Research On Detection Method Based On Integrated Magnetically Induced Impedance And Guided Wave

In the field of transportation,aerospace and other mechanical fields,the plate structure is widely used,such as aircraft wing,train body and so on.When affected by external factors,some parts of the structure will be easily damaged and the structural performance will be damaged.For plate structure,engineers put forward various health monitoring methods,such as detection of guided ultrasonic wave,electromechanical impedance and other methods.However,these detection methods have some limitations when they are used alone.The detection of ultrasonic Lamb wave has the advantages of far propagation distance,and accurate damage localization,but with low local sensitivity,it is not ideal for the identification of damage extent and type.Such as impedance detection,although with high local sensitivity,but the detection range is limited and the damage localization is difficult.In order to solve this problem,a kind of approach combining impedance with guided wave is proposed for realizing the accurate localization of damage and identifying the extent and type of damage qualitatively.In addition,the neural network data processing technology is applied to the signal analysis and processing of impedance detection in order to realize accurate quantification of the damage extent of plate.The results show that the neural network can improve the damage identification ability of impedance detection.The main work and results of this paper are as follows:(1)The propagation mechanism of Lamb wave is thoroughly analyzed,and the two basic forms of Lamb wave(symmetric mode and antisymmetric mode)are decomposed.The Raleigh-Lamb frequency equation is used to solve the dispersion curve of Lamb wave in the aluminum plate.The relationship between the group velocity,phase velocity and the variation mode with the frequency thickness product is illustrated by the diagram.(2)The magnetostrictive effect of ferromagnetic material is explained in detail.Based on this effect,the theoretical formula for calculating the magnetoresistance of the ferromagnetic material is derived.Then,the equivalent conversion of the theoretical calculation formula is carried out,and the measurement formula of the magneto-mechanical impedance for practical application is derived according to the properties of the transducer(such as resistance,inductance and magnetoelectric coupling coefficient etc.).(3)The mechanism,experimental instrument,test method and signal processing method of Lamb wave combined with magneto-mechanical impedance are described in detail,and a comprehensive damage index is proposed to achieve the data fusion of two kinds of test signals.The index integrates the respective damage characteristic exponents when the two methods detected separately.According to the numerical characteristics of the index,we can find out the relationship between the index and the damage attribute(such as type,extent,etc.).(4)The simulation study of Lamb wave and magneto-mechanical impedance detection is carried out respectively.By stimulating the aluminum plate with different types and size defects,the response signal is obtained and converted to calculate the damage index.The damage index is substituted into the algorithm of comprehensive damage index,and the comprehensive damage index is obtained.Finally,the comprehensive damage index can be used to distinguish the extent and type of the damage qualitatively.(5)To realize the quantitative identification of the damage extent of aluminum plate,neural network is applied to the analysis and processing of the magneto-mechanical signal.By establishing the numerical model of the aluminum plate,the magneto-mechanical impedance is obtained.And the acquired impedance data is substituted into the neural network model for training and testing.It is proved that the neural network can be used for the quantitative identification of the damage extent.