A study on reducing derailments at turnout

Date
2016
Journal Title
Journal ISSN
Volume Title
Publisher
University of Delaware
Abstract
The provision of a reliable railway system is necessary since the railway system is a transportation mode that communities depend on for freight and passenger purposes. Cost-effective maintenance methods are one of the traits of a reliable railway system. One of the major causes of accidents in the US railroad system is derailment. Derailment is the incident where the train wheels leave the rails and wheel climb is one of the common wheel climb reasons. Turnouts notably are one of the track areas where a significant number of derailments occurs. ☐ This research focuses on the evaluation of newly developed inspection gauges to reduce derailments at turnouts. As part of the review of the evaluated gauges, IDEA S-23 project, which is a Transportation Research Board project and a master thesis, was reviewed. Four gauges that were developed in this project are considered in this thesis. The four gauges are as follows: 1. Chipped point gauge 2. AAR 1B wheel contact gauge 3. Severely worn wheel profile gauge 4. Gage-face wear angle gauge. ☐ In this study, a discussion of the causes of derailments in general and wheel climb specifically is presented. This study used a methodology that has been proposed by a previous study1, where data taken from an Automated Switch Inspection Vehicle (ASIV) were utilized to overlay switch point profiles with wheels to determine critical points at turnouts. This method is used in this thesis to overlay the evaluated gauges with the switch point profile and then compare the results in terms of Lateral force (L)/Vertical force (V). ☐ As part of the evaluation, four gauges were selected on the basis of the results of IDEA S-23 as grounds for the field validation process. The field validation was completed as part of the IDEA S-28 project and this thesis. The discussion of the field evaluation process and the analysis made on the data obtained from the field evaluation are presented in this thesis. This thesis introduces two types of analysis, namely, agreement analysis and decision tree analysis, which are based on the data mining technique. As a result, the evaluation of the proposed gauges in terms of L/V and based on the IDEA S-28 analysis is presented. ☐ Moreover, the recommendation and modification of the gauges are presented.
Description
Keywords
Citation