Study on determination of parameters in classical SFCC model

Abstract

The variation of unfrozen water content with temperature significantly affects the heat and mass transport in frozen soil. The phase change-induced variations in unfrozen water content with temperature variation in frozen soils can cause frost heave and thaw settlement, consequently compromising the stability of infrastructure. The typical soil frozen characteristic curve (SFCC) depicts the variation of unfrozen water content at different temperatures; however, the lack of a universal computational method for ascertaining model parameters, which primarily depend on empirical fitting, considerably undermines the model's computational efficiency. To solve the problem, premelting theory and probabilistic ice formation were used to calculate the parameters in SFCC model. Based on the SFCC model presented by the Van Genuchten model, the relationship among model parameters n and α and temperature, equivalent particle size, Hamaker constant and other physical quantities was derived. By verifying the experimental data of 10 groups of different soil samples, the proposed parameter calculation method can effectively predict the unfrozen water content, and the calculated values are in good agreement with the experimental values. The results show that the parameters in classical SFCC model are correlated with ice-water interface free energy, the density of liquid solution, density of ice, Hamaker constant equivalent particle size, latent heat and temperature. The parameter calculation method proposed in this study addresses the limitation of existing empirical models that they cannot derive parameter values directly without compromising the original model's predictive accuracy, thereby offering a new approach for utilizing the SFCC model to predict unfrozen water content and conduct multi-physics coupling numerical simulations.