The spatial correlation between track roughness and ground-penetrating radar inferred ballast degradation

Abstract

Ballast degradation is considered to be a primary factor that contributes to the development of track roughness, and as such it is important to develop efficient techniques to assess the condition of the ballast. Ground-penetrating radar is one method that has been applied in a variety of railway foundation studies including those attempting to non-destructively assess ballast degradation. However, there has yet to be a large-scale study that attempts to correlate the ground-penetrating radar-based estimates of ballast degradation with the observed track roughness. This study investigates this correlation along a 335 km-long heavy-haul railway subdivision in Alberta, Canada. Track roughness is quantified from repeated track alignment and surface measurements spanning 15 months prior to the ground-penetrating radar data acquisition. Three sets of 400 MHz ground-penetrating radar measurements were performed in August 2012, one along each ballast shoulder and one along the track centreline. The results of this study reveal that significant correlations between the observed track roughness and the ground-penetrating radar-based interpretation of ballast degradation are rare and only exist when the data are compared at very small spatial scales. The absence of significant correlations between track roughness and the estimates of ballast degradation is primarily interpreted as being the result of ambiguous ground-penetrating radar data caused by local-scale variations in the track foundation unrelated to ballast degradation. To address these issues, potential improvements in the application of ground-penetrating radar as a ballast degradation detection tool are proposed.