Parameter-robust preconditioning for hybridizable symmetric discretizations
By: Esteban Henriquez, Jeonghun J. Lee, Sander Rhebergen
Potential Business Impact:
Makes computer math problems solve much faster.
Hybridizable discretizations allow for the elimination of local degrees-of-freedom leading to reduced linear systems. In this paper, we determine and analyse an approach to construct parameter-robust preconditioners for these reduced systems. Using the framework of Mardal and Winther (Numer. Linear Algebra Appl., 18(1):1--40, 2011) we first determine a parameter-robust preconditioner for the full system. We then eliminate the local degrees-of-freedom of this preconditioner to obtain a preconditioner for the reduced system. However, not all reduced preconditioners obtained in this way are automatically robust. We therefore present conditions that must be satisfied for the reduced preconditioner to be robust. To demonstrate our approach, we determine preconditioners for the reduced systems obtained from hybridizable discretizations of the Darcy and Stokes equations. Our analysis is verified by numerical examples in two and three dimensions.
Similar Papers
Preconditioning of a hybridizable discontinuous Galerkin method for Biot's consolidation model
Numerical Analysis
Solves tricky ground problems faster.
Parameter-robust preconditioner for Stokes-Darcy coupled problem with Lagrange multiplier
Numerical Analysis
Solves tricky water flow problems faster.
Distributed preconditioning for the parametric Helmholtz equation
Numerical Analysis
Speeds up computer math for tricky problems.