Modelling and Simulation of Train-Track Interaction including Wear Prediction

Doctoral thesis, 2003

Many problems in railway traffic arise as a result of the dynamic interaction between rolling trains and tracks. The problems include both instantaneous effects due to induced vibrations of track and vehicle, and more long-term processes of a detrimental nature. This thesis deals with the instantaneous dynamics of the train-track system, and its influence on rail corrugation development as a long-term phenomenon. For this purpose, both numerical and experimental studies are made although the emphasis is laid on numerical studies. A numerical method for the simulation of dynamic train-track interaction was devised. The method includes descriptions of the vehicle, track and their coupling. The dynamically coupled moving load problem is treated in the time domain; the method enables a three-dimensional motion of the vehicle traversing a flexible track. The compound formulation forms a basis for studies of trains on tangent tracks and in curves. The proposed train-track interaction model, together with a wear model, constitutes a framework within which the effects of the compound longitudinal, lateral and vertical dynamics on rail corrugation growth can be investigated. For a tangent track model, calibrated to experimental results from full-scale field measurements, numerical simulations indicate that a high corrugation growth rate at certain wavelengths corresponds to some specific vibrational modes of the coupled train-track system. The numerical studies also include issues related to computational efficiency. It is found that a staggered time-stepping approach is advantageous in train-track interaction simulations. For the same reason, model reduction techniques are also employed.