Study On Surface Microcrack Detection Based On Laser Ultrasound

During the manufacturing and use of the workpiece,various crack damages are easily generated on the surface.The appearance and expansion of cracks will significantly deteriorate the mechanical properties of the workpiece.If the defects cannot be detected in time,the workpiece may break.Eventually caused a major accident.Therefore,the detection of surface defects is an inevitable problem in the field of modern nondestructive testing.In this paper,using laser ultrasonic technology as the means and metal workpiece as the object,the location analysis of surface defects,the feature extraction of ultrasonic signals and the identification of defects were studied.The main research contents are as follows:(1)First introduced the principle of using the thermoelastic mechanism to excite ultrasonic waves;through comsol finite element simulation,the laser ultrasonic model of metal surface defects was established,the process of laser excitation of ultrasonic waves was simulated and the wave field propagation characteristics were analyzed.Angle analysis of ultrasonic signals laid the foundation.(2)Using the laser ultrasonic testing experimental platform to realize the scanning of metal surface defects in different directions,different depths and different angles,by analyzing the B scan of defects in different directions,the scope of surface defects is determined,and the defects are calculated Length and angle;by analyzing the time and frequency domain characteristics of the reflected and transmitted signals of the defect,the relationship between the time domain peak and valley time difference,the frequency domain energy and the wavelet packet energy and other feature quantities and the defect information are determined,and the characterizing defect depth is determined And angle identification parameters.(3)Using time domain peak and valley time difference,frequency domain energy andwavelet packet energy as input parameters,a defect recognition model based on BP neural network was established.Through training,quantitative recognition of metal surface defects was achieved.The research results of this paper provide theoretical basis and experimental methods for further research on surface defects detected by laser ultrasonic,and also promote the laser ultrasonic detection of surface defects from laboratory to industrial applications.