Deformation and Degradation of Clay fouled Ballast Subjected to Monotonic Loading

Abstract

Railways offer an efficient and economic transport mode in many countries including Australia, China and USA. Conventionally, rail tracks are positioned on ballast due to several potential benefits, including economy (availability and abundance), rapid drainage and high load bearing capacity. However, the ballast becomes contaminated due to intrusion of pumped subgrade material (e.g. clay and silt). This is one of the primary reasons for rapid track deterioration. In severe circumstances, fouled ballast needs to be cleaned or replaced to maintain the desired track resiliency, load bearing capacity and the track alignment, all of which influence the level of safety. In Australia, massive amount of funds have been invested in track maintenance. By employing an effective maintenance program, both cost and the ballast quarrying can be reduced with significant favourable environmental implications and improved productivity. In order to identify the risk associated with fouling, it is important to accurately assess the amount of fouling. In this paper, the current methods commonly used for evaluating the degree of ballast fouling were critically examined and a new parameter, Void Contaminant Index (VCI) was proposed to capture the role of different fouling materials in terms of volume based air-voids reduction. A series of isotropically consolidated drained triaxial tests using a large scale cylindrical triaxial apparatus were conducted on both clean and fouled ballast with varying VCI to establish the relationship between the extent of fouling and the associated strength-deformation properties. Based on the laboratory findings, an empirical relationship between the peak deviator stress and VCI is proposed to assist the practitioners for preliminary track assessment and in the mitigation of the risk associated with ballast fouling. A non-linear shear strength envelope for clay fouled ballast is presented in a non-dimensional form, based on the proposed empirical equations.