Skip to main content

A hybrid parallel solver for finite element computations

  • Conference paper
  • First Online:
High-Performance Computing and Networking (HPCN-Europe 1996)

Part of the book series: Lecture Notes in Computer Science ((LNCS,volume 1067))

Included in the following conference series:

  • 149 Accesses

Abstract

We present a parallel hybrid solver called dual Schur for finite element computations based on domain decomposition. Each subdomain is allocated to one-processor and the linear systems internal to the subdomains are solved with a direct method (LU or Cholesky). The interface problem is expressed in terms of dual variables and uses an iterative algorithm.

Typical results are illustrated for the solution of a Navier-Stokes flow on the Convex Exemplar SPP system. We also emphasize the influence of the decomposition quality on the parallel performances. A good decomposition must exhibit a high load balance with a low interface size and it should contain “compact” subdomains.

Presenting author. The Ph.D. work of Denis Vanderstraeten is supported by the Fonds National de la Recherche Scientifique of the Belgian State. This paper presents research results of the Belgian Programme on Interuniversity Poles of Attraction, initiated by the Belgian State, Prime Minister's Office for Science, Technology and Culture. The scientific responsibility rests with its authors.

This is a preview of subscription content, log in via an institution to check access.

Access this chapter

Institutional subscriptions

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. C. Farhat and F.-X. Roux. Int. J. Num. Meth. Engrg., Vol. 32, pp. 1205–1227, 1991.

    Google Scholar 

  2. C. Farhat and F.-X. Roux. Comput. Mechs. Advances, Vol. 2, pp. 1–124, 1994.

    Google Scholar 

  3. R. Aggarwal, F.-X. Roux, and R. Keunings. In R.F. Sincovec et al., editor, Proc. 6th SIAM Conf. on Parallel Proc. for Sci. Comput., pp. 79–82, 1993.

    Google Scholar 

  4. Saad Y. and M. H. Schultz. Math. of Comput., Vol. 44, pp. 417–424, 1985.

    Google Scholar 

  5. Saad Y. and M. H. Schultz. SIAM J. Sci. Comput., Vol. 7, pp. 856–869, 1986.

    Google Scholar 

  6. D. Vanderstraeten, C. Farhat, P.S. Chen, R. Keunings, and O. Zone. Comp. Meths. Appl. Mech. Engrg., (to appear).

    Google Scholar 

  7. D. Vanderstraeten and R. Keunings. Int. J. Num. Meth. Engrg., Vol. 38, pp. 433–450, 1995.

    Google Scholar 

  8. D. Vanderstraeten, R. Keunings, and C. Farhat. In D.H. Bailey et al., editor, Proc. 7th SIAM Conf. on Parallel Proc. for Sci. Comput., pp. 611–614, 1995.

    Google Scholar 

  9. D. Vanderstraeten, Ph.D. Thesis, Louvain-la-Neuve, 1996.

    Google Scholar 

Download references

Author information

Authors and Affiliations

Authors

Editor information

Heather Liddell Adrian Colbrook Bob Hertzberger Peter Sloot

Rights and permissions

Reprints and permissions

Copyright information

© 1996 Springer-Verlag Berlin Heidelberg

About this paper

Cite this paper

Vanderstraeten, D., Roux, F.X., Keunings, R. (1996). A hybrid parallel solver for finite element computations. In: Liddell, H., Colbrook, A., Hertzberger, B., Sloot, P. (eds) High-Performance Computing and Networking. HPCN-Europe 1996. Lecture Notes in Computer Science, vol 1067. Springer, Berlin, Heidelberg. https://doi.org/10.1007/3-540-61142-8_600

Download citation

  • DOI: https://doi.org/10.1007/3-540-61142-8_600

  • Published:

  • Publisher Name: Springer, Berlin, Heidelberg

  • Print ISBN: 978-3-540-61142-4

  • Online ISBN: 978-3-540-49955-8

  • eBook Packages: Springer Book Archive

Publish with us

Policies and ethics