The wheel profiles of a train affect the safety and dynamic performance of the train as well as economical aspects. This makes changes of the wheel profiles over time, mainly due to wear, an important area to study. 

 This study deals with wear simulations of the wheel profiles of the Regina train, modified for higher speeds up to 250 km/h but still equipped with passive bogies, on the railway line Stockholm - Hallsberg. 

 The vehicle-track simulations are performed with the multibody simulation software SIMPACK. The Hertzian theory has, together with Kalker’s simplified theory, been applied to describe the wheel – rail contact. 

 The methodology of the wear prediction procedure is based on the method developed by Jendel [8]. A simulation set design has been defined by parametric studies to resemble the railway network at hand. The simulation set design includes track design geometry as well as other parameters that influence the wear distribution of the wheel profiles; these parameters are varied in the time domain simulations. Considerations have been taken to the traffic conditions, braking, traction, wheel-rail friction coefficient and track irregularities. It is concluded that braking and traction could be disregarded in curves and are thus only included on tangent track.

 The Archard’s wear model is implemented into SIMPACK as a subroutine, created by Enblom. The subroutine enables wear calculations. The wear is calculated for each case in the simulation set design, after which the wheel profile is updated. The wear step is limited by a maximum wear depth of 0.1 mm and a maximum running distance of 1500 km. A wear map derived by Jendel from laboratory measurements has been used to determine the wear coefficients. Since these measurements were performed under dry conditions, a scaling factor has been introduced to simulate natural lubrication. 

 SIMPACK simulated wheel wear for about 28400 km. The results show wear distributed over the tread as well as concentrated to a point just before the flange. A diagram of the equivalent conicity confirms the expected increase. The simulation encountered a numerical problem to determine the contact point between the wheel and rail, due to a development of a false flange on the wheel profile. 

 Further work on improving the wear map and the simulation settings, since the simulation indicated a heavy rate of wear development, can be seen in the appendixes. Simulation results with the Regina 200 model was verified with measurements, which resulted in changes of the wear map, that specifies the wear coefficients. It can be stated that the changes of the wear map resulted in an improved wear distribution and development when compared to the measurements. The identical simulation set design was then simulated with the Regina 250 model to compare the models, which showed an expected and realistic difference between the bogies. Lastly, the changes of the wear map were applied to the simulation set design of the Regina 250 model, in an attempt to improve the simulation Wear of result. The result showed a smoother wear distribution and a realistic wear shape. However, the lack of flange wear is still a problem.