Formalisation of the multigrid computations
convergenceGauss-Seidel iterationgeometric multigrid methodNavier-Stokes equationnumerical experimentPoisson equationrobust multigrid algorithm
Laplace operator, Helmholtz equation (reduced wave equation), Poisson equation (35J05) Navier-Stokes equations (35Q30) Stability and convergence of numerical methods for boundary value problems involving PDEs (65N12) Multigrid methods; domain decomposition for boundary value problems involving PDEs (65N55)
Summary: The paper represents a robust multigrid algorithm for solving a large class of applied problems (from Poisson equation to Navier-Stokes equations) on the structured grids. The algorithm consists of the robust multigrid technique (variant of the geometric multigrid methods with the problem-independent transfer operators) and the smoothing procedure based on Gauss-Seidel iterations with block unknown ordering. Theoretical analysis and numerical experiments show that close-to-optimal convergence rate for many (non)linear problems is obtained with the same multigrid components. Proposed approach can be used in black box software.
- Understanding data flow graph for improving big data stream computing environments
- Numerical solving of boundary value problems on multiblock grids
- Black-box learning of multigrid parameters
- A formalism for designing robust multigrid methods for Helmholtz-like problems
- Robust multi-grid technique
- Formalization of computations at numerical solution of boundary value problems
- scientific article; zbMATH DE number 2063819 (Why is no real title available?)
- The robust multigrid technique. For black-box software
- Numerical methods for black-box software in computational continuum mechanics. Parallel high-performance computing
This page was built for publication: Formalisation of the multigrid computations
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2452133)