figure10

Domain decomposition On Unstructured Grids

What is DOUG?

DOUG is a black box parallel iterative solver for finite element systems arising from elliptic partial differential equations. Used in conjunction with a finite element discretisation code, DOUG will solve the resulting linear systems using an iterative method and provides a range of powerful domain decomposition preconditioners.

The code is designed to run effectively in parallel on virtually any machine that supports MPI. The matrix-vector operations arising in the iterative method are parallelised using graph partitioning software and additive Schwarz preconditioners can be automatically constructed by DOUG using only minimal input. In this first release a full additive Schwarz preconditioner with automatically generated coarse grid is provided in 2D only. A version without the coarse grid is available in 3D.

DOUG makes no assumptions whatsoever about the finite element mesh that the problem arises from; it may be as unstructured as necessary and only the basic output from the mesh generator and the finite element discretisation are required as inputs to DOUG. The preconditioner can then be used within a range of iterative methods. Currently CG and BiCGSTAB are implemented, other methods of this type can be added.

How does it work?

Basically DOUG is an implementation of the standard additive Schwarz overlapping subdomains algorithm. Requiring only a minimum of input from the user, DOUG is fully capable of automatically:

For more details on what DOUG can do you are referred to the userguide. This userguide is for version 1.98.

Getting DOUG

The DOUG package, currently in release version 1.98, is available by ftp. Installation is fairly straightforward, but if you wish to see what is involved then see the instructions here. This includes the necessary documentation for installation and use.

What computer can it run on?

DOUG has been succesfully used on a number of computer architectures, including SGI, SUN, SUN Solaris, IBM SP/2, Intel i860 and LINUX. It may well work on other systems.

The development of this code was part of the UK Engineering and Physical Sciences Research Council Project GR/J88616 on Parallel Methods for Elliptic PDEs.


Mark Hagger, Linda Stals
Department of Mathematics, University of Bath, Claverton Down, Bath, BA2 7AY, England
Email: parsoft@maths.bath.ac.uk


This page was last updated by Linda Stals on 30/11/98