| |
By Takafumi Kawamura and Takayuki Watanabe, Department of Environmental and Ocean Engineering, University of Tokyo, and Shin Hyung Rhee, Fluent, Inc.
View the pdf of this Supplement
Cavitation of marine propellers can cause many problems, such as vibration, noise, and erosion on the blades. However, with recent high-speed and shallow-draft ships, it is difficult to avoid cavitation without compromising the propeller efficiency. Propeller designers must therefore control the influence of cavitation rather than attempt to suppress its occurrence. As a result, the accurate prediction of cavitation is becoming increasingly important. While empirical and experimental methods are usually used for practical designs at present, a research group in the University of Tokyo has been using FLUENT for the prediction of steady and unsteady cavitation on marine propellers, with a special emphasis on the role that the ship's wake plays in the process.
Non-uniform inflow equivalent to a ship's wake; lines indicate the magnitude of the incoming axial velocity
In a recent study, the sliding mesh and cavitation models in FLUENT, along with the k-ω turbulence model were used to simulate the cavitating flow generated by a rotating propeller. About 1.8 million cells were used to mesh the computational domain around a five-bladed model propeller, 40 cm in diameter. The hybrid mesh with prismatic cells near the propeller surface and tetrahedral cells in the outer region was generated using GAMBIT.
Unsteady propeller cavitation in the wake of a ship; the ship wake is generated by a wire mesh placed on the upstream side of the tunnel
Variation of the pressure coefficient during a rotation at a point near the leading edge of the blade
When a marine propeller is operated in the wake of a ship, the angle of attack of the incoming flow relative to the blade varies as the propeller rotates. This causes the periodic growth and collapse of a cavity, and this fluctuating pattern is strongly related to the vibration and noise often associated with cavitation. In the present study, the measured wake distribution was specified at the inlet boundary using a user-defined function (UDF). When one of the blades is located at the top center position, it is in the central portion of the wake. The incoming axial flow is smallest at this position, and the angle of attack is largest. The wake is assumed to extend about 30 degrees on either side of the top center position. Outside the wake region, the incoming axial flow is uniform.
The time-dependent calculation with the nonuniform inflow reproduced the unsteady cavity pattern quite well. When one of the blades enters the wake region, leading edge cavitation is formed over a wide span and grows towards the trailing edge as the blade goes deeper into the wake. Subsequently, the cavity shrinks from the root to the tip when the blade exits the wake region. The whole process was captured correctly by FLUENT, and the variation of the pressure on the blade surface agreed very well with the measurement. However some discrepancies were also found. The shrink of the cavity was too quick, and the maximum cavity volume was about 50% smaller than in the experiment.
Further studies are being conducted for better agreement and improved computational efficiency. The present investigations are focused on the performance of the cavitation and turbulence models in the cavitating region. The cavitation model in FLUENT 6.1 is certainly a state-of-the-art model, and work will continue to clarify its range of application and limitations for reliable and quantitative predictions of cavitating flows.
References:
- T. Watanabe, T. Kawamura, Y. Takekosi, M. Maeda, S.H. Rhee, Simulation of Steady and Unsteady Cavitation on a Marine Propeller Using a RANS CFD Code. Proceedings of the 5th International Symposium on Cavitation, Osaka, Japan, November 1-4, 2003.
- T. Kawamura, T. Watanabe, Y. Takekoshi, M. Maeda, H. Yamaguchi, J. Soc. Nav. Architects Jpn. 2004, 195, 211-219.
- Y. Ukon, Y. Kurobe, T. Kudo, J. Soc. Nav. Architects Jpn. 1989, 165, 83-94.
|
|
|
