Railway crossings are one of the most important and vulnerable components in railway network. Nowadays, due to intensive use of the track together with higher train speeds and heavier axle loads, more and more problems associated with crossings are reported and it is continuing to be an important factor limiting its service life. In this MSc thesis, a realistic 3D finite element (FE) model of the crossing panel is developed to analyse the stress state arising from the impact event and, providing recommendations on how to effectively mitigate the impact loads. The scenario which is simulated and studied in this report is that of a train wheel passing a railway crossing in the facing as well as in the trailing direction. Prior to the FE modelling, first, 2D-geometric contact analysis is performed calculating all the contact properties at the wheel-rail interface. The obtained contact properties are then used as guidance during the FE modelling to implement adaptive mesh refinement at the running band of the wheel in order to get accurate solution of the rolling contact stresses. From the FE simulation results, high impact forces can be observed in the transition zone of the crossing. The detailed surface and subsurface stress analysis reveals that these forces generate high contact stresses subsequently causing yielding of the materials and intense plastic strain accumulation. Verifications and validations are carried out to examine whether the results from the FE model are correlating with the reality. From them, attention has been paid to minimize undesirable effects of the boundaries and to verify the convergence of the solution. Besides, the response of the FE model is validated against the field experiment of the axle box and the crossing nose accelerations. Thereafter, a parametric study is performed to investigate the influence of some interesting case studies on the magnitude of the impact loads. In this regard, a comparison of the contact properties utilizing elastic and plastic material models showed conformities as well as discrepancies in the stress state for these two material models. Besides, several cases studies have been carried out on the vertical substructure stiffness variation and geometric design modification of the crossing panel. From this it can be concluded that the investigated case studies provide interesting opportunities to reduce effectively the impact forces on the crossing. Moreover, a comparison between facing and trailing direction has managed to identify the impact force behaviour for these two different operational conditions. So in short, the FE model is thus capable to solve the wheel-crossing contact stress problem at the crossing and it is flexible enough to examine the influence of design modifications on the impact force and its resulting stress state under different operational conditions.