Defect identification in GRID-LOCK® joints

Abstract

Bonded metallic GRID-LOCK ® structures are being adopted for a variety of aerospace applications due to their structural efficiency and damage tolerance. The development of non-destructive evaluation (NDE) methods is necessary to identify bond defects that can lead to failures in these structures. However, this task is complicated by the lack of interior access and complex geometry of GRID-LOCK ® components. In this dissertation, the feasibility of various NDE techniques for detecting the existence, location, and extent of bond defects in GRID-LOCK ® joints is investigated. Experiments are conducted on customized test structures to compare the effectiveness of optical NDE, ultrasonic C-scans and vibration-based damage detection. Finite element analysis (FEA) is used to interpret experimental results and highlight the advantages of candidate methods. The qualitative effectiveness of optical NDE is further investigated using full-field surface slope measurements (shearography). Because accurate characterization of structural defects is critical to flight safety, a quantitative non-destructive evaluation (QNDE) method using artificial neural networks (ANNs) is developed. This method involves the use of radial basis function networks (RBFNs) trained and validated using FEA simulation data. The effectiveness of this QNDE approach is demonstrated using experimental data from a custom-built optical scanning system.ii My dad has never failed to inspire me and my mom has always supported me, so this one is for them.

Shearography technology and applications: a review

“…Other related investigations using shearography that have been reported in recent publications include the identification of defects in GRID-LOCK R joints [114] and analysis of strain distributions around the inserts in composite plates [115]. GRID-LOCK R is a novel method of joining structural components that was developed by the Goodrich Corporation and used recently in the F-15E fighter aircraft.…”

Simulation Dataset Preparation and Hybrid Training for Deep Learning in Defect Detection Using Digital Shearography

“…Since defect detection is a typical binary classification problem, artificial intelligence techniques are applied to locate the butterfly pattern. As an example, artificial neural network architectures have been combined in defect detection using digital shearography in some applications [18][19][20]. Moreover, with the great achievements of deep learning in image understanding and object detection in recent years, elaborate deep learning-based algorithms have also been explored in digital shearography, especially the most popular convolutional neural networks (CNNs) [21][22][23].…”