Experimental investigation of the superposition principle for a free surface roll damping tank

Roll stabilizing tanks are widely used on offshore supply vessels and fishing vessels to ensure a better and safer working environment. However, the free surface of a partially filled tank is subject to sloshing. On one hand this is necessary for a roll damping tank, on the other it can lead to undesired effects such as damage to surrounding structures or a high noise level. It is therefore crucial to design such a tank in a proper way in order to dampen the roll motions of a ship most efficiently. The focus of the present thesis lies in performing tests under pure roll and pure sway conditions with a model tank mounted on a Two degree of freedom vessel motion simulator . The tank was equipped with two damping grids. Each implemented experiment consisted of several tests with varying parameters. Three initial experiments were carried out to investigate the main characteristics of a free surface tank: testing the influence of different roll amplitudes with included damping grids, comparing the results with tests where the grids were absent, and testing the tank performance at different filling levels.The objective of the present work is to detect what could be the reason for discrepancies between tests MARINTEK performed with a tank on the vessel motion simulator and a tank installed in a ship model and tested in irregular waves. The problem is approached by checking first in pure sway at periods close to resonance, where the water movement in the tank behaves linearly. Further it is investigated whether a superposition of periods including the resonance period gives the same result as a combined test. The studies are conducted with and without grids. As a preliminary investigation, five periods are superposed in sway and in roll. To date, no detailed research in this area has been realised and experiments of this type are done for the first time compared to previous studies.The test results were plotted with a Matlab code and the damping moment curves were evaluated. For the different roll amplitudes the tank achieves the highest damping at the highest roll amplitude of six degrees. The results show that the occurring moments for the tank without damping grids are much higher than with installed damping grids. The advantage of using damping grids is that the damping is available over a larger range of periods. The comparison of different tank filling heights shows that higher moment amplitudes are achieved with a larger amount of water. However, a high water level gives very high damping at low periods, but only within a narrow range.Furthermore, a sequence analysis was carried out and compared to literature. The major finding was that due to the installed damping grids less water reaches the other side of the tank in time to raise a large counteracting moment. The occurring hydraulic jump is delayed. This leads to the conclusion that both roll amplitude and configuration of the grids have to be considered together when designing a roll damping tank.A superposition of roll and sway moments does not give the same result as a combination of roll and sway in one test. Three major findings resulted from the linearity checks and the superposition of different periods in sway: The damping grids although introducing nonlinearities help to keep the water movement linear in comparison to tests without grids. The resonance period can be included whereas the two neighbouring periods could not