摘要
讨论了大规模有限元并行计算需要解决的并行策略、大量数据的分布存储、方程组迭代求解和程序实现等问题。采用区域分解的“分而治之”的并行策略实现有限元并行。结合区域分解并行策略,将每个子区域的数据信息存储在相应的各个计算机上,实现存储局部化,大大减少并行计算中的通讯量,同时可以实现大规模计算。采用Schur补和共轭梯度法来实现方程组的并行求解,解决岩土有限元病态方程组的求解。采用面向对象的编程技术开发了并行有限元程序。对两个大规模算例进行了并行计算,得到了较好的结果。
The parallel strategies of finite element method, the distributed storage of abundant data, the convergence speed of iterative solution of large scale linear equations and the implementation of parallel finite element method are discussed. We use domain decomposition method （＂divide and conquer＂ strategy） to achieve the parallization of FEM. In combination of domain decomposition strategy, we store the data linking with each subdomaln in each processor. So the data locality not only can reduce the communication, but also increase the scale of the computation. We adopt Schur complement and preconditioning conjugated gradient （PCG） method to fulfill parallel solution of linear equations with bad condition number. We utilize object-oriented programming method to develop our parallel FEM software. Two examples are given to discuss the efficiency of the parallel software. The results show that the software has high performance.
出处
《岩土力学》
EI
CAS
CSCD
北大核心
2007年第4期684-688,共5页
Rock and Soil Mechanics
基金
国家973项目(No.2002CB412708)
国家自然科学基金重点项目(No.50334060)
岩土所开放基金(No.Z000505)资助
关键词
并行有限元
区域分解
岩土工程
parallel finite element method
domain decomposition
geotechnical engineering
