Microstructure and micromechanics each play a significant role in and are considered to govern the macroscale behavior of granular soils. The digital image analysis method is a popular approach to… Click to show full abstract
Microstructure and micromechanics each play a significant role in and are considered to govern the macroscale behavior of granular soils. The digital image analysis method is a popular approach to studying the microstructure of granular soils, but it is usually complicated and difficult to perform in the laboratory. In this article, a series of numerical models are developed to simulate the plane strain (PS) and conventional triaxial compression tests of granular soil. Based on the numerical simulations, a geometric algorithm to generate numerical ‘slicing’ images of three-dimensional (3D) particle assembly is proposed, which is directly analogous to the images extracted from the solidified-then-sectioned method in the laboratory. In this method, the same stereological calculations of the local void ratio distribution (LVRD) on 3D numerical specimens as on physical laboratory specimens are performed as are carried out on physical laboratory specimens. We developed a statistical model to analyze the LVRD of the PS specimen with different initial void ratios and confining stresses. Particle orientation distributions projected on a specific plane were investigated to simulate the common method in the laboratory. Our results show that macroscale (such as volume change) and microscale behaviors (such as homogeneity) as well as the fail modes (such as formation of shear band) could be revealed via analyses of LVRD and particle orientation distribution. The proposed numerical geometric algorithm method is proved to be a valid and more efficient approach for stereological analysis of microstructure of granular soils.
               
Click one of the above tabs to view related content.