CFD Simulation of the Air Flow Field around a Ship

Abstract

This Master’s thesis covers a project that aims to evaluate the effect of flow distortion on a wind sensor attached to the fjord cruise ship MS Bard using computational fluid dynamics. The simulations were performed with the Spalart-Allmaras turbulence model in steady-state. The incoming wind was modelled with a power law wind profile, and the horizontal wind angles ranged from 0° to 180°. The wind speed at the sensor height varied from 5 m/s to 30 m/s. The simulation method was validated by using it to recreate the results of a wind tunnel experiment performed on the Simple Frigate Ship Shape 2. The final simulation results indicated that the flow field around the wind sensor was generally simple, as the streamlines showed little sign of flow separation or turbulence. Furthermore, the flow around the wind sensor was found to be independent of the Reynold’s number. However, no relation between the flow fields at different wind wind angles was found. Furthermore, a linear relation between the predicted wind direction and the true wind direction was found. The study also concluded that the flow field predicted by different RANS turbulence models is similar on the windward side of MS Bard, but the turbulent structures on the leeward side can be quite different. At the end of the project, a correction factor was developed with which the wind speed and direction measured by the sensor could be corrected to retrieve the true incoming wind. It was impossible to estimate the correction method’s accuracy because there was no experimental data to compare.

This Master’s thesis covers a project that aims to evaluate the effect of flow distortion on a wind sensor attached to the fjord cruise ship MS Bard using computational fluid dynamics. The simulations were performed with the Spalart-Allmaras turbulence model in steady-state. The incoming wind was modelled with a power law wind profile, and the horizontal wind angles ranged from 0° to 180°. The wind speed at the sensor height varied from 5 m/s to 30 m/s. The simulation method was validated by using it to recreate the results of a wind tunnel experiment performed on the Simple Frigate Ship Shape 2. The final simulation results indicated that the flow field around the wind sensor was generally simple, as the streamlines showed little sign of flow separation or turbulence. Furthermore, the flow around the wind sensor was found to be independent of the Reynold’s number. However, no relation between the flow fields at different wind wind angles was found. Furthermore, a linear relation between the predicted wind direction and the true wind direction was found. The study also concluded that the flow field predicted by different RANS turbulence models is similar on the windward side of MS Bard, but the turbulent structures on the leeward side can be quite different. At the end of the project, a correction factor was developed with which the wind speed and direction measured by the sensor could be corrected to retrieve the true incoming wind. It was impossible to estimate the correction method’s accuracy because there was no experimental data to compare.