Molecular Dynamics Modeling of Hydraulic Conductivity of Soil Considering Variable Viscosity and Adsorbed Water

Siqi Zhang,Huafu Pei,Michael Plotze,Haochen Ying

DOI: https://doi.org/10.1016/j.clay.2022.106598

IF: 5.907

2022-01-01

Applied Clay Science

Abstract:Accurate prediction of the hydraulic conductivity of soils is crucial, e.g., to evaluate pollutant transport and design the disposal of radioactive waste. Previous investigations have focused on macroscopic seepage phenomena for soil to calculate the hydraulic conductivity with little attention to microscopic seepage behavior, leading to a difference between predictions and measurements. One of the reasons for this discrepancy is that the dynamic properties of water confined in pores affected by strong adsorption from constituent minerals are not necessarily the same as those in free water. In this paper, molecular dynamics (MD) is used to provide new insight into understanding the behavior of water transportation through the soil mesopores. The microscopic characteristics of water flow in quartz and Na-montmorillonite pores were investigated by MD simulations. The results indicate that the pressure-driven flow is laminar and follows Poiseuille's law. The dynamic viscosity of water confined in the pore was determined and found to show that the water viscosity decreases with increasing pore size and temperature. The thickness of the adsorbed water layer as a diffuse layer was quantitatively described by electrical double layer theory and MD simulations. For better application in engineering, the microscopic seepage behavior of constituent minerals was extended to the macroscopic characteristics of soils. Based on the MD results obtained, a modified Kozeny-Carman (KC) equation is developed by introducing the concept of variable viscosity and adsorbed water layer. The model also considers the effect of temperature on the hydraulic conductivity of soils. Furthermore, from a comparison of calculated and experimental data, the results confirm that the proposed equation shows higher performance in predicting the hydraulic conductivity of clays than the classical KC equation.