Novel Method for Quantitative Assessment of Reduced Chemical Mechanisms Based on the Inherent Similarity Evaluation: Case Study of n-Heptane

A quantitative assessment method of reduced chemical mechanisms based on the similarity evaluation is proposed in this study with special focus on the inherent chemical kinetic characteristics. Two reduced mechanisms, model R1 and model R2, are investigated with regard to ignition delay times under engine-relevant conditions, which are both derived from the same detailed n-heptane mechanism, model D. The uncertainty analysis and the global sensitivity analysis are first used to investigate each C0-C7 submechanism with the species containing the same maximum carbon number. Then, global sensitivity analysis and path sensitivity analysis are employed to study the reaction classes in the fuel-specific submechanisms. Finally, similarity coefficients are calculated between model R1 and model D, as well as between model R2 and model D, to realize the quantitative assessment of the two reduced chemical mechanisms. It is found that the chemical kinetic structure of the whole mechanism and the kinetic characteristics of the fuel-specific submechanism of model R1 are quite similar to those of model D, while lower similarity coefficients exist between model R2 and model D. The disadvantage of model R2 primarily arises from the elimination of some small species and the related reactions in the C1-C4 submechanisms, the excessive removal of some key reaction classes in the fuel-specific submechanism, and the unreasonable modification of the kinetic parameters. On the basis of the above findings, a new reduced n-heptane mechanism with small size and a high degree of similarity of the inherent chemical kinetic characteristics with the detailed mechanism was constructed. The reduced mechanism can accurately predict the fuel ignition, oxidation, and the flame propagation behaviors.

Automated summary

A quantitative assessment method of reduced chemical mechanisms based on the similarity evaluation was proposed in this study. Two reduced mechanisms were investigated and it was found that one model had a high degree of similarity to the detailed mechanism. A new reduced mechanism was constructed to accurately predict fuel ignition, oxidation, and flame propagation behaviors.