Micro-Acoustic Excition Fusion Detection Technology Of Metal And Nonmetal Bonding Structure

Metal and non-metallic bonding are widely used in aviation, aerospace andautomotive because of the light connection quality, stress distribution and good sealing.The mechanical properties of the metal and nonmetal bonded structure is a adirectreflection of its adhesive strength, therefore, the mechanical properties testing of thismaterial bonded structure reflects the bond quality, and we can prevent accidents andimprove product quality. Existing non-destructive testing focus on the detection ofdebonding of adhesive structure, can not effectively detect the bonding strength, and havesome defect,for example, the specimen is easily contaminated, the detection is slow andnot suitable for online testing, can not meet the real-time requirements of the production.The micro-acoustic excitation detection methods used in this paper is silencedincentive bonding site, bonding acoustic signal through the array of sensors to obtainwavelet packet transform stratification of the acoustic signal of a single sensor to calculatethe energy of the layers, select the appropriate energy layer of the wavelet coefficients asfeatures, And select the corresponding time-domain and frequency-domain characteristics;BP neural network on the amount of the selected features, the Rato power network training,the training of the sensor array as a reference model of the sensor array, the bondingstructure of the the Rato force data presented the data support the weighted fusionalgorithm, a comprehensive description of the bonding parts of the mechanical propertiesof. The algorithm is through mutual support function for creating a sensor data to the arrayof sensors each sensor added weights to avoid the impact due to anthropogenic factors suchas threshold or threshold, so that the the Rato force range and the true value of the forecastclose to the phase. The experimental results show that the fusion detection technology ofmicro-acoustic excitation can effectively obtain the Rato bonding parts of the force range,close to the true value can be achieved online fast, real-time detection.