Investigation of a High Enthalpy Cylinder Flow Using Ab Initio Air Chemistry Data

This study investigates the performance of recent ab initio air chemistry kinetic data to simulate high-enthalpy hypersonic flows, trying to address discrepancies observed in prior empirical models such as Park and Dunn-Kang, which were shown to underpredict surface heat flux at the strong non-equilibrium conditions obtained in experiments conducted in the High Enthalpy Shock Tunnel Göttingen (HEG) facility. Using the HEG cylinder experimental heating data at a freestream enthalpy of approximately 22 MJ/kg and Mach 9, simulations are conducted using NASA’s FUN3D computational fluid dynamics solver with a custom 38-reaction, five-species air model incorporating ab initio dissociation and exchange rates alongside Dunn-Kang recombination rates. A two-stage global sensitivity analysis framework, implemented in Julia, is employed: an initial binary screening identifies six dominant reactions influencing stagnation-point heat flux, followed by a detailed Sobol analysis of their Arrhenius parameters (pre-exponential factor, temperature exponent, and activation energy) within uncertainty bounds. Surrogate models, including random forests, XGBoost, Kriging, and Gaussian processes, facilitate efficient variance decomposition and uncertainty quantification, revealing key sensitivities in dissociation, exchange, and recombination processes. Calibration of parameters aligns predictions with experimental heat flux data, potentially enhancing fidelity over legacy empirical rates. The findings here underscore the advantages of ab initio rates in capturing non-equilibrium chemistry effects, with insights for refining future first-principles and empirical chemical reaction rates.