The Cross-Scale Interactive Simulation and Artificial Neural Network Failure Prediction of Temperature Field on Wet Friction Interface

Aiming at the temperature failure for wet friction interface within mechanical transmissions, this article constructs a single-pair asperities contact model, which considers the influence of rough surface micro-contact states on the wet friction interface temperature field, and it interacts with the macro thermal-mechanical coupled simulation model in real-time. Then, the interface temperature field is analyzed by cross-scale interactive simulation model, and two types of temperature field failure characterization parameters are proposed. The accuracy of the cross-scale interactive simulation model is validated through the wet clutch accelerated life test. Furthermore, this article constructs a failure prediction model focusing on critical circumferential and radial temperature differences. Subsequently, it investigates the influence patterns of pressure and speed on the failure characteristic parameters based on the controlled variable method. The results indicate that the cross-scale interactive simulation model exhibits higher reliability and accuracy compared to the macro thermal-mechanical coupled simulation model. The average error of the failure prediction model is 3.71%. Additionally, the failure characteristic parameters are more sensitive to pressure variations.