| Austenitic stainless steel is widely used in many important areas because of its excellent performance, during the welding process, fluctuation of welding parameters may lead to various defects: such as slag inclusion, cracks, gas porosity, lack of fusion,lack of penetration. Nondestructive technique is adopted to test base metal and weld metal in order to ensure the security and reliability of welding structure. Ultrasonic testing is widely used in nondestructive testing of weld because its portability, security,simple operation and reliable detection ability. However, due to the influence of anisotropic structures of coarse columnar grains in the austenitic stainless steel weld,Ultrasonic wave in the weld will be severe attenuation, scatter, acoustic beam had been distorted. These adverse effects lead to lower signal-to-noise ratio of detection signal,and the characteristics of defects are hard to be distinguished. Therefore, the development of new nondestructive testing technique for austenitic stainlss steel weld must receive adequate attention.Ultrasonic time-of-flight diffraction(TOFD) is a method of ultilizing the diffraction wave of defect’s ends to evaluate the status of defects. Compared with conventional ultrasonics inspection, ultrasonic TOFD have higher reliability and accuray. Ultrasonic TOFD is applied into stainless steel weld, which need to face the attenuation and noise problems caused by weld structure. In the paper, the characteristic of ultrasonic TOFD detection signal is analyzed. On this basis, an adaptive filtering algorithm is designed to process the detection signal of austenitic stainless steel weld,for the purpose of improving the detection signal to noise ratio; finally, ultrasonic TOFD imaging detection based on the processed signal is adopted to improve resolution and avoid misrecognition and undetected error.First, the characteristics of ultrasonic TOFD diffracted waves are analyzed. In this reserch, according to several ultrasonic TOFD diffracted wave of an artificial defect at different test conditions, the distribution of the diffracted wave amplitude which is affected by the effect of the test parameters and shape of the end of the defect is analyzed. The test result shows that: diffraction wave of different end shape have the same amplitude distribution. The probe arrangements have no effect on distribution of the amplitude. In addition, according to comparing the signal amplitude distribution ofdifferent material, the result shows that the signal amplitude distribution is not affected by the materials, and the diffraction wave of aluminum and steel have the same amplitude distribution.Secondly, the adaptive filtering algorithm and threshold algorithm are designed to carry on self-adaptive filtering processing of detection signal of ultrasonic TOFD. The threshold is defined according to the squares of the difference between the output result and the desired signal. Considering the signal to ratio and less information distortion,the optimal threshold is obtained. Filtering processing and image for ultrasonic TOFD signal at the optimal threshold. The result shows that noise have been decreased, the signal to noise ratio of the defect has been enhanced, and defects have been identified simple. The detecting image quality based on the self adaptive filtering has been greatly improved.Finally, the optimal adaptive filtering parameters have been obtained. Among all the parameters of the adaptive filter, the step size, the legth of desired signal and filter order have great influence on the filtering result. Comprehensive analysis on amplitude SNR, energy SNR and image obtained by the adaptive filter, the optimal parameters are obtained: 1) the optimum step-size is 0.4*A(A is the reciproca of maximum eigenvalue of correlation matrix of input signal); 2) the optimal filter order is 7; 3) filtering effect is optimal when the length of the desired signal was 0.7μs; 4) the desired signal should be chosen when the signal were similitude with the amplitudes distribution of ultrasonic TOFD diffraction wave.In this study, characteristic signal of ultrasonic diffracted wave in austenitic stainless steel weld is studyded. An efficient adaptive filtering algorithm for the detection signal is designed and image analysis after processing ultrasonic TOFD signal by adaptive filter. The result shows that: although the diffracted wave seriously enterfere by the noise, but its distribution of the diffracted wave amplitude is stable. The adaptive filtering algorithm is designed based on chosing optimal desired signal to improving the SNR. Imaging processing based on the filtered signal, the result shows that: noise wave have been greatly filtered, image feature of defect was more obvious, avoiding the error re cognition, misjudgment and undetected problem caused by noise. Combined with advance signal processing and analysis techniques, ultrasonic TOFD is hopeful to be applied in engineering applications of austenitic stainless steel weld. |
PDF Full Text Request