The use of ammonia and natural gas blends in Homogeneous Charge Compression Ignition ( HCCI) engines offers a promising pathway for efficient ammonia utilization while reducing emissions such as soot and NO x . In this study, a detailed kinetic mechanism for NH 3 and CH 4 combustion is first constructed by combining the KONNOV mechanism with Aramco Mech 3.0. The mechanism is then progressively reduced under HCCI conditions using the Directed Relation Graph with Error Propagation (DRGEP), Directed Relation Graph with Path Flux Analysis (DRGPFA), and Full Species Sensitivity Analysis (FSSA). To enhance prediction accuracy, sensitivity analysis is conducted to identify the elementary reactions that have the greatest influence on ignition delay time (IDT) and laminar flame speed (LFS). Using response surface methodology (RSM), the Arrhenius parameters including the pre-exponential factor A, the temperature exponent b, and the activation energy Ea of the key reactions R9 ( H 2 O 2 + M ⇄ 2 OH + M ) and R5 ( O 2 + H ⇄ OH + O ) are optimized. The optimal Arrhenius parameters A, b, and Ea for R9 and R5 are 9.162 × 10 13 , 0.536, and 5.4 × 10 4 cal/mol, and 2.25 × 10 14 , 0.22, and 1.88 × 10 4 cal/mol, respectively. The resulting optimized mechanism, which contains 48 species and 192 reactions, shows significantly improved predictive accuracy. The detailed, reduced, and optimized mechanisms are subsequently coupled with an HCCI engine model to evaluate combustion and emission characteristics. Among the three mechanisms, the optimized mechanism achieves the best agreement with simulated ignition timing and heat release behavior. Its reliability is further confirmed through comparison with two independent natural gas fueled HCCI experimental datasets, where it provides the closest match to measured in-cylinder pressure and heat release rate. These findings show that the optimized mechanism offers a robust kinetic foundation for NH 3 and CH 4 HCCI combustion modeling, even when dedicated dual fuel experimental data are limited.
Zhang et al. (Wed,) studied this question.