Experimental reproduction of an extreme sea state in two wave tanks at various generation scales

Wave structure interaction experiments require the generation of design sea states at model scale in wave tank environments. A vast majority of the industrial practices rely on the stochastic approach. Long free surface elevation timeseries (realizations) are generated by the wave maker of the tank. The input signals are built in the Fourier space, based on a design wave spectrum and random phases. The number of realizations is large enough to contain the events leading to the extreme responses. The quality of the wave field is checked at the structure position (target position). The two main quantities of interest are i) the wave spectrum and ii) the crest height distribution (characterizing the occurrence of the extreme events). As the wave propagation is not linear, those quantities vary from the wavemaker to the target position. To correct the spectrum, existing wave generation procedures iterate on the wavemaker motion. Using such a procedure, the present experimental study focused on the accurate reproduction of an extreme unidirectional design sea state. The later induced numerous and strong breaking events. It was generated in two tanks at two different model scales, at the same target position. The wave height statistics varied from one configuration to an other. The induced limitations for the industrial practices were studied. A particular attention was paid to the effects of the generation scale on the breaking events.