Abstract
A reliable multi-component surrogate fuel model needs to be able to represent both physical properties and chemical kinetics of a real fuel. However, enhancing the fidelity of a model with detailed description of physical and chemical behavior of all fuel components found in real fuels is limited by the prohibitive computational load to calculate the combustion chemistry of the fuel. Hence, it is desirable to achieve computational efficiency by reducing the number of chemical surrogates at the minimum expense of prediction accuracy. The objective of this work is to develop a model that can simulate the oxidation of multi-component fuels by representing the ignition characteristics of physical surrogate components with fewer chemical surrogates and achieve both computational efficiency and prediction accuracy. The main advantage of the model, called the ReactivityAdjustment (ReAd) combustion model, is to accurately predict the reactivity of the physical surrogate components that the reaction mechanisms of which are not included in the reaction kinetics model employed in the simulation. The reactivity variation of local mixtures with different compositions is modeled by adjusting the reaction rate constants of selected control-reactions in the reaction mechanism of the representative chemical surrogates. An initial version of the model has been developed employing a single chemical surrogate to represent the combustion of diesel fuel which is modeled as multiple surrogate components to capture the physical properties of the real fuel. The model was extended to consider two more chemical surrogate components to represent the ignition characteristics of other chemical families than n-alkanes. This enabled to avoid the excessive adjustment of reaction rate constants that were necessary when a single chemical surrogate is used to represent the oxidation kinetics of entire multi-component fuels. The model was extensively tested for simulating oxidation processes of many fuels with a variety of fuel reactivity and in various combustion regimes. The results demonstrated that excellent accuracy of the ignition/combustion prediction was achieved while ensuring computational efficiency.