Investigation on the Wheel/Rail Contact and Longitudinal Train/Track Interaction Forces
thesisposted on 15.04.2014, 00:00 authored by Ali Afshari
Previous analytical and experimental investigations have shown that the wheel/rail contact forces have a significant effect on the nonlinear dynamics, ride comfort, and stability of railroad vehicle systems. The wheel/rail contact force can be partitioned into two main components; the normal and tangential components. While the direction of the normal contact force is well defined, different directions for the tangential contact forces have been proposed in the literature. In some of the wheel/rail creep theories used in railroad vehicle simulations, the direction of the tangential creep forces is assumed to be the wheel rolling direction (RD). When Hertz theory is used, an assumption is made that the rolling direction is the direction of one of the axes of the contact ellipse. The RD and CE directions can also be different from the direction of the rail longitudinal tangent (LT) at the contact point. In this investigation, the differences between the contact frames that are based on the RD, LT, and CE directions are discussed. In order to better understand the role of geometry in the formulation of the creep forces, the relationship between the principal curvatures of the rail surface and the curvatures of the rail profile and the rail space curve is discussed in this investigation. Numerical examples are presented in order to examine the differences in the results obtained using the RD, LT and CE contact frames. The effect of the tangential forces becomes more significant in the case of multiple contact points between the wheel and the rail. It is important to predict accurately these forces in order to evaluate their effect on the dynamic behavior of the railroad vehicle system. However, in order to correctly determine these contact forces, including the normal forces, it is important to determine first all possible wheel/rail contact points. Nonetheless, most existing wheel/rail contact formulations can be used to predict only one or two points of contact between the wheel and rail. The lack of accurate three-dimensional multiple point wheel/rail contact formulation represents a serious limitation when derailments and accidents are investigated. In this thesis, a new multipoint contact search algorithm that can be applied to general three-dimensional wheel/rail contact problems is developed. Longitudinal train forces resulting from coupler and braking forces are other types of forces that must be considered in derailment and accident investigations. Nonetheless, the integration of an accurate air brake model with a nonlinear train dynamic model remains a challenging problem. One of the goals of this investigation is to integrate an air brake model with efficient train longitudinal force algorithms based on the trajectory coordinate formulations. The air brake model, developed in this investigation consists of the locomotive automatic brake valve, air brake pipe, and car control unit (CCU). A detailed model of the locomotive automatic brake valve is presented in this investigation and used to define the inputs to the air brake pipe during the simulation. Then, the detailed CCU formulation is presented. Different computer simulation scenarios are considered in order to investigate the effect of the air brake forces on the train longitudinal dynamics in the case of different braking modes.