パルス加熱後の時系列温度分布のサーモグラフィ計測ならびにデータ解析による接着部材の非破壊検査

抄録

Currently, the automobile industry is promoting weight reduction of car bodies by using high-strength thin steel sheets to improve fuel efficiency. On the other hand, since the rigidity of high-strength steel sheets alone is reduced, the weld bond method is attracting attention as a way to compensate for this. The weld bond method is a joining method that uses both adhesives and spot welding to improve the rigidity of steel plates while keeping them lightweight. The weld bond method can improve the rigidity only when the adhesive is applied to the required area. In this study, a new non-destructive inspection method using infrared thermography is developed as a high efficiency and high quality inspection method. The adhesive section is detected by measuring the difference in temperature fluctuation between the adhesive and non-adhesive section that occurs after heating the inspection target. However, the adhesive cannot be detected clearly with the time-series infrared data measured by temperature fluctuation. To solve this problem, we proposed PCA processing for time-series infrared data. The PCA process extracts each of the feature values present in the data as independent variables, and the changes in the feature values can be confirmed by images. It was found that the adhesive section, which cannot be confirmed by the infrared temperature image, can be clearly detected by the PCA processing.