Railway Ballast Aggregates under Drained Triaxial Compression: Experiments and Calibration of NorSand Model with Critical State Uncertainties

The performance of railway ballast aggregates is investigated via element laboratory tests and numerical simulations considering critical state uncertainties. To this end, drained triaxial monotonic compression tests on natural gabbro aggregates and electric arc furnace (EAF) materials are conducted. Additionally, stress–strain data pertaining to various ballast materials from different geologic environments available in literature are collected and critically reviewed. Using this data, a detailed calibration procedure is presented for the NorSand model accounting for critical-state uncertainties. Two distinct techniques are proposed to calibrate the elastic shear ( I r ) and plastic hardening parameters ( H ) by presenting explicit analytical formulae for these parameters. The impact of uncertainties on the critical state line (CSL) is examined on the calibrated parameters. Additionally, stress–strain response conditioned on a specific model parameter is examined through the creation of a parameter covariance matrix. The findings indicate that a dependable determination of I r is crucial to reduce the model uncertainties when the conjugate shear strain ( ε q ) is smaller than approximately 0.04%. Besides, dilatancy parameter ( χ tc ) contributes to model uncertainties when 0.04% < ε q < 10%, while the critical state parameters have little effect until ε q exceeds 10% or so. The study suggests that uncertainty in the critical state line has a limited influence on modeling near-peak stress–strain behavior of dilatant ballast aggregates within the NorSand framework.