Determining the effect of velocity on sensor selection and position in non-destructive testing with magnetic flux leakage method: a pipe inspection gauge design study with Ansys Maxwell

Abstract

The magnetic flux leakage (MFL) method, which is among the non-destructive testing (NDT) methods, is frequently used in the determination of discontinuity in ferromagnetic materials. The method focuses on the detection of the MFLs in the discontinuity regions of magnetized samples with the use of magnetic sensors and the analysis of the detected MFL signals. The characteristic of the MFL signal is directly related to the selection of sensors. Sensor selection, on the other hand, depends on the magnitude of the MFL in the discontinuity region. The variation of MFL in the discontinuity region varies with the distance of the discontinuity to the magnetizer (yoke). In this case, the characteristic of the MFL signal is also closely related to the position of the sensor during measurement. Furthermore, the magnetization time of the material has no impact on the characteristics of the MFL signal, but affects its amplitude. As such, the primary goal of this study is to determine the sensor position for the MFL method, which yields the optimal MFL signal at a constant magnetization velocity for natural gas pipes, a ferromagnetic product, using the ANSYS Maxwell simulation software on a 2-D transient model. Subsequently, the effects of crack geometry and magnetization velocity on the MFL signal were examined in the model. Last, the study focused on the selection of potential sensors which can be utilized in the model for the optimal sensor position determined. ANSYS Maxwell simulation results suggest that the sensor needs to be positioned at the exact center of the magnetizer and at the closest distance to the material. It was determined that increased magnetization velocity results in a decrease in MFL signal amplitude. Accordingly, a pipeline inspection gauge (PIG) device was 3D-designed and produced using the ANSYS Maxwell simulation software, with the optimal magnetization velocity (PIG velocity), optimal sensor position, and sensor selection. By using the produced PIG and the developed measurement system, experiments were carried out at different velocity values, and the velocity-dependent variation of the MFL signal was experimentally investigated.