Evaluation Of Structural Characteristics Of Metal Materials Based On Electromagnetic Ultrasonic Transducer

With China’s reform and opening up and the continuous development of the world economy,people’s use of metal products continues to rise.Due to the limitation of raw materials,processing technology and other technical factors,metal defects such as holes and cracks are inevitable in the production process.In addition,many metal products work in high temperature,high pressure,wet environment,in the process of use is extremely vulnerable to corrosion and wear.Therefore,regular nondestructive testing of metal structure products can prevent the occurrence of safety accidents and cause great economic losses.Although electromagnetic ultrasonic technology has many advantages,its energy conversion efficiency is very poor.How to reasonably design the permanent magnet structure and coil type of the electromagnetic ultrasonic transducer to improve the energy exchange efficiency and the signal-to-noise ratio to make the received signal more pure and improve the detection accuracy of the system are the severe challenges faced by in the current research field of EMAT.Initially,three kinds of energy transfer mechanism of electromagnetic ultrasonic transducer are introduced in this paper.Based on the energy transfer mechanism,the propagation process of shear wave and Rayleigh wave is analyzed when the EMAT is placed on the edge of the metal plate,and the influence on the echo signal of a certain tilt angle at the bottom of the metal plate and the different position of EMAT edge is also analyzed.Then,superheterodyne orthogonalization sensitive detection technique is used to analyze the received shear and Rayleigh wave packets.Finally,the measurement accuracy of metal materials is evaluated by experiments.A butterfly coil with a "mirror" symmetrical structure is designed in this paper to enhance the amplitude of the echo signal.By analyzing the static flux density,particle vibration and the Lorentz force distribution in the x direction,it is concluded that the energy transfer efficiency of the new transducer is significantly higher than that of the traditional transducer.In the frequency domain,the resonant frequency relationship between the transducer with "mirror image" symmetrical structure butterfly coil and the traditional transducer is analyzed.Finally,according to the experimental data and SNC method,the echo containing noise is processed to improve the detection effect of the new transducer.According to the fact that there is no absolutely smooth object in nature,this paper focuses on the analysis of the influence on the propagation of ultrasonic wave when the metal surface has a certain roughness.A two-dimensional Gaussian finite element model of rough surface was established to analyze the transmission,reception,reflection and transmission of ultrasonic wave on different roughness surfaces and roughness surface positions.Finally,on the basis of SVM,5000-SNAP system was used to detect rough surfaces with Ra of 0.8μm,1.6μm,3.2μm and 6.3 μm.Data processing and signal feature extraction were carried out on the time-domain echo signals and frequeny-domain signals,and then the trained SVM classifier was used for classification and recognition.