Uncertainty Quantification in Rolling Contact Fatigue Reliability of M50 Steel via Microstructure‐Informed Modeling

ABSTRACT Rolling contact fatigue (RCF) failure in M50 bearing steel exhibits pronounced life scatter, arising from both microstructural variability and surface‐condition randomness. However, conventional life prediction approaches primarily rely on empirical statistics and cannot quantitatively attribute reliability dispersion to its physical origins. This study develops a microstructure‐informed UQ framework to predict M50 bearing steel's RCF reliability and identify life variability sources. Using crystal plasticity–continuum damage models and Voronoi‐generated microstructures with randomized grains, combined with stochastic friction coefficients for surface uncertainty, Monte Carlo sampling generates diverse microstructure–friction ensembles. Results reveal load‐dependent reliability mechanisms: At low contact pressures, microstructural randomness governs life dispersion via grain‐scale stress localization; at higher loads, surface friction uncertainty becomes dominant. This transition correlates with increasing Weibull shape parameters, indicating reduced life scatter under heavier loads. The framework physically identifies reliability‐controlling factors, providing a basis for microstructure‐driven bearing design and batch consistency evaluation in aerospace components.