Automated, Fast Computational Fluid Dynamics (CFD) Solver Technologies for Hypersonics

The main goal of this effort is to develop a capability for fully-automated fast Computational Fluid Dynamics (CFD) solvers for computation of Boundary Layer Transition (BLT) high-quality laminar base flows.Ý To achieve this goal, the full workflow of the path to the CFD solution must be considered. First, the base flows are typically considered steady in their temporal nature, enabling the use of implicit methods to accelerate convergence.ÝSecond, if a 10x speed up in time to solution is achieved, then the generation of the computational mesh becomes a major impediment to a fast workflow especially in a Multi-Disciplinary Analysis and Optimization (MDAO) framework where the design space of the geometry is explored. To address these problems, SE and its partners at UF will develop accelerated implicit methods for solving the flow using the Arrowhead Matrix Decomposition (AMD) approach along with improved polyhedral spline modeling for parametric geometry to automatically generate hexahedral meshes that can be adaptively refined.