面向大规模科学计算的三维Delaunay快速插点算法

该文提出一种快速、稳定的Delaunay插点算法.这一算法提高了单机有元建模的规模,可在PC计算机生成千万级有元四面体网格.算法通过点与点之间位置关系,建立对位置信息;据这些信息在查找BASE单元时,提高＂walk-through＂点定位算法的速度.而在生成新单元和建立邻接关系过程中,算法利用CORE表面的三角网格,在性时间内完成CORE附近的新旧单元更新操作,并出算法时间复杂度证明.本文以分别以空间任意点集、正文体删格和机械模三角面片为例,测试应用Delaunay逐点插入算法.算例表明,本算法在一台Intel（R） Core（TM）2 Duo CPU E7200@2.53GHz,1.98GB内存的PC上可生成千万单元量级四面体网格,生成速度达11-15万单元/秒.

Rapid 3D Delaunay inserting method towards large scale computing

A rapid and robust 3D Delaunay inserting method is proposed in this paper.This method enhances the element generating scale and achieves more than 10 million elements generation in one PC.In the aspect of searching the BASE element,the topological relationship of nodes is utilized to improve the speed of walk-through algorithm.In terms of the process of new element updating,a liner algorithm is proposed according to the triangular mesh of the CORE.At the end of this paper,point sets such as random points,voxel points and triangular mesh points are employed to present the effectiveness of the rapid 3D Delaunay inserting method.In an Intel（R） Core（TM）2 Duo CPU E7200 @ 2.53 GHz,1.98 GB Memory PC with Windows XP operation system,the method generates tetrahedral mesh in 10 million scale.And its speed is 110–150 （thousand elements/second）.