Modern railway systems are a beneficial way of transport. However, the operation requires maintenance to compensate for rail degradations. Key issues of degradation regard severe plastic deformation and rolling contact fatigue, which lead to the formation of cracks. A better understanding contributes to improve maintenance and operations. Numerical simulations offer ways to predict these degradations.
State of the art for the simulation of solid mechanics is the Finite Element Method. An inherent limitation, though, is the basic assumption that the material is a continuum. Discontinuities like cracks violate this assumption. Thus, alternatives like Discrete Element Method models and Peridynamics were developed recently. These methods describe the material by a modular assembly of elements. This arrangement allows for a better modelling of discontinuities. Though, modelling of rolling contacts is a case where compressive loads dominate. This fact challenges discrete modelling approaches due to the common assumption that the failure of an element implies its removal. Another challenge models of this type face regards the validation of rolling contact fatigue.
The aim of this work was to develop a discrete element model, which is optimised for the rolling contact application. For this reason, it was called the "Discrete Element Rolling Contact" model. Based on a linear-elastic model, a fatigue capability was introduced. This capability was transferred to the set-up of the rolling contact, which highlighted the first limitation stated above. In order to address this issue, a solution was developed and validated to calibrate the model to crack closure.
Experimental results indicate that the fatigue crack growth behaviour of rails is governed by the severe plastic deformation the materials are subject to. For this reason, data of undeformed and deformed materials was parameterised and validated. A method to interpolate materials of varying degrees of deformation was introduced. Further, a method to transfer the material parameters to other materials of similar strength was developed. The result was a material library which is available for the model.
In order to transfer the fatigue parameters to the rolling contact condition, a set-up to model the behaviour of highly shear deformed material was introduced. In this way, a drawback of the adopted fatigue law was highlighted, which emphasised the requirement to weight the influence of compressive strains. Finally, a procedure for the validation of the rolling contact model was proposed.
In this work, a novel rolling contact fatigue model and a procedure for validation are presented, that consider the anisotropy of severe plastic deformation. With regard to the validation of discrete rolling contact models, the proposed procedure delivers a contribution that addresses key aspects that are not matured yet. These aspects, however, are critical for the development of dependable discrete models to predict rolling contact fatigue.
