ABSTRACT For enhanced detection of flaws in engineering components using magnetic flux leakage (MFL) technique, measurement of the leakage magnetic field components along the three perpendicular directions is beneficial. This article presents the three dimensional-magnetic flux leakage (3D-MFL) modeling and experimental studies carried out on carbon steel plates. Magnetic dipole model has been used for the prediction of MFL signals and images. Sensitivity of the MFL signals peak amplitudes of tangential (HX), circumferential (HY), and normal (HZ) components with respect to flaw length, width, depth and lift-off have been studied. A 3D-GMR sensor has been used for simultaneous measurement of all the three components of leakage magnetic fields from surface flaws in 12 mm thick carbon steel plates. The experimental MFL images have been compared with the model predicted MFL images. The sensor has shown the capability to detect and image 0.9 mm deep surface flaws with a signal to noise ratio of 8 dB. Principal component analysis (PCA)-based image fusion has been performed for fusion of the 3D-MFL images to obtain a geometrical profile of the flaws. Study reveals that 3D-GMR enhances the capability for detection of flaws having irregular geometries.

中文翻译：

---

用于增强碳钢板缺陷检测的 3D 漏磁检测建模与实验研究

摘要 为了使用漏磁 (MFL) 技术增强对工程部件缺陷的检测，测量沿三个垂直方向的漏磁场分量是有益的。本文介绍了在碳钢板上进行的三维漏磁 (3D-MFL) 建模和实验研究。磁偶极子模型已用于预测 MFL 信号和图像。已经研究了 MFL 信号切向 (HX)、周向 (HY) 和法向 (HZ) 分量相对于缺陷长度、宽度、深度和剥离的峰值幅度的敏感性。3D-GMR 传感器已被用于同时测量 12 毫米厚碳钢板表面缺陷的漏磁场的所有三个分量。已将实验 MFL 图像与模型预测的 MFL 图像进行了比较。该传感器已显示出能够以 8 dB 的信噪比检测和成像 0.9 毫米深的表面缺陷。已执行基于主成分分析 (PCA) 的图像融合以融合 3D-MFL 图像以获得缺陷的几何轮廓。研究表明，3D-GMR 增强了检测具有不规则几何形状的缺陷的能力。