Effect of solid contaminants on ratcheting and crack nucleation in wheel-rail contact

Solid contamination is a key issue in recent studies on wheel-rail contact. In fact, these components are more and more called to resist to the severe damage due to the presence of sand at the contact interface, caused either by the intentional addition as friction modifier, or by operation in sandy environments, such as deserts. In this work, a fast numerical method for studying ratcheting in solid contaminated contact is presented. It is based on an analytical plane strain model for determining pressure distributions in solid contaminated contact, coupled with the non-linear kinematic hardening model of Chaboche-Lemaitre for the material cyclic plasticity. As a case study, the method was used for assessing the results, previously published, of experiments carried out by means of a bi-disc test bench, where three wheel steels were coupled with the same rail steel in sand contaminated contact. The steel constitutive laws were obtained by approximating with the Chaboche-Lemaitre model the previously published data. The pressure distributions were obtained considering round equally-spaced particles of solid contaminant, with three different sizes within a realistic range. The calculations showed that, at a depth comparable with the particle size, the stresses are mainly influenced by the contact between the particles and the main body (disc), determining stress peaks much higher than the case of clean (non-contaminated) contact; furthermore, each load passage causes multiple stress cycles, due to the number of particles entrapped at the contact interface. As far as the depth increases, the role of the particles decreases and the stresses are similar to the case of clean contact. Depending on the material cyclic yield stress, this stress field can determine ratcheting at a twofold level: at a surface layer, mainly driven by the local contact between the contaminant particles and the discs, and at a sub-surface layer, mainly driven by the global contact between the contacting discs. Ratcheting cannot practically be avoided at the surface level in contaminated contact; on the contrary, it can be limited at the sub-surface level by increasing the steel cyclic yield stress. Furthermore, the contaminant particle size appears very important in determining the severity of the damage: the smaller the particles, the thinner the highly damaged surface layer. The calculated results were consistent with the experimental ones: softer material showed ratcheting involving both the surface and sub-surface layer, whereas in hard materials the damage was confined to the surface layer.