nearest neighbor search (nns) algorithm

Evaluation of Solid-Fluid interaction is of high significance in all engineering fields. In this paper, the Smoothed Particle Hydrodynamics (SPH) and the Discrete Element Method (DEM) were employed to simulate solids and fluids, respectively. As for the simulation of water behavior, first, a single-phase “dam break” experiment is modeled by the SPH. In this model, variable smoothing length and an almost novel boundary method were utilized which resulted in a tremendous boost in its resemblance to the experiment. To couple these methods (SPH and DEM), three approaches were proposed and validated against a famous experimental test named “dam break with an elastic gate test”. Finally, to be sure of the correctness of these approaches, an elastic plate under a water column in the hydrostatic condition was simulated and the error was 2 percent in comparison with the analytical solution. In pursuit of having short run-times, parallel computing on GPU (CUDA) was employed, and a robust nearest neighbor search (NNS) algorithm was modified and developed.