【摘要】 地质学及矿山研究对象主要特点是结构复杂、非均质、各向异性,四面体网格生成面临的关键技术难题是生成的四面体网格必须保证各个单一地质体边界的一致性.为解决以上问题,本文提出了一种基于面扩展的快速D elaunay四面体网格生成的分治算法,算法首先将地质模型分解为多个简单地质体,利用给定的边界数据样点将这些单一地质体边界进行三角剖分,然后从边界三角形开始递归生成各个简单地质体的四面体网格,最后再将各个简单地质体网格合并为一体.本算法生成的四面体网格可保证多个地质体边界一致,避免四面体单元重叠,且大部分四面体网格符合D elaunay法则.更多还原

【Abstract】 Complex structure, non-homogeneity, anisotropy and multiple constraint conditions are typical characters of 3D data modeling of geo-science and mining. It is a key problem for algorithm of generating tetrahedrons to ensure that every simple geological body’s boundary is conformance. To resolve this problem, a fast divide & conquer algorithm is proposed to efficiency generate Delaunay tetrahedron mesh from a given point set. Firstly, to separate a geo-science model into one or many simple geological bodies whose boundaries then are intersected into Delaunay triangles. Secondly,for every geological body, starting with a seed boundary triangle, the algorithm grows recursively tetrahedron meshes of every geological body by selecting a suit point from given point cloud. Finally, all tetrahedron meshes of every geological body are joined together. In this algorithm the boundary conference is ensured, overlapped tetrahedrons are avoided, and most generated tetrahedrons are Delaunay.更多还原