The primary aim of this study is to develop a more efficient reaction mechanism for accurately simulating methane combustion, with a specific focus on ignition delay, laminar flame speed, and 2-D simulated flames, while also reducing computational time. Ten reduced reaction mechanisms for methane combustion were evaluated, with only one, "SK30," meeting the required accuracy standards. However, SK30 proved to be computationally intensive when simulating a 2-D premixed flame at a microscale. To address this challenge, a two-step reduction process was implemented. Firstly, an automated algorithm utilized direct relation graphs and sensitivity analysis with ignition delays as a reference to streamline the mechanism while maintaining accuracy. Subsequently, the second step involved identifying key reactions that had a more significant impact on flame speed than ignition delay through sensitivity analysis. Any missing reactions were then added judiciously, prioritizing the retrieval of the missing but important reactions to overall accuracy while minimizing computational cost. This process resulted in a novel mechanism comprising 25 species and 132 reactions for methane-air combustion. The validity of this mechanism was confirmed through comparison with a benchmark model, demonstrating satisfactory agreement in 1-D flame speed and 2-D premixed flame modeling. Most notably, the new mechanism substantially reduced processing time, achieving a 50 % speedup compared to SK30.
Read full abstract