Uncertainty Quantifications in Hypersonic Simulations

Engineers are lacking adequate tools to manage the ensemble of simulation and test data that is generated when applying verification, validation, and uncertainty quantification (UQ) standards and methods. This results in excessive time spent to decipher, utilize and manage the provenance of simulation-based workflows. Commercial industry needs a solution that bridges the gap between existing uncertainty quantification toolkits and advanced computer simulations with a foundation in strong data management. As part of an effort to develop Spectre, a new full-stack application developed by Intelligent Light through a Department of Energy (DOE) SBIR, the purpose of this project is to develop a series of computational fluid dynamics (CFD) simulations and perform preliminary UQ analysis. Due to the unique complexities associated with hypersonic flight, such as complex boundary layer profiles due to viscous layers and high temperature gas dynamics, hypersonic simulations are not only more time consuming but also more difficult to achieve and validate; thus, a hypersonic case was chosen. The target of UQ analysis was the boundary layer of an axisymmetric ogive cylinder at Mach 7. Discretization error was analyzed via Richardson extrapolation on a set of 5 grid refinement levels, and each grid was simulated as fully laminar, then fully turbulent using the Spalart-Allmaras turbulence model. Uncertainty in the parameter space was analyzed via polynomial chaos. Original: POSTER PRESENTATION; AUDIO added when event became online only ARIZONA SPACE GRANT AWAR