This paper presents a numerical study of the effects of inlet supply starvation on film thickness in EHL point contacts. Generally this problem is treated using the position of the inlet meniscus as the governing parameter; however, it is difficult to measure this in real applications. Thus, in this paper an alternative approach is adopted whereby the amount of oil present on the surfaces is used to define the degree of starvation. It is this property which determines both meniscus position and film thickness reduction. The effect of subsequent overrollings on film thickness decay can also be evaluated. In the simplest case a constant lubricant inlet film thickness in the Y direction is assumed and the film thickness distribution is computed as a function of the oil available. This yields an equation predicting the film thickness reduction, with respect to the fully flooded value, from the amount of lubricant initially available on the surface, as a function of the number of overrollings n. However, the constant inlet film thickness does not give a realistic description of starvation for all conditions. Some experimental studies show that the combination of side flow and replenishment action can generate large differences in local oil supply and that the side reservoirs play an important role in this replenishment mechanism. Thus the contact centre can be fully starved whilst the contact sides remain well lubricated. In these cases, a complete analysis with a realistic inlet distribution has been carried out and the numerical results agree well with experimental findings.

No abstract

Rolling contact bearing life is calculated using stresses calculated at the surface and in the volume. Surface stresses account for profile and misalignment as well as asperity deformations. Sub-surface stresses are calculated beneath the asperities (for defining the life of the surface) and deeper in the volume for calculating the life of the volume. The stress-life criterion adopted is the Dang Van one in which the local stabilized shear stress is compared to the material endurance limit defined as a function of the hydrostatic pressure (itself a function of the contact pressure) but also residual stresses and hoop stresses (due to fit). A stress-life exponent c, of the order of 4 (instead of 34/3 in the standard Lundberg and Palmgren model) is used for respecting a local load-life exponent of 10/3 at typical load levels. Life of any circumferential slices of the inner, outer, and roller is defined for obtaining the final bearing life. Trends showing how the bearing life varies as a function of the applied bearing load and Λ ratio (film thickness/RMS roughness height) are given.