Ultrasonic Imaging Systems

The use of ultrasonic imaging systems for non-destructive evaluation is increasing, with particular interest being paid to research into real time and quasi-real time imaging systems. Photos are shown which were taken using an electronically scanned and focused real time ultrasonic imaging system. The system can be operated with longitudinal waves, shear waves, Rayleigh waves, and lamb waves in the 1.5 MHz to 3.5 MHz frequency range, and has been successfully used on composite materials (boron fiber epoxy on titanium) and on a number of metals (steel, aluminum, and titanium}. This system has been operated in both transmission and reflection modes; examples of each are shown. · Results Obtained on an Electronically Scanned and Focused Imaging Sys tern During the last three years, a new type of electronically scanned and focused acoustic imaging system has been demonstrated for use in nondestructive testing applications. The system has been operated in A-scan, B-scan, and C-scan testing modes and has demonstrated increases in speed of as much as 100 times that of an equivalent mechanically scanned system while producing images which are easily recognizable by a human operator and which produce information on the· two-dimensional location of its size and shape. As far as we are aware, this is the first electronically scanned and focused imaging system that has been used in NDT applications. Consider the application of the system in a C-scan mode. With a normal mechanical scan, illustrated in Fig. 1, a focused receiver and transmitter would be used to obtain definitions of the order of 1 .mm. It might take many hours to scan a raster to produce a visual display of a large object. In an equivalent electronically scanned system, illustrated in Fig. 2, arrays of as many as 100 transducer elements are employed in the transmitter and receiver. By electronic signal processing, the beam can be focused and scanned along a line parallel to the array in less than .60 llsec. The method employed makes use of surface acoustic wave technology, and employs a surface acoustic wave delay line with one tap per element of the array. By this means, a phase reference is introduced which is equivalent to the action of a physical lens. The electronic lens behaves like a physical lens which scans a line in the x direction parallel to the array at· a rate comparable to the acoustic velocity in the medium; the focal length of this lens can be changed electronically, at will. At the present time, the definition .in the perpendicular direction, they direction (parallel to the array surface) is controlled either by a slit in front of the· object or by a physical lens, or, alternatively, by the width of the beam itself. Scanning in the vertical direction is still arranged mechanically, but now, because of the use of 100 elements in the array, the basic scanning speed has been increased by 100 times. An acoustic image of a boron fiber reinforced epoxy laminate laid down on titanium, supplied to us by E. Caustin of the los Angeles Division, Rockwell International, is shown in the upper display of Fig• 3 with an illustration of the voids in the original panel. It will be seen tha.~ the device picks out most of the defects very clearly. The system can also be operated with electronic scanning in both the x and y directions by using crossed arrays. We have done relatively little with this mode of operation. For completeness, a result taken of the letter "S" cut into a piece of rubber is shown in Fig. 4. Most. of our work has been devoted to B-scan i.aging. The basic operation of the system is illustrated i.n Fig. 5. In this mode of operation, the array focusing system can be regarded as a moving lens which is scanned at a velocity comparable to the acoustic velocity in the medium itself. The system is first operated as a transmitter. It is focused on a line a distance z from the array, and, thus, scans this line. At a fime 2T later, where T = z/va and Va is the velocity of an acousti.c wave in the medium, the device is operated as a moving receiver lens, so that it picks up a signal from a point which has been illuminated at 359 a time 2T before. A signal arriving from a more distant point will arrive after the lens has moved past the received beam, so that there will be good range definition. Similar~y. a signal arriving from a different point on the scan line will not be seen, because of the good transverse definition of the lens. Thus, the device has good transverse and range resolution. In practice, the system is used to scan along a line, as shown in Fig. 6, then refocused and the time delay between transmit and receive changed so as to scan a line at a different distance z from the array until a complete raster normal to the array has been scanned. A simple picture of a step block in water is shown in Fig. 6. It will be noted that.both the transverse posit1on and longitudinal position of the steps can be clearly seen; the steps being approximately 8 mm apart and 5 111111 wide, the. results being taken at an acoustic frequency of 2 M.Hz. During the last year the system has been used to take images with various types of waves in metals. As an illustration, we have used 3 MHz longitudinal waves in aluminum to obtain an image of a flat bottomed hole. The results are shown ill Fig. 7, where it wi 11 be seen that the top surface of the hole can be observed, as well as the front and back wa 11 s of the meta 1 samp 1 e. The hole has a diameter of 3 mm and a length of 25 mm. It will be noted that the diameter and length of the hole can be measured directly by this technique, without further interpretation of the results. Because the surface of the hole is a specular reflector, we only observe its top surface, a point of great importance in NOT problems. A further point which is important to realize is that, in contradistinction to our earlier imaging system results, we are now able to ob.tain images of small reflectors very near to large reflectors, like the wall of a sample; the basic difficulty in doing this is that the relat~vely weak sidelobes of the image from strong reflectors can "swamp out" the weak reflection from a small neighboring flaw.