Automated chemical mechanism generation for extinction strain rates predictions with applications in flame stabilization and combustion instabilities

Despite the massive gains in computing power, predictive CFD codes for combustion - be it DNS or LES - are still constrained to use relatively small chemical kinetic mechanisms. Kinetic mechanisms used for these simulations are generally derived by reducing comprehensive models based on a target variable such as flame-speed or ignition delay that does not necessarily capture the chemistry of flame stabilization and combustion instabilities. In this paper, we generate skeletal kinetic mechanisms for flame stabilization using the Reaction Mechanism Generator (RMG) software, an automated rate-based tool for generating detailed chemical kinetic mechanisms. First, we provide examples of phenomena that scale with the extinction strain rate, highlighting the need for it to be a target variable for mechanisms describing flame stabilization in gas turbines. Next, we develop a skeletal mechanism for methane-air systems consisting of 21 reactive species and 148 reactions, and compare its performance with comprehensive literature models and experimental data. An efficient methane oxy-combustion mechanism is developed as well. We conclude by providing skeletal mechanism generation heuristics for larger fuels such as n-alkanes and show good agreement for extinction strain rate predictions for n-C5H12, n-C7H16, n-C10H22, and n-C12H26compared to experimental results and a comprehensive model. The presented heuristics provide a novel and automated procedure to generate skeletal n-alkane mechanisms with about 27-36 reactive species each using the skeletal methane-air mechanism as a sub reaction network.