A New Quasi-Three Dimensional Combustion Model for Prediction of DI Diesel Engines' Performance and Pollutant Emissions

Abstract

The fundamental understanding of mixture formation and combustion process taking place in a Dl diesel engine cylinder is an important parameter for engine design since they affect engine performance and pollutant emissions. Multi-dimensional CFD models are used for detailed simulation of these processes, but suffer from complexity and require significant computational time. The purpose of our work is to develop a new quasi-dimensional 3D combustion model capable of describing the air fuel mixing, combustion and pollutant formation mechanisms, on an engine cycle by cycle basis, needing reasonably low computational time compared to CFD ones, while describing in a more fundamental way the various processes compared to existing multi-zone phenomenological models. As a result, a number of problems associated with the application of multi-zone models are resolved. The combustion chamber is divided into a number of computational cells and the equations of mass, energy and species conservation are solved to calculate temperature and species concentration at each node. The gas flow field is estimated using a newly developed semi-empirical gas motion model, based on the assumption that the in-cylinder pressure is uniform. The finite volume method is used for the solution of the conservation equations. The implicit temporal and hybrid central upwind spatial differencing scheme is used for the discretization of the conservation equations. Spray trajectory, fuel vaporization and combustion are simulated using simplified sub-models based on semi-empirical correlations. A first application of this new model is made on a high speed DI diesel engine. The effect of operating conditions on the combustion and pollutant formation mechanism is examined. Information concerning local air fuel ratio, temperature distribution and species concentrations is derived. To validate the model experiments have been conducted at the authors' laboratory on a DI diesel engine. Comparing experimental with calculated results, a relatively good agreement is observed. This reveals that the proposed model produces qualitatively reliable predictions of the in-cylinder processes and engine performance at reasonable computational time. Furthermore it appears that it can be used to study pollutant formation and the effect of various engine design parameters on the combustion mechanism, providing results that are a compromise between phenomenological models and detailed CFD ones.