针对原有细分四面体切割算法复杂、易生成质量较差新体元的缺点,提出一种新的四面体剖分算法,对原切割算法中的三类不完全切割四面体进行剖分,转换为另两类完全切割四面体类型,使原有的基于最小四面体单元生成算法中的四面体切割类型由五类化简为两类,并对四面体切割过程中可能出现的边界切割和共面切割的特殊情况进行了分类处理,大大简化了四面体切割的分类。同时,为提高新生成的四面体单元的质量,提出了适用于此算法的切割点平移策略。实验结果表明,在一个由2 490个四面体构成的人脑模型上进行切割算法模拟,模型的平均体元质量仅下降0.000 062,说明该算法能够较好地维持模型质量的稳定。
To overcome the defects of the prior method of tetrahedron subdivision, such as complex to implement and low quality of new tetrahedron, this method proposed a new method. To simplify the procedure and improve the efficiency, this method represented some rules to convert the 3 types of uncompleted subdividing tetrahedron into the other 2 types of completed subdividing tetrahedron. It discussed and solved the special cases that might happen during the subdivision procedure. In order to improve the tetrahedron quality, this paper represented a method of moving vertexes. The simulation results show that the average element quality decreases by 0. 000 062 on a brain model using 2 490 tetra meshes ,which proves that the algorithm can keep the quality of model.