Experimental study on the crack detection with optimized spatial wavelet analysis and windowing. (1st May 2018)

• Record Type: Journal Article
• Authors: Ghanbari Mardasi, Amir; Wu, Nan; Wu, Christine
• Is Part Of: Mechanical systems and signal processing. Volume 104(2018)
• Page Start: 619
• Page End: 630
• Publication Date: 2018-05-01
• Subjects: Crack detection -- Crack localization -- Spatial wavelet transformation -- Optimization -- Edge effect -- Windowing; Structural dynamics -- Periodicals; Vibration -- Periodicals; Constructions -- Dynamique -- Périodiques; Vibration -- Périodiques; Structural dynamics; Vibration; Periodicals; 621
• DOI: 10.1016/j.ymssp.2017.11.039
• Languages: English
• ISSNs: 0888-3270

Abstract

Highlights: Spatial wavelet optimization progress is stated for high sensitive crack localization. Small crack size (<10% of structure thickness) can be experimentally realized. Window functions are studied to detect small crack close to the measurement edges. The method can precisely detect small local perturbation in spatial signal. Abstract: In this paper, a high sensitive crack detection is experimentally realized and presented on a beam under certain deflection by optimizing spatial wavelet analysis. Due to the crack existence in the beam structure, a perturbation/slop singularity is induced in the deflection profile. Spatial wavelet transformation works as a magnifier to amplify the small perturbation signal at the crack location to detect and localize the damage. The profile of a deflected aluminum cantilever beam is obtained for both intact and cracked beams by a high resolution laser profile sensor. Gabor wavelet transformation is applied on the subtraction of intact and cracked data sets. To improve detection sensitivity, scale factor in spatial wavelet transformation and the transformation repeat times are optimized. Furthermore, to detect the possible crack close to the measurement boundaries, wavelet transformation edge effect, which induces large values of wavelet coefficient around the measurement boundaries, is efficiently reduced by introducing different windowing functions. The result shows that a small crack with depth of less than 10% of the beam height can be localized with a clear perturbation. Moreover, the perturbation caused by a crack at 0.85 mm away from one end of the measurement range, which is covered by wavelet transform edge effect, emerges by applying proper window functions.