Evaluation of Soil Water Retention Curve Models for Fouled Ballast

The objectives of this study were to identify the optimal soil water retention curve (SWRC) equation for unsaturated fouling ballast and to evaluate unsaturated fouled ballast peak shear stress. Ballast fouling degrades the track and results in poor performance. However, there is limited research on the sensitivity of the mechanical behavior of ballast fouling to variations in water content. In this research, clay and coal fouling materials were mixed with the clean ballast to varying degrees of fouling. For the hydraulic behavior, a custom large-scale system was developed to measure the specimen volumetric water content at various suctions during drainage. A nonlinear least squares method was used to determine best-fit parameters for four common empirical equations for water content versus suction of five fouled ballast specimens. All soil water retention models were subjected to mathematical correction techniques to incorporate high soil suctions up to 10,00,000 kPa. For the mechanical behavior, large-scale direct shear tests were performed on fouled ballast at the volumetric water content beyond air-entry value and over much of the unsaturated region. The study highlights how SWRC characteristics and different fouling materials impact peak shear stress. As the degree of fouling increased, the percentage of fines increased which extends the fouled ballast capillarity. However, the highest peak shear stress was observed at the dry condition for all experiments, likely because the wet fouling material lubricates the ballast aggregates. Unsaturated coal fouling resulted in greater peak stress compared to clay fouling. The parametric model used to determine the residual water content reduces uncertainty in describing the results. Additionally, the selected model will be used to develop a constitutive model to describe the unsaturated strength of fouled ballast. Such a model will improve the fundamental understanding of ballast degradation characteristics so that the likelihood of fouling can be predicted and managed versus identified and repaired.