Parametric Study Comparing Rolling Contact Fatigue Prediction Models Using Multibody Simulations

Abstract

Rolling contact fatigue (RCF) is a major damage mechanism in wheel-rail systems, with strong implications for safety and life-cycle costs. This paper presents an automated framework that combines multibody simulations in Simpack with Python-based pre- and post-processing to quantify RCF and wear under realistic operating conditions. The workflow generates parametric Simpack models from a passenger coach, runs large numbers of simulations in parallel on mainline routes with realistic speed profiles and measured track irregularities, and applies RCF damage models to the resulting contact data. Global RCF models, namely the Whole Life Rail Model and the Surface Fatigue Index approach, are implemented for rails and wheels and compared to an empirical grinding guideline. Parameter studies vary one quantity at a time: rail profile (60E1 versus 60E2), friction coefficient (0.1–0.5), and carbody mass (five loading states), to isolate the influence of design parameters and operating conditions. The results show that the 60E2 rail profile systematically reduces RCF damage and wear compared with 60E1. Higher friction and carbody mass increase damage, with the Surface Fatigue Index model exhibiting pronounced sensitivity to friction. Comparing simulated damage with the grinding limit according to the Heyder model indicates that visible cracks are expected on some sections before grinding, whereas on other sections extended grinding intervals may be feasible. The framework is generic and suitable for future studies and calibration of RCF models.