Rail transport has become a key element in our society, offering a sustainable and affordable solution for mobility while meeting global climate goals. Nevertheless, the increasing number of passengers and growing availability of trains have reduced the time for maintenance, increasing the risk of failures. Among various rail failure mechanisms, the ¿squat¿ has become a challenge for railway operators worldwide, characterised by a rapid increase in its incidence. This thesis is dedicated to an extensive investigation of the squat, aiming to elucidate its initiation and propagation mechanisms and analyse the critical influencing factors by means of numerical analysis and experimental methodologies. The residual stress state of rails and its influence on squats has been investigated. To this end, two measurement techniques, the contour method (CM) and the X-ray diffraction (XRD) method, have been applied to damaged rails with squats. It was found that the near-surface residual stresses can only be captured by XRD, but the overall distribution can be visualised by CM due to its underlying resolution. However, no obvious features of the residual stress distribution due to the squats in the surrounding area could be detected. Extensive metallographic studies have also been carried out to analyse squats at different stages in different materials. The findings are largely in agreement with expectations, but there are indications that subsurface cracks may also act as initiating or accompanying points for squats. The main part of the work was the development of a numerical model for the investigation of squats. As a first step, a cyclic rollover model was developed and validated with experiments on the full-scale wheel¿rail test rig. The results of this global finite element model serve as input for a submodel that contains penny-shaped cracks as initiation points of squats. In conclusion, this thesis contributes to the current knowledge of squats and their initiation mechanisms. The proposed numerical approach allow a better understanding of the mechanisms controlling squat initiation and may act as basis for further research in order to finally prevent squats in rails. Solving the puzzle of squats will increase rail life and reduce costs for railway operators.