RANS analyses of cavitating propeller flows: Licentiate thesis

This publication presents validation studies for the cavitation model implemented in the Reynolds-averaged Navier-Stokes equation solver FINFLO. The validation studies relate to ship propellers in uniform and non-uniform inflow conditions. The main physical phenomena involved in cavitation are first introduced. Then, the cavitation phenomena related to marine applications are presented, and the physics behind sheet and vortex cavitation are explained. As cavitating flows are strongly related to turbu-lence, the physics behind turbulence and its simulation methods are also introduced. The benefits and uncertainties related to cavitation tests are described. It is important to understand the drawbacks of experimental methods when comparing the simulation results with the test observations. A brief description of the existing cavitation models is also given, and the utilized cavitation model and its numerical implementation are described in detail. The validation cases are introduced and the simulation results are compared to the out-come of the cavitation tests. The simulation results generally showed good correlation with the experiments. Sheet cavitation was observed in the tests on both the suction and pressure sides of the blades in the validation cases, which was also found in the simulations. The cavitating tip vortices were also found to be similar in the experiments and simulations. The propeller slipstream must be discretized with a high resolution grid in order to predict the cavitating tip vortices and the wakes of the blades with reasonable accuracy. A verification and validation analysis was performed for the global propeller performance characteristics according to the methodology recommended by the ITTC. The influence of the empirical constants in the utilized mass transfer model on the cavitating tip vortices is studied. Finally, explanations for the similarities and differences between the results of the ex-periments and the simulations are discussed. The main differences are found to be caused by laminar flow separation at the leading edge of the blades in the tests, and the limitations of the turbulence and cavitation models utilized in the present simulations.