(6)

It is necessary to fully understand the settlement of high-speed railway subgrade induced by train loading to ensure the operation safety of high-speed trains. A 1:7 reduced-scale model test was designed to investigate the settlement of subgrade under two loading methods: continuous and intermittent cyclic loading. The testing results show that an increase in load amplitude enhances the load transmission effect to the bottom of the subgrade. After 105 cycles of continuous loading, the cumulative settlement of the subgrade at depth of 0, 20, and 40 cm directly below the loading range is 3.247, 1.05, and 0.09 mm, respectively, showing significant decreases with depth. A significant rebound can be observed when the applied load is removed during the intermittent loading process, which is quite different from the results under condition of continuous loading. Thus, the intermittent effect of train load on the cumulative deformation of the subgrade cannot be ignored. In addition, to better predict the cumulative settlement of the subgrade, a prediction method based on the state evolution model was proposed and used to quantitatively analyze the testing observations. Based on the state evolution model, the predicted cumulative strains at depths of 0, 20, and 40 cm were 1.218%, 0.457%, and 0.047%, respectively, which are in good agreement with the experimental results of 1.099%, 0.48%, and 0.045%, indicating that the theoretical model can accurately predict the cumulative strain of the subgrade caused by train load. Additionally, the parameters of the state evolution model can be updated in a timely manner by applying the updated monitoring data to enhance the prediction accuracy. The current work provides an alternative method for predicting the long-term cumulative settlement of subgrade induced by the train loading, and also a basis for the optimization of high-speed railway subgrade design.

来源平台：INTERNATIONAL JOURNAL OF GEOMECHANICS