Improved railway wheelset–track interaction model in the high-frequency domain

Martínez-Casas, J., Giner-Navarro, J., Baeza, L. and Denia, F.D. (2017) Improved railway wheelset–track interaction model in the high-frequency domain. Journal of Computational and Applied Mathematics, 309, 642-653. (doi:10.1016/j.cam.2016.04.034).

Abstract

As it is well known, there are various phenomena related to railway train–track interaction, some of them caused by the high frequency dynamics of the system, such as rolling noise when the vehicle runs over the track, as well as squeal noise and short-pitch rail corrugation for curved tracks. Due to these phenomena and some others unsolved so far, a large effort has been made over the last 40 years in order to define suitable models to study the train–track interaction. The introduction of flexibility in wheelset and rail models was required to have a more realistic representation of the wheel–rail interaction effects at high frequencies. In recently published train–track interaction models, the rails are modelled by means of Timoshenko beam elements, valid up to 1.5 kHz for lateral rail vibration and up to 2 kHz for vertical vibration. This confines the frequency range of validity for the complete train–track model to 1.5 kHz.

With the purpose of extending the range of validity above 1.5 kHz, a 3D track model based on the Moving Element Method (MEM) is developed in this paper to replace the Timoshenko beam considered in earlier studies, adopting cyclic boundary conditions and Eulerian coordinates. The MEM approach considers a mobile Finite Element (FE) mesh which moves with the vehicle, so the mass of the rail ‘flows’ with the vehicle speed but in the opposite direction through the mesh. Therefore, the MEM permits to fix the contact area in the middle of a finitely long track and to refine the mesh only around the contact area, where the forces and displacements will be more significant. Additionally, a modal approach is adopted in order to reduce the number of degrees of freedom of the rail model. Both strategies lower substantially the computational cost. Simulation results are presented and discussed for different excitation sources including random rail roughness and singularities such as wheel flats. All the simulation cases are carried out for a Timoshenko beam and a 3D MEM track model in order to point out the differences in the contact forces above the range of validity of the Timoshenko beam.

This record has no associated files available for download.

Contributors

Author: L. Baeza

Download statistics