Robust reconstruction of the catalytic properties of thermal protection materials from sparse high-enthalpy facility experimental data

• Published in: Experimental thermal and fluid science Vol. 96; pp. 482 - 492
• Main Authors: Sanson, Francois, Panerai, Francesco, Magin, Thierry E., Congedo, Pietro M.
• Format: Journal Article
• Language: English
• Published: Philadelphia Elsevier Inc 01.09.2018; Elsevier Science Ltd; Elsevier
• Subjects: Atmospheric entry; Bayesian analysis; Bayesian inference; Catalysis; Ceramic matrix composites; Enthalpy; Experimental data; Experiments; Fluid Dynamics; Ground tests; Heat flux; Heat transfer; Molecular chains; Nitrogen atoms; Physics; Properties (attributes); Random variables; Recombination; Reconstruction; Spacecraft; Thermal protection; Thermal protection systems; Uncertainty; Uncertainty quantification; Uncertainty quantification; Thermal protection systems; Catalysis; Bayesian inference; Uncertainty Quantification; Bayesian Inference; Thermal Protection Systems
• ISSN: 0894-1777; 1879-2286
• DOI: 10.1016/j.expthermflusci.2018.03.028

•We apply a new approach to quantify uncertainties in reusable TPS catalytic property.•For the experimental data analyzed, this property cannot be accurately estimated.•The new method makes no assumption on the reference calorimeter catalytic property.•Assumptions on the reference calorimeter leads to a bias in the TPS catalytic property. Quantifying the catalytic properties of reusable thermal protection system materials is essential for the design of atmospheric entry vehicles. Their properties quantify the recombination of oxygen and nitrogen atoms into molecules, and allow for accurate computation of the heat flux to the spacecraft. Their rebuilding from ground test data, however, is not straightforward and subject to uncertainties. We propose a fully Bayesian approach to reconstruct the catalytic properties of ceramic matrix composites from sparse high-enthalpy facility experimental data with uncertainty estimates. The results are compared to those obtained by means of an alternative reconstruction procedure, where the experimental measurements are also treated as random variables but propagated through a deterministic solver. For the testing conditions presented in this work, the contribution to the measured heat flux of the molecular recombination is negligible. Therefore, the material catalytic property cannot be estimated precisely. Moreover, epistemic uncertainties are rigorously included, such as the unknown reference calorimeter catalytic property.