約半年ぶりに論文掲載通知が来ました1。南京航空航天大学のグループとの共著での「肉厚評価における低周波渦電流探傷法とパルス渦電流探傷法の詳細比較（Experimental investigation of low-frequency and pulsed eddy current testing in thickness measurement）」と題した論文がNondestructive Testing and Evaluaitonに掲載決定です。
この論文はコロナ時に出した「パルス渦電流探傷法の周波数リッチは情報リッチと言えるのか～数値解析による分析」の続報的なもので、筆頭著者は当時学振ポスドクとして当研究室に在籍しておりまた当該論文の共著者でもある南京航空航天大学のGe准教授です。「パルス渦電流探傷法の周波数リッチは情報リッチと言えるのか～数値解析による分析」では雑音がない理想的な信号での比較検討ということで数値解析だけで議論したのですが、今回は実際に測定試験も行って、また非磁性材料のみならず磁性材料に対しても検討を、という内容になります。また、パルス渦電流探傷法では様々な信号特徴量が提案されているのですが、どれがいいのだろう、ということにもちょっと踏み込んでいます。
何はともあれ、論文の概要は以下。
Title: Experimental investigation of low-frequency and pulsed eddy current testing in thickness measurement
Abstract: This paper experimentally compares low-frequency and pulsed eddy current testing techniques in their performance of measuring plate thickness. The experimental setups for both low-frequency and pulsed eddy current testing were built using general commercial devices. A single polarity rectangular wave with a 10 Hz pulse repetition frequency and 50% duty was utilised as the excitation wave in pulsed eddy current testing. Ferromagnetic steel and nonferromagnetic aluminium alloy plates of 1-, 3-, 5-, and 10-mm thicknesses were tested at various liftoff distances to discuss the capabilities and characteristics of both methods. Using peak amplitude, valley amplitude, time to attenuate, logarithmic slope, and reciprocal square root of the logarithmic slope as features, pulsed eddy current testing could effectively evaluate the steel plates thicker than 3 mm and all the aluminium alloy plates. However, the selection of threshold values in the time domain determining the features significantly affects the performance of pulsed eddy current testing. In contrast, low-frequency eddy current testing using a 10 Hz excitation frequency could evaluate both steel and aluminium plates using signal phase. This indicates that redundant frequency components may impair the effectiveness of pulsed eddy current testing. Moreover, low signal-to-noise ratio and simple signal processing methods struggle to detect minor signal variations. Signal-to-noise ratios should be considered in pulsed eddy current testing. Additionally, the linearity of phase versus thickness for steel is better than for an aluminium alloy, potentially indicating that ferromagnetic materials enhance the coupling between the coil and the plate.