Wind Turbine Blade Damage Detection Via 3-Dimensional Phase-Based Motion Estimation

Abstract

As a typical non-contact sensing and monitoring capability, optical measurements have been gaining tremendous attention recently in structural dynamics identification and health monitoring. This non-contact methodology usually supplies full-field measurement, and is able to out-play traditional transducers such as accelerometers, laser vibrometers and strain gauges. As a new computer vision based approach, 2-D Phase-Based Motion Estimation (PME) is extended in this paper to 3-D applications for structural health monitoring (SHM) purposes. Compared to the 2-D PME, which only captures the in-plane motions on the selected field of view of the camera, 3-D PME removes these restrictions and estimates the out-of-plane motion partially if the required conditions are met for the proposed algorithm. The modal information such as natural frequencies of the test wind turbine blade are extracted and compared with the results obtained via 2-D PME, as well as other conventional measurement approaches. With all the modes decoupled on the testing turbine blade, a contrived damage is added to the structure, and the 3-D PME proposed in this paper successfully extracts the modal frequencies for both conditions, baseline and damaged, and therefore detects damages in a non-contact and full-scale fashion.