We design and analyse an iterative method, which uses a specific block smoother for the multigrid cycle. Among many possibilities we choose a few multigrid iterations as the smoother’s blocks. The result is a multilevel procedure that works for regular saddle point problems and features all good properties of the classical multigrid for elliptic problems, such as the optimal complexity and convergence rate independent of the number of levels.
CITATION STYLE
Krzyzanowski, P. (2004). A class of block smoothers for multigrid solution of saddle point problems with application to fluid flow. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 3019, pp. 1006–1013). Springer Verlag. https://doi.org/10.1007/978-3-540-24669-5_130
Mendeley helps you to discover research relevant for your work.