Study on Validation Conditions of Rail Corrugation Prediction Models

Rail corrugation not only reduces the ride comfort of passengers, but also increases damages of tracks and vehicles, and even affects the safe operation of trains. In order to verify the correctness of rail corrugation prediction models, firstly the occurrence probability of rail corrugation was calculated based on field investigation data into a metro line and railway main-lines. Secondly, aiming to the shortcomings of the validation method of traditional rail corrugation prediction models, and based on the regularity of rail corrugation occurrence, the benchmark conditions for the verification of the rail corrugation prediction models were proposed, which include the first condition: rail corrugation on the low and high rails at a tight curved tracks whose radii are less than 350 m;the second condition: rail corrugation on the low and high rails at mild curved tracks whose radii are larger than 650 m without Colong-egg fasteners; the third condition: rail corrugation on the two rails of tangential tracks without Colong-egg fasteners. A case study on the prediction of rail corrugation on an actual metro line was performed. Finally, based on whether the creep force is saturated or not, a fast method for predicting rail corrugation was proposed. The research result shows that existing validation conditions for rail corrugation are not universal, that the traditional rail corrugation models neglect the effect of track radii; and that the rail vibration evolution from a new rail without corrugation to the rail with corrugation was ignored in the model validation procedure, which lead to a low model prediction accuracy of rail corrugation. The prediction accuracy of the proposed fast method for predicting rail corrugation reaches 85.00%.