高级检索

基于多重网格尺寸场的非结构自适应网格生成

Unstructured Adaptive Mesh Generation based on the Multi-grid Size Field

  • 摘要: 针对复杂外形自适应四面体网格生成问题, 以递归细分的规则网格——多重网格作为三维尺寸场构建的背景网格, 提出一种基于多重网格的尺寸场构建算法. 首先从网格密度和几何贴合度可控的几何网格上提取曲率自适应网格尺寸; 然后利用简化的尺寸源将单元尺寸约束信息传递给背景网格, 将规则网格细分递归地应用到尺寸远大于单元尺寸约束的单元格; 最后提出一种高效的双向渐进光顺方法, 限制背景网格上的尺寸场梯度. 采用规则网格、多重网格等网格结构计算了多个工业模型的尺寸分布. 数值算例结果表明, 多重网格在单元形状质量方面具有较好的效果, 使用更少的时间和约14%的内存构建了与规则网格几乎相同的渐变尺寸场.

     

    Abstract: For the generation of adaptive tetrahedral meshes, a new algorithm is presented to construct a three-dimensional size field taking the recursively subdivided regular grid, multi-grid, as the background grid. First, curvature adaptive mesh sizes are extracted from the geometric mesh controlled by refinement angle and global mesh size. Then, size constraint information is passed to the background grid using simplified size sources, and a regular grid is inserted into those cells where the element size requirement is much smaller than that of the current cell. Finally, a bi-directional progressive smoothing method is proposed to limit the size gradient. Various grid constructions such as regular grid, multi-grid are employed to represent the size distribution of a number of industrial models. The numerical examples show that multi-grid is superior in terms of element shape quality. Compared to regular grid, multi-grid required less time and about 14% memory to construct an almost identical gradient size field.

     

/

返回文章
返回