Short Spray Penetration for Direct Injection Gasoline Engines With Secondary-Drop-Breakup Simulation Integrated With Fuel-Breakup Simulation

Abstract

Direct injection gasoline engines have both better engine power and fuel efficiency than port injection gasoline engines. However, direct injection gasoline engines also emit more particulate matter (PM) than port injection gasoline engines do. To decrease PM, fuel injectors with short spray penetration are required. More effective fuel injectors can be preliminarily designed by numerically simulating fuel spray. We previously developed a fuel-spray simulation. Both the fuel flow within the flow paths of an injector and the liquid column at the injector outlet were simulated by using a grid method. The liquid-column breakup was simulated by using a particle method. The motion of droplets within the air/fuel mixture (secondary-drop-breakup) region was calculated by using a discrete droplet model (DDM). In this study, we applied our fuel-spray simulation to sprays for the direct injection gasoline engines. Simulated spray penetrations agreed relatively well with measured spray penetrations. Velocity distributions at the outlet of three kinds of nozzles were plotted by using a histogram, and the relationship between the velocity distributions and spray penetrations was studied. We found that shrinking the high-speed region and making the velocity-distribution uniform were required for short spray penetration.

Adapting diesel large-eddy simulation spray models for direct-injection spark-ignition applications

“…Recent in-nozzle flow simulations of multi-hole DISI injectors show spray structures that do not correspond to typical diesel sprays, including hydraulic flip and atomization at the nozzle exit. [41][42][43] Other researchers simulating DISI sprays using a RANS turbulence model 40,[44][45][46][47][48][49] have modified their drop initialization, typically specifying an initial dropsize distribution, and it may be that this type of change to the spray initialization procedures is necessary to further improve the accuracy of DISI spray models in either a RANS or LES context. Further work is necessary to better understand how DISI sprays form, how the in-nozzle processes may be affecting the downstream atomization, and how this information is best incorporated into DISI break-up models.…”