With the discrete element method and the proposed virtual slicing technique for three‐dimensional discrete element model, random pore‐structural models of soil‐rock mixtures (SRMs) are constructed and voxelized. Then, the three‐dimensional lattice Boltzmann method (LBM) is introduced to simulate the seepage flow in SRMs on the pore scale and the influences of rock content, rock size, relative density on the simulated permeability of SRMs are comprehensively investigated. Finally, several theoretical models are introduced to predict the permeability of SRMs, and comparison is made between the results given by these prediction models and the numerical results obtained in this study. The results show that the permeability of SRMs presents a gradually increasing trend with the increase of rock content, and the rate of increase is also getting larger. When the other conditions remain unchanged, the permeability of SRMs decreases with the increase of relative density. When the rock content is lower than a threshold value, the permeability decreases as the rock size increases; however, this trend gets reversed when the rock content exceeds the threshold value. The permeability values of SRMs simulated with the LBM are lower in different levels than those calculated with the HS model, EMA model, and Zhou model, but most of them are very close to those predicted by the Daigle model.

中文翻译：

---

土石混合物渗透性的三维格子玻尔兹曼模拟及与其他预测模型的比较

利用离散元方法和拟议的虚拟切片技术对三维离散元模型进行建模，构造并体素化了土石混合体的随机孔结构模型。然后，引入三维格子玻尔兹曼法（LBM）模拟了SRMs中的渗流对孔隙尺度的影响，并全面研究了岩石含量，岩石大小，相对密度对SRMs模拟渗透率的影响。最后，介绍了几种理论模型来预测SRM的渗透率，并将这些预测模型给出的结果与本研究获得的数值结果进行了比较。结果表明，随着岩石含量的增加，SRMs的渗透率呈现出逐渐增加的趋势，并且增加的速率也越来越大。当其他条件保持不变时，SRMs的渗透率会随着相对密度的增加而降低。当岩石含量低于阈值时，渗透率会随着岩石尺寸的增加而降低；但是，当岩石含量超过阈值时，这种趋势就会逆转。用LBM模拟的SRM的渗透率值在不同的水平上要比用HS模型，EMA模型和Zhou模型所计算的值低，但它们中的大多数与Daigle模型所预测的值非常接近。