Meso-scale investigation on the permeability of frozen soils with the lattice Boltzmann method

Complex composition and intricate pore-scale structure of frozen soils poses significant challenges in reliably and efficiently obtaining their permeability. In this study, we propose a modified quartet structure generation set (QSGS) numerical tool for generating frozen soils and present the development of a computational simulation code based on the multiple-relaxation-time lattice Boltzmann method (LBM). In the modified QSGS, the arc-shaped water-ice interface is depicted, and the influence of pore-scale geometry on freezing temperature is considered. The validity of combining the proposed QSGS model and the LBM code is proved by comparing calculated results to analytical and experimental results of porous media. Our objective was to investigate the effects of soil features, including porosity, grain diameter, shape anisotropy of soil particles, and ice content on the intrinsic permeability of frozen soil. Additionally, we examined the relationship between these features and the specific surface area and tortuosity. Numerical results show that the intrinsic permeability of frozen soils increases with increasing porosity, larger granular diameter, and anisotropy, which is identical with the pressure gradient. The presence of ice led to clogging flow pathways and drastically decreased the intrinsic permeability, which is significantly less than unfrozen soil with same effective porosity. This study provides a useful tool to investigate the intricate interplay between the pore-scale structure and the intrinsic permeability of frozen soils.