Ultrasonic soundings of targets by processing of signals from plural receivers

Ultrasonic pulse echo flaw detection systems have been widely used in the field of nondestructive testing for quite long time. Theoretical studies made in the 1970’s and early 1980’s suggest that ultrasonic NDT using pulse compression technique can overcome some of the drawbacks of classical pulse echography. The limitation of currently available ultrasonic instruments hardly lies on the property of hardware but it may lie on the lack of sufficient signal processing techniques. At present, the ultrasonic A-scan instrument is most commonly used, which is in fact the fundamental part of other advanced techniques such as ultrasonic imaging. It is believed that the received A-scan signal may carry a lot of information on material property and defect but the information appears in various guises of noise, which is yet to be deciphered completely. So further research in ultrasonic signal processing techniques is essential to enhance the testing ability of ultrasonic instruments. During the 1990’s, a lot of research has been done on ultrasonic signal processing and still now the endeavor is going on in search for more reliable and versatile signal processing techniques. Generally the flaw signals measured in ultrasonic NDT include the effects of the measurement system and are corrupted by different kind of noise. The highly complex interaction between the defect geometry and the back-scattered ultrasonic wave inside the test piece may not be assumed as a linear process. So the signal processing techniques, which require a priori knowledge of noise statistics are subject to fail in many situations. Therefore the approach of signal processing should be involving the noisy signal itself in constructing the signal processing method. A study on choice of coding signal demonstrates that M-sequence is the best choice if the medium is subject to motion or movement. So the application of M-sequence approach should get more attention. The use of inverse filter in ultrasonic signal processing has given a significant degree of success over rival methods during the last one decade with different people using it in different ways. In these methods, probably the potential of further improvement lies in utilizing the received data itself to design the inverse filter or even cascaded inverse filters. Conventionally three types of scanning, A-, B-, and C-scan, systems are used for ultrasonic flaw detection or target sensing. In each case of the scans, the transceiver simply transmits a pulse signal in the medium and then receives the echoes from the target. Then the resulting echoes are shown on the display where A-scan works by fixing the transceiver system whereas both B-scan and C-scan necessitate moving it over the test specimen. Here the amplitude of the reflected signal plays an important role for the display mechanism. These traditional methods usually involve single transceiver system, which must use a scanning mechanism in order to detect and locate the target. It would have been a simple and convenient way, if the target could be detected and located by keeping the transmitter and receiver at fixed places, which does not need thorough scanning of the investigating medium. For an instance, it may be useful to locate a school of fish inside shallow water in a bay without sounding by boat. In this paper, we propose a method, which involves a single transmitter at a fixed place, and several receivers, which are also fixed at some predetermined positions, using the correlation method for acoustical diagnosis of targets or flaws. To make the method workable under any noise condition, several noise reduction techniques are devised, which can successfully retrieve very weak signal buried under high noise. In this paper, a new approach of designing an inverse filter (mismatched filter) is introduced, which calculates the parameters based on the part of the system response. So it is able to adjust its parameters according to the system response in minimizing the system noise. In addition, a simple elementary geometrical method has been proposed which can efficiently pinpoint the location of faults. The method is based on the travel time (TT) of the signal from transmitter to the receiver via target. Primarily the air medium between two acrylic plates has been chosen to illustrate the feasibility of the method. The system has been successfully operated to detect artificial flaws consisting of small targets in air.