Correlation of track irregularities and vehicle responses based on measured data

Track geometry quality and dynamic vehicle response are closely related, but do not always correspond with each other in terms of maximum values and standard deviations. This can often be seen to give poor results in analyses with correlation coefficients or regression analysis. Measured data from both the EU project DynoTRAIN and the Swedish Green Train (Grona Taget) research programme is used in this paper to evaluate track-vehicle response for three vehicles. A single degree of freedom model is used as an inspiration to divide track-vehicle interaction into three parts, which are analysed in terms of correlation. One part, the vertical axle box acceleration divided by vehicle speed squareds shown to be the weak link with lower correlation coefficients than the other parts. Future efforts should therefore be directed towards investigating the relation between axle box accelerations and track irregularity second derivatives.