Effect of microinclusions on the fatigue life of rolling bodies

The goal of this study is to understand how the microinclusions affect the fatigue life of rolling bodies. In order to do this, a numerical method was considered, and its results were correlated with experimental ones, investigating whether a microinclusion threshold parameter could be defined, to be related to the life bearing requirements. Many phenomena are involved in damage of rolling elements of bearings. Rolling contact fatigue is the main cause of failure, along with contact pressure related fatigue and dimensional instabilities effect. Most of those are well known, and are described by wide experimental, analytical and numerical literature. Damage phenomena are related to material properties and manufacturing processes. Nowadays, steels are well-suited in bearings manufacturing, and they represent the most established materials for these applications. Each of these is not free of non metallic inclusions because of the manufacturing process. These inhomogeneities have an import role on the fatigue performances, since the damage evolution of the rolling elements of the bearings is affected by the microinclusions present in the material. This study is focused on the AISI 52100 (100Cr6), that is the main steel in bearings production. Non metallic inclusions like sulfides, carbides and oxides act like stress concentrators, and this can originate cracks that lead to failure. Inclusions have different composition, dimension, configuration and distribution so their effect on the fatigue life changes depending on this parameters. For this reasons, different types of non metallic inclusions were analyzed, in order to have an overview of the phenomena. The stresses around the non metallic inclusions are evaluated thanks to a MATLAB code based on the Eshelby solution for this type of problem. Then, this stresses are counted in order to evaluate a microinclusion threshold parameter. A correlation between stress computed by the model and experimental failures (performed thanks to fatigue test rigs and a rotating bending machine) has been shown, with a good matching of the data.