Uncertainty quantification for characterization of high enthalpy facilities

The postflight analysis of a space mission requires accurate determination of the free-stream conditions for the trajectory. The Mach number, temperature, and pressure conditions can be rebuilt from the heat flux and pressure measured on the spacecraft by means of a Flush Air Data System (FADS). This instrumentation comprises a set of sensors flush mounted in the thermal protection system to measure the static pressure (pressure taps) and heat flux (calorimeters). Knowing that experimental data suffer from errors, this methodology needs to integrate quantification of uncertainties. Epistemic uncertainties on the models for chemistry in the bulk and at the wall (surface catalysis) should also be taken into account. To study this problem it is necessary to solve a stochastic backward problem. This paper focuses on a preliminary sensitivity analysis of the forward problem to understand which uncertainties need to be accounted for. In section 2, the uncertainty quantification methodologies used in this work are presented. Section 3 is dedicated to the one-dimensional (1D) simulations of the shock layer to identify which chemical reactions of the mechanism need to be accounted for in the Uncertainty Quantification (UQ). After this triage procedure, the two-dimensional (2D) axisymmetric flow around the blunt nose was simulated for two trajectory points of EXPERT (EXPErimental Reentry Test-bed) is simulated and the propagation of the uncertainties on the stagnation pressure and heat flux has been studied. To do this study, the open source software DAKOTA from Sandia National Laboratory [1] is coupled with two in-house codes: SHOCKING that simulates the evolution of the chemical relaxation in the shock layer [2], and COSMIC that simulates axisymmetric chemically reacting flows [3]