Numerical Study of Two Sparse AMG-methods

From MaRDI portal





The author studies the construction and the properties of sparse algebraic multigrid methods (AMG) for Schur complement approximation. The motivation of this study is that sparse AMG is a cheap and accurate way to compute approximations of Schur complements of matrices arising from discretizations of some symmetric and positive definite partial differential operators. The results from numerical experiments are presented in order to give an accurate qualitative information about the properties of the method. Basing on comparison of the sparse AMG method with the sparse BPX method of \textit{J. H. Bramble, J. E. Pasciak} and \textit{J. Xu} [Math. Comput. 55, 1-22 (1990; Zbl 0703.65076)], the author advice to use the sparse BPX for Laplace-like problems on structured meshes and AMG method for Schur complement approximation and for boundary Lagrange multiplier preconditioning for some more complex operators and problems on unstructured grids.











This page was built for publication: Numerical Study of Two Sparse AMG-methods

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4423107)