Web Exclusive Articles
« back to listing
How do you test composites? More than just ‘black metal’, these new materials require, and deserve, a whole new approach that takes into account their unique properties
The rise and rise of composites is one of the stories of the last decade. Initially confined to the military due to their cost and perceived risk, the use of composite materials is now firmly entrenched within the commercial arena. In 10 years, Boeing has leapt from 12% composite weight in the 777 to 50% in the 787. In an industry where environmental issues, the price of oil and an increasingly competitive market are all paramount, the high strength-to-weight ratio of dual materials is seen as irresistible.
“Introducing a new solution into an industry as conservative as aerospace takes a very long time,” says Faye Smith, director of Avalon, a UK-based composites consultancy, which includes BAE Systems and Saint-Gobain among its clients. “We are in phase one at the moment, where there is a lot of direct substitution. We call it black metal, where the design replaces the metal part with composites but with only minor modification. Though not the best way to use composites, this had to happen in such a conservative sector as aerospace. Since the parts look the same, the industry is happy. The next phase will be a complete redesign such that you’re using composites in the optimum way. The best thing about composites is that you can create different properties within the same section, enabling a completely alternative kind of design.”
Although the process of redesign continues to be a slow one, the current benefits of composites are still clear to see: fatigue and corrosion resistance, lower pressure tooling, lightness, the possibility of complex shapes and the ability to tailor a part to your own specifications. These generally offset the higher cost and damage tolerance associated with these low-density stiff materials. The question remains, however: what are the key differences when testing composites as compared with standard isotropic materials?
Slade Gardner is a Fellow at Lockheed Martin (LM) Space Systems Company. He is the man behind Apex, an LM-developed thermoplastic nanocomposite consisting of a short fiber and nanofiber-reinforced blend of ultrapolymer. It is produced using injection molding and extrusion and is being implemented on the next generation of LM’s A2100 satellite buses. It is also earmarked for future designs of its missile defense products such as THAAD.
“We’ve developed a very significant design database for the Apex material system and this was created through extensive mechanical and physical testing,” says Gardner. “All the typical laboratory tests you might think of, we have done – tension, compression, sheer, impact, flexure, fatigue – along with all the physical properties you would consider, such as galvanic corrosion, thermal expansion, electrical and thermal conductivity, and density. All that testing feeds into a database that we offer to our program customers in order to build implementation articles. Sometimes our customers tell us there is a gap in the database for their particular design needs. They may ask for very high temperature mechanical test data, in which case we will return to the lab and generate the data for that customer. The tests we do on composites are the same as their aluminum equivalent, but with different results.”
With both the A2100 satellites and missile defense products, Apex is used for secondary internal structures such as clips, brackets and cable trays. As Apex is a new material, Gardner and his colleagues are starting out with secondary structures to build their competence base before embarking upon higher load case conditions. For the A2100 satellite, they are deploying over 1000 clips that stabilize the craft’s main structure, while for missile interceptors they are considering a handful of composite components internal to the missile, such as brackets, clips and cable trays.
“We test the components in different ways,” says Gardner. “For satellite clips, they are bonded in place then connected using a fastener. When we test those in a simulated environment, we bond them into test hardware that simulates the bonded condition then test that adhesive bond as well as the mechanical integrity of the clips. For the missile application, the test requires mechanical testing at higher temperatures and different vibration regimes so those parts are tested for mechanical rather than a bonded installation.”
Bill Hooper is manager of Composite Research and Development at ATK, a leading aerospace manufacturer. His focus varies from orbit vehicles to military and commercial aircraft.
“Fundamentally, the biggest thing to understand about composites is that you’ve designed your material, and designed it to have some sort of mechanical response,” he says. “Composites are anisotropic – they possess very different properties in each direction. This drives the way you get data, and the data you get. With composites, there are enormous benefits but also more degrees of flexibility that you have to understand.”
According to Hooper, much of the actual equipment used is essentially the same as that used for testing metal components: tensile test machines or sheer tests or compression tests. What varies is the type of test specimen used: “There are a lot of standard test specimens defined for tensile testing,” he says. “Another common test is open hole compression. These are industry test techniques, so everyone is generating data that is comparable, regardless of who does it. From a test machine standpoint, it is essentially the same sort of technology and equipment you’d use with other materials. The main difference is that you often have bigger load cells and bigger actuators.”
