The influence of Λ ratio and surface roughness on the initiation and progression of micropitting damage

Abstract

Micropitting is a major failure mode in gears and rolling bearings. Despite its practical importance, there currently exist no universally accepted design guidelines for its prevention primarily due to the great number of influencing parameters. In the absence of an established criterion, the Λ-ratio (the ratio of lubricant film thickness to surface roughness) is often used as a simple way to assess the risk of micropitting. In this paper we present new data to establish and decouple the individual effects of two of the most important influencing parameters in micropitting: the surface roughness amplitude and Λ-ratio. The experiments are conducted on a triple disc contact fatigue rig and carefully designed to decouple the effect of roughness from that of the Λ-ratio. To isolate the influence of the Λ ratio, we use specimens with the same and very tightly controlled surface roughness and then vary the film thickness by blending different viscosities of the same PAO base oil while keeping all other influential parameters constant. To isolate the influence of surface roughness, we keep the Λ ratio the same using the same PAO oils with appropriate viscosities while changing the RMS roughness of the specimen and keeping other roughness parameters as constant as possible. All tests use case carburised 16MnCr5 gear steel specimens. Results show that higher Λ ratio at fixed roughness produces less micropitting as may intuitively be expected. More importantly, at a fixed Λ ratio the amount of micropitting damage was extremely sensitive to the actual surface roughness, with lower roughness both lengthening the micropitting incubation phase and decreasing the rate of material loss in the micropitting wear phase. This shows that Λ ratio on its own cannot be used to assess the risk of micropitting because it is not able to account for the effects of different roughness levels on micropitting, despite including roughness in its definition. Furthermore, in relative terms micropitting was more sensitive to surface roughness than to Λ. In addition, we show that there exists a sufficiently low surface roughness below which micropitting is unlikely to occur even at extremely low Λ ratios. The results presented here can help in design of gears and other mechanical components to mitigate against micropitting.