On The Use Of Response Conditioned Focused Wave And Wind Events For The Prediction Of Design Loads

Reducing the time required to predict design loads and responses would result in significant efficiency improvements to the floating offshore wind turbine (FOWT) design process. Present methods of predicting design values require at least six, 1-hour sea condition simulations. It would therefore be highly beneficial to identify what conditions lead to the extremes of a range of responses of interest to reduce the length of the time series required. The results of physical, 1:70 scale model experiments using the UMaine VolturnUS-S platform and IEA 15MW reference wind turbine are presented. A Real-time hybrid testing approach, where wind loading is calculated via a surrogate model and reproduced by two on board fans is used to approximate the aerodynamic loads at model scale. The response spectra for several responses of interest are used to construct short wave and wind time series using a 'conditional random response wave' approach developed in ship design. These time series, the responses they produce, and their resulting design value estimates are compared to those from 1-hour-long, irregular wave, and turbulent wind time series. This comparison is conducted for operating conditions at rated wind speed corresponding to design load case (DLC) 1.6, and it is performed at a location in the Celtic Sea.