Pitting Life Prediction Based on a 3-D Line Contact Mixed EHL Analysis and Subsurface von Mises Stress Calculation

Abstract

Surface pitting due to contact fatigue is a major failure mode of many mechanical components , such as various gears and rolling-element bearings. Pitting life prediction, therefore, is vital to design and performance/reliability improvements. Conventional prediction methods, commonly found in industrial standards, are based on the Hertzian contact theory under the assumptions that surfaces are ideally smooth with no lubrication. Today, the trend of high power density, high reliability compact design requires the life prediction to consider severe operation conditions in mixed lubrication, and the effect of surface roughness and topography. Also, it has been well known that ductile material failures in concentrated contacts are better correlated with the subsurface von Mises stress, rather than the normal Hertzian pressure. The present study aims to develop a pitting life prediction approach for line-contact components based on a 3D line-contact mixed elastohydrodynamic lubrication (EHL) model recently developed by Ren et al. (2009, "A Three-Dimensional Deterministic Model for Rough Surface Line-Contact EHL Problems," ASME J. Tribol., 131, p. 011501), which is capable of simulating the entire transition from full-film and mixed EHL down to dry contact of real machined rough surfaces under severe operating conditions. The pitting life evaluation employs the fatigue life model developed by Zaretsky (1987, "Fatigue Criterion to System Design, Life and Reliability," J. Propul. Power, 3(1), pp. 76-83) and extended by Epstein et al. (2003, "Effect of Surface Topography on Contact Fatigue in Mixed Lubrication," Tribol. Trans., 46, pp. 506-513) using the von Mises stress field calculated based on the rough surface mixed-EHL results. Sample cases are analyzed for 15 sets of transmission gears, and the life prediction results are compared with available experimental data. With optimized material constants in the life model, predicted pitting life results well agree with the test data.