Coupled modeling of dynamic ice-structure interaction on offshore wind turbines

Abstract

The offshore wind industry is growing at high pace and is already one of the major renewable energy sources in Europe. To support this growth, costs must be further decreased, amongst others by efficiently designing the support structure of the offshore wind turbines (OWT). This requires accurate prediction of environmental loading such as dynamic ice-structure interaction. From industry practice it is known, that current phenomenological models show results that are highly sensitive to the used input parameters. It is investigated if the methodology used in industry practice, can be improved by coupled phenomenological modeling of dynamic ice-structure interaction. The phenomenological model proposed by Hendrikse and Metrikine (2015) is implemented and extended with the creep and buckling failure modes. The input parameters of the model are derived from reference data that is obtained from full-scale measurements. A verification study is performed to demonstrate that the reference data is matched. Furthermore, a qualitative verification is performed on the failure modes and ice-induced vibration regimes. The phenomenological model is applied to a structural model of an OWT and investigated in two case studies. For the implementation of the structural model, firstly the phenomenological model for ice-structure interaction is adjusted for cylindrical structures. Aerodynamic damping and thrust are included to enable simulations of the ice-structure interaction during the production of electrical energy. The coupled model is applied in two cases. The first case considers an offshore wind turbine model with ice loading as the only subjected load to the structure. The second case considers the combination of ice and wind loading. A comparison is made between coupled and uncoupled models for ice-structure interaction. The uncoupled models use an external ice load series to represent the ice-structure interaction and are therefore based on methodology used in industry practice. In case of ice-only loading, it is concluded that the uncoupled model is not capable of capturing intermittent crushing and frequency lock-in behavior due to inconsistency between the load signal and the response. In case of combined ice and wind loading, it is concluded that the uncoupled model is incapable of capturing intermittent crushing behavior and ice-induced damping. The same inconsistency is used as an explanation.