Discrete modeling of coupled flow-deformation response of granular soil media

Abstract

The behavior of soils under loading exhibits different characteristics due to their inherent structures. The nature of soil behavior varies at both macro and micro scales. To investigate the microscale behavior of soils that influences their macroscopic response, numerical methods capable of particle-based calculations are required. The Discrete Element Method (DEM) is an important tool for studying such soil behavior at the particle level. Additionally, since soils contain pore water, numerical methods that incorporate fluid dynamics principles need to be coupled with DEM to analyze this behavior at the microscale. By combining laboratory tests to determine the mechanical properties of soils and the numerical methods, the behavior of soils under dynamic excitations can be examined. Numerical modeling of soils using computational methods provides a robust tool for predicting future soil behavior and potential damage resulting from dynamic effects. This study discusses the use of the DEM and fluid dynamics numerical methods to investigate behavior at the particle scale. The working mechanisms and underlying principles of these methods are explained. In order to examine the granular media containing solid and fluid phases, examples are presented in this thesis considering the cases of the solid phase only, the fluid phase only, and the interaction between solid and fluid phases. Mechanical soil properties obtained previously are utilized to model laboratory pressure tests of soil samples and analyze their stress-strain behavior using the PFC2D software. Dynamic response of soil grains at microscale investigated using the GiD pre and post processor and Kratos MultiPhysics software. A basic flow model is investigated using the MechSys software for the fluid phase only. In the examined model, the time-dependent velocity vectors and fluid densities of the flow are analyzed using the Lattice Boltzmann Method (LBM), which is the numerical method employed within the related software. Subsequently, the simulation of a pore water flow containing solid particles is modeled using the coupled DEM and LBM. Overall, it is demonstrated that coupling of the DEM and the LBM is a powerful tool for modeling solid-fluid interactions at the microscale for soils.