Ballast Degradation Modeling for Turnouts based on Track Recording Car Data

Abstract

Turnouts play a central role in the railway infrastructure since they enable increased network capacity and allow for minimal impact of train delays. Their performance is of paramount importance for infrastructure managers, who face large maintenance cost in order to secure proper turnouts operability. Railway turnouts are complex mechanical systems, whose dynamic performance depends on the health state of the different components of the superstructure and substructure. A key component is the ballast as it provides the elastic support to the track and the sleepers and it largely contributes to the safety and reliability of the infrastructure. Ballast degradation can be a root cause of excessive failures in other components. A track recording car is typically used to collect geometry data that is used to assess the quality of the railway tracks; however this type of data has not been widely used for ballast quality evaluation in turnouts. One reason is that maintenance decision for turnouts are dominantly made based on visual inspections and/or manual measurement of track geometry, as turnouts are significantly more complex that traditional railway track.  This study presents the application of fractal dimensioning of track longitudinal level for the monitoring of ballast degradation in railway turnouts. In other words, the irregularities of the track vertical profile related to the ballast degradation are quantified as a ballast quality index. The ballast quality index is the basis for developing ballast degradation models in different sections of the turnout based on a segmentation scheme. Using track geometry data of 88 turnouts in the Danish railway network for the period 2012-2017, this study develops and compares ballast degradation models based on regression analysis and stochastic processes (lognormal and Gamma processes). The models are estimated for different sections of the single turnout, for different turnouts at distinct geographical locations. The proposed method provides an efficient tool for the analysis of the effect of tamping on ballast degradation rate. Moreover, the effects on ballast degradation of track loading rate, train speeds and seasonal changes of weather conditions are quantified.