Extracting full-field dynamic strain on a wind turbine rotor subjected to arbitrary excitations using 3D point tracking and a modal expansion technique

Abstract

a b s t r a c tHealth monitoring of rotating structures such as wind turbines and helicopter rotors is generally performed using conventional sensors that provide a limited set of data at discrete locations near or on the hub. These sensors usually provide no data on the blades or inside them where failures might occur. Within this paper, an approach was used to extract the full-field dynamic strain on a wind turbine assembly subject to arbitrary loading conditions. A three-bladed wind turbine having 2.3-m long blades was placed in a semi-built-in boundary condition using a hub, a machining chuck, and a steel block. For three different test cases, the turbine was excited using (1) pluck testing, (2) random impacts on blades with three impact hammers, and (3) random excitation by a mechanical shaker. The response of the structure to the excitations was measured using threedimensional point tracking. A pair of high-speed cameras was used to measure displacement of optical targets on the structure when the blades were vibrating. The measured displacements at discrete locations were expanded and applied to the finite element model of the structure to extract the full-field dynamic strain. The results of the paper show an excellent correlation between the strain predicted using the proposed approach and the strain measured with strain-gages for each of the three loading conditions. The approach used in this paper to predict the strain showed higher accuracy than the digital image correlation technique. The new expansion approach is able to extract dynamic strain all over the entire structure, even inside the structure beyond the line of sight of the measurement system. Because the method is based on a non-contacting measurement approach, it can be readily applied to a variety of structures having different boundary and operating conditions, including rotating blades.

Estimation of full‐field dynamic strains from digital video measurements of output‐only beam structures by video motion processing and modal superposition

“…Furthermore, the experimental modal coordinates q i ( t ) estimated from the video measurement of the operating structure, hence the dynamic strain ϵ ( x , t ) (by Equation ), carry the damping and excitation or operational information of the structure, which may not be obtained from the FEM itself. Therefore, by making full use of the full‐field experimental (video) measurement and integrating it with the high‐fidelity FEM updating process, the proposed method distinguishes itself from other existing modal‐ and model‐based full‐field strain estimation methods such as Baqersad et al that involves only a limited set of experimental measurements using a few optical markers on the structure.…”

Video-based multiscale identification approach for tower vibration of a cable-stayed bridge model under earthquake ground motions

“…The KLT algorithm is not robust in a complex scene, although the KLT algorithm is accurate under a stable scene. The DIC (digital image correlation) and the three‐dimensional point‐tracking techniques have been used in experimental modal analysis . DIC approach can provide full‐field displacement of a planar structure using two cameras by comparing a pair of digital images taken before and after deformation.…”

“…The DIC (digital image correlation) and the three-dimensional point-tracking techniques have been used in experimental modal analysis. [16][17][18][19][20] DIC approach can provide full-field displacement of a planar structure using two cameras by comparing a pair of digital images taken before and after deformation. The advantage of this algorithm over other algorithms is the full-field measurement with high accuracy, whereas the disadvantage of the algorithm is its sensitivity to image noise.…”

Spatiotemporal video-domain high-fidelity simulation and realistic visualization of full-field dynamic responses of structures by a combination of high-spatial-resolution modal model and video motion manipulations

“…As an alternative noncontact measurement method, digital video cameras are relatively low cost, agile, and provide simultaneous, high-spatial resolution measurements. Combined with image processing algorithms (e.g., image correlation [8] or point tracking and optical flow [9] ), video camera-based measurements have been successfully used for structural dynamic response measurement [10][11][12][13][14][15] and experimental modal analysis [16][17][18][19][20][21] to obtain full-field mode shapes. One issue associated with these methods is that they typically require speckle pattern or high-contrast markers to be placed on the surface of structures, which induces the surface preparation and target installation issue and is less feasible when the measurement area is large or inaccessible.…”