HOME  

 

TABLE OF CONTENTS  

 

          

 

Story of the Tapper

 

When the composites group in CNDE took on the job of modernizing the old practice of coin tap four years ago, they had no idea that they would one day use the instrument they developed to map out damages on a rudder of an Airbus A300 failed by its composite tail. Neither had they thought that the device would attract attention of inspectors for racing sail boats from overseas.

Tap test is an age-old technique for inspecting composite parts on aircraft for flaws and damages, especially on movable control surfaces such as rudders, spoilers and flaps. In this technique one takes a hand-held mass, such as a coin or a machined piece of metal, and tap on the surface. A good region without defects or damages would produce a crisp and solid sound whereas a bad or damaged region would give a dud sound. The method is simple and cheap, but is also subjective, inaccurate and operator dependent. Over the years there have been a number of attempts to instrument the tap test and remove the dependence on human hand and ear. These efforts resulted in several instrumented tap test devices on the market; however, they have stopped short of bringing in another powerful tool in nondestructive evaluation: a visual image. When an area suspected of defect or damage is systematically tapped and the results are displayed as a visual image, the inspector can identify the defect or damage, as well as the normal substructures. The size, shape and location of the damages can  be assessed much more easily with the aid of an image. An image-capable tap test device would also produce an electronic archivable record for later reference.

So the question is: what should be the quantity to display in a tap test image?  The answer lies in an important attribute of tap test: a tap test samples the local mechanical property of the structure, specifically the contact stiffness. This is quite different from a global test such as listening to the “ring” of a railroad wheel after it is struck with a hammer. In a tap test the time of contact (impact duration) between the impactor mass and the structure depends on the local stiffness of the structure. An impactor would bounce off quickly from a solid, un-damaged structure (short contact time), but would ride with the surface of a disbonded or broken structure and produce a long contact time. The duration of the impact is therefore the most fundamental quantity to measure and record in a tap test. The common practice of measuring the impact duration is to use an accelerometer that contains a piezoelectric crystal. As an example, figure 1 shows a 8-inch diameter repaired region on a honeycomb composite sandwich with 7-ply CFRP facesheet. The image shows short contact time (blue color) attributed to increased stiffness around the scarfed circle and potting at the center, and long contact time (orange to black) due to imbedded flaws and induced damages.

Figure 1  Tap test image of a repair and engineered flaws on a CFRP honeycomb sandwich

The next issue in producing a tap test image is position encoding. To generate an image, one needs to acquire contact time data as a function of position over the area of inspection. Although there are numerous motorized, computer-controlled scanning systems that are capable of acquiring such data, the experience of the CNDE composite group has been that the airline inspection community tends to shy away from complex automated systems due to cost, speed and operator training considerations. The composite group has therefore opted for a simple method of position encoding:  a thin plastic sheet with a printed square grid taped over the area of inspection.  To take data, each grid box was tapped once by hand. The tapping mass was an accelerometer fitted with a hemispherical tup. The output voltage signal of the accelerometer was processed in an accelerometer-computer interface circuit and the contact time data were then fed into a laptop PC for generating the tap test images. A considerable and sustained effort was made to develop versatile software for data acquisition and processing. To initiate a scan, the operator would specify the width and length of the scan area and the row and column convention on the PC and then proceed with the tapping. Figure 2 shows a manual tap test conducted on the wing of an MD-80 over an aluminum honeycomb structure known as a heater blanket. The scan image shows small disbonds in a 2-foot by 3-foot inspection area. Although the manual mode of tap test is intrinsically tedious, it still remains a valuable technique for inspecting curved, irregular parts and in hard to reach places.


Figure 2  Grid overlay used as a position encoder in manual tap test on an MD80 wing.  The resulting image shows a 2-foot by 3-foot area.

Although the manual tap test system using the grid overlay was successfully applied to on-aircraft inspections and produced useful images, the slow speed and tedious data taking clearly needed improvements. To make the tapping faster and more uniform, Dan Barnard and John Peters, two engineers in the FAA-funded project, invented an elegant “magnetic cam” that used the repulsive force between small but strong permanent magnets to throw the accelerometer toward the part surface. As shown in figure 3 and in the linked movie, (requires plug-in support) magnets embedded in the circumference of a wheel pushed a magnet in a rocker arm, and hence the accelerometer, up and down as the wheel rolls over the surface. Because the magnetic cam is based on action-at-a-distance, it is largely free from wear and alignment problems and hence superior to a mechanical tapper. The magnetic cam makes uniform, equally spaced taps along the rolling direction and only requires an overlay with a set of parallel lines to guide its indexing motion as it scans an area. The spacing between taps can be changed by having different number of magnets on the wheel. With a pitch of ¾-inch, a 6-foot2 area can be mapped out in less than 10 minutes with the magnetic wheel.

Figure 3 A magnetic cam drives the accelerometer (left) up and down as the wheel rolls over the area of inspection.

In a tap test the contact time depends on the local stiffness of the part as well as the mass of the impactor used. In the computer aided tap test system developed at CNDE, a simple spring model is used to compute the local stiffness from the measured contact time and the known mass of the impactor. The stiffness so deduced from the tap test data is a meaningful mechanical property of the structure. It was further demonstrated that the stiffness deduced from tap test and the stiffness measured directly in a load test showed good agreement for a variety of composites over a range of stiffness.

Over a period of three years, the composite group at CNDE has completed the development of the Computer Aided Tap Test (CATT) system and applied it to a wide range of composite materials and structures. It had been used on carbon and glass composite control surfaces of aircraft, helicopter blades with Kevlar facesheet, aluminum honeycomb sandwich with perforated skins, foam- and balsa-core boat building structures, and adhesively bonded fiber boards. The system remained simple and portable, as shown in figure 4, and has survived more than 15 field trials at a number of airline maintenance facilities, aircraft manufacturers, and military bases.

Figure 4 The Computer Aided Tap Test (CATT) system

Each test in the field has invariably led to some improvements in the software or hardware. The development of the technique and the instrument has benefited greatly from interactions with industry. Betasite tests were conducted at American Airlines, Northwest Airlines (figure 5), United Airlines and the Royal Air Force (RAF). The RAF has provided valuable human factors data for the system. Finally, this R&D effort has also benefited substantially from the contribution of students. The software development was done by Brian Danowsky and Nordica Hudelson , the laboratory experiments were assisted by Zach Nielsen and Trent Simpson, and all students have participated in field tests and gained real world experience.

For further information, contact Dave Hsu

BACK TO TOP