Solution Algorithm on Unstructured Grids Using Hamiltonian Paths and Strand Grids
Speaker: Jay Sitaraman, Parallel Geometric Algorithms, LLC
Presentation
Abstract
An approach for achieving line-structures in purely unstructured grids is presented in two and three dimensions. The method entails the identification of Hamiltonian paths created by quadrilateral subdivision of unstructured meshes to represent two distinct surface coordinate directions and strand grids to represent the wall normal direction. This method allows for use of numerical schemes typical to traditional structured curvilinear grids, viz stencil-based discretization schemes and line-implicit operators that use approximate factorization, on unstructured meshes. Several canonical test cases will be presented in two dimensions (airfoil flows, flat-plate) followed by three dimensional applications on wing, sphere and the robin helicopter fuselage. The Hamiltonian path based solver (HAMSTR) developed as part of this work is expected to become one of the near-body solvers in the Helios framework.
Biography
Jay Sitaraman is currently one of the developers of the U.S Army Helios framework and works as a contractor through Parallel Geometric Algorithms LLC, a small scientific computing firm based in Sunnyvale, CA founded by him. Jay holds a PhD in Aerospace Engineering from the University of Maryland, earned in 2003, and has held various research positions in government, industry and academia. Before moving to NASA ARC, he was an Associate Professor at the Department of Mechanical Engineering at the University of Wyoming.