A new method for nonlinear dynamic modeling of wheel/rail multiple contacts

The current trend in railroad industry is the development of reliable nonlinear computational dynamic algorithms that can be used in the simulations of vehicle behavior under different operating conditions. Accurate modeling of the wheel/rail contact, and dealing with the nonlinearities associated with the kinematic constraints and track geometry are perhaps the most challenging problems encountered in the development of accurate and robust specialized railroad vehicle dynamic simulation algorithms. In this thesis, a velocity transformation method is proposed for railroad vehicle systems in order to deal with the nonlinearity of the constraint equations when the vehicles negotiate curved tracks. In this formulation, two different sets of coordinates are simultaneously used. In addition to the problem of nonlinear constraints, the wheel/rail contact problem becomes very complex when the vehicle traverses rail switches or worn rail segments. Complexities arise due to the continuous change in the rail geometry along its space curve, which makes multiple points of contact very probable in such scenarios. This thesis proposes a geometric finite element based method for the representation of the rigid rail surface when the profile is allowed to change as a function of the arc length of the rail space curve. Modeling such geometric changes in railroad vehicle applications requires the development of a wheel/rail contact search method that is capable of capturing multi-point contacts that characterize wheel/switch interaction. The proposed methods in this thesis can be widely used in specialized multibody railroad vehicle simulation codes. This will eventually contribute to more sophisticated dynamic simulation algorithms that are essential for the development of new, safe, and economical designs in railroad industry.