In previous publications, Singh et al. [1, 2] have shown that direct integration of CFD with a detailed chemistry auto-ignition model (KIVA-CHEMKIN) performs reasonably well for predicting combustion, emissions, and flame structure for stratified diesel engine operation. In this publication, it is shown that the same model fails to predict combustion for partially premixed dual-fuel engines. In general, models that account for chemistry alone, greatly under-predict cylinder pressure. This is shown to be due to the inability of such models to simulate a propagating flame, which is the major source of heat release in partially premixed dual-fuel engines, under certain operating conditions.
To extend the range of the existing model, a level-set-based, hybrid, auto-ignition/flame-propagation (KIVA-CHEMKIN-G) model is proposed, validated and applied for both stratified diesel engine and partially premixed dual-fuel engine operation. The newly proposed model gives good predictions of trends in cylinder pressure and exhaust NOx emissions for both the engines. Finally, both the KIVA-CHEMKIN and the KIVA-CHEMKIN-G models are validated using available measurements of diesel flame liftoff lengths. The models give good qualitative predictions of trends in liftoff; however, the predicted liftoff is shorter than the experimental measurements.