The flooding after damage of a warship with complex internal compartments - experiments on a fully constrained model in calm water and regular beam seas

In order to provide data to assist in developing and validating a numerical code to simulate the flooding immediately following damage scale model experiments were conducted on a fully constrained model to investigate the progressive flooding through a complex series of internal compartments within a generic destroyer type hull form.

A 3.268 metre long model of a generic destroyer hull form with a simplified, typical internal arrangement was constructed to cover the configuration of greatest interest. A very rapid damage opening scenario was simulated by rupturing a taut membrane covering an opening. The model was instrumented to measure the levels of water and the air pressures in various compartments. In addition, video footage was obtained of the flooding process from both internally and externally of the model.

Previous work presented by Macfarlane et al. (2010) showed the results for the unconstrained model. This paper reports on the outcomes from the experimental program where the model was fully constrained in all six degrees of freedom. Firstly, tests were conducted in calm water with damage opening extents ranging from 50% to 100%. When the damage opening was only 50% the rate of rise of water in each of the compartments was only marginally slower than for the 100% damage extent case.

Secondly, the test results in calm water were compared against results from tests in regular beam seas. A 'set-up' of water inside each of the compartments on the 2^nd Deck was found during the wave tests. The result of this is that the mean equilibrium water level in each compartment in the regular beam sea cases is noticeably higher than the equivalent calm water case, particularly for the two compartments on the port side, away from the damage. Finally, analysis of the data from further calm water and beam sea tests suggests that a similar result also occurs when the model is fixed at various non-zero heel angles.