Abstract
The guide shoe, as a core guiding component of elevators, plays a crucial role in determining the safety, smoothness, and reliability of the elevator system. To quantitatively compare the speed-dependent tribological responses of candidate polymer guide-shoe materials, four representative polymers—polyurethane, natural rubber (NR), polytetrafluoroethylene (PTFE), and PA6 nylon—were investigated under controlled ball-on-disc sliding conditions against Q235 steel. The speed exponent n and speed wear factor Kv of each material were quantitatively characterized, clarifying the speed effects on the wear behavior of the four polymers. To complement the physics-based wear analysis, a condition-level machine-learning framework was further evaluated for interpolating the late-stage coefficient of friction at withheld intermediate sliding speeds using prescribed velocity and physically motivated contact descriptors. The data-driven analysis is treated as a proof-of-concept for within-range interpolation rather than as a general predictive model for untested materials or service conditions. The results provide experimentally derived parameters and comparative data for laboratory screening of polymer guide-shoe materials with respect to sliding-speed sensitivity. The fitted speed exponents ranged from 0.4726 for PA6 nylon to 2.2790 for NR, while GPR achieved an R2 of 0.97536 and an MAE of 0.02559 on the eight held-out material–speed conditions.
