Research On Digital Image Correlation And Applications In Health Diagnosis Of Wind Turbine Blades

Wind power is a reliable and clean energy that is quickly developing around the word for improvement of the environment,as well as for further development of the economy and society.However,the maintenance of wind turbines is costly.In general,health diagnostics of wind turbines is based on contact measurement.Non-contact detection method could significantly reduce maintenance costs.One option is the digital image correlation(DIC)technique,which is a full-field,non-contact,optical method used to measure the surface displacement and strain of an object as it deforms.DIC has many distinct advantages such as the simple experimental arrangement,high accuracy and wide range of resolution.As such,it has been widely used in many fields that include structural problems in the auto industry,aerospace field,bioengineering,etc.However,there are numerous challenges in the application of DIC for evaluating wind turbine blade condition because some key technical challenges need to be solved.This thesis aims to study the application of DIC to wind turbine blade health diagnosis.A real-time DIC technique for measuring strain and a full-field DIC method for overcoming large rotation are developed.Using the proposed method,rotating wind turbine blades are measured and the full-field deformation is obtained successfully.The main contents of this thesis include the following:(1)A fast,real-time DIC algorithm for displacement or strain measurement for dynamic tests is proposed.The method can reach a rate of 60 fps(frames per second)for strain measurement under cyclic loading with equivalent accuracy compared with the conventional DIC algorithm.Furthermore,a real-time 3D-DIC method is obtained by combining a fast correlation matching strategy.(2)A DIC algorithm considering rotation for overcoming de-correlation in the case of deformation with large rotation is proposed.The method could measure full-field deformation with arbitrary rotations.(3)In order to meet the need of measurements with large field of view,a calibration method based on the speckle pattern of the object is described in detail.This calibration method is convenient and reliable to calibrate the dual-camera system used in DIC.It is suitable for large field measurements because it only needs to match the speckle pattern between the adjacent left and right image.(4)A fault detection method based on three-dimensional motion for wind turbine blades is proposed.Two kinds of imbalance faults of blade are simulated in the laboratory.The feasibility of the method is verified by measuring three-dimensional motion and its spectrum analysis.(5)A fault detection method based on the relative deformation of wind turbine blades is proposed.A crack fault is simulated in the laboratory and the relative full-field deformation of the blade is measured.Moreover,the strain characteristics are analyzed and extracted.The method is applied to measurement and health diagnosis of 5kw wind turbine blades.When used together,the proposed methods have great potential to play an important role in the on-line detection of large-scale wind turbine systems.