Fuel Sensitive Ignition Delay Models for a Local and Global Description of Direct Injection Internal Combustion Engines

Abstract

Models for ignition delay are investigated and fuelspecific properties are included to predict the effects of different fuels on the ignition delay. These models follow the Arrhenius type expression for the ignition delay modified with the oxygen concentration and Cetane number to extend the range of validity. In this investigation, two fuelsensitive spray ignition delay models are developed: a global model and a local model. The global model is based on the global combustion chamber charge properties including temperature, pressure, and oxygen/fuel content. The local model is developed to account for temporal and spatial variations in properties of separated spray zones such as local temperature, oxidizer, and fuel concentrations obtained by a quasidimensional multizone fuel spray model. These variations are integrated in time to predict the ignition delay. Often ignition delay models are recalibrated for a specific fuel but in this study, the global ignition delay model includes the Cetane number to capture ignition delay of various fuels. The local model uses Cetane number and local stoichiometric oxygen to fuel molar ratio. The model is therefore capable of predicting spray ignition delays for a set of fuels with a single calibration. Experimental dataset of spray ignition delay in a constant volume chamber is used for model development and calibration. The models show a good accuracy for the predicted ignition delay of four different fuels: JP8, DF2, nheptane, and ndodecane. The investigation revealed that the most accurate form of the models is from a calibration done for each individual fuel with only a slight decrease in accuracy when a single calibration is done for all fuels. The single calibration case is the more desirable outcome as it leads to general models that cover all the fuels. Of the two proposed models, the local model has a slightly better accuracy compared to the global model. Results for both models demonstrate the improvements that can be obtained for the ignition delay model when additional fuelspecific properties are included in the spray ignition model. Other alternative fuels like synthetic oxygenated fuels were included in the investigation. These fuels behave differently such that the Cetane number does not provide the same explanation for the trend in ignition delay. Though of lower accuracy, the new models do improve the predictive capability when compared with existing types of ignition delay models applied to this kind of fuels.