The efficiency of contact search is one of the key factors related to the computational efficiency of three-dimensional sphere discontinuous deformation analysis (3D SDDA). This paper proposes an efficient contact search algorithm, called box search algorithm (BSA), for 3D SDDA. The implementation steps and data structure for BSA are designed, with a case study being conducted to verify its efficiency. The data structure also has been improved for parallelizing the computation in contact search. For the demonstration of the proposed algorithm (BSA), six cases with various sphere numbers are simulated. Simulation results show that the time consumed in contact search using BSA (CTofBSA) is much less than that by the direct search algorithm (DSA) (CTofDSA). For the case with 12,000 spheres, CTofBSA is 1.1[Formula: see text]h, which is only 1.3% of CTofDSA (84.62[Formula: see text]h). In addition, the proportion of the computation quantity of contact search in the entire computation (Pcs) is 91.3% by using DSA, while this value by BSA is only 12.4%, which demonstrates the contribution of BSA. The efficiency brought about by BSA (time consumed and computation quantity) may enable 3D SDDA to simulate large-scale problems.
Read full abstract