Characterizing Mechanical Behaviors of Railroad Ballast Based on Large-Scale Triaxial Tests

Ballast, typically composed of large unbound aggregates with uniform gradation, is essential for the structural integrity of railroad tracks, facilitating load transfer, drainage, and ensuring stability and resilience. The mechanical behaviors of ballast, including shear resistance, resilient modulus, and permanent deformation, are critical for effective railroad design and maintenance, directly impacting track geometry. Degradation of these mechanical properties can lead to significant track-related issues, such as excessive settlement and train derailments, resulting in substantial property damage, injuries, and fatalities. This dissertation aims to characterize mechanical behaviors of railroad ballast from perspectives of aggregate interaction, cyclic loading pulses and climatic conditions, based on large-scale triaxial tests integrating discrete element method (DEM) analysis and SmartRock sensing technology. The findings indicate that the stiffness of railroad ballast is influenced by compaction methods due to variations in initial particle arrangements. Notably, monotonic loading tests may underestimate the shear strength of ballast by as much as 72.7%, depending on the confining pressure, with increased shear strength attributed to enhanced anisotropy of soil fabric resulting from particle rearrangement during loading-unloading cycles. A modified failure criterion that incorporates fabric anisotropy is proposed and validated. To characterize the effects of loading waveforms with varying rest periods on ballast behavior, a new index termed the Cyclic Loading Duration Ratio (CLDR) is introduced. As CLDR increases, the resilient modulus initially decreases before rising, with a CLDR of 0.20 yielding the lowest resilient modulus. An empirical correlation model for resilient modulus that incorporates CLDR is developed and validated. Furthermore, resilient modulus is significantly influenced by particle settle acceleration under cyclic loading, and the Cyclic Loading Acceleration Index (CLAI) effectively unifies the effects of varying loading conditions and settle acceleration, facilitating the development of a practical resilient modulus model. As ballast becomes fouled, a higher Fouling Index (FI) correlates with increased permanent deformation, irrespective of water content. The disparity in deformation between high and low FI samples expands during incremental wetting. Although cumulative permanent deformation of fouled ballast continues to rise with increased water content, samples maintain stabilized permanent strain rates, a phenomenon attributed to cyclic densification effects. The progressive wetting process accelerates the loss of matric suction in fine/coarse soil mixtures while simultaneously enhancing the density and interlocking of ballast aggregates under cyclic loading, thus mitigating the adverse effects of wetting. To support effective railroad track maintenance in the context of varying rainfall events, this dissertation proposes a novel predictive model for the permanent deformation of fouled ballast that incorporates both water content and FI.