Data processing methods for thermographic NDT with localised cryogenic cooling

Infrared thermographic non-destructive testing (NDT) is an effective technique for detecting defects in composite materials, particularly when paired with active thermal stimulation. This study investigates the use of localised cryogenic cooling as an alternative stimulation source in infrared thermographic NDT of fibre-reinforced polymer (FRP) composites. A low-cost handheld cryosurgical device (Histofreezer) was used to deliver precise, repeatable cooling, generating distinct thermal gradients recorded by a micro-bolometric camera. While localised cooling offers unique advantages, its strong non-uniformity complicates data processing and interpretation. This work focuses on identifying effective techniques for defect detection under such conditions. Experiments on carbon fibre reinforced polymer (CFRP) specimens impacted at ambient and cryogenic temperatures used contrast-to-noise ratio (CNR) as the performance metric. Wavelet transforms and polynomial background subtraction proved most effective in isolating defect signals; the former excelled when defect and cooling signals overlapped, while the latter worked best when cooling occurred away from defects. Combining polynomial background subtraction with principal component analysis (PCA) further improved contrast and segmentation quality. Localised cooling demonstrates strong potential as a precise, controllable, and low-cost stimulation method for thermographic NDT, supporting the development of portable and operator-safe field inspection systems.