Impact of heterogeneity in stabilised clay on the settlement of ballasted track

Abstract

Accurately predicting accumulated track settlement and the resulting deterioration of vertical track geometry is a major challenge in railway infrastructure maintenance. This issue is particularly critical for railways constructed on soft clay soils, where repeated train loading can lead to excessive and differential settlements, increased maintenance demand, reduced ride comfort, and decreased operational reliability. An additional complexity arises from the spatial variability of soil conditions and ground improvement measures, which can strongly influence track response and long-term performance. This thesis investigates the dynamic vehicle–track–soil in teraction and long-term settlement behaviour of ballasted railway tracks founded on soft clay, with particular focus on the influence of spatial variability in lime-cement column (LCC) improved ground. A numerical modelling framework is employed by coupling dynamic finite-element simulations with empirical settlement models for ballast and subgrade layers. Spatial variability in LCC properties is incorpor ated through random field modelling. Different LCC reinforcement configurations are evaluated to investigate their influence on track stiffness, stress distribution, and accumulated settlements. In addition, sensitivity analyses are performed to identify the most influential parameters governing settlement development and track response. The results indicate that spatial variability significantly affects the distri bution of track stiffness and the development of differential settlements, while LCC reinforcement effectively reduces long-term settlements and improves overall track performance. The developed framework also enables systematic evaluation of mitig ation strategies for railway infrastructure subjected to spatially varying soft ground conditions over time. The findings contribute to a better understanding of railway behaviour on soft soils and provide a basis for more reliable prediction of long-term track performance and maintenance planning.