Clay-rich media are critical to the stability and long-term performance of natural and engineered systems, yet their compression-induced microstructural evolution remains poorly quantified at the molecular scale. Existing molecular dynamics studies are largely restricted to simplified platelet geometries, equilibrium or fixed-stress boundary conditions, and limited stress ranges, which prevent them from capturing stress-driven consolidation pathways, pore collapse, platelet reorientation, and the development of fabric anisotropy from dispersed colloidal states to densely compacted clay soils. In this study, we present a large-scale molecular dynamics simulation framework, inspired by the classical oedometer test used for geotechnical engineering to investigate the compression behavior and structural transitions of a nanoscale clay platelet assembly. The model incorporates 100 randomly distributed platelets (∼430,000 atoms) and captures a broad range of physical states − from highly porous colloidal dispersions (void ratio up to 18.25) to densely compacted clay soils (void ratio down to 0.31) − under vertical stress conditions ranging from 10 -7 to 10 2 GPa. Moisture content (0–40% by weight) and cation exchange capacity (0–100 meq/100 g) are systematically varied to simulate realistic environmental conditions. The simulations reveal strong coupling between compression behavior and microstructural evolution, characterized by changes in pore sizes and their distributions, platelets orientation angles, and scalar order (S-order) parameters. Structural transitions occur in three distinct stages (i.e., colloid, mud and soil stages), driven by progressive platelet reorientation and densification. Moisture enhances rotational freedom, while CEC promotes tighter packing and restricts reorientation. We further derive two closed-form equations, inspired by the soil–water characteristic curve (SWCC), to capture the stress-dependent evolution of void ratio and S-order. This study offers new insights into the mechanics of clay-rich media under stress and lay the groundwork for improved design and analysis in geotechnical, environmental, and infrastructure applications.
Du et al. (Tue,) studied this question.