收录:
摘要:
Harmonic eddy current testing (HECT) is an electromagnetic non-destructive testing method, which can realize non-contact detection of metal pipeline defects. Eddy current will be produced at the metal pipeline affected by harmonic excitation with multiple frequencies. Meanwhile, the eddy current generated would be uneven due to the disturbance of the defects. In order to detect the abnormalities effectively, the relationship model among the diameter of the pipeline the lift-off height of the probe and the angle of excitation coils was established within this paper. The effect of the magnetic field focusing using different excitation coils models were studied by finite element simulation and the optimal structure of the harmonic excitation coils model was finally achieved. The experimental platform was established, and three types of defects were designed on the top and left side of the pipeline, including circular holes, rectangular grooves, and 45-degree rectangular grooves. The harmonic eddy current magnetic field signals were acquired by the 3-axis Tunnel Magneto-Resistance (TMR) sensors array. The frequency bands of interference and original excitation signal were removed by using Multi-Resolution Analysis (MRA), and thus, the useful signal frequency bands were selected for fusion. The fused signals were input into the bistable Stochastic Resonance (SR) system to realize the noise energy transfer, which was analyzed by the fourth-order Runge-Kutta algorithm and the scale transformation, so as to enhance the energy of the signals at the pipeline defects. The results show that the designed harmonic excitation coils model improves the excitation effect of both sides of the pipeline and expands the coverage of the flaw detection of the pipeline. Besides, the signal processing method proposed within this paper can extract the pipeline defects signal effectively and improve the ability of defects identification. The above research can be helpful on the non-contact non-destructive testing of metallic materials. © 2019 IEEE.
关键词:
通讯作者信息:
电子邮件地址:
来源 :
ISSN: 1559-9450
年份: 2019
卷: 2019-June
页码: 700-704
语种: 英文