A Multi-Scale System Analysis and Verification for Improved Contact Fatigue Life Cycle of a Cam-Roller System

Abstract

Surface distress in the form of contact fatigue is encountered in cam-roller systems. The contact fatigue appears to be initiated at micrometer-scale subsurface region. High stress is a result of the macro-scale requirement on the cam-roller motion event that produces high contact loads due to inertia of the roller and its follower link. Sliding of the roller and its impact onto the cam surface further compounds the detrimental effect of contact load. While conventionally a Hertz contact stress analysis can be used in ascertaining contact stress and maximum subsurface von Mises stress, it generally underestimates the stress when compared to the micrometer-scale subsurface stresses due to the presence of surface roughness. Contact analyses of cam and roller with rough surfaces are performed to examine the effects of two surface treatments. These involve surface finishing process in which a surface is rendered smooth, and the addition of a coating to the roller surface. Measurements of such cam and roller surfaces are used in micro-contact analysis module of a Surface Distress Analytical Toolkit to examine the effect of surface finish and coating on maximum subsurface stress. It is found that smooth surface provides a 53% reduction in maximum subsurface stress. The analysis also shows that the addition of coating further reduces subsurface stress nearly 7%. The impact of the combined treatment of the surface is an increase in contact fatigue life of the cam-roller system by nearly two orders of magnitude. The above findings are confirmed by laboratory tests using six rollers with various degrees of finishing processes, and with and without addition of coating to the surfaces. Examination of the rollers indicates a general improvement in roller performance due to addition of coating. Most notably, the combination of finishing process and coating was found to provide the best contact fatigue life since the corresponding rollers showed no observable wear even after testing for 2161h, or the same number of cycles accumulated over about 500,000 truck miles.