Ultrasonic Non-Destructive Testing Based On DSP Technology

In the present work the advanced DSP technology has been applied to the transform and processing of signals acquired from ultrasonic non-destructive testing. The purpose is to find useful information from the ultrasonic signal, which can show characteristics of materials. Thus, the reliability of the ultrasonic non-destructive testing may be improved. In this project copper-steel friction-welds and car body's coated steel spot welds have been studied. In the study of ultrasonic non-destructive test signals for the copper-steel friction-welded joint, immersion method and focusing sensor were used. Wavelet analysis method was used to decrease the noise, and wavelet package analysis method was used to decompose the ultrasonic A-signal. In addition an "energy-defect"analysis method was used to study the relation between the bonding quality and the energy of ultrasonic reflective signal at the joints interface. The results show that the bonding quality and shear strength may be evaluated through energy spectrum distribution from the wavelet package transform of the ultrasonic A signals. In the study of ultrasonic non-destructive test signals for car body coated steel spot -welded joint, direct contacting method was used to pick-up the A-signal. FFT was used to study the characteristics at frequency domain of the ultrasonic signals. It is indicated that the high frequency amplitude of the signals is decreased with the increase of spot size and strength of joints. Through the FFT, the high frequency amplitude of ultrasonic testing signal not only are used to evaluate the spot size and tension strength, but also the specimens with perfect nut can be distinguished from that with poor nut. The study for the ultrasonic testing signal of the two kinds of welded joints shows that proper DSP technology can improve the reliability of ultrasonic testing.