Adaptive spectral band integration in flash thermography: Enhanced defect detectability and quantification in composites

In flash thermography, the maximum inspectable defect depth is limited when only the raw thermographic sequence is analyzed. The introduction of pulsed phase thermography (PPT), in which phase (contrast) images at different thermal wave frequencies are obtained, significantly improved the maximum inspectable depth while reducing the effects of non-uniform heating and non-uniform surface properties. However, in a practical environment, the evaluation of many phase images per inspection is a cumbersome procedure. In this paper, a novel Adaptive Spectral Band Integration (ASBI) procedure is introduced for the post-processing of flash thermographic datasets, which yields a unique damage index map. ASBI integrates the most useful spectral information for each pixel individually, obtaining a maximized defect detectability and an almost zero-reference level. The performance of ASBI with respect to defect detectability as well as defect sizing and depth inversion is evaluated thoroughly with both experimentally and numerically generated datasets. The ASBI procedure is successfully applied on various composite coupons with flat bottom holes and barely visible impact damage, as well as on a stiffened aircraft composite panel with a complex cluster of production defects. The ASBI procedure is compared with existing data-processing techniques in literature, illustrating an enhanced performance.