首页> 外文期刊>Applied Mathematical Modelling >Efficient a priori pivoting schemes for a sparse direct Gaussian equation solver for the mixed finite element formulation of the Navier-Stokes equations
【24h】

Efficient a priori pivoting schemes for a sparse direct Gaussian equation solver for the mixed finite element formulation of the Navier-Stokes equations

机译:Navier-Stokes方程的混合有限元公式的稀疏直接高斯方程求解器的有效先验枢轴方案

获取原文
获取原文并翻译 | 示例
           

摘要

A sparse direct solver for the Navier-Stokes is designed. The equation matrix is generated with a priori pivoting. The a priori pivoting is obtained by sorting the nodes prior to assembling the finite element matrix. Three ordering schemes of nodes are investigated. The nodes are sorted with respect to a point far away from the mesh, sequential ordering, the nodes are sorted with respect to the center of the mesh, circular ordering, and the node ordering is randomized, random ordering. The matrix is stored in two one dimensional vectors, containing the upper and lower triangular parts of the equation matrix. The addressing to the coefficient in the matrix is performed through addressing vectors. The complete structure of the equation matrix is obtained by symbolic factorization. The priori pivoting schemes are compared with regard to storage requirements and time efficiency.
机译:设计了Navier-Stokes的稀疏直接求解器。方程矩阵是通过先验枢轴生成的。先验枢轴是通过在组装有限元矩阵之前对节点进行排序而获得的。研究了节点的三种排序方案。相对于远离网格的点对节点进行排序,进行顺序排序,相对于网格的中心对节点进行排序,进行圆形排序,然后将节点排序进行随机化,随机排序。矩阵存储在两个一维向量中,其中包含方程矩阵的上下三角形部分。通过寻址向量对矩阵中的系数进行寻址。通过符号分解获得方程矩阵的完整结构。比较了先验数据透视方案的存储要求和时间效率。

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