Relationship between track geometry defects and measured track substructure condition using emerging data analysis methods
Date
2018
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
University of Delaware
Abstract
Railroad infrastructure Reliability, Availability, Maintainability, and Safety (RAMS) is crucial in rail traffic operation. One of the approaches for improving the RAMS of railroad track assets is by enhancing either: inspection methods, or analysis methods followed by proper on-time action. Late, or inappropriate action will reduce RAMS by a reoccurrence of old defects and amplification (escalation) of new occurred defects. Besides, it will make maintenance works expensive, inefficient with unavailable track and less safe for the users and those around the track. ☐ This dissertation was based on research sponsored by the Federal Railroad Administration (FRA), which seeks to improve the interrelated analysis of track superstructure and substructure. Thus, it aims to find and develop relationships between inspection parameters of track geometry as measured by Track Recording Vehicle (TRV) and track substructure as measured by the Ground Penetration Radar (GPR). Those relationships have been widely discussed in the literature, but neither defined nor confirmed. The satisfactory results will allow the Infrastructure Managers (IM) to extract knowledge and learn from their massive multivariable inspection datasets and to act in preventative maintenance. Preventative maintenance is essential especially in track locations where there is a potential for development, growth, and propagation of track geometry defects due to the subsurface condition. The research was performed using emerging multivariate comprehensive analytical methods combined with track engineering insight. The data include Class 1 US Railroad companies’ inspections. The variables that were analyzed in the research include; Ballast Fouling and Ballast Layer Thickness, from GPR inspection and Profile (measured over a 62-foot chord) from Geometry Inspection. The track engineers chose those parameters to evaluate the condition of the track most effectively. ☐ The desired outcome is to develop a research methodology including a provisional empiric model that predicts the probability of co-occurrence of deviations of track geometry associated with given substructure conditions. Ultimately the objective is to allow the maintenance manager to take advantage of the existing massive amount of data and to make better decisions for the carrying out of preventative maintenance. After analysis, the results showed that there are significant relationships between geometry deviations and poor subsurface condition that were validated statistically. The subsurface conditions were defined by the GPR inspection parameters, such as Ballast Fouling Index (BFI) and Ballast Layer Thickness (BLT). Furthermore, the research results include several predictive models of the probability of a profile deviation as a function of the mentioned GPR parameters.