Study On Fatigue Damage Monitoring Of Composite Wind Turbine Blade

The thesis focuses on study of PZT-based fatigue damage monitoring technique for composite wind turbine blade, and ICA-based Blind Source Separation method with respect to multiple damage signals and noise. The main contents are organized as follows:Firstly, a displacement-controlled mechanical loading system is designed and used to implement fatigue tests of full size small scale wind turbine blades. During the experiments, structural strain and acoustic emission data are collected by strain gauges and PZT sensors respectively. Analysis results on the strain data show that there is an evidently increasing strain near the cracking area before failure of the blade. So an approach based on strain monitoring is proposed for fatigue damage monitoring of wind turbine blades.Secondly, conventional means is used for analysizing the acoustic emission data collected by the PZT sensors. Results show that some useful AE parameters containing counts, energy, amplitude and average frequency can describe characteristics in different stages of fatigue damage. Furthermore, some approaches based on the AE parameters schedule monitoring is proposed to monitor and identify fatigue damage types, such as matrix cracking, delamination or fiber fracture in wind turbine blades.Thirdly, in order to denoise and extract meaningful information of different structural damage sources from their mixed damage signals generated by different damage mechanism in Structural Health Monitoring, a method called BSS(Blind Source Separation) Based on ICA(Independent Component Analysis) is introduced in this thesis. Denoising function of this method is proved by separating of a group of signals mixed by sine signal, gaussion white noise and uniform noise. Effectivity of processing mixed structural damage signals is demonstrated by denoising and extracting independent AE signals generated with respect to different fatigue damage mechanisms of wind turbine blades.