Defect Visualization and Depth Quantification in Scanning Induction Thermography

Abstract

Nowadays, scanning induction thermography (SIT) could be efficient and practical for the inspection of elongated ferromagnetic specimens in the manufacturing process. Owing to the relative movement between the specimen and excitation coil, defect visualization and depth quantification in SIT remain challenging. Focusing on these two challenges, surface-breaking notches with different depths in a steel plate at different scanning speeds have been studied in this paper. A visualization method, including image reconstruction and Wiener filter, is proposed to realize defect visualization within 600 mm/s. Then, thermal features are extracted to realize depth quantification for notches within 3.0 mm. In addition, the effect of scanning speed on the SIT detectability is investigated and it is found that the signal-to-noise ratio (SNR) decreases significantly with increasing scanning speed.