Three-dimensional numerical reconstruction method for irregular structures of granular geomaterials

Abstract

Granular geomaterials are comprised of many geometrically irregular and randomly distributed particles on multi-scales that strongly affect their physical and or mechanical properties. However, providing an accurate representation or characterization of the complex structures of granular geomaterials remains a challenge. In this study, we propose a novel computerized method to numerically represent a three-dimensional (3D) irregular granular structure based on a natural conglomerate sample. To represent the 3D granular structure, the reconstruction algorithm incorporates a two-point probability function along both the principal and diagonal directions, as well as a linear-path function along the principal directions of the model, and a multithread parallel computing algorithm is used to accelerate the 3D reconstruction process of the irregular structures. To validate the accuracy of the proposed algorithm, the characteristics and mechanical properties of the reconstructed 3D granular structure are analyzed and compared with those of the prototype conglomerate sample. This comparison shows that the proposed algorithm improves the reconstruction efficiency of previous algorithms, and the reconstruction model exhibits greater similarity to the natural conglomerate structure than previous reconstruction models. This proposed computational method can also be used to reconstruct other granular structures.