The purpose of this paper is to study the effects of cavitation models on the simulation of full-scale propeller cavitation.

The minimum rotational speed of the propeller required for cavitation is predicted by the Bernoulli equation, and the propeller's cavitation condition is observed by full-scale experiment. After determining the boundary layer mesh thickness and establishing the hydrodynamic model, meshes are generated. In the numerical simulation, the turbulence model is selected from either the standard $k-\epsilon$ model or Realizable $k-\epsilon$ model, and the cavitation model is selected from either the Schnerr−Sauer (S−S) model or the Zwart−Gerber−Belamri (ZGB) model. The maximum gas volume fraction on the surface of the propeller is tracked and recorded. In the meantime, the reliability of the numerical simulation is evaluated by observing the gas distribution on the surface of the propeller. And the simulation accuracy of the related models is compared.

The simulation results show that compared with the standard $k-\epsilon$ turbulence model, the cavitation region obtained with the Realizable $k-\epsilon$ model is significantly more consistent with the experimental results. By monitoring the maximum gas volume fraction on propeller's surface, it is found that the choice of turbulence model has little effect on the cavitation intensity, and the cavitation intensity obtained by S−S model is significantly higher than that obtained by the ZGB model.

This study shows that the turbulence model has a great influence on the area of the cavitation region, and cavitation models have a great influence on cavitation intensity. In terms of the simulation of propeller cavitation, the accuracy of the Realizable $k-\epsilon$ model is higher than that of the standard $k-\epsilon$ model.