Dynamic modelling of floating wind turbines

Thesis

Floating Wind Turbines (FWTs) offer a promising solution to harnessing substantial offshore wind energy in deep waters. However, accurately modelling their complex dynamics is challenging due to the complex coupling between aerodynamics, hydrodynamics, structural elasticity, and controls. This thesis proposes an integrated FWT simulation framework in Simulink to investigate these dynamics, with a particular focus on nonlinear wave-platform interactions induced by large platform motions.

A comprehensive FWT model is developed in Simulink, featuring efficient state-space representations for linear wave radiation and excitation effects, a multibody formulation to account for large geometrically nonlinear blade deformations, a modified blade-element momentum method for unsteady aerodynamics, and a multi-loop ROSCO control strategy for power regulation and platform stabilization. Validation against the widely-used OpenFAST simulation tool demonstrates the ability of the developed Simulink model to capture the dominant FWT dynamic couplings under realistic environmental conditions.

To address the limitations of the conventional linear hydrodynamic model, which assumes small platform motion around the equilibrium position, an analytical solution to the 2D linear wave-platform boundary value problem (BVP) linearized at an arbitrary platform pose is proposed. The analytical method is compared against a numerical boundary element method, and the influence of platform pose on the hydrodynamic behaviour is investigated.

Additionally, a novel linear parameter varying (LPV) modelling framework is developed to capture the geometrically nonlinear wave radiation effect. The LPV system is constructed using state-space models identified from the BVPs linearized at various platform positions, with the instantaneous platform pose used to interpolate the state-space matrices. To ensure state-basis coherency across all linear models, a black-box method based on the balanced realization and a gray-box method that ties the state vector to physical parameters are proposed. Both LPV models are validated on the benchmark van Daalen floating cylinder, and they are then integrated into the Simulink FWT model. Results indicate that the LPV models effectively capture nonlinear hydrodynamic effects and offer significant computational efficiency, making them practical for early-stage FWT design and optimization.

Authors: Meng, J

• DOI: 10.5287/ora-vy86kjg5y
• Type of award: DPhil
• Level of award: Doctoral
• Awarding institution: University of Oxford
• Language: English
• Keywords: wave-platform interaction; Matlab/Simulink; dynamic modelling; linear parameter-varying system; floating wind turbines
• Subjects: Structural dynamics; System identification; Fluid-structure interaction