Probabilistic assessment of high-speed vehicle-induced cumulative settlement of railway soft soil subgrade based on direct probability integral method

Abstract

Due to the variability of train loads, soil properties, and other factors, the cumulative settlement of high-speed railway soft soil subgrade induced by cyclic trains holds strong randomness and evolutionary characteristics. Thus, the prediction of cumulative settlement and assessment of the subgrade status is important to ensure the safety of high-speed trains. For this purpose, a vehicle-track-subgrade interaction model is established for calculating the dynamic stresses of the subgrade, and then the cumulative settlement and its evolution characteristics are predicted. Furthermore, the direct probability integral method (DPIM) is introduced to implement the uncertainty quantification and reliability assessment of cumulative settlement, especially spatial distribution, 3D-probability density function and the variation of reliability index with load cycles. A comparison between Monte Carlo method and DPIM is performed to demonstrate the accuracy and efficiency of stochastic analysis with respect to cumulative settlement. The sensitivity and reliability assessment of the cumulative settlement under the cases of different speed, train load and soil strength are investigated by numerical examples to illustrate the robustness of the proposed method. The results show that cumulative settlement is most sensitive to soil strength, followed by train axle loads. Besides, a reliability assessment method proposed in this work, which considers cumulative settlement, provides guidance for the design and maintenance of subgrade in this paper.