A Simplified Blade Model for Reliable Seismic Assessments of Wind Turbines

Abstract

Wind turbines are dynamically complex structures. They entail slender towers, flexible foundations, and heavy rotor-nacelle-assembly (RNA). In the context of earthquake analysis, the RNA is often simplified as a rigid point mass which can suppress the modal contribution of blades from the global system dynamics and may further affect the reliability of seismic response of wind turbines. Other high-fidelity finite element (FE) blade models are difficult to implement due to the complex layup of composite materials and reduced computational efficiency. Thus, there is a need for an intermediate solution that allows the realistic and accurate consideration of blades in the seismic assessment of wind turbines. This study presents a meta-heuristic, problem-independent optimisation method, known as the genetic algorithm (GA), to identify simplified material and cross-sectional properties of a typical wind turbine blade. The properties are used to construct a simplified FE blade model that can be implemented in global wind turbine models. The optimised design solutions are examined with regards to the mechanical and dynamic response of a reference 5MW wind turbine having 61.5 m long blades. The accuracy of the modal behaviour validates the presented optimisation and simplified blade modelling approach and signifies the potential of its application in seismic assessments of wind turbines.