Recursive spectral algorithms for automatic domain partitioning in parallel finite element analysis

Abstract Recently, several domain partitioning algorithms have been proposed to effect load-balancing among processors in parallel finite element analysis. The recursive spectral bisection (RSB) algorithm [1] has been shown to be effective. However, the bisection nature of the RSB results in partitions of an integer power of two, which is too restrictive for computing environments consisting of an arbitrary number of processors. This paper presents two recursive spectral partitioning algorithms, both of which generalize the RSB algorithm for an arbitrary number of partitions. These algorithms are based on a graph partitioning approach which includes spectral techniques and graph representation of finite element meshes. The ‘algebraic connectivity vector’ is introduced as a parameter to assess the quality of the partitioning results. Both node-based and element-based partitioning strategies are discussed. The spectral algorithms are also evaluated and compared for coarse-grained partitioning using different types of structures modelled by 1-D, 2-D and 3-D finite elements.

[1]  Frank Harary,et al.  Graph Theory , 2016 .

[2]  Horst D. Simon,et al.  Fast multilevel implementation of recursive spectral bisection for partitioning unstructured problems , 1994, Concurr. Pract. Exp..

[3]  Charbel Farhat,et al.  A simple and efficient automatic fem domain decomposer , 1988 .

[4]  Gene H. Golub,et al.  Matrix computations , 1983 .

[5]  Alex Pothen,et al.  PARTITIONING SPARSE MATRICES WITH EIGENVECTORS OF GRAPHS* , 1990 .

[6]  Joe Padovan,et al.  Hierarchically Parallelized Constrained Nonlinear Solvers with Automated Substructuring , 1994 .

[7]  H. Saunders,et al.  Finite element procedures in engineering analysis , 1982 .

[8]  Duc T. Nguyen,et al.  An algorithm for domain decomposition in finite element analysis , 1991 .

[9]  Horst D. Simon,et al.  Partitioning of unstructured problems for parallel processing , 1991 .

[10]  Roy D. Williams,et al.  Performance of dynamic load balancing algorithms for unstructured mesh calculations , 1991, Concurr. Pract. Exp..

[11]  N.R. Malik,et al.  Graph theory with applications to engineering and computer science , 1975, Proceedings of the IEEE.

[12]  Steve W. Otto,et al.  Optimal mapping of irregular finite element domains to parallel processors , 1987 .

[13]  Gary L. Miller,et al.  Automatic Mesh Partitioning , 1992 .

[14]  Jerome F. Hajjar,et al.  Parallel processing of central difference transient analysis for three-dimensional nonlinear framed structures , 1989 .

[15]  Shang-Hsien Hsieh,et al.  Parallel processing for nonlinear dynamics simulations of structures including rotating bladed-disk assemblies , 1993 .

[16]  Steven J. Fenves,et al.  A two‐step approach to finite element ordering , 1983 .

[17]  T. D. Morley,et al.  Eigenvalues of the Laplacian of a graph , 1985 .

[18]  Hajjar,et al.  Parallel processing for transient nonlinear structural dynamics of three-dimensional framed structures , 1988 .

[19]  David S. Johnson,et al.  Computers and Intractability: A Guide to the Theory of NP-Completeness , 1978 .

[20]  R. Cook,et al.  Concepts and Applications of Finite Element Analysis , 1974 .

[21]  Narsingh Deo,et al.  Graph Theory with Applications to Engineering and Computer Science , 1975, Networks.

[22]  M. Fiedler.  A property of eigenvectors of nonnegative symmetric matrices and its application to graph theory , 1975 .

[23]  M. Fiedler.  Eigenvectors of acyclic matrices , 1975 .

[24]  J. G. Malone.  Automated mesh decomposition and concurrent finite element analysis for hypercube multiprocessor computers , 1988 .