Key points are not available for this paper at this time.
Abstract Assessment of the extreme responses that offshore structures will experience within their life-cycle, involves numerous long-duration calculations for various weather and service conditions. The response distribution of any quantity driven by wave-structure interaction can be obtained efficiently with linear theory. However, the accuracy of that method is known to quickly deteriorate in severe sea states and for highly nonlinear quantities. In those cases, simulations with a high-fidelity solver or experiments are necessary which, although accurate, imply a significant cost (computational or experimental time) and in several cases become unfeasible. For this reason, the concept of response-conditioned waves (RCW) has been developed, as a way of investigating solely wave patterns of short duration, which can be considered tantamount to brute-force simulations. This approach consists of computing short-time wave episodes, that induce a given target response of the studied quantity. At a later stage, these waves can be generated and propagated in a computational domain, so that the fully nonlinear responses are obtained, by using only brief high-fidelity numerical simulations. The present paper demonstrates the relevance of such an approach, designed as a two-step method. First, the RCW is computed with a low-fidelity surrogate model, based on a nonlinear High Order Spectral (HOS) potential flow solver for wave modeling and a linear wave-structure interaction model. The method has been validated for two distinct response cases, namely the prediction of the vertical bending moment (VBM) of a zero-speed containership and the nacelle acceleration of a floating wind turbine. An overview of those applications is provided and the principal strengths, as well as the potential limitations of the method, are identified and discussed.
Bouscasse et al. (Sun,) studied this question.