Combustion model development of future DI engines for carbon emission reduction

Abstract

Application of low- and zero-carbon fuels is a prevailing trend to reduce engine emissions, while fuel direct injection (DI) is an important way. But the increasing fuel type diversity and injection strategy complexity are beyond the application scope of existing model-based prediction tools. To fill in the gaps, this study developed a unified turbulent mixing and combustion model applicable for various gaseous and liquid fuels. The jet model could accurately capture the penetration under varying injection strategies. Considered the effects of fuel thermo-physical properties and dynamic states, a probability density function (PDF)-particle collision map coupling model was proposed and applied to various fuels. The particles’ discrete method greatly cut down the number of particles processed, remarkably speeding up the computation; and the regions of effective particles collision regimes were refined, vividly mimicking the real-time fuel–air mixing mode, with the specific binary particles exchange mechanisms proposed. Besides, the influence factors from practical engine operations on fuel–air mixing rate were thoroughly considered. Finally, the model accuracy was validated by various fuels with different injection strategies. The results showed that, regardless of the minor difference due to the assumption of constant jet cone angle, the prediction error was generally within 5% for spray tip penetration (Stip) and heat release rate (HRR), with a prediction accuracy over 96% offered for the combustion efficiency (CE). The high accuracy and robustness guarantee a trustworthy prediction and analysis on the spray combustion behaviors of flexible-fuel DI engines with variable injection strategies, which greatly benefits the engine builders to deliver quick and pragmatic industry feasible solutions for engine performance optimization.