Meso-scale kinematics in shear bands and impact of material heterogeneity on shear band development in sand

This study aims to improve our understanding of strain localization in sands from the meso-scale perspective. First, this research contributes to characterizing and quantifying experimentally the micro- and meso-scale kinematic mechanisms associated with force chains and vortex structures in shear bands in real granular materials. The knowledge gained from this research is relevant for other dense granular flow problems (e.g. silo flow), where force chains control the transition from jammed to flowing states. Also, the results serve to reveal microstructural explanations for different global material responses between softening and critical state. ❧ The Digital Image Correlation (DIC) technique is used to obtain the two-dimensional, incremental surface displacements from digital images of deforming plane strain sand specimens. The aim herein is to use DIC results as a means to 1) evaluate meso-scale kinematic behavior inside shear bands through the course of large deformation, 2) infer the kinematic manifestation of force chains and vortex-structures and study their evolution and dissolution in sheared sands, and 3) conduct a preliminary investigation of length scales in granular shear. ❧ The obtained results reveal that softening/critical state transition is microstructurally defined by a coordinated, multi force chain collapse event which induces coherent vortices. The critical state microstructure, on the other hand, is characterized by the continual buildup of force chains at the conflux of opposing displacement vectors between adjacent vortices, then force chains collapse and formation of new vortices. The temporal changes in meso-scale kinematics are seen to additionally correlate with fluctuations in macroscopic shear stress. Auto correlation analysis and wavelet analysis are used to quantify spatial periodicity of vortices and the size of them. Our results suggest that the correlation length or the size of vortices is in the order of 6 to 7 grains and the spatial periodicity of the observed pattern is not only a function of median grain size, more research is required to explore the role of morphological and mechanical parameters in this regard. ❧ Second, we studied the physical mechanism underlying shear band formation in a supposedly uniform soil specimen, apart from the boundary influences or loading non-uniformities. Specifically we investigate how meso-scale sand density variations affect the onset of strain localization, providing insight as to why persistent shear bands form where they form in the soil specimen. The technique of X-Ray Computed Tomography (CT) is used to capture meso-scale density variations in plane strain specimens of sand. Digital image processing techniques are then used to transfer the CT results as input to the three-dimensional FE models. Finally, Digital Image Correlation (DIC) enables tracking of the in plane displacement of the sand specimen throughout plane strain compression tests performed on the CT specimens. The laboratory results are then compared with numerical predictions. This direct comparison should enable refinement in numerical models to achieve better predictability. ❧ The experimental technique that we used here could precisely capture the role of minor density variation on pre-peak response of specimen and shear band formation. The resulting FE predictions of shear band location further prove that spatial density variations have an influential role in the development of persistent shear band.