Numerical Simulation Study of Cavitating Nozzle Flow and Spray Propagation with Respect to Liquid Compressibility Effects 2014-01-1421

The paper addresses aspects of modeling cavitating flows within high pressure injection equipment while considering the effects of liquid compressibility. The presented numerical study, performed using the commercial CFD code AVL FIRE^®, mimics common rail conditions, where the variation in liquid density as a function of pressure may be relevant owing to very high pressure injection scenarios. The flow through the injector has been calculated and the conditions at the outlet of the nozzle orifice have been applied as inlet condition for subsequent Euler-Lagrangian spray calculations to investigate the effects of liquid compressibility treatment on spray propagation. Flows of such nature are of interest within automotive and other internal combustion (IC) related industries to obtain good spray and emissions characteristics. In the development process of the injection equipment, predictive methods using Computational Fluid Dynamics (CFD) contribute to lower development costs, improved engine efficiency, decreased emissions and nevertheless shorter development cycles. Notable yet not drastic differences in the flow structure evidenced by vapor volume fraction distribution, mass flow rates have been observed while drawing comparisons between an incompressible and a weakly compressible assumption. A realistic transient pressure level of 1800 bar has been applied at the inlet of the injector to mimic realistic time dependent inlet conditions. The study is crucial due to the fact that the local compressibility effects can cast a grave consequence on the engine performance (through enhanced spray break-up) and also on the injector behavior due to cavitation erosion process. Spray simulations have been performed on a simple spray bomb geometry using the nozzle file interface as the particle introduction method. Differences in spray propagation are also notable but not as large as one would expect.