Parked and idling are two different operating strategies when the wind turbine is shut down. When the wind inflow velocity is bigger than the cut-out velocity, the wind turbine would enter either one of the two conditions depending on the prescribed control strategy by the manufacturers to protect itself from the strong wind. Herein, to ensure excellent aeroelastic stability behavior of a wind turbine is of great significance.In this thesis, an aeroelastic computation tool named GAST written with FORTRAN programming language is utilized to carry out time-domain numerical simulations and aeroelastic stability analysis for full-scale wind turbines. By observing the load amplification range and its corresponding local angle of attack and local lift and drag coefficient, it is found that most of the angle of attacks are among the stall region, which is strong indication of stall-induced vibrations. Then quasi steady-state aerodynamics and unsteady aerodynamics are respectively incorporated into the aeroelastic stability analysis to derive the solution of logarithmic decrement and aerodynamic work, both of the two representing the damping of the modes. Two methods are adopted for the determination of the aeroelastic instability: eigenvalue method and aerodynamic work method. By carrying out Fast Fourier Transformation of the time-domain loading signals, the correctness of the two independent calculations, namely time-domain simulations and aeroelastic stability analyses, can be validated. Since the aerodynamic work method has taken the non-linear effects of the unsteady ONERA aerodynamic model into account, therefore it is more reliable than the eigenvalue approach.This aeroelastic computation program is able to carry out fast, comprehensive and effective prediction of aeroelastic stability behavior of a full-scale wind turbine in parked and idling conditions, which could offer guidelines for the aeroelastic stability optimization. |