Diesel ignition and combustion modeling with comparisons to in-cylinder flame imaging

An integrated numerical model was developed for Diesel engine computations based on the KIVA-II code. The model incorporates a wave breakup spray model, the Shell ignition model, the laminar-and-turbulent characteristic-time combustion model, a spray/wall impingement model and other improved submodels in the KIVA code. The model was validated and applied to model successfully different types of Diesel engines under various operating conditions. These engines include a single-cylinder version of the Caterpillar 3406 heavy duty diesel engine, a Tacom engine, and a Cummins engine. Good levels of agreement in cylinder pressures and heat release rate data were obtained using the same computer model for all engine cases.; An image acquisition-and-processing system was developed for in-cylinder flame imaging of the above Caterpillar engine. The imaging system incorporated an endoscope, an optical linkage, an intensified camera, and a frame grabber along with a computer. Experiments included both single and split injection cases with simulated turbocharging. For the single injection cases, results show that the fuel spray ignited very close to the injector nozzle exit followed by the rapid flame spreading. The ignition timings and locations were very repeatable. For split injection cases, as the amount of fuel in the first injection pulse decreased, ignition occurred further downstream of the nozzle. When only a small amount of fuel was injected, ignition did not occur until the fuel puff reached the piston crown. The second injection fuel pulse was injected into a high temperature environment containing the high temperature combustion products of the first injection and it ignited near the nozzle exit without detectable delay. The possible reason that some split injection schemes can reduce the soot and NOx emissions is because the in-cylinder temperature was maintained at a suitable level throughout the combustion process to prevent massive NOx production and to improve soot oxidation.; The comparisons of flame images with the numerical results show that the present improved KIVA model predicted the ignition timings and locations very well. Shortly after ignition the computed flame propagation was not as rapid as that observed in the experiments. The computed and experimental flame structures agree well again roughly 6 crank angle degrees after ignition for the single injection cases. The numerical model did not simulate adequately the combustion of the small amount of fuel puff in the split injection cases. As the amount of injected fuel increased, the agreement was improved. Future studies are needed to improve the predictions of split injection combustion.