Latest Developments in Real-Time 2D Ultrasound Inspection for Aging AircraftDr. Marvin Lasser, Mr. Bob Lasser, Mr. John Kula, Mr. Gene RohrerImperium, Inc., 2092 Gaither Road, 220F, Rockville, Maryland 20850 Dr. George Harrison University of Maryland Medical School 6-015 Bressler Research Building, 655 West Baltimore Street, Baltimore, Maryland 21201 Corresponding Author Contact: Email: blasser@PO.MCTEC.COM |
TABLE OF CONTENTS |
Nondestructive testing of structural components through ultrasonic testing is well established as one of the industry's benchmark techniques. It's capability to penetrate both thin and thick material provides arguably the best information to inspectors on subsurface faults. However, there are two basic drawbacks to its use: its difficulty to employ and its slow speed. Real-time C-scan solves both of these issues while maintaining high quality subsurface information. Cracking, corrosion, voids, delaminations and impact damage can be observed in 1/30 second. The basis for this technology is a novel two-dimensional imaging array that creates immediate, high-resolution images of subsurface faults. The latest developments of the technique include commercial introduction of a through-transmission scanning product which can inspect large structures, as well as significant progress in the development of a hand held device which produces instantaneous high quality imagery of defects in reflection over an area as the user simply holds a probe up to the target. This work is funded in part by the Navy SBIR "Fasttrack" program.
The system that embodies this technology, called Acoustocam, is best understood by describing state-of-the-art imaging techniques. Imaging arrays have been used for many years in both the visible and infrared. With infrared, heat from a scene is radiated. A lens is used to collect the radiated energy onto an infrared sensitive two-dimensional imaging array. This array is termed a 'hybrid', made up of two different components. The first is an infrared sensitive material. This material creates voltages when heat is applied. A microelectronic chip specially designed to acquire these voltages reads out the voltages as they are created. These chips are generally up to one-centimeter square and are made of thousands of very tiny pixels. Today's arrays typically have 64 x 64 pixels up to 1024 x 1024 pixels, each pixel measuring anywhere from 20 to 150 microns (1 to 6 mils). This is the spatial resolution that these arrays provide. Figure 1 shows this structure in a commercial infrared camera and a typical IR image.
Fig 1
Fig 2: The use of an ultrasound sensitive array to generate an image of impact damage in a composite material (coverage of one inch by one inch). The image appears on the LCD.
|
Fig 3: Imaging array with piezoelectric layer and several thousand pixels.
| |
To create a similar quality of imagery using ultrasound, Imperium adapted these infrared chips to operate for ultrasound. Simply put, we removed the infrared material on these imaging arrays and replaced it with ultrasound sensitive material. Therefore, this chip is only a receiver.
To generate C-scan images, ultrasound is introduced into the target through a large, unfocused commercial transducer. The pressure wave strikes the target and is scattered. This scattered energy is collected by an acoustic lens and focused onto the array, identical to the infrared. The use of a lens provides a simple, inexpensive alternative to complex beam forming often employed in ultrasound imaging. The user simply focuses by adjusting a lens while looking at the image. Furthermore, it provides a means to trade off resolution and area coverage, or zoom in and out. Standard video electronics and image processing are used to format the image for presentation to the user; either on a PC, monitor, or hand held LCD. Each time ultrasound is sent into the material (up to several thousand times/second), information can be collected at the array. To remain compatible with standard video equipment, the imagery is typically presented at 30 frames/second, again similar to infrared or visible.
The array in Figure 2 below shows a similar structure to Figure 1. Note that now there is an ultrasound receiving (piezoelectric) layer deposited onto the chip instead of infrared. The result is a system that can image subsurface characteristics. The chip is responsive over a wide range of frequencies, although most images are created between 1 MHz and 10 MHz.
Figure 3 is a picture of our current imaging array.
The analogy for the visible case is a standard video camcorder. Within a camcorder, a two-dimensional array and zoomable lens produce high quality imagery of a scene. In this case, the array was designed to be sensitive to light, not infrared or ultrasound.
