Random Field Modeling of Railway Track Irregularities

Abstract

Track irregularity is considerably important in track-vehicle studies. Recorded data of these irregularities form a basis to develop track standards. This also helps specify track inputs to the vehicles. In this paper random field models for vertical irregularity data obtained from Indian railways has been presented. It is shown that irregularity in the vertical profile can be modeled as stationary Gaussian random field, which can be defined in terms of the power spectral density (PSD) function. Such a model can predict the values of peak amplitudes in a given track length. Further, the absolute vertical profile of the left and right rails belong to the isotropic random field. The implication of this in describing the input matrix of multiaxle vehicles is discussed.