A collection of aircraft parts available for NDI research
By Barb Zuehlke
A lending library of aircraft parts? Yes, the Structures Laboratory of the Institute for Aerospace Research, part of the National Research Council Canada, based in Ottawa, has developed a collection of more than 700 aircraft parts for use in aging aircraft and nondestructive inspections (NDI) research.
The Institute for Aerospace Research (IAR) conducts research on issues related to the design, performance, use, and safety of aircraft. The institute performs research and product development studies with an eye on the OEM, maintenance, and the airworthiness authority (TC, CAA, or FAA). R&D includes teardown and experimental investigations, NDI, computational and experimental analysis technologies, life enhancement technologies and repair, and proactive maintenance practices. The Structures group performs materials properties tests on coupons, built-up structures, and full-scale tests: currently a F-18 wing test. It has developed enhanced visual inspection techniques for aerospace and forensic applications, and has created a library of parts to support its research, the Aircraft Specimen Library (ASL).
The library collection was driven by IAR's early work on D Sight™, a technology adapted to detect barely visible impact damage on composites. It was thought at the time that D Sight would also be good for detecting corrosion pillowing on metallic construction. That led to a discussion of whether to build specimens to support the experiments. "We quickly realized that there are so many changes in construction throughout the aircraft that it would be a nightmare to manufacture exact replicas," says Ron Gould, IAR technical officer and the ASL custodian. The need for specimens for research started as a small collection to support one program and from there grew into a library supporting many programs.
Starting the collection
Ron Gould and his partner Jerzy Komorowski, now the director of the Structures, Materials, and Propulsion Laboratory, went off into the desert and started collecting pieces. The first specimens collected were small and "all fit in a box that we carried home." Since that start in 1993, IAR has collected more than 700 fuselage and wing pieces from more than 50 aircraft, both civilian and military. Most specimens are 10 square feet or more in size but the largest piece is the "Aging Aircraft NDI/Maintenance Test-bed Vehicle" a complete Boeing 727-90C which is being used for corrosion and aging wiring studies.
The collection includes both corroded and noncorroded structures from withdrawn-from-use aircraft. In situations where naturally corroded specimens have not yet been collected, IAR disassembles and selectively removes the protective coatings in the joints of undamaged specimens, reassembles them and, through a salt fog exposure program, studies corrosion from the onset. Careful protective packaging ensures that only the targeted surface is exposed to the corrosive environment.
"That's the kind of thing we offer through the library," Gould says. "We supply a number of external labs and internal projects that have a specific construction they want to interrogate or develop a technique on. If we don't have an example of a naturally corroded structure in the library or can't find it in the bone yards, we will grow it. We've gone to the expense of chemically analyzing the corrosion product that results and comparing it to what we find in the naturally corroded joints to prove that we do produce the same results."
There is no charge to borrow specimens, just the shipping costs that are incurred. While the initial goal was to have the parts available for internal projects, they can also be borrowed or purchased.
"Any piece in the library has value added to it," Gould says, "because we've done the nondestructive inspections to quantify the damage. In those cases where we've sold pieces, we charge for that added value. When we loan specimens for, say, the development of laser ultrasonics, we ask that the data collected be added to our NDI database."
All the parts get a close visual inspection, and are inspected with D Sight and documented photographically. CAD drawings are created for many. IAR's NDI group has a robot and instead of taking spot measurements with eddy current (EC) or ultrasonics (UT) to check thickness loss or corrosion damage, it can provide maps of the entire joint. There is also in-house and collaborative research into multi-frequency and pulsed eddy current (PEC) which are able to better tolerate lift-off and quantify corrosion damage. Whenever a specimen is being screened for selection, inspections will be carried out with both traditional and experimental NDI techniques and these results are added to the database. Thermography has recently been added so that inspection data will be included, where appropriate, as well.
"Ultimately the researchers want to know how well their new NDI technique is performing or they want to accurately measure the damage. We will then take a section of the specimen, and very carefully disassemble it," Gould says. "We've developed techniques to remove the corrosion product chemically rather than mechanically so we clean it down to the point where only the remains of original alloy are left. Then we take a high resolution X-ray image along with a master calibration tool and digitize the X-ray film." The result is a highly accurate image representing the thickness loss. Each skin is processed separately and then the joint is put back together and the total thickness loss is determined. If a new NDI technology is being developed, specimens with quantified damage can be selected for use in determining the capability or sensitivity of the technology.
