Modeling fidelity in floating offshore wind turbine simulation: from load assessment and uncertainty quantification to integrated control co-design

Wind turbines have become essential contributors to power grids around the world. This development was enabled in large part by a continuous growth of the heights and diameters of the turbines. Today, this trend is still ongoing. The sheer size of modern wind turbines challenges many aspects related to their development involving manufacturing, logistics, environmental considerations and, not least, their design. The introduction of floating offshore wind turbines enables energy harvesting in deep waters, but at the same time it compounds these challenges. One central novelty related to floating turbines is their responsive nature to wind and wave loads, causing displacements and rotations of the entire system that are not known from turbines clamped to the ground. The ability to accurately predict their coupled system response to wind and wave excitation is essential to enable cost-effective but robust designs. A complete model must incorporate the elastic response of the rotor blades and tower, the drive train dynamics, the motion of the substructure, the dynamics of the mooring system and the controller actuation. It is the complexity of this coupled system that, in combination with the turbine sizes, suggests that traditional assumptions made by industry standard design tools are no longer valid. In consequence, designs that rely on such methods do not fully exploit their potential. The early stage design process traditionally relies on reduced-order methods that only capture the global response of the system. These methods are accompanied by engineering models that resolve the nonlinear time-domain response. Such methods have recently been integrated into design frameworks to include time-domain simulation into the design process. Currently, such engineering tools usually rely on blade element momentum aerodynamics combined with multi-body structural dynamics that utilize diagonal beam elements. This beam formulation does not reflect the anisotrope characteristics of modern, aero-elastically tailored wind turbine blades, while the momentum based aerodynamics fail to fully represent the increasingly unsteady environment in which these floating turbines operate. Instead, lifting-line free vortex wake aerodynamics, with its explicit treatment of the wake, is able to accurately model the influence of more modern design features of wind turbine blades and yields less conservative estimates of fatigue and ultimate loads. Timoshenko beams that use fully populated stiffness and inertia matrices resolve structurally coupled degrees of freedom and thereby improve the capture of the structural response of blades. The QBlade wind turbine simulation code includes both methods and combines them with a recently developed hydrodynamic module. This dissertation encompasses relevant demonstration cases of these methods relevant for simulating the next generation of wind turbines. In addition, a research gap concerning the impact of aero-elastic modeling fidelity on design outcomes is addressed. In particular, QBlade is interfaced with an open-source optimization framework, which enables the application of free vortex based aerodynamics and the advanced beam model in context of control co-design of floating wind turbines. In a case study that uses the developed interface the impact of varying aero-elastic fidelity on the sizing of a substructure and the simultaneous optimization of the wind turbine controller shows that better design candidates are feasible if higher fidelity methods are used.