Prediction model for soil-water characteristic curve of loess under porosity change

The loess has obvious macropore structure. The internal pores change continuously in the process of water immersion and collapsibility, resulting in the dynamic change of soil-water characteristic curve in the process of water immersion. In order to simply predict the soil-water characteristic curve of loess under different pore conditions, taking the pore volume function as the breakthrough point, assuming the influence law of the change of soaking pore on the pore volume function, and introducing the translation factor ξ_1i and the compression factor ξ_2i, the relationship among the translation factor, the compression factor and the pore index e is established through the average pore radius, and the different pores are derived based on the initial pore state. Based on the Gardner model and introducing the pore index, the model for the soil-water characteristic curve of loess considering the change of soaking pore is established. When the influences of pore change are not considered, the model degenerates to the Gardner model. The model contains six parameters, which can be determined by experiments. The parameters of the model are calibrated by the initial state, and the soil-water characteristic curve of loess with pore change is predicted. The combination shows that the predicted results by the model are in good agreement with the test ones. The evolution law of the soil-water characteristics of loess under different conditions of water soaked pore changes can be accurately predicted by using this model, which may provide preference for establishing the constitutive model for loess.