Trending and correlation analysis of performance indicators for measuring and monitoring rail profile wear

Loading...
Thumbnail Image
Date
2019
Authors
Kashi Mansouri, Mohammadreza
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
This thesis develops a comprehensive understanding of rail profile performance indicators (PIs), their temporal trends and the relation between them. The increasing demand for freight and passenger rail transportation accentuates the need for regular and timely rail maintenance, particularly rail grinding. To enhance the efficiency and effectiveness of maintenance activities, meaningful PIs must be developed and monitored. Despite general recognition of the benefits of adopting performance-based rail monitoring and management programs, knowledge gaps remain in terms of: (1) the selection of relevant indicators of rail condition and performance; (2) deterioration rates and the thresholds that trigger maintenance interventions; and (3) the effectiveness of rail grinding in prolonging the life of rail assets. This research partially fills these knowledge gaps. This research develops a new algorithm in MATLAB© to: (1) automate the extraction, compilation and screening of historical rail profile data, (2) calculate multiple rail profile PIs over multiple years, (3) store the calculation results, and (4) analyse them using qualitative (i.e., temporal trending graphs) and statistical tools (i.e., Spearman correlation technique). The algorithm enables user flexibility in the definition of temporal periods to evaluate performance before and after maintenance interventions. Moreover, it improves analytical efficiency and enables customization of analysis steps and results. The trending and correlation analyses integrate industry-standard PIs (head loss, gauge wear, vertical wear, and grind quality index) with newly-developed PIs (average rail profile, lateral contact position, and contact radius). There appears to be a strong agreement ii between head loss and vertical wear; however, other performance indicators truly measure unique aspects of rail profile performance and should be considered alongside each other. The findings provide some evidence of the value of maintenance interventions—quantified in terms of the lower grind quality index over time. However, additional information on rail maintenance (time and level of effort) and operations (e.g., tonnage and number of passes) is required to develop more conclusive insights. Also, the trends for certain PIs reveal the pending need for replacement when the PIs approach relevant condemning limits. This information supports more proactive and effective rail maintenance intervention decisions.
Description
Keywords
Rail Profile Wear
Citation