The research presented in this thesis outlines the development of a simplified hydrocarbon fuel ignition model, and its application to the prediction of diesel engine combustion. The model is referred to as the EMIT (Elementary Mechanism Induction Time) ignition model. It is an empirical induction time ignition model based on correlations obtained from predictions from a detailed elementary reaction mechanism for n-heptane fuel. The detailed chemical kinetics mechanism used was based on the Sandia CHEMKIN code. The new autoignition model not only incorporates the effects of pressure, temperature, and equivalence ratio, but also includes the effect of exhaust gas recirculation (EGR) on the ignition process. EGR is currently the principle method of reducing NOx, emissions from diesel engines. Modifications of the ignition model to consider the effect of a representative radical species (OH), and to account for the influence of the turbulent flow field on the ignition delay are also presented. The model is formulated such as to allow the effect of the scalar dissipation rate on ignition delay to be included in the ignition model.
The EMIT ignition model has been integrated with a multidimensional engine CFD code (KIVA-3V), which features improved submodels for turbulence, sprays and combustion. The code is applied to simulate both engine and spray bomb experiments. The model has also been applied to consider the effects of split and single injections, with and without EGR for a variety of diesel engines. In addition, a detailed evaluation of the performance of the ignition model is presented for a heavy-duty class diesel engine (Detroit Diesel Series 50). The model performance is shown to provide acceptably accurate predictions of the diesel ignition process over a range of operating conditions. In addition, comparisons with available experimental results show that the integrated model provides adequate predictions of cylinder gas pressure, heat release histories, and with NOx and soot emissions trends. Additional comparisons were made against measured in-cylinder soot concentration (KL) and temperature spatial distributions, and the model predictions were found to be in good agreement with the measurements.