Cavitation in Diesel Fuel Injector Nozzles

Schmidt, D. P. Cavitation in Diesel Fuel Injector Nozzles. University of Wisconsin-Madison, 1997.

Improvements in the fuel injection systems of internal combustion engines can substantially reduce the emission of harmful pollutants. The fuel injection system produces the spray, which directly affects the combustion of the fuel, which in turn determines the production of pollutants. However, the details of this causal relationship remain unclear. The goal of this project is to understand the flow inside fuel injector nozzles and the implications for the downstream spray.

Preliminary simulations of steady, incompressible, nozzle flow were run with a commercial computational fluid dynamics code. These results suggested that cavitation was likely occurring inside many fuel injector nozzles. Next an experiment was performed to identify cavitation in working Diesel fuel injectors. The results of this investigation were consistent with cavitation in the injector nozzles. The project then turned towards understanding and modeling cavitation in nozzles. We also validated an existing analytical cavitation model. This model was extended to provide analytical estimates of the nozzle exit flow profile.

For additional insight into cavitating nozzle flow, a fully compressible, two-dimensional, computational model was constructed. The model, known as Cavalry, assumed that the fluid is homogeneously mixed on the sub-grid scale and that the two phases are in homogenous equilibrium. The model was used to study cavitation in a variety of nozzle shapes. Cavalry was validated with experimental data for nozzle flow and exit velocity. The model was also used to study cavitation between the injector needle and seat. In order to study the effects of asymmetry, the computer model was used to simulate an asymmetric cavitating slot. The asymmetric calculations demonstrated strong, cyclical, oscillations.

Cavalry emphasized the inertial features of cavitation. In some cases thermal effects could be important and the mechanical nature of cavitation not as dominant. Another model, called Moby, was constructed to include thermal non-equilibrium effects. Moby was used to predict the flow of hot, two-phase, nozzle flow. Moby is applicable to the injection of hot fuel, which may be superheated with respect to the downstream conditions. Moby’s predictions have been compared to experimental data for flashing and subcooled nozzle flow.