Title:UQ of 2D Slab Burner DNS: Surrogates, Uncertainty Propagation, and Parameter Calibration

Abstract:The goal of this paper is to demonstrate and address challenges related to all aspects of performing a complete uncertainty quantification (UQ) analysis of a complicated physics-based simulation like a 2D slab burner direct numerical simulation (DNS). The UQ framework includes the development of data-driven surrogate models, propagation of parametric uncertainties to the fuel regression rate--the primary quantity of interest--and Bayesian calibration of critical parameters influencing the regression rate using experimental data. Specifically, the parameters calibrated include the latent heat of sublimation and a chemical reaction temperature exponent. Two surrogate models, a Gaussian Process (GP) and a Hierarchical Multiscale Surrogate (HMS) were constructed using an ensemble of 64 simulations generated via Latin Hypercube sampling. Both models exhibited comparable performance during cross-validation. However, the HMS was more stable due to its ability to handle multiscale effects, in contrast with the GP which was very sensitive to kernel choice. Analysis revealed that the surrogates do not accurately predict all spatial locations of the slab burner as-is. Subsequent Bayesian calibration of the physical parameters against experimental observations resulted in regression rate predictions that closer align with experimental observation in specific regions. This study highlights the importance of surrogate model selection and parameter calibration in quantifying uncertainty in predictions of fuel regression rates in complex combustion systems.