Multi-grid solution of three-dimensional problems with discontinuous coefficients (Q790595): Difference between revisions
From MaRDI portal
Latest revision as of 18:03, 30 July 2024
scientific article
Language | Label | Description | Also known as |
---|---|---|---|
English | Multi-grid solution of three-dimensional problems with discontinuous coefficients |
scientific article |
Statements
Multi-grid solution of three-dimensional problems with discontinuous coefficients (English)
0 references
1983
0 references
The multigrid method for elliptic boundary value problems with strongly discontinuous coefficients [cf. \textit{R. E. Alcouffe, A. Brandt, J. E. Dendy} jun. and \textit{J. W. Painter}, SIAM J. Sci. Stat. Comput. 2, 430- 454 (1981; Zbl 0474.76082)] is extended to the three-dimensional case. An interpolation operator from a coarse grid to the next finer grid is defined using the coefficients of the problem on the finer grid. Coarse grid problems are then defined variationally as Galerkin approximations. Smoothings by ILU decomposition and plane relaxation are considered. The algorithm is compared with an efficient incomplete Cholesky conjugate gradient algorithm. Numerical results show that the multigrid method in three dimensions does not perform as good as in two dimensions on similar problems.
0 references
multigrid methods
0 references
strongly discontinuous coefficients
0 references
Coarse grid problems
0 references
Galerkin approximations
0 references
Smoothings
0 references
ILU decomposition
0 references
plane relaxation
0 references
incomplete Cholesky conjugate gradient algorithm
0 references
Numerical results
0 references
0 references
0 references
0 references
0 references