A Preliminary Study on the Effect of Low Temperature Kinetics on Engine Modeling

Modeling autoignition in diesel engines is a challenging task because of the wide range of equivalence ratios over which it takes place. A variety of detailed autoignition models has been proposed in literature for different fuels. Since these models include about one thousand chemical reactions and more than one hundred species, their application to CFD engines simulations requires a very high computational time, so that they are of no practical interest. In order to lower the computational time, a number of reduced models has been developed including the shell model, which is one of the most used. This model does not take into account low temperature kinetics and consists of seven reactions and three radicals. The use of this model in engine simulations shows its limits when applied to delayed injections because of the predominant influence of the low temperature kinetics. A modified version of the shell model is proposed in the present study. It includes the effect of low temperature kinetics by the addition of two more radicals and three new kinetics reactions. The model has been implemented in a modified version of the KIVA3V code. The performance of the new kinetic scheme has been investigated by computing the ignition delay at different operating conditions in bomb like simulations. The investigated operating conditions were obtained by changing gas temperature and pressure, air to fuel ratio (i.e., oxygen concentration). The effect of the pre-exponential factor in the rate of production of the intermediate agent has also been investigated. The model has also been applied to predict autoignition of a commercial small bore direct injection diesel engine for different injection timings. Results showed that, for delayed injections, the new model was able to better predict the heat released and the pressure traces. The influence of the modified shell model on engine emissions has also been analyzed. INTRODUCTION In compression ignition engines, the combustion process starts from the autoignition of the fuel-oxygen mixture and the ignition delay has a strong influence on engine performance and pollutant emissions. In fact, the two main parameters characterizing the autoignition process are the initial temperature at which autoignition can develop and the time delay before ignition. The detailed kinetic mechanism of the autoignition process includes about one thousand chemical reactions and over a hundred species. Thus, its application to CFD simulations codes requires high computational times. A group of researchers from Shell Research, Ldt. developed a reduced kinetic scheme introducing generic species with kinetic rate constants deduced from experimental data [1] in order to predict knock in spark ignition engines. This model was modified to assure mass conservation [3] and to adjust kinetic rates to fit experimental data [2]. Theobald [4] applied the shell model to diesel engine autoignition phenomena by modifying the kinetic rate. Since then, the shell model has been widely applied in CFD simulations of diesel engine [5] and has been shown to successfully predict ignition processes happening before top dead center.On the other hand, this reduced scheme fails to predict both the ignition delay and the pressure rate of change in the case of delayed injections. This reveals that the shell model does not adequately describe the ignition phenomena which take place at low temperatures. To improve the prediction capability of the model, Cox et al. [6] and Hu et al [7] increased the number of reactions introducing the chemistry controlling the hydrocarbon oxidation process at low temperatures. But the main advantage of the shell model is the low number of species and reactions, which makes this model suitable for CFD applications. Thus, improvements should not substantially increase the complexity of the model. To achieve this twofold goal (i.e., a simple scheme able to predict low temperature hydrocarbon ignition) Gaballo [8] developed a modified version of the shell model by adding two more species and three kinetic reactions. In the present investigation, this modified version of shell model has been implemented in the KIVA3V code [11] and its capability to predict the low temperature autoignition process has been checked. Experimental data from a small bore direct injection diesel engine have been used for comparison. Moreover, the numerical pressure traces obtained with the modified version of the shell model have been compared with the results of the original shell model and the effect of the low temperature chemistry on the soot and NOx engine emissions has been also investigated.