Towards the Identification of Wheel-Rail Contact Forces

The accurate determination of the contact forces between rail and wheel is essential inthe analysis of different kinds of rolling contact fatigue. As the contact forces cannot bemeasured directly in the field, one approach is to measure the strain at certain positionsof an instrumented wheel. Upon employing signal processing techniques, the forces can beestimated. However, such a scheme typically involves restrictions in terms of the choice ofspatial and temporal discretization of the underlying equations of motion or a neglect ofthe inertia terms.In this work, the vertical contact force is determined by the solution of an inverse problem.A minimization problem is considered in which the time-history of the contact forceis sought such that the discrepancy between the predicted and the measured strains isminimized. A particular feature of this formulation is that the discretization of the pertinentstate equations in space-time, the sampling instances of the measurements and theparameterization of the sought contact force are all independent of each other. Additionally,a well-established mathematical framework can be exploited to analyze the effect ofmeasurement noise as well as devise suitable strategies for regularization. Moreover, theconvergence of the spatial and temporal discretization of the model and the time parameterizationof the contact force history are investigated.In the first paper of this work, the proposed strategy is evaluated for a simplified 2D discwith focus on the temporal discretization. In addition, sensitivity to noise and improvementsdue to proper regularization are investigated. In the second paper, the identificationstrategy is modified by applying virtual calibration in order to compensate for spatial meshsensitivity