Fatigue and aging
In other studies aging degradation of material properties is also being analyzed. Through a reheating process named retrogression and re-aging, the fatigue and corrosion-resistant properties of various aluminum alloys can be re-established. According to Gould, this heat treatment process is about to be demonstrated on C-130 sloping longerons and the process should "save people a lot of money." One of the C130 Hercules specimens from the library is being used to qualify the technique.
The library includes specimens of both upper and lower wing skin material. A number of ongoing projects use the upper wing skin material to assess the structural significance of exfoliation corrosion damage, assess NDI techniques to detect and quantify the damage, and assess the effects of maintenance actions carried out to remove it.
IAR is comparing pristine, naturally corroded, and artificially corroded test coupons to understand how the structures will respond when they get old. Instrumented specimens are fatigue tested; these may be coupons of the material or built-up joint structures. IAR has designed built-up specimens that do not suffer from edge effects.
"When we first got involved in composites, we manufactured our own specimens from various materials and layups," Gould says, "and then we inflicted them with impact, delamination, and disbond damage. Over time the impact damage wasn't as spectacular as it used to appear in our enhanced visual inspections." This drove IAR to investigate further.
Indent depth and the number of damage sites in a given area are criteria when evaluating damage to composite structures. When an aircraft is taking off and a stone is thrown up by the wheels against a composite trailing edge flap, the impact damages that surface. The manufacturer's service manual stipulates how many or how deep an indent can be before repair is mandated.
"It's a concern as far as the structural integrity," Gould says, "and everybody bases that on being able to visually detect it and associate the internal damage to the measured indent depth. What we discovered was that even statically within a very short period of time or within one or two load cycles, the indent depth may be reduced by up to 40 percent. So any relationship you had between being able to observe, measure, and imply the significance of the damage is very much compromised by that relaxation. It's all driven by the fact that you have all these layers stuck together, you impact them, the adhesive lets go between the layers, and now they're free to move and the indent decreases."
Composite specimens are collected into the library. The latest include aluminum honeycomb core trailing edge flaps with moisture intrusion and corrosion plus an Airbus A320 vertical stabilizer recovered from a landing accident.
So how does this information affect mechanics? What should they be looking for?
One of the things predicted with finite element analysis of the different kinds of joint construction in the library was that they each respond differently when corrosion starts to build up. "Soon after it starts, the product buildup will push hard enough to force the material through the yield point and the skin becomes permanently deformed," Gould says.
The usual procedure is to look for cracks that appear at a fastener hole. "What IAR predicted and discovered was," Gould says, "rather than cracks running from rivet to rivet, we've detected nonsurface-breaking cracks that take off in all directions and they don't start at the rivet hole, they start away from the rivet hole and work both toward it and away from it. It is only much later that they come through the surface of the material and you can detect them from the outside."
Another thing IAR discovered is that when so much pressure is being exerted by this buildup of corrosion, the "guys in the field will notice when a rivet head pops off and they'll punch that one out and often replace it with a blind fastener because they can't get access to the inside," Gould says.
When taking specimens apart, rather than drilling out the rivets, IAR machines off the shop head and pushes them out, which allows for examination of the rivets. Adjacent to the replaced rivets are other rivets that are also suffering from this pressure and they're distressed and have cracks, according to Gould. In some cases, at total thickness, losses are between only 3 and 8 percent; and 40 to 75 percent of the adjacent rivets have been found to be damaged.
When dealing with corrosion removal in fuselage joints that may have permanently deformed skins, technicians are allowed to de-rivet a section, normally no more than 20 inches at a time, wedge the joint open, and reach in and mechanically remove the corrosion. Corrosion prevention compounds are added to try and stop the process and then it's buttoned back up. A lot of damage can occur to the remaining alloy when removing the corrosion. "And you only get to see that when you have the luxury of taking the joint completely apart.
"It's kind of a catch-22. You've been asked to do maintenance and do it quickly, and so people tend to use power tools," Gould says. "The SRM instructions don't say anything about power tools or rigid abrasive discs. But with the limited amount of space to work in, how else are you going to accomplish the task?" And the research continues.
Work is underway to complete an online database of all the parts in the ASL. The ASL Database Search program identifies aircraft model, aircraft history, specimen ID, type of joint and fastener, skin thickness, specimen properties, damage found, and the records from the NDI and teardown inspections conducted. A total of 27 parameters are available to conduct database searches.
Institute for Aerospace Research
Ottawa, Ontario, Canada