The Research On Damage Probability Imaging Detection Technology Of Plate Structure Based On Acoustic-Ultrasonic

With excellent quality of processability,weldability,electroplating,thermal conductivity,and toughness,metal sheets such as aluminum,steel and other metal plates are widely used in aerospace,military weapons,transportation,shipbuilding,wind power plants,pressure boilers,architectural decoration,etc.In the process of manufacturing or service of large metal sheet plate parts,due to the influence of many factors,there are inevitably inclusions,cracks,and other defects in the sheet parts,and corrosion holes will appear on the surface of the sheet parts.When these defects accumulate and expand to a certain scale,the service life of the structure will be greatly shortened,and even cause a huge safety accident.In order to detect the possible defects in the structure of large metal plates in a timely and efficient manner,and then determine the location,size and damage degree of the defects,it is very necessary to conduct online nondestructive testing of in-service plates.The traditional visualization detection method of point-to-point scanning is time-consuming and not suitable for large components.The existing imaging methods based on time reversal,tomography and transit time can guarantee high-quality imaging results,but the signal data amount is large,which cost a lot of time to calculate and analyze,so the imaging efficiency is not high.The characteristics of long-distance propagation and small energy attenuation of the acoustic-ultrasonic signal enable it to quickly detect large plates,which has a great advantage in saving imaging time.Probabilistic imaging method USES the probabilistic index of damage existence as the source of image pixel value,and the algorithm can achieve rapid imaging under the condition of certain imaging quality and has no high requirements for signal quality.In this paper,the acoustic-ultrasonic probabilistic imaging identification method for structural damage of plates is studied.The main work is as follows:Based on sound-ultrasonic technology,ultrasonic Lamb wave theory,and the Lamb wave frequency equation,a sound of wave mechanics model which numerically simulates the propagation of acoustic-ultrasonic wave in the sheet using MATLAB was established to obtain the phase velocity frequency curve,and found that the phase velocity of the same mode wave varies with the frequency,and there are many modes of waveforms at the same frequency.The dispersion characteristics and multi-mode variation characteristics of sonic-ultrasonic propagation in thin plates lay a foundation for the following study of the basic theory of sonic-ultrasonic probabilistic imaging and the simulation of sonic-ultrasonic damage probabilistic imaging detection.A finite element model of four piezoelectric wafer rectangular arrays for detecting aluminum plate defects was established by using COMSOL Multiphysics.The acoustic-ultrasonic signals were excited and acquired using the model-based acoustic-ultrasonic technique,and processed by a probabilistic imaging algorithm based on the differential coefficient method.By comparing bright region in the damage probability imaging detection map with the actual defect position,the feasibility of the acoustic-ultrasonic probabilistic imaging damage detection method was verified.Experiments on the probabilistic imaging detection of plate structure damage were carried out.Eight rectangular wafer arrays were used to detect aluminum plate defects,and two damage indicators based on the difference coefficient method and wavelet transform method were used to calculate the damage probability and map.The results of the defects map showed that The image quality whose damage index is based on wavelet transform method is higher than that based on the difference coefficient method.The influence of different values of size parameter? on the imaging results is analyzed.It is concluded that the increase of the sensor to a certain extent can significantly improve the imaging quality without causing the too long processing time of the data and the waste of the sensor resources.The research results show that the method of acoustic-ultrasonic damage probability imaging detection can realize fast and high-quality on-line imaging detection of large plates.