Failure stress modification in fatigue life models for rolling bearings

Abstract

The critical subsurface shear stress related to rolling contact fatigue is modified to model the effects of residual stress common in case hardened materials, such as M50-NiL. The role of hoop stress, generated due to race rotation and shrink fits, is also modeled. It is shown that even relatively low levels of compressive residual stress could contribute to notable increase in bearing life. An equivalent life modification factor is dependent on both residual stress and applied load. Model predictions are in agreement with available experimental life data obtained with a 40-mm angular contact ball bearing with M50-NiL races and silicon nitride balls. The stress modification approach is also applied to model the role of any fatigue limiting shear stress, such that the solutions converge to validated Lundberg–Palmgren solutions as limiting stress reduces to zero. However, bearing life predictions at light loads, under any reasonable limiting stress, are unreasonably high. As an alternate approach, the empirical constant in the limiting stress model, with a prescribed limiting stress, is determined by least-squared regression between model predictions and available experimental life data. With such an approach, the least-squared deviation between model predictions and experimental data shows a monotonic increase as a function of the limiting stress with a minimum at no limiting stress. This observation suggests that simple failure stress modification in the current subsurface stress-based models may not be suitable to implement any fatigue limiting stress for rolling contacts.