Comparison of the Homogeneous Relaxation Model and a Rayleigh Plesset Cavitation Model in Predicting the Cavitating Flow Through Various Injector Hole Shapes

Validation of a three-phase Eulerian CFD model to account for cavitation and spray atomization phenomena

“…E. Winklhofer et al [49] provided a value of CCN = 2.33 for this case, and proved that it depends on the geometry. Nonetheless, based on the results of F. Brusiani et al [7] and the ones presented here, it can be assumed that it also depends on the working fluid. The values obtained by F. Brusiani et al [7] are CCN = 2.4 − 2.5, but for the present study they would be between 3 and 4.…”

“…The first validation case corresponds to the two-dimensional throttle arrangement developed and tested by E. Winklhofer et al [49]. It has been previously used as validation case by other researches, for example, F. Brusiani et al [7]. Even though they tested three different geometries with three different contraction levels, only one of them with a contraction of 5%, named "throttle U", is selected for validation purposes.…”

“…Another relevant difference is the slight over-prediction (about a 2.3%) of the chocked mass flow value. These two differences can be attributed to the difference in the saturation pressure value characterizing the experimental diesel and the surrogate adopted for these simulations (which has about 4 times higher saturation pressure than standard diesel) [7]. Mass flow is plotted also versus the cavitation number in Figure 4b.…”

“…As shown in Figure 5a, for the section closest to the throttle inlet, a velocity peak is found in the experiments close to the walls of the throttle. According to the explanation of F. Brusiani et al [7], vapor has lower density than liquid, and so, zones where there is vapor suffer from an increase in velocity by conservation of momentum. This behavior is not captured by the simulations because of three reasons.…”

“…They found out two types of shedding: primary (large-scale cloud cavities) and secondary (detachment of small-scale cavity structures). The Shingal cavitation model was compared to the Homogeneous Relaxation Model (HRM) by F. Brusiani et al [7] for predicting the experimental results obtained by E. Winklhofer et al [49] in a two-dimensional nozzle (throttle). Even though both models reproduced the experimental evolution of mass flow as function of pressure drop, Shingal model seemed to better predict the axial and radial distributions of the vapor region.…”