A Computationally-efficient Frequency Domain Model of a Floating Wind Turbine with Hull-based Tuned Mass Damper Elements

Floating offshore wind turbines (FOWTs) offer a means to capture an abundant offshore wind resource in intermediate to deep waters where wind energy density is stronger and there are fewer competing use issues. However, FOWTs often require substantial floating foundations and mooring systems to support the wind turbine which increases the levelized cost of energy (LCOE) produced by the unit. As such, foundation technologies that minimize turbine motions and structural loads while enabling smaller, more economical platforms and mooring systems are of great interest and a key ingredient in achieving a commercially-competitive LCOE. One such technology that can result in a lighter foundation is the use of tuned mass dampers (TMDs). These dampers can be feasibly achieved in a number of ways, for example, by employing on-board ballast water as the mass element of the TMD. However, designing a system of this type with additional degrees-of-freedom (DoF) is not trivial, and new simulation tools are needed that can be used to not only analyze such configurations, but also facilitate rapid optimization of these designs. To that end, this paper presents a computationally-efficient frequency domain model for FOWTs with multiple hull-based TMDs. DoF for the platform translational motions, TMD motions and tower deformation are included. Wave forcing is computed via a response surface model created using simulation results from the potential flow solver, WAMIT. Wind forcing is applied using wind speed to aerodynamic load transfer functions derived from the time-domain FOWT simulator, OpenFAST. Using response amplitude operators (RAOs) derived from the Frequency Domain model in conjunction with wind and wave forcing spectra, one can quickly compute response spectra, fatigue and ultimate load estimates for the FOWT. Comparisons of the frequency domain model-computed values with those obtained from a validated OpenFAST model capable of simulating the effect of hull-based TMDs shows very good agreement. Additionally, the frequency domain model is shown to take significantly less computational effort to obtain reliable preliminary design values than the time-domain OpenFAST simulator.