Detection of debonding defects in honeycomb sandwich composite structures using low-power ultrasound excited thermography optimized by post-processing techniques

Honeycomb sandwich composite structures (HSCS) often develop debonding defects during manufacturing and service. Therefore, detecting these defects is crucial for ensuring structural safety, but it remains challenging. This study explores the use of low-power ultrasound excited thermography (LUET) t...

Full description

Saved in:
Bibliographic Details
Published in:Journal of thermal analysis and calorimetry Vol. 150; no. 11; pp. 8123 - 8145
Main Authors: Zhang, Yubin, Xu, Changhang, Liu, Pengqian, Liu, Rui, Zhao, Qing, Wang, Longbo, Xie, Jing
Format: Journal Article
Language:English
Published: Cham Springer International Publishing 01.06.2025
Springer Nature B.V
Subjects:
Accident prevention
Adhesives
Algorithms
Analytical Chemistry
Chemistry
Chemistry and Materials Science
Composite materials
Composite structures
Debonding
Deep learning
Defects
Finite element method
Independent component analysis
Inorganic Chemistry
Least squares method
Measurement Science and Instrumentation
Metal fatigue
Modal analysis
Nondestructive testing
Physical Chemistry
Polymer Sciences
Principal components analysis
Sandwich structures
Statistical analysis
Structural safety
Thermography
Honeycomb sandwich composite structures
Debonding defects
Low-power ultrasound excited thermography
Thermographic data post-processing techniques
ISSN:1388-6150, 1588-2926
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Honeycomb sandwich composite structures (HSCS) often develop debonding defects during manufacturing and service. Therefore, detecting these defects is crucial for ensuring structural safety, but it remains challenging. This study explores the use of low-power ultrasound excited thermography (LUET) technique for detecting debonding defects in HCSC and investigates suitable post-processing techniques to further enhance detection effectiveness. Initially, the detection mechanism of LUET for HSCS debonding defects is theoretically analyzed. A new post-processing framework, convolutional denoising autoencoder (CDAE) combined with typical post-processing techniques, is described. Subsequently, a finite element model is developed for modal analysis to estimate the local defect resonance (LDR) frequency. Then, based on the LDR frequency estimates obtained from numerical simulations, the developed experimental system was used to detect debonding defects in the fabricated HSCS sample, and the effect of the positioning of the ultrasonic transducer on the detection results was explored. Finally, various techniques, including principal component analysis (PCA), independent component analysis (ICA), and partial least squares regression (PLSR), as well as versions combined with CDAE (CDAE-PCA, CDAE-ICA, and CDAE-PLSR), are applied to process the experimental data, and then analyzed qualitatively and quantitatively. Statistical analysis demonstrates that the combination of CDAE with typical post-processing methods shows superior performance compared to the original algorithms, achieving the best CNR values in 10 out of 12 defect detection scenarios. Notably, CDAE-PLS1 showed the most substantial improvement with a 127% increase in CNR ( p  = 0.02), while CDAE-PC1 and CDAE-IC1 demonstrated improvements of 89% ( p  = 0.01) and 37% ( p  = 0.01), respectively. The results indicate that LUET can efficiently and reliably detect the debonding defects of HCSC, and further optimize the detection results through post-processing techniques. In particular, the combination of CDAE with typical post-processing methods shows superior performance compared to the original algorithms.
Bibliography:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ISSN:1388-6150
1588-2926
DOI:10.1007/s10973-025-14138-3