The Detection Of Deep Weld Defects And Algorithm Simulation Of MUSIC In Austenitic Stainless Steel

Deep austenitic stainless steel is widely used in nuclear power industry,petrochemical industry and machinery manufacturing industry.During the welding process,it is very easy to produce the invisible pores defects,and the microstructure of this kind of welding contain coarse grains,so it has obvious anisotropy characteristics.Therefore,it is difficult to achieve accurate and reliable inspection information by traditional nondestructive testing(NDT)methods.It is necessary to search a suitable NDT method to improve the accuracy of locating the deep thickness weld defects of austenitic stainless steel and to improve the detection rate of defects.The development of ultrasonic phased arrays technology is very fast,because of its advantages of high reliability,high detection resolution and high detection efficiency.So it is able to be applied in the detection of austenitic stainless steel weld deep defects.It attracted wide attention of world wide researchers.They had obtained a series of achievements.This dissertation includes the experiments of ultrasonic phased array,multi signal classification in the field of array signal relatively mature(Multiple Signal Classification MUSIC)algorithm,combined with the relatively mature field of sonar and radar time reversal(Time Reversal TR)algorithm,re-calculation defects positions based on ultrasonic phased array detection results.(1)Based on the principle of ultrasonic phased array,simulation of acoustic field distribution of the ultrasonic pulse signal was carried out using the Rayleigh integral in the MATLAB platform.The simulation phased array is a ultrasound radiation sound field of 16 element with the center frequency of 5MHz.And the evaluation of defect detection reliability index-detection rate was introduced.(2)Based on the existing ultrasonic phased-array equipment on austenitic stainless steel for testing blocks(70 mm,the number of defects 6)two defects depth at 10 mm of ultrasonic phased array detection,with the same defect depth,weld defects deteriorate the signal-to-noise ratio(SNR)to 2~3dB.Followed by the austenitic stainless steel on test block(90 mm,the number of defects 5)three defect depths of 30 mm and 50 mm,70 mm of ultrasonic phased array detection,the SNR significantly reduced with the increment of depth,and the speed of decline trend is obvious.(3)On the basis of above experiments,study the influence of weld defect detection rate.The conclusion is at the confidence level of 95%,for the no welding area defects,the maximum detection depth is less than 33.5 mm.And when the detection rate is more than 90%,for the welding defects,the maximum detection depth is less than 25.5 mm.(4)By comparing the TFM and TR-MUSIC algorithms,the TR-MUSIC algorithm is more suitable for the resolution of similar defects in distance.On the basis of simulation,using time reversal algorithm simulate the defect location,through TR-MUSIC algorithm,positioning of the two defects are simulated and verified by experiment,finding can only detect the defect with strong signal.