Skip to main content
NASA Advanced Supercomputing (NAS) Division World-class modeling, simulation, and supercomputing in support of the NASA mission

Implicit Preconditioning for Explicit Multigrid Solvers on Cut-Cell Cartesian Meshes

AMS Seminar Series

Speaker: Jonathan J. Chiew, Computational Aerosciences Branch, NASA Ames Research Center, Moffett Field, CA

Presentation

Seminar Slide Deck (PDF-3.4MB)

Abstract

Cut-cell Cartesian methods enable automated mesh generation and adaptation which are essential for vehicle design and optimization. However, optimizations may slow down or terminate if the solver is not able to converge. In this talk, we discuss an investigation into implicit preconditioning for strengthening the explicit multigrid solver utilized in the Cart3D simulation framework. A Jacobian-free Krylov subspace method is employed to solve the linear preconditioning equations, demonstrating improved multigrid convergence rates on a variety of problems. Sequential hybrid preconditioning is also studied as a method to reduce computational cost. The last portion of the talk will discuss a novel formulation of dual time stepping which encompasses both fully and diagonally implicit Runge-Kutta methods as well as BDF schemes in a unified framework.

Biography

Photo of Jonathan Chiew
Jonathan Chiew is a Research Aerospace Engineer in the Computational Aerosciences Branch of the NASA Advanced Supercomputing Division at NASA Ames Research Center in Moffett Field, California. He has been part of the Cart3D development team since 2013, primarily working on algorithm development, time-accurate methods, and optimization framework integration. Prior to joining NASA, Jon was an aerodynamics engineer at The Boeing Company performing aerodynamic analysis for various rotorcraft programs including V-22 and H-47. He received an M.S. and Ph.D. in Aeronautics and Astronautics from Stanford University as well as a B.S. in Aerospace Engineering from Georgia Tech.