Our current array is 128 x 128 pixels, totaling 16,384 pixels. Each pixel is 0.080 mm, or about 3 mils. The array is one centimeter by one centimeter. However, the resolution and area coverage may vary. The frequency that is employed, which has varying wavelengths, will impact resolution. Also, the placement of the lens will determine both resolution and coverage. For example, we could look at one-centimeter square with 80-micron resolution, or move the lens and look at one-inch square with approximately 200-micron resolution, depending on the need.
The implementation of this technique can be either in transmission or reflection. In transmission, a source transducer sends ultrasound through a target and is collected on the far side by the acoustic lens and array. Figure 4 shows this implementation for the inspection of faults in a manufacturing environment as compared to a typical C-scan system. The section below describes the potential timesaving by employing this implementation.
In reflection, the ultrasound source, lens, and array are packaged together for those applications where only one side of the target is accessible for manufacturing and in-service inspections. To perform this, we employ an acoustic beam splitter that cuts the beam by 50%. This approach, used commonly in optics and lasers, provides a means to package all the components into a small hand held probe. Figure 5 shows two views of this beam splitter structure that has the added advantage of keeping all beams in parallel.
For reflection, where depth information is critical, a small pulse echo thickness gauge is installed directly into the probe and provides a very precise measure of the thickness at any point. Note this depth finder on Figure 5.
Fig 4: Current C-scan and Acoustocam C-scan in manufacturing implementation | Fig 5a: 3-D of hand held probe using beam splitter | Fig 5b: Side view | Fig 6: Acoustocam hand held system |
The resulting system being developed is a two handed device which a user places against the target to inspect. A non-attenuating membrane creates a couplant between the probe and the target. A small amount of gel or water may be needed to perform the inspection. The user moves the probe over the target and imagery appears on the LCD. If a fault is observed, a crosshairs on the LCD is moved over the fault and the depth measurement is made at that point. All information is to be recorded, potentially with voice annotation. Figure 6 shows the final system.
Real-time C-scan can potentially be applied to any inspection that now uses ultrasonic inspection. Surface reflections and attenuation are identical. In other words, if today's ultrasound cannot image a fault, the new technique cannot either. Currently, a C-scan image is generated as a single point probe is mechanically moved across a target. After the scan is complete, a two-dimensional image is produced. Real-time C-scanning will produce the same image instantaneously.
During the development efforts of this technology, Imperium has recorded real time images of faults in metals, composites, and microelectronic circuits (voids inside chips). The following images were generated in 1/30 second.
Fig 7: Subsurface images of a) corrosion with under 1% weight loss and b) a crack in a rivet (coverage of both is one centimeter by one centimeter). |
Specific applications that offer a substantial potential for this technique include:
Users who could benefit from this technique include:
This technique has now been developed for commercial implementation for through transmission applications. This is typically done using immersion or a water squirter setup (see Figure 4). The next section describes the justification to use such a system.
The pulse echo device is operating on a benchtop and is not yet useable commercially. The beam splitter setup is functioning and the proper front coupling is being evaluated. A large test evaluation camera system is generating images. Later in the development effort, the system will be redesigned into the smaller hand held device shown in Figure 6.
Fig 8: Comparative coverage within a single acquisition time |
In contrast to typical ultrasound imaging techniques, Acoustocam offers several orders of magnitude in speed over what is currently available. Figure 8 shows the amount of data that would be covered in a single capture frame of a current C-scan versus two-dimensional imaging for the same resolution.
As an example, if an inspection is performed of a 10-foot by 10-foot component, with a minimum spatial resolution of 0.5-mm (21 mils) and a scanning rate of 12 inches per second, the difference in scanning time is as follows:
Current C-scan | Real-Time C-scan |
100 (ft) 2 | 100 (ft) 2 |
6096 passes | 48 passes |
16.9 hours | 8 minutes |
Real-time C-scan technology offers an exciting next generation of ultrasound imaging. The use of semiconductor technology, and standard "optical" techniques such as lenses to create real time imagery means that commercial systems can be delivered at costs less than or equal to today's slower C-scan systems. For the hand held device, no formal training or certification would be required to observe subsurface faults. For in-service inspections, this technique can be used to quickly quantify very small defects.
NDT.net |