This project seeks to advance a pore-scale micromechanical modeling framework for quantifying capillary-induced intergranular stress in unsaturated soils. Focus is placed on refining a statistical upscaling methodology used to differentiate portions of the pore network partially filled with water in the form of interparticle liquid bridges and portions of the pore network completely filled with water in the form of saturated pockets. A corresponding capillary stress function may be computed from relatively simple measurements of the soil-water characteristic curve, quantified over the complete range of pore water saturation, and cast into conventional effective stress formulations for saturated soil behavior. Specific objectives are to refine the modeling framework for capturing generalized pore geometries, account for surface adsorption effects in fine-grained materials, and account for wetting-drying hysteresis effects. Laboratory measurements of water retention behavior, particle and pore morphology, shear strength, tensile strength, and small-strain dynamic behavior will be used to evaluate the model performance. Education and outreach activities designed to stimulate graduate and undergraduate research, diversity, and more effective implementation of unsaturated soil mechanics into practice are integrated with the research plan. Improved basic understanding of the pore-scale forces and processes that govern unsaturated soil behavior is expected to yield insight into problems in geotechnical engineering practice that may occur under unsaturated soil conditions, including precipitation-induced landsliding, foundation and excavation stability, and the dynamic response of compacted earthworks.
A micromechanical framework has been developed for quantifying the fluid fabric of unsaturated coarse-grained soils directly from more conventional macroscale measurements of water retention behavior. A key attribute of the approach is capability to decompose the total volume of water retained by soil into independent fractions adsorbed as thin films on particle surfaces, as liquid bridges retained between individual particles, and as saturated pockets retained in small pores. Because water in each of these forms plays a very different role in affecting bulk soil behavior, the approach for independently quantifying each contribution, and how each contribution evolves with changing saturation or wetting direction, has important implications toward understanding more general unsaturated soil behavior. The framework is applicable over the complete range of saturation, and thus is a significant advancement beyond existing modeling frameworks and particle-based simulations for unsaturated soils, which have historically been limited to very low, or very high, saturations. The framework has been implemented to better understand soil thermal properties, effective stress in unsaturated soils, and unsaturated soil hydraulic properties. Specifically: (1) we are able to effectively capture wetting-drying hysteresis behavior in water retention curves. Specifically, if the drying branch of the SWCC is measured, the new approach allows us to effectively predict the subsequent wetting branch. This is significant because most techniques currently available for measuring SWCCs are applicable along a drying path, yet many geotechnical problems of interest involve soil behavior along a wetting path (e.g., precipitation-induced landsliding). We are also able to capture surface tension lowering and air-water interface area. (2) We have developed a new approach for predicting the SWCC from measured grain size distribution. (3) We have developed a formulation for quantifying capillary-induced stress in a conventional effective stress framework. (4) We have used understanding gained from the model to interpret fall cone tests for unsaturated sands. (5) We have developed a model to estimate saturated and unsaturated hydraulic conductivity from grain size distribution. (6) We have developed a formulation for predicting thermal conductivity dryout curves from pore-scale theory.