It is the mix of different properties in different directions that drives the complex testing. As a result, characterization programs for composite materials can become very expensive. According to Hooper, it is not uncommon for someone making branded parts, for instance for a hinter stage on a rocket, to spend upward of US$3 million on testing. Rocket motor case materials are tested at the fiber and resin level to provide design input properties. ATK also does subscale cylinder and bottle winding, with tests on coupons or burst tests of the subscale bottles to verify their design input properties. All its full-scale rocket motor cases have a pressurized ‘proof test’ to verify they will perform in operational conditions.
“We do not have rocket motor cases fail from the composite failing because we have proof tested them all,” says Hooper. “Overall, with composites you have to understand your response and stress-strain levels. That’s why you have to do the test and that comes at a cost. Metals have more standard, established data design. Composites vary depending on how you choose to lay them up. That’s really the trade-off you get there – added cost versus customized response. The pay-off comes back by reduced weight.”
“Composites have a much more complex structure involving a lot of potentially weak interfaces,” says Dr Dan Kells, research manager at the National Composites Centre in the UK. “Although metals can have a very complex microstructure, there is less variation. In addition, environmental testing can be completely different. Thus testing against hot, wet conditions and various chemicals and oils, such as Skydrol, is usually carried out, whereas metals would be tested against corrosion.”
According to Kells, the main challenge when testing composites is the number of variations and stages in manufacture. In order to be able to fly, the material must pass tests at all levels: starting with an assessment of the fiber/resin matrix and interface between them, moving through the testing as coupon laminates to structural elements and then full structures. In order to change material, it is necessary to go through the entire pyramid again, at a cost of several million pounds. This means that aircraft are not always designed on the latest resin/fiber combination. Metals are not immune from this hierarchy but it is much less complex.
“Composites can be very sensitive to specimen configuration, such as notch sensitivity, or sensitivity to changes in thickness,” he says. “In addition, specimen preparation is much more difficult and the tests themselves are much less consistent. This of course means that more tests are needed. Furthermore, in service a large amount of sub-surface damage can be caused by relatively minor impact and this needs to be detected.”
For aircraft structures, the most common method is ultrasonic c-scan, Kells explains. This allows large structures to be fully assessed and will identify voids or other visible structural defects.
“This is accurate and accepted but it is relatively slow and requires water coupling between the probe and the specimen just like a pregnancy scan,” he says. “X-rays are also used which are faster but not so good for large areas. In the field, various ‘coin tapping’ techniques are often used to detect delamination damage. Two other techniques are eddy current testing, based on the conducting properties of carbon fiber, and thermography. In the lab, similar mechanical testing will be done for composites and metals but the jig and fixtures are often different. One of the developing areas is the incorporation of sensors into composite structures to sense internal damage of loss of performance. This is often referred to as Structural Health Monitoring and is most commonly based on incorporating fiber-optic sensors into the structure during manufacture. This has been shown not to adversely affect the structural performance.”
Most aerospace certification is historical. As a result, current standards are often based on years of understanding of how metal performs. Some within the industry have pointed out that certification requirements now need to be rewritten to account for, and take advantage of, the unique properties of composites.
“This is happening,” says Hooper. “NASA has a clear set of requirements to qualify composite structures for use in rocket launches. They have very specific requirements for going through a proof test cycle due to the importance of safety with astronauts on board. There will be extensive non-destructive evaluation of parts before proof test and after, plus a qualification requirement. That’s how the rocket people do it. For aircraft it is different. Subscale parts will be tested, and some structure assemblies, but they certify a lot more based on analysis of that data. For instance, it is not practical to take the airplane and break a wing off, as compared to NASA requiring a failure test of a perfectly good interstage.”
“There is less standardization in basic materials,” says Kells. “Thus, aluminum alloys have a detailed alloy classification and theoretically anyone could reproduce, say, 7050 aircraft alloy. In composites, each resin and fiber combination will be proprietary to a manufacturer and it would be virtually impossible for anyone else to reproduce it. The starting point in most industries has been the metal standards. So, for example, aircraft standards have been designed around the most common aircraft material, aluminum, which is particularly vulnerable to creep. This has meant that design of composite aircraft components has been unnecessarily conservative. Although this is changing, it has been a matter of catching up.”
Saul Wordsworth is a freelance aerospace journalist for Aerospace Testing International.