Detection of defects in metallic materials using an improved post-processing method of pulsed thermography

Abstract

A low-cost and easy-to-implement method based on the first-order forward difference (FOFD) and infrared thermal signal reconstruction is proposed in this work. The proposed method first adopts the FOFD method to effectively improve the image signal-to-noise ratio (SNR) and detect the defects at deeper locations. After that, an improved thermal signal reconstruction (ITSR) method is proposed to be applied to the long pulse thermography. The numerical calculation and experiment were adopted to study the effectiveness of the method at the detection of defects of metallic materials. The results indicate that the proposed FOFD method improves the SNR by 80.92% over the raw image. The error of the proposed ITSR method is less than 0.4oC and the relative error of the experimental depth is only 9.65%. Comparing with existing methods, the proposed method holds the advantage of high efficiency and simple in implementation and is therefore highly recommended for the practice use of the pulsed thermography.