The flow of immiscible fluids through a deformable porous granular medium occurs naturally in many engineering and scientific applications, including soil mechanics and oil and gas extraction. It entailed comprehending the distribution and interactions of all constituent phases. Significant advances in micromechanical modelling, including the discrete element approach, had been made since the earlier work in developing macro-continuum models for solid-fluid interactions in porous granular material. The macro-continuum models had grown to include hydro-mechanical coupling, desiccation cracking, and other observed phenomena. However, these approaches had inherent limitations as they relied on phenomenological assumptions, such as water retention behaviour and often lack the ident...[ Read more ]

The flow of immiscible fluids through a deformable porous granular medium occurs naturally in many engineering and scientific applications, including soil mechanics and oil and gas extraction. It entailed comprehending the distribution and interactions of all constituent phases. Significant advances in micromechanical modelling, including the discrete element approach, had been made since the earlier work in developing macro-continuum models for solid-fluid interactions in porous granular material. The macro-continuum models had grown to include hydro-mechanical coupling, desiccation cracking, and other observed phenomena. However, these approaches had inherent limitations as they relied on phenomenological assumptions, such as water retention behaviour and often lack the identification of a generally accepted effective stress. Furthermore, the majority of micromechanical research was restricted to the so-called pendular regime, in which the water phase existed as highly localised, disconnected capillary bridges.

In this thesis, these constraints of both macro-continuum and micromechanical models were tackled by extending a pore-scale approach. This pore-scale coupled hydro-mechanical model’s purpose was to emulate the quasi-static drainage and imbibition processes in a deformable porous granular media. The pore-scale approach was incorporated with the discrete element method (DEM) for fluids and grains, respectively. In these methods, the pore spaces were discretized as a collection of pore bodies and throats, and local flow rules were provided such that the fluid displacement was resolved locally at the pore-scale level. A local pore-scale criterion was set up to model the displacement of a fluid-fluid interface. This was done to model the effect of pore-scale heterogeneities on capillarity, and more specifically, to quantify the amount of force exerted on a solid grain that was partially surrounded by a wetting and non-wetting phase. Water retention curves were compared with the experimental results to assess the model’s validity. The model also analysed Bishop’s effective stress parameter and its link to micro-scale contact stress.

The controlled suction and constant water content triaxial compression tests for the partially wet granular material exhibited a unique failure envelope, which further corroborated Bishop’s effective stress theory. A pore based fabric tensor based on the Minkowski moment tensor were explored to understand the pore anisotropy for the shearing process. The anisotropy of the pore orientational tensor closely resembled the evolution of the contact normal anisotropy, though a correlation was not established following a lag in the evolution of two anisotropies. However, an interesting correlation was identified between the average shape anisotropy, β_avg and porosity and supported its usage as an indicator for macroscopic density. The pore based fabric for the partially wet granular materials revealed that highly localized saturated pore-units aligned in the direction of the loading, whereas the pores with relatively lower saturation equal to or below a threshold saturation were isotropic.