Material Point Methods Applied to Grain Compaction*

The material point method is applied to investigate compaction and crushing of rock sugar spheres undergoing large deformation and failure. The method is chosen because of its capability of simulating complex geometry and of avoiding mesh tangling and distortion often encountered in a Lagrangian mesh-based methods. The application, however, introduces two new challenges about pre-processing and post-processing. The first challenge is the generation of a suitable initial material point configuration from a complex experimental geometry to describe initial geometry of spheres. A second challenge is how to identify surfaces from material point data, which is important in identifying cracks of the grains. In this work, a fast method to generate initial material point locations given a stereolithography file (STL) that is scalable to generate billions of material points using a single computer node is introduced to address the first challenge. To address the second challenge, a method to generate isosurfaces from the material point data is developed by extending the Marching Cubes 33 algorithm (Chernyaev, CERN Technical report CN/95-17, 1995). These tools allow for better comparison between experiments and simulations. The techniques are applied to investigate the large deformation and breakup of grains under compaction. Using these tools, numerical results are compared with X-ray tomographic 3D imaging experiments during early compaction.