Voronoi图在移动粒子半隐式法中的应用

To optimize the algorithms for establishing and analyzing spatial topology in the moving particle semi-implicit (MPS) method, the Voronoi diagram to three key components is applied: neighbor particle search, surface particle detection, and particle shifting. The specific approaches are as follows: a particle Voronoi diagram is constructed to facilitate neighbor particle search by leveraging its inherent topological relationships; a novel method based on Voronoi cell structure is proposed to address the challenge of surface particle detection; and a particle shifting scheme derived from the Voronoi diagram is developed by integrating the Lloyd relaxation algorithm with the Green-Gauss formula. The results demonstrate that: the Voronoi-based particle search achieves an eiiicu-ncy iniprovemenl oi over 159 times compared to the traditional all pair search method: the Voronoi-based surface particle detection is more accurate than the conventional particle number density method and computes 43. 3 times faster than the geometric method; furthermore, the Voronoi-based particle shifting effectively reduces system disturbance while preserving the smoothness and non-overlapping characteristics of the free surface, also yielding a computational efficiency advantage exceeding 6 times compared to traditional shifting approaches. In conclusion, the proficient spatial partitioning properties of the Voronoi diagram present promising potential for its application within the MPS framework.