CSP-driven optimization of a 16-species skeletal mechanism for methane ignition at high pressure

The present work aims at providing a novel kinetic mechanism optimization algorithm based on the computational singular perturbation (CSP) theory. The innovative feature of the optimization process lies in the preliminary sensitivity analysis of the chemical reactions present in the original kinetic scheme, which hinges on calculating application-tailored CSP importance indices. This way, importance indices provide a reliable measure of the contribution of each reaction to the phenomenon of interest, ensuring an overall speed-up of the optimization process. The method is tested on a 16-species skeletal mechanism for the ignition of methane-oxygen mixtures at high pressure, typical of liquid rocket engines (LREs) operating conditions, derived from the C1-C4 version of the detailed mechanism by Zhukov. Compared with the parent detailed kinetic scheme, the predictive capability of the 16-species kinetic mechanism about the ignition delay time is far from adequate in the first place. Once the most relevant reaction to pre-ignition chemistry is identified via CSP analysis, the optimization of its pre-exponential factor over a wide range of mixture initial temperatures, T0, and equivalence ratios, Phi, is carried out by resorting to the multi-objective genetic algorithm (MOGA-II). Consequently, the optimal behavior of the pre-exponential factor as a function of T0 and Phi is condensed into a response surface model (RSM), which equips the optimized 16-species mechanism. Lastly, the performance of the RSM-featured skeletal scheme is validated over the entire set of reactor configurations employed within the optimization process, showing close agreement with the ignition delay estimates provided by Zhukov’s detailed mechanism.