Comprehensive efficient vertical and lateral track dynamic model to study the evolution of rail corrugation in sharp curves

Abstract

• A highly accurate and fast model is developed for the study of rail corrugation. • Experimental receptances are precisely fitted using RFP and genetic algorithms. • Lateral system dynamics is added to the model preserving short computational times. • The model has been validated with experimental measurements. This paper addresses a time/space domain model, extremely efficient in terms of computation, which combines the vertical and lateral dynamics of the track to predict undulatory wear depending on track and vehicle parameters. As it is known, this is a source of noise and vibration that can provoke very high levels. The model is obtained from the transformation of a model in the frequency domain to the time/space domain by means of the rational fraction polynomials method. Multiobjective genetic algorithms are used to minimize the error in fitting receptances and to ensure the stability of the system. This model includes the vertical wheel-rail dynamic interaction with a Hertzian spring, and the FASTSIM algorithm is used to study tangential contact. The model is validated with the experimental corrugation measurements taken on a metro line.