Solving elliptic problems by the domain decomposition method using preconditioning matrices derived by multilevel splittings of the finite element matrix

  • R. D. Lazarov
  • P. S. Vassilevski
  • S. D. Margenov
Session 9A: Algorithms, Architectures And Performance I
Part of the Lecture Notes in Computer Science book series (LNCS, volume 297)


We study some aspects of domain decomposition technique applied to a multilevel iterative method in solving second order elliptic problems by the finite element method. This iterative method makes use of a preconditioning matrix which is constructed by using a sequence of nested discretizations of the considered elliptic problem. The main computational task in solving problems with this preconditioning matrix is solution of problems, on each fixed level, with matrices which have condition number bounded on hk (— the discretization parameter on the level k). Here we propose to use, in connection with parallel implementation of this multilevel iterative method, domain decomposition technique in order to solve these problems with constantly conditioned matrices. Numerical tests are presented and they are compared with tests in which the solution method is the preconditioned conjugate gradient method with preconditioning matrices derived by approximate (blockwise) factorizations of the constantly conditioned matrices. The last methods suffers of the fact that it is not well parallelizable but it is very fast convergent.


Elliptic Problem Domain Decomposition Conjugate Gradient Method Multigrid Method Domain Decomposition Method 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    Axelsson, O.: A survey of vectorizable preconditioning methods for large scale finite element matrices. In: Colloquium Topics in Applied Numerical Analysis, J.G. Verner (Ed.), CWI syllabus 4&5, Amsterdam 1983.Google Scholar
  2. 2.
    Axelsson, O. and I. Gustafsson: Preconditioning and two-level multigrid methods of arbitrary degree of approximation. Math. Comput. 40 (1983), pp. 219–242.Google Scholar
  3. 3.
    Axelsson, O. and B. Polman: On approximate factorization methods for block-matrices suitable for vector and parallel processors Lin. Alg. Appl., 77 (1986), pp. 3–26.Google Scholar
  4. 4.
    Bank, R. and T. Dupont: Analysis of a two-level Scheme for solving finite element equations. Report CNA-159, Center for Numerical Analysis, The University of Texas at Austin, 1980.Google Scholar
  5. 5.
    Bjorstad, P.E. and O.B. Widlund: Iterative methods for the solution of elliptic problems on regions partitioned into substructures. SIAM J. Numer. Analysis.Google Scholar
  6. 6.
    Bramble, J.H., J.E. Pasciak and A.H. Schatz: An iterative method for elliptic problems on regions partitioned into substructures. Math. Compt., 46 (1986), pp. 361–369.Google Scholar
  7. 7.
    Bramble, J.H., J.E. Pasciak and A.H. Schatz: The construction of preconditioners for elliptic problems by substructuring. I. Math. Comput., 47 (1986), pp. 103–134.Google Scholar
  8. 8.
    Bramble, J.H., J.E. Pasciak and A.H. Schatz: The construction of preconditioners for elliptic problems by substructuring, II. Preprint, BNL — AMD — 1986, Cornell University, Ithaca, New York, 1986.Google Scholar
  9. 9.
    Chandra R.: Conjugate gradient methods for partial differential equations. Research Report #119, 1978, Dept. Comput. Sci., Yale University.Google Scholar
  10. 10.
    Concus, P., G.H. Golub and D.P. O'Leary: A generalized CG method for the numerical solution of elliptic partial differential equations. In: J.R. Bunch and D.J. Rose, Eds: Sparse Matrix Computations, pp. 309–332, New York, Academic Press, 1976.Google Scholar
  11. 11.
    Dryja, M.: A capacitance matrix method for Dirichlet problem on polygon region. Numer. Math., 39 (1982), pp. 51–64.Google Scholar
  12. 12.
    Dryja, M.: A finite element-capacitance method for elliptic problems on regions partitioned into substructures. Numer. Math., 44 (1984), pp. 153–168.Google Scholar
  13. 13.
