Reduced-Order Modelling of Nonlinear Wind Turbine Blade Vibrations in a High-Fidelity Unsteady Aeroelastic Framework

Abstract. Accurate modelling of nonlinear wind turbine blade dynamics is essential for the reliable design and operation of modern wind turbines. Wind turbine blades have complex shapes with spatially varying sectional properties, twists and curvatures. The dynamics of long, flexible blades are strongly influenced by geometric nonlinearities, centrifugal effects, and material anisotropies. Conventional linear beam models are inadequate for capturing these complexities, since they are not designed to handle large deformations and nonlinear couplings. This study presents a high-fidelity nonlinear reduced-order model (ROM) for predicting the structural dynamics of modern wind turbine blades. The ROM uses a biorthogonal modal basis derived from the intrinsic mixed formulation of geometrically exact beam theory. The structural model is coupled to a high-fidelity unsteady vortex lattice solver in a two-way fluid-structure interaction framework, improving on the traditional blade element momentum method. The proposed model was validated using numerical and experimental results from the NREL 5-MW and SNL CX-100 wind turbine blades. Benchmark tests demonstrate that the ROM successfully captures the wind turbine blade’s nonlinear frequency response to harmonic loads and dynamic response to transient loads, including extreme wind gust, wind shear, and turbulence. Among the wind load cases tested, the ROM’s maximum out-of-plane tip displacement error was 0.77 % of reference tip deflection, compared with 20.90 % for conventional linear time-invariant models. This framework has applications in wind turbine control, structural health monitoring, fatigue life studies, and digital twins.