Parameter Sensitivity for Uncertainty Quantification in Hypersonic Non-equilibrium Flows.
Brian Lockwood, University of Wyoming
The simulation of hypersonic flow is a subject of interest in many engineering fields and is important for evaluating the performance of atmospheric re-entry vehicles. Hypersonic flow is typically characterized by the presence of chemical reactions and the excitation of internal energy modes, such as vibrational or electronic energy. Additionally, these flows typically occur at low density, which requires that these effects be treated as non-equilibrium phenomena. The modeling of these non-equilibrium effects depends on elaborate empirical models and experimentally measured constants. Quantifying the effect these model parameters have on engineering quantities of interest is the basis of sensitivity analysis. In addition to adaptation and optimization applications, this sensitivity information can be used to determine the uncertainty of an output due to the uncertainties associated with model parameters. The sensitivity approach outlined in this poster is a discrete adjoint approach. This adjoint approach is beneficial as the sensitivity derivatives of a large number of parameters can be computed with minimal computational expense. These sensitivity derivatives can then be used to directly quantify uncertainty with moment methods or used in conjunction with traditional sampling based techniques. Within this poster, the results of a five species, two temperature non-equilibrium calculation are presented and a parameter sensitivity study is performed. In particular, the sensitivities of integrated surface heating to parameters relating to reaction rates, energy coupling and transport coefficients are presented.
Abstract Author(s): Brian A. Lockwood and Dimitri J. Mavriplis