Modeling strategies and experimental comparison for large-eddy simulations of a laser-ignited rocket combustor

One of the main goals of the Stanford Predictive Science Academic Alliance Program (PSAAP) III Center is to predict the reliability of laser ignition in a methane and oxygen-fueled rocket combustor under in-flight conditions. This prediction relies on multifidelity ensembles of numerical simulations running on exascale-class, heterogeneous supercomputers (CPUs and GPUs). To evaluate the accuracy and reliability of the Hypersonics Task-based Research (HTR) solver (Di Renzo et al., Computer Physics Communications, 2020), which was developed as part of the Stanford PSAAP-II and PSAAP-III Centers, parallel experimental validation efforts are being conducted in a rocket combustor configuration at the Zucrow Propulsion Laboratory at Purdue University. This study aims to evaluate modeling strategies for performing Large Eddy Simulations (LES) of laser-ignited rocket combustors by comparing numerical results with experimental data. The focus is on three key aspects of the ignition process: 1) the non-reacting coaxial jet in the pre-ignition stage, 2) the hot kernel dynamics resulting from laser energy deposition, and 3) the post-ignition combustion process. First, we quantitatively evaluate the performance of the LES by comparing the mean flow statistics to PIV data. Next, we assess the sensitivity of the hot kernel dynamics, characterized by the formation of an ejecta, to various modeling parameters. By comparing the numerical results with experimental Schlieren imaging, we establish the uncertainty range of these parameters, accounting for shot-to-shot variability. Finally, we demonstrate the benefits of using a dynamic thickened flame combustion model when considering the flame expansion and pressure rise within the combustor.