| In NDT(Non-destructive Testing) of weld flaw, UT(Ultrasonic Testing) is an important method. In Ultrasonic NDT, the background noise by the unresolved scatterers such as grain boundaries and other microstructure, and electrical noise by the ultrasonic defect detector often masks the flaw signal, creating a hindrance to detection. The wavelet analysis(WA) is the most recent technique for processing signal with time-varying spectra and is the a powerful tool to analyze ultrasonic signal. A method based on wavelet transform is put forward to analyze and process ultrasonic detection signal.A digitized ultrasonic detection system is constructed with personal computer(PC), ultrasonic defect detector and high-powered digital oscilloscope. In the system, high-powered digital oscilloscope perform data acquisition. It has advantages of strong antijamming capability and high reliability. In the system, ultrasonic signal digitization is implemented.A software of analyzing and processing ultrasonic detection signal with wavelet is developed with MATLAB6.1. MATLAB6.1 is a very useful and available tool for wavelet analysis and processing of ultrasonic signal.Based on searching wavelet and wavelet transform theory, analyzing mathematical model of impulse ultrasonic signal and discussing denoising arithmetic with wavelet transform, denoising experiments are done for a lot of ultrasonic detection signals. The results show that wavelet has time-frequency analysis capability. It takes on upstanding effect in frequency band separateing, restraining jamming and denoising, etc. It has well denoising effect not only to punctuate flaw but also strip flaw.
|
PDF Full Text Request