One of the main problems of carbonate sands is the fragile nature of particles and their susceptibility to breakage. Carbonate sands are affected by volumetric strain even at low stress levels, which is not the case with silicate sands. By defining a simple breakage model, the current study develops an elastoplastic critical state constitutive model that considers the impact of particle breakage on the mechanical behavior of carbonate sands. The particle breakage model depends on mean effective stress and critical breakage stress, which is assumed to correspond with the precompression pressure of soil in the oedometer test. In the proposed model, critical state line movement with the breakage parameter (alpha) considers the particle breakage effect. Based on the unified clay and sand model (CASM), a novel dynamic yield surface with a shape parameter affected by particle breaking has been created. Certain modifications are made to the modified Cam-Clay stress dilatancy to predict the behavior of carbonate sand. The current model has only ten parameters that simulate the carbonate sands' behavior even at high-stress levels without any breakage test. Experimental data with different soil densities, loading stress paths, and stress levels were compared with the model, and the results demonstrated satisfactory conformance.

Particle gradation effect on the shear-dilatancy of granular soils was studied through a series of drained triaxial tests using the discrete element and finite difference methods (PFC3D-FLAC3D). Spherical particles with different coefficients of uniformity Cu and median particle sizes D50 were assembled to exclude the strong size-shape correlation in natural sands. Four groups of Cu and four groups of D50 at the same void ratio ecprior to shearing, and seven more groups with the same void ratio e0 prior to isotropic compression were tested. Various mechanical behaviors were analyzed, including the stress-strain response, the stress-dilatancy response, friction angle, and fabric anisotropy. Cu significantly influences the peak friction angle, the maximum dilation angle, and the anisotropies of normal contact force and contact normal, whereas these are almost independent of D50. The contribution of the maximum rate of dilation to the excess friction angle is largely independent of Cu and D50 for spherical particles. (c) 2024 The Society of Powder Technology Japan. Published by Elsevier BV and The Society of Powder Technology Japan. All rights are reserved, including those for text and data mining, AI training, and