Modelling contact in the wheel-rail interface
- Professor Mats Berg (Principal supervisor)
- Dr. (Supervisor)
- Lic.Tech. Matin Sh. Sichani (PhD student)
The interface between wheel and rail is vital to the dynamic behaviour of a rail vehicle, and a feasible description of the interaction is necessary in a simulation context. In a transient simulation, this evaluation is performed at every contact location in each time step. Thus, fast and simple numeric algorithms are required.
In connection with recent research related to wheel and rail performance prediction, the traditional methods for calculating contact stresses and slip have been found insufficient. For simulation of deterioration phenomena like wear and fatigue, the resolution of the contact evaluation in terms of shape of the contact area and its stress distribution needs to be improved.
The aim of this research is to arrive at a model practically applicable in the context of vehicle dynamics simulation, resting on a firm scientific foundation and answering to modern requirements regarding precision and numerical efficiency.
Limitations related to traditional methods, for instance geometrical constraints, elastic identity, or half space assumption, are expected to be overcome.
The small and highly stressed contact patch is the interface to the infrastructure to be evaluated at each time step in a transient analysis. Thus, the model has to be numerically efficient. Traditional methods often used in this context are Hertz' method for the normal contact and Kalker's simplified model for the tangential solution.
The starting point of this project is a survey of recent pertinent research and related modelling ideas. Evaluation of approaches like multiple ellipses, discretisation by strips, various amendments to Kalker's methods, Winkler-type elastic foundations, and more is anticipated. The feasibility of modern numerical methods like boundary element discretisation should be investigated as well.
Some important steps are listed below:
- Determination of the shape and size of the contact patch and the contact pressure distribution. With the traditional half space assumption, the normal contact becomes well defined. In case of small radii or close to conformal contact, this condition may be violated. Thus, an improved model shall be able to handle non-elliptic contact areas on curved surfaces.
- Assessment of the shear stress distribution. With the traditional assumptions of quasi-identical contacting bodies, the normal and tangential problems can be solved independently. Analysis of more general contacts may however require simultaneous solution.
- Selection of numerical algorithm and implementation. With modern computer power, more sophisticated numerical methods than traditionally may be realistic. A competing consequence of the improving computer capacity is however increasing expectations on model size.
- Validation. Since the research target is some kind of simplified model it is possible to verify it by more detailed calculations like finite element analysis. Experimental verification is desirable and ultrasound measurements may be an option.