Comparison of fatigue model results for rough surface elastohydrodynamic lubrication

Abstract

The paper compares the results of applying different fatigue failure models to surfaces that experience elastohydrodynamic lubrication (EHL) and have surface roughness features which are large compared with the equivalent smooth surface film thickness. The surface profiles used for the comparisons reported are taken from test gear surfaces used in an FZG gear test. The surface has a roughness of 0.31 μm Ra, with peak to valley dimensions of the order 1.5 μm. The tips of the asperities have been modified by running-in during the gear test. A transient EHL analysis was carried out using contacting components having this profile for a number of sliding speeds and nominal viscosity values. From these analyses, time-dependent pressure and shear stress distributions at the contacting surfaces were calculated. The resulting subsurface stress field was obtained relative to axes fixed in the moving surfaces using an elastic analysis so as to give the stress history at each point in a representative test section of the contacting component. A number of multi-axial fatigue criteria based on a critical plane approach were applied to the test section and the results compared. In addition, a varying amplitude multi-axial fatigue theory based on shear strain cycles was also applied to the section. The cycle counts were obtained using the rainflow counting method and the accumulated damage in a single pass through the contact area was calculated. In comparing the results obtained, the various fatigue models were found to identify the same asperity features as being those most prone to fatigue.