The discrete element method (DEM) is employed to investigate the impact of coupling between volumetric and axial strains on the flow liquefaction vulnerability of 3D cubic particulate specimens. The virtual testing program conducted here encompasses a wide range of initial states and varying degrees of coupling between volumetric and axial strains. Utilizing data obtained from DEM simulations, the evolution of micro- and macroscale variables, including coordination number, contact fabric anisotropy, redundancy index, strong force networks, invariants of the effective stress tensor, and excess pore-water pressure, is examined. Results from DEM tests indicate that coupling expansive volumetric strain with axial strain leads to a gradual loosening of the load bearing microstructure, a decrease in coordination number, and a faster change in contact anisotropy. DEM simulations demonstrate that the triggering of flow liquefaction instability is followed by a sudden increase in contact fabric anisotropy and abrupt drops in coordination number and redundancy index. Moreover, a detailed analysis of the findings suggests that the stress ratio at the onset of post-peak softening decreases with increasing expansive volumetric strains.

In recent years, data-driven approaches have gained considerable momentum in the scientific and engineering communities, owing to their capacity to extract complex patterns from high-dimensional data. Despite their potential, these approaches often require extensive high-quality datasets, may exhibit limited extrapolation capability beyond the training domain, and lack a rigorous foundation grounded in physical and thermodynamic principles. To overcome these limitations, physics-informed neural networks have been introduced, embedding governing equations directly into the learning process. Building upon this paradigm, this study presents a novel thermodynamically consistent constitutive model for granular soils, developed within the framework of geotechnically- and physics-informed neural networks (GINN). The model integrates physical laws with data-driven learning via a composite loss function. These include: (i) strictly non-negative material dissipation rate to ensure thermodynamic admissibility, (ii) an admissible range for the predicted stress state, and (iii) bounds on the predicted void ratio. The material dissipation rate is calculated using the total work input and a free energy potential expressed in terms of stress invariants. The model is validated against monotonic drained triaxial test data for specimens prepared with diverse initial void ratios and stress states. The model accurately simulates both the shear strength and dilative response of granular soil samples. Its predictive performance is further benchmarked against two widely adopted constitutive models from the literature, demonstrating comparable accuracy while maintaining consistency with thermodynamic laws.