A functional-group-based approach to modeling real-fuel combustion chemistry – II: Kinetic model construction and validation

Abstract

Construction of kinetic models to predict real-fuel combustion properties requires significant human and computational resources. In the first of this two-part study, a functional group correlation approach called FGMech was proposed for predicting the stoichiometric parameters in lumped pyrolysis reactions. The stoichiometric parameters were implemented in a recent real-fuel kinetic model, HyChem (Xu et al., 2018), and the validity of this approach was demonstrated for simulating real-fuel combustion. The present work extends the FGMech approach for developing surrogate and real-fuel kinetic models. Our approach is fundamentally different from the HyChem development approach in that no parameters are tuned to match actual real-fuel pyrolysis/oxidation data, and all model parameters are derived only from functional group data. Along with the stoichiometric parameters obtained in the first part of this study, the thermodynamic data, lumped reaction rate parameters and transport data were predicted in this work based on the functional group characterization of real fuels. The Benson group additivity method was adopted to estimate the thermodynamic data of real fuels, while rate rules developed for pure fuels were used to estimate the rate constants of lumped reactions in real-fuel models. For transport data, normal boiling point, critical temperature and pressure (estimated using the Joback group contribution method) were used to obtain Lennard-Jones parameters. The format of lumped reactions in FGMech followed the HyChem approach, and the base mechanism was adopted from the AramcoMech 2.0 and USC Mech II, respectively, to compare the model performance with different base mechanisms. Fourteen surrogate and twelve real-fuel models were developed based on this approach; they were validated against the experimental data in the literature. FGMech's performance was also compared with detailed and reduced models available in the literature. FGMech reasonably captures the experimental data in the literature, indicating that the present modeling approach is promising for modeling the combustion behavior of fuel, including surrogate mixtures and real fuels.