Research On The Method Of Detecting Metal Internal Stress Based On The Linear And Nonlinear Method Of Surface Ultrasonic Wave

Ultrasonic surface wave stress detection technology has become one of the hot spots in the field of nondestructive testing because of its non-destructive property,low time consumption and no harm.It is widely used in aerospace,hardware equipment and construction and other aspects.At present,ultrasonic stress detection methods are mainly divided into linear parameter detection and nonlinear parameter detection,but the corresponding research lacks the analysis and modeling of signal components,and the stress detection parameters and schemes are relatively single.In order to solve the above problems,the following works are carried out in this thesis.First,starting from the characteristics of production and propagation of ultrasonic surface wave,surface wave will be concluded experiment by the temperature,the surface of the propagation path and tested the effects of thickness,the second of ultrasonic detection mechanism of the linear and nonlinear parameters,has been clear about the both measurement parameters on stress testing and changing with the stress.Finally,starting from the signal composition,the echo signal is decomposed into random noise,ultrasonic fundamental frequency and ultrasonic distortion signal,and mathematical modeling is carried out to complete the mathematical simulation of the echo signal.Second,a multi-band ultrasonic surface wave detection system is designed.For the detection of linear parameters,the excitation part adopts capacitor charge-discharge mode.Using large capacitor charge-discharge technology,it can generate 500V excitation high voltage,and the instantaneous voltage can reach 50A.For nonlinear parameter detection,the excitation part adopts the transformer mode,and the design uses continuous pulse+high-frequency pulse expansion technology,which can achieve 5MHz continuous pulse excitation,the maximum excitation voltage peaking value can be up to 150V,and the energy conversion efficiency is about 80%.In the echo conditioning part,pre-amplification+six-order high-pass filtering processing mode is used to achieve 12x amplification and 95%suppression of 370kHz frequency.Thirdly,the ultrasonic echo signal processing scheme is designed.The sample library of echo signal under various conditions is established,and its composition is analyzed.The noise is composed of impulse noise and electronic noise.The combined filtering of wavelet packet and EMD modal decomposition is designed to suppress the noise,and the SNR of the echo signal within the range of 2.5MHz to 5MHz is increased from 5.08dB to 22.27dB.Fourthly,the correctness of the surface wave is verified by propagation experiments.In the detection of linear parameters,surface waves with two frequencies of 2.5MHz and 5MHz are used to measure.The experimental results show that the absolute acoustic time difference of the two frequencies is positively correlated with the stress,and the fitting error rate of the 5MHz relation curve is 2.91%,which is a small deviation from the theory.In the nonlinear parameter detection,the nonlinear spectrum difference was defined to measure the stress,which showed a proportional relationship,and the average error rate of fitting was 3.46%.The average sensitivity coefficient of the calculated nonlinear spectrum difference is 77%higher than that of the 5MHz linear sound velocity detection,which indicates that the nonlinear parameters are more sensitive to the change of stress.