Session: 04-01: Emerging Techniques & Technology

Paper Number: 147459

147459 - Extracting Sub-Pixel Displacement Measurements Using Visual Vibrometry for Defect Detection 

Abstract: 

This paper focuses on the field of visual vibrometry and its application to NDE. The capabilities of visual vibrometry were explored by using standard off-the-shelf cameras to measure the displacement of a structure when excited. Visual vibrometry works by measuring the relative changes in pixel values across a field of view. It is the use of cameras to collect spatially-dense full-field measurements that makes visual vibrometry attractive as a method for rapid, non-contact, large area SHM.

This study adopts a phase-based motion estimation technique to extract precise sub-pixel displacement measurements, thus enabling the assessment of the visual system’s sensitivity and its defect detection capability. Visual vibrometry as a method for NDE makes use of the fact that damage alters vibrational modes. The presented work demonstrates how subtle structural changes in structural response, imperceptible to the human eye, can be discerned through image processing.

We show through experimental approaches and finite element analysis how visual vibrometry methods can effectively detect and quantify damage on structures by measuring sub-pixel displacements. Furthermore, enhancements to the visual system's capabilities are proposed by optimising test parameters and sample rates to capture higher frequency excitations at sub-Nyquist frame rates. This approach aims to broaden the system's capabilities by facilitating the detection of higher-order mode shapes, thereby advancing its ability in modal defect detection applications.

In summary, this paper explores the potential of visual vibrometry as a versatile tool for NDT. The presented work offers insights into its applications in damage detection and showcases strategies for enhancing its performance in capturing structural dynamics.

Presenting Author: Lucy Dougill University of Bristol

Presenting Author Biography: A research student with an interest in Non-Destructive Testing.
My current work involves the application of visual systems, computer vision algorithms and vibration techniques to utilise visual vibrometry for NDT applications.
Particularly looking at how the sensitivity of the visual methods and their sampling capabilities may be pushed to their limit in order to measure small displacements arising from useful ultrasonic frequencies for the purpose of defect detection.
Based at the University of Bristol and part of the Future Innovations in NDE CDT (FIND).
Supervisors Prof. Paul Wilcox and Prof. Anthony Croxford.

Authors:

Lucy Dougill University of Bristol