Details of Research Outputs

The Reynolds-averaged Naviers-Stokes (RANS) method coupling with cavitation model is still a practical tool to predict cavitating flows, particularly in industrial applications, due to its computational efficiency. However, the compressibility effects induced by cavitation are not well considered in conventional RANS methods, which often causes the blockage of the reentrant jet and the total steadiness of the simulated cavity. To this end, modeling of compressibility effects becomes critical to predict the characteristics of unsteady cavitating flows. An empirical eddy viscosity correction [Reboudet al., "Two phase flow structure of cavitation: experiment and modeling of unsteady effects,"in 3rd International Symposium on Cavitation CAV1998, Grenoble, France (1998), Vol. 26.] was proposed to consider the compressibility effects induced by cavitation. Although this modification is able to capture unsteady behaviors of cavitating flows in various configurations, it is still not fully analyzed in terms of the turbulent quantities, e.g., Reynolds shear stress. In this work, we investigate the effects of this compressibility correction on the Reynolds shear stress, by comparing with x-ray experimental data in a small Venturi channel. It is shown that the Reboud correction reduces the eddy viscosity in the entire cavity region, which improves the prediction of Reynolds shear stress near the wall significantly. However, the correction depends only on the simulated mixture density, leading to poor predictions near the phase interface where the simulated mixture density has large discrepancies. Based on the results, we propose an empirical eddy viscosity limiter to confine the original correction beneath the cavitating layer and demonstrate the merits of the proposed correction by comparing with experimental measurements.