Local Fourier analysis of multigrid methods with ILU smoother for triangular grids using aggressive coarsening
摘要
This work investigates multigrid methods on triangular grids with aggressive coarsening and incomplete factorization smoothers. Two-grid local Fourier analysis is employed to optimize smoothing parameters and ensure rapid convergence under aggressive coarsening. A two-dimensional diffusion problem serves as a test case, demonstrating that aggressive coarsening is a viable alternative to standard coarsening. The results highlight substantial computational and memory savings, making aggressive coarsening an efficient choice for multigrid method.