Study On Signal Processing Of Defects Detection Using Ultrasonic Phased Array

Ultrasonic phased array technology is widely used for its advantages such as high scanning speed, flexible beam control, high sensitivity and wide scope of application. This technology is becoming an international industry research in NDT filed. In this article, researches are all about the related signal processing technology, and we start with the achievement of ultrasonic signal both on hardware method and software method, aiming to lay is the foundation for independent research and development of ultrasonic phased array equipment. Then the signal processing methods have been studied and verified by experiment, and finally the fractal theory is used to extract the feature of near-surface defect signal and to realize defect detection.Firstly, this study overview the principle of acoustic emission and reception of ultrasonic phased array. Then two ultrasonic phased array testing model are set up, which testing with wedge and aiming to realize the acoustic beam deflection and focus. With these models, the delay law is calculated. Explaining the effect of the delay law which makes the phase changed and comes about beam synthesis by simulation. Finally, different imaging method of ultrasonic phased array flaw detection is analyzed.As FPGA is one of the key technologies of ultrasonic phased array detection, an embedded system based on FPGA is proposed for ultrasonic data acquisition. That is a system on a programmable chip, termed system on a programmable chip(SOPC), which is built with Nios ? soft-core processor inside FPGA and aiming to control the A/D chip and USB chip to realize signal acquisition and transmission. An embedded ultrasonic data acquisition platform is built.When conducting an inspection we can choose to collect the full matrix capture(FMC) data, which involves the collection of the transmit-receive time domain signals from all element combinations. In order to get full matrix data form device memory, this paper makes a development of software termed Multi2000, and post-process the data by total focusing method(TFM) to form an image. Then the simulation of FMC is done to validate the TFM. The experiment results show that the TFM algorithm can improve the image resolution.Due to the undetected of near-surface defects in ultrasonic phased array detection, this article use fractal theory to process A-Scan signal, and the calculation result of signal's box-counting dimension and statistical analysis prove that the fractal theory can be used in the recognition of near-surface defects. Finally, the factors which affect the recognition are analyzed.