Improving defect visibility in square pulse thermography of metallic components using correlation analysis. (15th March 2018)

• Record Type: Journal Article
• Authors: Xu, Changhang; Xie, Jing; Huang, Weiping; Chen, Guoming; Gong, Xumei
• Is Part Of: Mechanical systems and signal processing. Volume 103(2018)
• Page Start: 162
• Page End: 173
• Publication Date: 2018-03-15
• Subjects: Defect visibility -- Square pulse thermography -- Metallic component -- Correlation function; Structural dynamics -- Periodicals; Vibration -- Periodicals; Constructions -- Dynamique -- Périodiques; Vibration -- Périodiques; Structural dynamics; Vibration; Periodicals; 621
• DOI: 10.1016/j.ymssp.2017.09.030
• Languages: English
• ISSNs: 0888-3270

Abstract

Highlights: A novel approach was developed to improving defect visibility in SPT of metallic components. Correlation function of thermal contrast (CFC) was defined and investigated for the first time. A peak lag time image is constructed by the proposed approach. The efficacy of the proposed approach was verified on experimental data. Abstract: Infrared (IR) thermography has gained wide applications as an important non-destructive testing (NDT) technique. Improving defect visibility is critical to achieving an accurate detection result through IR thermography. In this study, we propose a novel approach to improving defect visibility in square pulse thermography (SPT) of metallic components. In the proposed approach, the correlation function of contrast (CFC) is defined for the first time. Based on the theories of heat conduction and of correlation analysis, the differences of CFC between defects and sound regions are determined. We found that the peak lag time of the CFC is an effective feature for discriminating defects and sound regions in SPT. A new image is then constructed using the peak lag time of the CFC to improve defect visibility. To verify the efficiency of the proposed approach, an experiment was conducted on a steel specimen, and the principle component analysis (PCA) and the presented approach were compared. The results show that through the proposed approach, defects in metallic components can be indicated more clearly and detected more accurately.