Non-Euclidian discrete geometric modeling of granular soils

This paper suggests a new geometric modeling approach for discrete representation of a realistic pore-scale morphology using an amorphous non-Euclidian geometry of soil grains’ shape. The soil skeleton composition process follows a statistically-based consistent algorithm, which applies random non-repetitive soil skeleton morphology while keeping the same macro-scale geotechnical properties. The model parameters are determined such that both micro-scale indexes (as angularity, roundness and roughness) and macro-scale
outcomes (as phase and grading parameters) are correlated. The stability and the convergence of the solution is demonstrated in this paper upon 2D model. In addition, the model allows a direct representation of the amorphous pore space and can be used for further issues related to a non-homogeneous distribution of various materials (liquids or solids) in the porous medium.