Validation of a model of a railway vehicle based on CFD function comparison

Abstract

Recent improvements of computers and numerical tools gave engineers a possibility to develop and use very complex numerical models. One of the essential steps in the model developing is a process of verification and validation. Traditionally, comparisons of time histories on a graph have been used to make subjective engineering judgments, how well the histories agree or disagree. In order to introduce an objective assessment of agreement between simulation and experimental results of two validation metrics were presented in this paper. The Area validation metrics and Kolmogorov-Smirnov test, based on the comparison of the cumulative probability functions, were used for validation of developed model of a rail vehicle. The obtained results have shown that presented validation metrics may be used for the model validation. However, the subjective judgment in the process of the validation cannot be completely avoided. The future work should be focused on development of a new validation metric which will eliminate subjective judgment in the process of the model validation.