Numerical sealing performance assessments of PTFE rotary lip seals based on the elasto-hydrodynamic analysis with the consideration of asperity interactions and accelerated wear experimental validations

Tribological experiments were conducted by using a plate-on-disc tribometer to determine the wear rate of polytetrafluoroethylene (PTFE) filled with 5 % glass fiber and 5 % MoS₂ sliding against SKD11 counterparts. An experimental approach is proposed to accelerate the wear of PTFE rotary lip seals leading to earlier leakage failure. Numerical simulations based on the wear model with the consideration of the fracture of PTFE composite caused by asperity interactions are introduced to describe the dependency of roughness parameters of both the seal lip and shaft on the dry wear rate. The elasto-hydrodynamic analysis (EHA) integrated with the convolutional neural networks (CNN) model in the U-net architecture is implemented into sealing performance assessments to significantly enhance its computational efficiency. Simulation results show that the predicted minimum sealed pressure required to induce the leakage failure of seals are in fair agreement with the corresponding experimental measurements.