A pilot study using flying spot laser thermography and signal reconstruction

Active thermography (AT) has emerged as a critical non-destructive testing and evaluation (NDT&E) technique for identifying subsurface defects in diverse industrial materials. Despite its widespread application, AT faces challenges such as inadequate heat input, noisy thermal signals, and non-uniform heating, which can obscure defect detection. This study introduces an enhanced approach leveraging flying spot laser thermography combined with advanced signal processing techniques to address these challenges. A meticulously designed calibration block, embedded with 180 spherical and rectangular notch defects of varying depths and dimensions, was fabricated using 3D printing to serve as the experimental model. The laser-induced thermal data were acquired at three distinct scanning speeds and underwent temporal alignment to synchronize heating events across all pixels. Principal Component Analysis (PCA) and Independent Component Analysis (ICA) were subsequently applied to the aligned datasets to extract and isolate defect-related thermal signatures. PCA effectively reduced data dimensionality and highlighted major thermal diffusion patterns associated with significant defects, particularly notched anomalies. However, its sensitivity diminished for smaller or deeper defects. In contrast, ICA provided more refined separation of thermal signals, enhancing defect visualization and contrast, especially at slower scanning speeds where higher heat input improved thermal differentiation. Notably, ICA demonstrated superior performance in isolating notched defects compared to spherical ones due to pronounced thermal gradients. The findings underscore the potential of combining flying spot thermography with PCA and ICA to enhance defect detection and characterization in NDT&E applications. Future work will focus on optimizing scanning parameters through simulation models and integrating machine learning algorithms to further improve the detection of smaller and shallower defects, thereby advancing the precision and efficacy of thermal analysis techniques.