Signal Processing Technique Research Of Ultrasonic Testing For Pipe With Rough Surface Based On Virtual Instrument

The surface condition of the pipe becomes a major factor which affect the result when using Ultrasonic Testing(UT) to detect the pipe. The ultrasound will be reflected or refracted in the surface of the material if the surface condition of the pipe is poor, which causes a huge attenuation of the ultrasonic echo signal and makes a lot of noise among the useful signal. The ultrasonic testing sensitivity and signal to noise ratio(SNR) will be reduced. The subsequent UT defect signal discrimination, qualitative and quantitative analysis will be affected seriously. It will influence the accuracy of the UT results.Considering the research purposes and methods of the subject, using the laboratory equipment in hand, we have set up a virtual UT signal processing system by ourself. A virtual UT signal processing platform is designed using Lab VIEW software which is developed by National Instruments(NI) in the Windows operating system. The virtual UT signal processing system can stimulate the probe transmitting and receiving ultrasonic signal. Data acquisition card was used to converse the analog ultrasonic echo signal to digital signal. The USB bus was used to complete the transmission of the ultrasonic digital data to computer. The virtual UT signal processing platform developed on the computer can be used to set the control states and parameters of the data acquisition card and complete signal processing of the ultrasound echo signals.The UT echo signal of pipe with rough surface was processed using the virtual UT signal processing systems. The UT echo signal of pipe with rough surface has been processed based on the Fourier Transform(FT) and Wavelet Transform(WT) respectively. The signal processing work that based on the FT is mainly refers to:①The UT echo signal of pipe with rough surface was processed using Fast Fourier Transform(FFT) to complete spectrum analysis.②The UT echo signal of pipe with rough surface was filtered using Finite Impulse Response(FIR) window filter based on the spectral analysis and the filtering effects of different window functions was compared.③The UT echo signal of pipe with rough surface was processed using Short Time Fourier Transform(STFT) to complete time frequency analysis. The signal processing work that based on the WT is mainly refers to:①The UT echo signal of pipe with rough surface was processed using WT to complete time frequency analysis and the time frequency analysis results of STFT and WT was compared.②The UT echo signal of pipe with rough surface was de-noised using wavelet thresholding method. The denoising results using soft threshold and hard threshold were compared and analyzed.The results has showed that it is difficult to satisfied the detecting required (SNR should generally above6dB-8dB in industrial flaw detection) without any denoising processing to the UT echo signal of pipe with rough surface. The SNR of detection signal has been improved obviously after appropriate filtering and denoising processing. Some of the useful signals has been eliminated while using the FIR window filter technology. It has led to the distortion of waveform. However, it almost has no useful signal losing while using WT to process the UT echo signal of pipe with rough surface. The characteristic morphology of the defect signal was maintained effectively. The Wavelet denoising technology can get a more desirable result. This research has provided a good way to resolve the UT signal processing of pipe with rough surface in the future.