Probability of detection of deep defects in steel samples using Barker coded independent component thermography

Abstract

Barker coded independent component thermography (BCICT) approach known for exploiting the pulse compression properties with independent component analysis method was used to examine a mild steel sample with drilled flat-bottomed holes at various depths. This Letter emphasises the application of the probability of detection (POD) as a predictive assessment tool for the efficacy of the different methods to detect the defects at varying depths. The results exhibited that the aspect ratio of the deep defect r90 with 90% POD and r90/95 for 90% POD with 95% confidence that can be detected using the BCICT approach are 3.161 and 3.835, respectively. Besides, a relation between the POD of the defect at different depths as a feature of the defect aspect ratio (diameter/depth) has also been presented