Investigation of the Correlation Between Track Geometry Defect Occurrence and Wood Tie Condition

Abstract

This paper presents the results of a statistical correlation analysis between the occurrence of track geometry defects and tie condition as measured by an automated tie inspection system on a major US Class 1 railroad. This activity focuses on tie condition, as defined by the automated tie inspection system and calibrated to railroad standards, its frequency and distribution, and corresponding track geometry condition as defined by FRA and railroad standards. In this paper, multiple correlation analyses are presented to include basic correlation analyses, Wilcoxon rank-sum test, cluster dendrogram, cumulative tie segments analysis (greater than defined tie condition), and development of mixed distribution models. There was evidence of a correlation between tie condition and the occurrence of several, but not all, of the track geometry defect categories. While there was some variation between the different analyses results, in general, the following geometry defect categories were found to have a defined relationship to tie condition: Cant, CLIM (Cross-Level Index Modified, a parameter that examines the extent of multiple sequential cross-level defects.), Wide Gage, Tight Gage, Warp 62, Profile, and Crosslevel. Overall, the study showed that there is strong evidence of a correlation between tie condition and the occurrence of certain classes of track geometry defects.