Abstract
Non-destructive testing (NDT) plays a crucial role in ensuring the structural integrity and safety of industrial facilities and components. Long pulse thermography (LPT), a form of active thermographic testing (ATT), has gained attention for its ability to detect subsurface defects efficiently. However, non-uniform thermal excitation and environmental noise often degrade the accuracy of defect detection. This study proposes an advanced thermographic inspection technique incorporating a halogen array (HA) lamp and a compensation methodology to enhance the reliability of defect detection. Two compensation methods, namely absolute temperature compensation (ATC) and temperature rate compensation (TRC), were developed to correct non-uniform thermal loads and improve the defect contrast. Experimental validation was conducted on A-type and B-type mock-up specimens with artificial subsurface defects (10-90% depth). The results demonstrated a significant enhancement in the signal-to-noise ratio (SNR), reaching up to a 42 dB improvement in severe defects. Furthermore, a quantitative evaluation method was proposed using SNR-based defect depth estimation models, improving the accuracy of defect sizing. This approach eliminates the need for complex amplitude and phase transformations, enabling direct defect assessment from temperature thermograms.