首页> 外文会议>Computational Mechanics >Efficient Surface Grid Generation Based on Remeshing
【24h】

Efficient Surface Grid Generation Based on Remeshing

机译:基于网格划分的高效表面网格生成

获取原文

摘要

Surface mesh generation is still one of the most challenging and difficult prerequisites for various numerical simulations, and especially for large-scale volumetric mesh generation. Existing approaches can be essentially categorized into two kinds: direct and indirect [1]. In the direct approach, mesh is generated directly on the surface in 3D such as the Octree-based method and the Advancing Front method. The indirect approach relies on the generation in the parametric domain using Advancing Front method, Delaunay-based technique, or any 2D mesh generation technique [1]. The mesh is then mapped onto the surface using the reverse mapping. The direct methods obviously have the difficulty in mesh validity checking, while the indirect methods are facing the trouble to control the sizing and shapes of the elements to be generated in the 2D region. Accordingly, it is still very time-consuming for generating large-scale surface meshes of sufficient quality.In this paper a new surface mesh generation method is described for efficient automatic generation of high-quality surface grids based on remeshing technique [2]. Supported geometry import formats include composite parametric surface with analytic definition or discrete composite parametric Ferguson curve and surfaces. Initially, for each independent surface patch, a boundary triangulation is generated in the parametric domain, and these patch meshes are mapped onto the surface and assembled into the beginning surface mesh conformingly. Next, remeshing techniques including edges splitting and contraction are utilized for the refinement or coarsening of this initial surface mesh. Metric controlled edge length is computed with respect to surface curvature or coupled with user-specified spacing, both for isotropic or anisotropic cases. Grid points are added accurately through edges splitting using the surface definition; edges contractions are performed robustly with help of determinants computation of triangles in the parametric domain related to the contracted edge. After the refinement, the surface mesh is optimized through edges swapping and points relocation. All the operations are locally performed and high efficiency is resulted. Above one million elements with high quality can be generated in less than one minute in a PC with reasonable CPU speed. Furthermore, small-angle elements due to neighboring patches interface can be avoided by the remeshing techniques naturally. Also, the existing method can be used directly for remeshing surfaces with SGL, STL, IGES formats as geometry input. At the end of the paper, various complex geometry configurations, such as fighter-type aircraft, human organs such as heart and kidney are tested as examples to demonstrate the capabilities of this approach.
机译:表面网格生成仍然是各种数值模拟中最具挑战性和最困难的先决条件之一,特别是对于大型体积网格产生。现有方法可以基本上分为两种:直接和间接[1]。在直接方法中,网格直接在3D的表面上生成,例如基于Octree的方法和前进的前线方法。间接方法依赖于使用前进的前阶段,基于Delaunay的技术或任何2D网格生成技术的参数域中的生成[1]。然后使用反向映射将网格映射到表面上。直接方法显然具有网格有效性检查的困难,而间接方法面临着控制在2D区域中要生成的元素的大小和形状的麻烦。因此,对于产生足够的质量的大规模表面网来仍然非常耗时。 在本文中,描述了一种新的表面网格生成方法,用于基于倒闭技术的高质量表面网格的高效自动生成[2]。支持的几何导入格式包括具有分析定义或离散复合参数弗格森曲线和表面的复合参数表面。最初,对于每个独立的表面贴片,在参数域中产生边界三角测量,并且这些贴片网格被映射到表面上并符合其组装到开始表面网格中。接下来,利用包括边缘分裂和收缩的倒闭技术用于该初始表面网格的改进或粗化。根据表面曲率计算度量控制边缘长度,或者与用户指定的间隔耦合,用于各向同性或各向异性案例。使用表面定义,通过边缘分裂精确地添加网格点;利用与收缩边缘相关的参数域中的三角形的计算的帮助,难以求解边缘收缩。改进后,表面网通过边缘交换和点重新定位进行优化。所有操作都在本地执行,导致高效率。高于100百万元素,在PC中不到一分钟,具有合理的CPU速度。此外,可以通过自然的回忆技术来避免由于相邻贴片接口引起的小角度元件。此外,现有方法可以直接用于用SGL,STL,IGES格式回忆曲面,作为几何输入。在纸张结束时,各种复杂的几何构造,如战斗机型飞机,人体器官,如心脏和肾脏被测试为示例以证明这种方法的能力。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号