Combustion and Flame 2018-03-02

Critical kinetic uncertainties in modeling hydrogen/carbon monoxide, methane, methanol, formaldehyde, and ethylene combustion

Yujie Tao, Gregory P. Smith, Hai Wang

文献索引：10.1016/j.combustflame.2018.02.006

摘要

In view of the critical role of the underlying uncertainties of the reaction model in future progress of combustion chemistry modeling, Foundational Fuel Chemistry Model 1.0 (FFCM-1) was developed with uncertainty minimization against available fundamental combustion data of H2, H2/CO, CH4, CH2O, and C2H6. As a critical feature, FFCM-1 not only reconciles a large body of fundamental combustion data, it also has rigorously evaluated uncertainties for the rate coefficients, the combustion experimental targets used for model optimization and uncertainty minimization, and most importantly, an optimized reaction model with quantified uncertainties. In the present work, the remaining kinetic uncertainties of FFCM-1 are examined using a perfectly stirred reactor (PSR) as the relevant model platform for which reliable experiments under the conditions tested are unavailable. The key questions to address include the level of improvement from model optimization in the prediction uncertainties of PSR residence times at extinction and ignition and the rate coefficients of reactions that must be improved in order to reduce the prediction uncertainties. Computational tests are made for H2/CO , CH2O , CH4 , CH3OH and C2H4–air mixtures over the pressure range of 10–100 atm and PSR inlet temperatures that would yield residence times comparable to the time scales typical of fuel combustion in practical combustors. The results show that although model optimization reduces the prediction uncertainties of residence time at extinction and ignition, the remaining uncertainties remain rather large. Key reactions for which reduced rate uncertainties would greatly improve the reaction model quality and accuracy have been identified and discussed in detail.