Assessment of RANS turbulence models and Zwart cavitation model empirical coefficients for the simulation of unsteady cloud cavitation

Abstract

The numerical simulation of unsteady cavitation flows is sensitive to the selected models and associated parameters. Consequently, three Reynolds Average Navier-Stokes (RANS) turbulence models and the Zwart cavitation model were selected to assess their performance for the simulation of cloud cavitation on 2D hydrofoils. The experimental cavitation tests from a NACA65012 hydrofoil at different hydrodynamic conditions were used as a reference to tune the modeling parameters and the experimental tests from a NACA0015 were finally used to validate them. The effects of near wall grid refinement, time step, iterations and mesh elements were also investigated. The results indicate that the Shear Stress Transport (SST) model is sensitive to near wall grid resolution which should be fine enough. Moreover, the cavitation morphology and dynamic behavior are sensitive to the selection of the Zwart empirical vaporization, Fv, and condensation, Fc, coefficients. Therefore, a multiple linear regression approach with the single objective of predicting the shedding frequency was carried out that permitted to find the range of coefficient values giving the most accurate results. In addition, it was observed that they provided a better prediction of the vapor volume fraction and of the instantaneous pressure pulse generated by the main cloud cavity collapse.