CFD modelling of natural gas combustion in spark ignited engines

Natural gas is gaining popularity as a fuel for automobile applications. In order to procure all the benefits of using natural gas, dedicated engines optimised for this fuel are needed. It is necessary to validate multi-dimensional models of combustion of natural gas in SI engines as such models provide a powerful tool in designing engines and optimising them for this fuel and for studying the emission of pollutants. The multi-dimensional CFD code, KN A-IT, has been used in this study, with necessary modifications and additions. Several combustion models commonly used for SI engine modelling in CFD environments were evaluated. It was concluded that flamelet combustion models are more suitable for engine combustion modelling as engine combustion, in almost all cases, lies within the flamelet regime of combustion. One such flamelet model, the Fractal Flame Model (FFM), was improved, implemented and validated for simulating the turbulent phase of the combustion process in a spark ignited, natural gas fuelled engine. The model validation was done by comparing predictions with experimental data for the base line case. This validated model was then used to study the effects of engine operating parameters and combustion chamber geometry on engine performance and emissions thus providing a suitable analytical background for the designing of dedicated engines optimised for natural gas. This parametric study was performed over a range of air/fuel ratios, with ignition timing, compression ratio and engine speed as the variable parameters. The trends predicted were used to assess the predictive capabilities of the FFM. Eddy break up model (EBUM) was used to compare the predictions by the FFM and these comparisons indicated that the FFM resulted in better agreement with experiments. The predicted trends of engine performance and pollutant formation with the FFM agree well with those found in the literature. This indicates the strength of the FFM as a combustion model for natural gas fuelled SI engines. The effects of the combustion chamber geometry and exhaust gas recirculation on engine performance and pollutant formation were numerically studied and the trends obtained agree with experimental data available. The Renormalisation Group Theory (RNG) k-ε turbulence model was compared with the standard k-ε model and no significant effect on results was found.