Microstructure characterization of granular materials

This paper presents techniques and algorithms to compute microstructure properties of irregular-shaped granulate assemblies utilizing 3D images. The techniques are capable of extracting microstructure properties of particles such as centeroid, particle size distribution, shape indices (i.e., sphericiy and angularity), number of contacts (i.e., distribution of coordination numbers), contact network, packing efficiency, distribution of local void ratio and radial distribution function. Such properties are critical parameters for micromechanical-based numerical models to capture micro- and macromechanical behavior of geomaterials. X-ray microtomography was used to reconstruct high-resolution 3D image of a natural sand system to represent granular materials. Microstructure properties of the sand system were computed and compared with properties of a computer-simulated image of periodic random spheres. Findings indicate that the use of simplified systems of idealized spheres to model micro- and macromechanical behavior of granular systems can lead to inaccurate results due to the differences in the microstructure between both systems. Methods presented in this paper enabled capturing a more realistic microstructure that can be incorporated in micromechanical models to better simulate, understand, or explain macroscale behavior of granular materials based on their actual microstructure.