Patrick M. Knupp Distinguished Member Technical Staff Applied Mathematics & Applications Dept.

Office of Science, Applied Math Research

Title: Mesh Quality Optimization via the Target-matrix Paradigm

Patrick M. Knupp, PI, Sandia National Laboratories
Lori Diachin, Lawrence Livermore National Laboratory
Tim Tautges, Argonne National Laboratory

Executive Summary:

Application scientists can often achieve improved solution accuracy in their PDE simulations at a reduced computational cost by applying mesh quality improvement or optimization techniques to adjust mesh geometry or topology. A barrier preventing these scientists from applying mesh optimization methods is the complexity associated with identifying appropriate quality metrics, effective solution strategies, and the lack of guidelines for when and how to apply the many mesh manipulation tools in order to achieve a demonstrable improvement in accuracy. By focusing on the mathematical formulation, analysis, and solution of mesh quality improvement problems, our team will reduce this barrier so that the full potential of mesh improvement techniques can be realized by application scientists. This research leverages our previous successful research that created the Target-matrix Paradigm and addresses the many recurring issues in mesh quality which span a wide variety of applications of importance to the Department of Energy.

© Sandia Corporation | Return to the Sandia Home Page | Optimization & Uncertainty Estimation | Site Map | Privacy and Security