· To eliminate this difficulty, we have adopted a method known as "gating the pulse," which ma.kes it possible to look only at the image corresponding to the main lobe of the reflector of interest .. This slows the scan rate considerably, but proves feasibility for the rapid scanning system which we are now developing, and is a valid technique for examining small flaws and is still much quicker and inore accurate than mechanical scanning. As a second illustration, in Fig. 8 we show how a shear wave is excited in an aluminum sample, with the shear wave propagating along the axis of the sample, and focused and scanned in a direction perpendicular to the paper. A series of holes cut into the sample are shown in the same figure, and it can be seen that both the transverse and longitudinal position of these holes can clearly be observed. A similar technique can be used to excite Rayleigh waves and Lamb waves in target plates, as shown in Fig. 9. Some results obtained for Lamb waves are shown in Fig. 10. Another set of results for Rayleigh waves is shown in Fig. 11. In the first photo, the top and bottom edge of the plate can be seen, for the Rayleigh wave passes over the top surface of the plate and around the edge to the bottom surface. By changing the image intensity, three small 0.5 mm D holes 9 mm from the edge can also be seen. Finally, we show Rayleigh wave images of an artificial crack 1 em long, -100 ~m side, eloxed in a metal sample. This sample can be rotated in the surface wave field, so that the crack can be aligned at an angle to the transducer array. lihen the crack is para 11 el to the array {a = 0), the crack is clearly seen, as shown in Fig. 12(a). Because the crack is a specular reflector, when it is turned at an angle no signal will be returned to the array from the middle of the crack; only the two ends of the crack will be observed. The results for a = o•, 45", and 90" are shown in Figs. 12(a), 12(b), and 12(c), respectively. The scattering from the crack ends can clearly be seen. A similar set of results is shown in Fig. 13 for two holes whose diameters are the same as the width of the crack. The results will be seen to be almost indistinguishable from those of a crack, except for the e = 0 case. But more precise analysis does show differences in amplitudes between the two cases, so that one might 360 expect that a more detailed study would show up considerable differences between a crack and two holes. These studies confirm the results obtained in theoretical work on th1s program of scattering from cracks. The two ends of the cracks behave like sources, and the imaging technique shows the presence of the sources directly, rather than resting on an indirect inference based on the interference phenomenon between the two sources, as would be obtained with simple plane wave transducers. The Design of a New Imaging System After working with the present imaging system in NOT applications during the last three years, we have obtained considerable insight into its advantages and disadvantages. The advantages are the basic ones to be expected; good definition in two directions, high speed, and large field of view in the direction nonnal to the array. The disadvantages are, a relatively limited field of view in the direction parallel to the array because of the limited number of array elements used in our research system, complexity, higher sidelobe levels than we would like, at least in the high speed operating mode, a relatively small aperture, so that the range of viewing angles is relatively small thus tending to give ppor images of specular reflectors, and operation at a lower frequency than optimum (2.5 MHz). In order to eliminate most of these difficulties, we have been examining and experimenting with component parts of a new imaging system. The new system is eventually intended to employ a very long array of elements, perhaps several hundred, or, alternatively, a small hand held array with perhaps 20-30 elements which can be moved by hand over the surface of an object. In each case, only 20-30 elements will be used at a time, thus making the electronic problems of the control circuitry, amplifiers, and other components of the system much simpler. In the purely electronic system, the equivalent of mechanical scanning will be carried out by a multiplexing arrangement to switch the electronic circuits along the way. The system will be arranged so that from any 30 elements a focused beam will be obtained, which scans over a circular sector of the order of +40° or more, and yields a complete dynamically focused image of this sectorial region of radius up to 20 em. The process will then be repeated for each ·position along the long array, or each mechanical position. Equivalent points in the multiple images will be computed and placed at the same point on the cathode ray screen. This is like the operation of the present medical B-scan imaging system, systems which are unfocused and purely mechanically scanned. In that case, the radial sector scanning is obtained by tilting the transducer and relying on the flexibility of the body surface. The operating system will emit a pulse, or train of pulses, from a single highly efficient transducer element, receive the reflected echoes back at the same transducer, and then use an A to D converter to digitize the received signals. The device will then store the digital :!