(3)

The geometric structure and topological configuration of particles exert a significant influence on the pore structure characteristics of soil and the permeability of particle-pore media. Utilising the spatial random convexity growth algorithm, soil particle models with varying anisotropic shapes were generated. By employing PFC(Particle Flow Code) software, the accumulation behaviour of soil particles under natural gravity was simulated.Subsequently, AVIZO software, an advanced software for three-dimensional image analysis and scientific visualization, was used to extract the characteristic parameters of the pore structure, and numerical simulations of seepage in porous media were conducted. The permeability test was carried out to verify the numerical model.The grey correlation analysis was carried out based on the normalization method. The results showed that the correlation between particle size and permeability coefficient was 0.82, which was dominant. The particle size mainly affects the channel cross-sectional area of the pore structure. When the particle size increases, the absolute permeability increases from 0.4 to 1.2 mu m. The correlation coefficients between sphericity and fractal dimension and tortuosity are 0.848 and 0.758, indicating that particle shape will seriously affect the complexity and connectivity of pore structure. When the particle size distribution exhibits a right-skew, the peak equivalent diameter is predominantly centered within the range of 0.6-0.7 mu m. Additionally, the occurrence of larger pores is more prevalent compared to normal and uniform distributions; however, the impact of this phenomenon on permeability remains relatively constrained.

来源平台：GEOTECHNICAL AND GEOLOGICAL ENGINEERING