    Dryja, M. and W. Proskurowski: Iterative methods in subspases for solving elliptic problems using domain decomposition. Technical report CRI-86-10, 1986, Department of Mathematics, DRB 306, University of Southern California, Los Angeles.Google Scholar
  14. 14.
    Dryja, M., W. Proskurowski and O.B. Widlund: A method of domain decomposition with cross points for elliptic finite element problems. In: Proceedings from an International Symposium on Optimal Algorithms held in Blagoevgrad, Bulgaria, April 1986.Google Scholar
  15. 15.
    Golub, G.H. and D. Mayers: The use of pre-conditioning over irregular regions. In: Proceedings of the Sixth International Conference on Computing Methods in Science and Engineering, Versailles, France, December 1983, pp. 3–14.Google Scholar
  16. 16.
    Keyes, D.E. and W.D. Gropp: A comparison of domain decomposition techniques for elliptic partial differential equations and their parallel implementation. Research Report YALEU/DCS/RR-448, December 1985.Google Scholar
  17. 17.
    Kuznetsov, Yu.A.: Computational methods in subspaces. In: Computational Processes and Systems, vol. 2, Nauka, Moscow, 1985, pp. 265–360. In Russian.Google Scholar
  18. 18.
    Lebedev, V.I. and V.I. Agoshkov: Poincare-Steklov operators and their applications in analysis. Department of Computational Mathematics, USSR Academy of Sciences, Moscow, 1983. In Russian.Google Scholar
  19. 19.
    Maitre, J.E. and F. Musy: The contraction number of two-level methods; An exact evaluation for some finite element subspaces and model problems. In: Multigrid Methods, W. Hackbusch and U. Trottenberg, (eds.)., Lect. Notes Math., #960, Springer-Verlag, Berlin-Heidelberg-New York, 1982, pp. 535–544.Google Scholar
  20. 20.
    Matsokin, A.M.: Methods of fictitious components and alternation in subregions. Preprint 612, 1985. Computing Center, Sibirian Branch of the USSR Academy of Sciences, Novosibirsk, In Russian.Google Scholar
  21. 21.
    Nepomnyashtchich, S.V.: On application of bordering method to mixed b.v.p. for elliptic equations and on network norms in W21/2 (S). Preprint 106, 1984. Computing Center, Sibirian Branch of the USSR Academy of Sciences, Novosibirsk. In Russian.Google Scholar
  22. 22.
    Stuben, K. and U. Trottenberg: MG methods: Fundamental Algorithms, Model Problem Analysis and Applications. In: W. Hackbuch and U. Trottenberg. Eds.: Multigrid Methods. Lecture Notes Math., #960, Springer-Verlag, Berlin-Heidelberg-New York, 1982, pp. 1–172.Google Scholar
  23. 23.
    Vassilevski, P.S.: Nearly optimal iterative methods for solving f.e. equations based on multilevel splitting of the matrix. (Submitted) 1987.Google Scholar
  24. 24.
    Vassilevski, P.S.: On some ways of approximating inverses of banded matrices in connection with deriving preconditioners based on incomplete blockwise factorizations. (Submitted), 1987.Google Scholar
  25. 25.
    Vassilevski, P.S.: The role of inaccurate residuals of the Schur complement in solving f.e. elliptic problems by substructuring. (Submitted), 1987.Google Scholar
  26. 26.
    Vassilevski, P.S.: On some ways of constructing preconditioners for solving f.e. elliptic problems by substructuring. (Submitted), 1987.Google Scholar
  27. 27.
    Vassilevski, P.S. and S.I. Petrova: A note on construction of preconditioners in solving 3D elliptic problems by substructuring. (To appear), 1987.Google Scholar
  28. 28.
    Widlund, O.B.: A guide to the literature on iterative substructuring. Courant Institute, Dept. Comput. Sci., New York University, August 1986.Google Scholar
  29. 29.
    Widlund, O.B.: An extension theorem for finite element spaces with three applications. Technical Report #223, 1986, New York University, Dept. Comput. Sci., Courant Institute.Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 1988

Authors and Affiliations

  • R. D. Lazarov
    • 1
  • P. S. Vassilevski
    • 1
  • S. D. Margenov
    • 2
  1. 1.Bulgarian Academy of Sciences Institute of Mathematics with Computing CenterSofiaBulgaria
  2. 2.Bulgarian Academy of Sciences Center of Informatics and Computer TechnologySofiaBulgaria

Personalised recommendations