Block triangular preconditioner for static Maxwell equations (Q428373): Difference between revisions

The authors explore block triangular preconditioning techniques applied to the iterative solution of saddle point linear systems arising from the discretized Maxwell equations. In more detail, the paper deals with numerical solution of saddle point linear systems of the form \[ \begin{bmatrix} A & B^T \cr B & 0 \\ \end{bmatrix} \begin{bmatrix} u \\ p \\ \end{bmatrix} = \begin{bmatrix} g \\ 0 \\ \end{bmatrix} \] where \(A\in\mathbb{R}^{n\times n}\) is symmetric positive semidefinite with nullity \(m\) and \(B\in\mathbb{R}^{m\times n}\) has rank \(m\). Inspired by the paper by Rees and Greif [see \textit{T. Rees} and \textit{C. Greif}, SIAM J. Sci. Comput. 29, No. 5, 1992--2007 (2007; Zbl 1155.65048)], the authors propose three new types of block triangular preconditioners, analyze the spectral properties of the preconditioned matrices, and present numerical examples. They conclude that the new preconditioners are slightly more efficient than that of Rees and Greif.