A note on spectrum analysis of augmentation block preconditioned generalized saddle point matrices (Q1758405)

This note deals with augmentation block preconditioners for solving generalized saddle point systems whose coefficient matrices have singular \((1,1)\) blocks. Results concerning the eigenvalue distribution and forms of the eigenvectors of the preconditioned matrices and their minimal polynomials are given. Among a set of augmentation preconditioners, an optimal one is derived. The theoretical results presented include and extend previous ones in the literature. Numerical tests demonstrate that the optimal preconditioner is very effective.