Abstract
In the present work, the step heating thermography technique has been used to detect disbond defects at the edge of the adhered carbon/epoxy patch on the aluminum plate. The four common post-processing algorithms including principle component analysis (PCA), thermal signal reconstruction (TSR), Fourier transform (FT), and one-dimensional wavelet transform (1D WT) have been applied to the infrared thermal images obtained from the composite patch thermography inspection. A novel post processing method based on the area under temperature-time curve (Thermal Signal Area: TSA) has been proposed for the infrared thermal image analysis. The proposed post-processing method is fast and non-sensitive to noise. A comparison of the new proposed method with the four common post-processing methods has been carried out on disbond detection in a composite patch in a quantitative and qualitative manner. Furthermore, a data fusion of the proposed method with the 1D wavelets transform has been applied to disbond defect characterization which improves the accuracy by decreasing noise level.
Original language | English |
---|---|
Article number | 102153 |
Journal | NDT&E International |
Volume | 107 |
DOIs | |
Publication status | Published - 2019 |
Externally published | Yes |
Keywords
- Infrared thermography
- Image processing
- Wavelet transforms
- Composite patch
- Data fusion