a fixed motion frequency, both parameters increase linearly with motion amplitude under different interface thermal contact conductance. The prediction model for heat flow distribution coefficient and surface contact temperature shows average relative errors of 0.45 % and 3.53 %, respectively. This research provides a new efficient way to analyze heat transfer in reciprocating sliding contacts and predict the contact surface temperatures.