Investigating Statistical Modelling of Excitation Forces in Inhomogeneous Waves

Abstract In coastal areas, and especially fjords, the wave field can present inhomogeneities that affect the responses of floating bridges among other very large floating structures. Recent studies have begun to include the effects of such inhomogeneities in the design phase. To that end, it is necessary to estimate not only the spectrum of the loads on the different pontoons, but also their coherence with each other. However, such information is usually not directly available everywhere, especially for coherence, as hindcast data does not contain phase information, and buoy data is sparse. Thus, assumptions need to be made for the coherence function. This study presents a benchmark of the various assumptions made in recent studies of fjord-crossing floating bridges, on the coherence of linear excitation forces on pontoons. Synthetic inhomogeneous wave fields were generated from an analytical solution, allowing to compare the forces obtained from the various assumptions to true values, not accessible otherwise at sea. This was done by placing “islands”, modelled by large circular cylinders, in an infinite ocean with incident homogeneous short-crested waves. It was found that most assumptions perform well for two pontoons either very close or far from each other. A new assumption is also proposed, adapted to measurements from heave-pitch-roll buoys. It performed systematically better than the historical assumptions for wavelengths larger than the pontoons, and for any distance between them.