An advanced pore-scale model for simulating water retention characteristics in granular soils

Abstract

Accurately estimating the soil water characteristics curve (SWCC) has become indispensable in successfully implementing unsaturated soil mechanics into engineering practice. The objective of this paper is to present a pore-scale model for the prediction of the hysteretic SWCC of granular soils using basic properties like grain size distribution (GSD) and porosity. The Discrete Element Method (DEM) is used to model the soil structure by generating a stable assembly of spherical particles to attain a predefined grain size distribution and porosity. A network of pores and throats is extracted from this assembly using a medial axis-based network extraction technique, and adequate pore-scale mechanisms are availed to simulate two-phase flow along drying and wetting paths. The current work is further expanded to model SWCC under different void ratios. In addition, SWCCs exhibit stress-dependent behavior, and a numerical framework for modeling SWCCs under various stress conditions is proposed. The effect of particle shapes used for generating DEM samples on SWCC is also investigated. The modeled SWCC compares well with measured curves obtained for a range of granular soils from the literature. The model effectively estimates the hysteretic drying-wetting response and can capture the variation in SWCC with change in net stress and initial void ratio.