A method of predicting variable speed rail corrugation growth using standard statistical moments

Wear-type rail corrugation is a significant problem in the railway transport industry. Some recent work has suggested that speed control can be used as an effective tool to minimize the rate of corrugation growth. This has brought about the need to model corrugation growth under variable passing speed. Variable speed rail corrugation growth modelling normally consists of either numerical simulation of a sequence of varied speed wheel passes or direct integration of a probabilistic passing speed distribution function; both of which are computationally expensive. This paper investigates the use of the statistical moments of the speed probability density function to greatly improve the computational speed of variable speed corrugation growth models and compares results of changing standard deviation and skewness to numerical integration models. It also identifies the effects of individual statistical moments on corrugation growth to provide better insight into control methods. The new modelling method correlated well with the numerical integration models for small standard deviations in speed (less than 10%) and highlighted a need to consider kurtosis in predicting the performance of speed control based corrugation mitigation schemes. For larger standard deviations in speed, higher than 4th order effects need to be considered.