Artificial Time Histories of Wind ActionsFor Structural Analysis of Wind Turbines

Abstract

A computational method for generating artificial time histories of wind loads on wind turbine towers is presented, based on wind turbine aerodynamics and a wind field model. First, the forces acting on the blades parallel and perpendicular to the rotor’s plane are calculated according to aerodynamics theory for given mean wind velocity. Due to the blades’ aerodynamic behavior the inflow wind velocity is transformed into the relative wind velocity, depending on the axial and tangential flow induction factors, the blades’ angular velocity and their radius. Then, the forces acting on the blades are calculated combining relative velocity with two-dimensional aerofoil coefficients depending on the blades’ geometry and crosssection. The flow induction factors are estimated by an iterative process taking into account the flow angle between the relative wind velocity and the rotor’s plane and the aerofoil coefficients. Next, the turbulence component of the wind is determined by the stochastic theory, in order to describe the total wind field model and compute more realistic wind induced actions on the blades. Each fluctuating component is modeled as Gaussian, stationary stochastic process with zero-mean value and is completely characterized by the correlation matrix in time domain or the power spectral density matrix in frequency domain. Wind time histories are simulated by the decomposition of the power spectral density matrix. Then, finite element models of the wind turbine tower are subjected to the load time histories derived above and dynamic analyses are performed. Ultimate objective of this research is to study fatigue at bolted and welded connections between adjacent parts of wind turbine towers and to investigate the importance of dynamic effects on local buckling of the tower shell.