三维实体网格自适应划分算法

高质量的网格划分是三维建模研究的关键。根据对三维形体的几何特征和物理特征进行分析，给出三维网格划分的加密规则。通过研究网格加密区域和网格节点算法，设计基于Delaunay剖分的动态节点单元一体化三维网格自适应生成算法。在对对象的加密区域、区域布点等前期处理后，对选取的节点集进行Delaunay三角剖分，选取最优节点，并对特征集中区域进行局部加密，从而完成对三维体的网格划分，与传统划分方法相比较更具精确性和高效性。通过对机械零件进行网格划分，根据加密规则实现网格的疏密分布，划分结果能够准确描述出三维体的几何形体特征和物理特征分布，与传统划分方法的网格图形进行对比验证该方法的有效性，为更深入地研究形体提供基本保证。

Three Dimensional Entity Mesh Generation Algorithm

The high grade grid division is the key of the three dimensional modeling research. Based on analysis to the three dimensional geometry and the physical characteristics, a three dimensional grid division encryption rule is introduced. Through the research of grid refinement region and the grid node algorithm, a dynamic node unit integration three dimensional grid adaptive division algorithm is designed on the basis of on Delaunay triangulation. With the preprocessing of the object, the nodes are optimized by using Delaunay triangulation and characteristic centralized region is refined locally. Compared with traditional method, this algorithm has obvious superiority in computation and accuracy. Through grid division to the machine parts and optimized grid density distribution, the results can accurately describe the three dimensional entity geometrical physique characteristic and the physical characteristic distribution. Experiment verifies the validity of this method, thus providing the basic guarantee to more in-depth researches of entity characters.