Composites require very different testing to metals – and this must be taken into account when assessing material properties. Lou Reade reports. Composites and metals are very different in terms of their mechanical behaviour. A simple example of this is how they react to a bump. While an aluminium panel may pick up an obvious dent it can easily be knocked back into shape, with no effect on its properties. But because a composite springs back to shape, it can sustain ‘internal' damage – despite there being no external evidence.
“It will undergo some internal delamination, which reduces strength,” begins Instron's composites market manager, Ian McEnteggart, which develops an array of testing machinery.
Composites are also far more susceptible to heat and humidity: while metal can rust – over a long period – it does not absorb moisture in the same way as a composite. With the ever-increasing use of composite, standards in the aerospace industry are among the most stringent. In addition to ASTM and ISO, manufacturers also have their own standards, which are usually administered through auditing organisation Nadcap. Its rules – which go way beyond just mechanical testing – are very strict, and this is unlikely to change.
Thankfully, a comprehensive set of tests is available for composite materials, whether in aerospace or any other industry.
“For composites, you need a lot more testing than you do for metallic materials – because their properties are a lot more complex,” he adds.
In general, he says, metals have the same properties in both directions, so a single tensile test will often be enough. But composites are anisotropic – meaning their properties are different in the axial and transverse directions. “That's a huge advantage in a material,” he says. “You can put strength in the direction that you need.”
However, it does make testing more difficult. A second ‘disadvantage' of composites is their inhomogeneity: by their very nature, they are composed of two or more components – which are prone to delamination under enough force.
“This means that a composite's properties are determined by factors such as adhesion and the strength of the matrix,” he says.
Instron has developed or updated a number of testing procedures, which it showcased at JEC in Paris recently. For example, it has extended the capability of its Bluehill 3 testing software with a new Test Profiler module, which gives more ‘real' results.
“It allows you to control the forces and deflections that you apply very flexibly,” McEnteggart says. “This simulates what might happen to a component in real life, rather than just doing a straightforward tensile or compression test.”
Standard tests are still necessary, according to McEnteggart in order to determine the basic properties of a material. It is more appropriate for ‘qualification' testing, at the early stages of qualifying a material for use in components. Once a material has been passed, it will be used to make actual components – which are then tested using more ‘realistic' QC tests.
Test Profiler can put a sample through complicated cycles and loading patterns to mimic real-life conditions. This includes fluctuating temperature and a change in data acquisition frequency.
McEnteggart likens aerospace testing to a huge pyramid, with the complete airframe at the top and basic materials at the bottom. Just off the bottom would be a material with a hole in it – a very common component in aerospace.
Meeting rigorous standards
Strain testing is crucial for this kind of component – and for many others. Strain is the relationship between the applied load and the extension that it causes. It can be tested using contact methods, or non-contact methods like video extensometry. Instron says that its recently improved AVE2 is the first video extensometer to meet the most rigorous composites testing standards – particularly ISO 527 and ASTM D638.
Strain can be measured by applying dots to the surface and measuring their relative movement under load. This tends to be done in one direction only, and a second test is required for perpendicular displacement.
But an extension to AVE2, called Digital Image Correlation (DIC), allows strain measurements to be made in both directions at once. Rather than analysing arbitrary fixed points, DIC can look at a speckle pattern on the surface of the composite. For something like a panel with a hole in it, this is a vital tool.
“Strain concentrates around the hole,” notes McEnteggart. “With the DIC option, you get a full field strain map and can measure along both axes at once.”
DIC is an advanced technique that helps engineers get under the skin of why a part is failing, by revealing non-uniform strain.
“It can help you identify the failure mode and gives you the research understanding, rather than just a number.”
Although DIC is a recognised technique, it has historically been available only as a high end research instrument. Now, says Instron, it is available in the commercial sphere – and is fully integrated with other test equipment, allowing results to be gathered quickly and easily. Another advantage for non-contact methods is to protect the instrument itself.
“When composites fail, they release lots of energy and can completely disintegrate,” says McEnteggart. “If you use a contact extensometer, there's a risk that it will be damaged.”
In addition to these quasi-static tests, Instron has made improvements to its dynamic and fatigue testing. In another example of the different testing regimes for metals and composites, it has developed the Specimen Self Heating Control (SSHC). SSHC optimises fatigue testing by allowing faster testing of composites. When a metal is fatigue tested, it can withstand the subsequent temperature rise of a few tens of degrees. But composites cannot.
“They are sensitive to heat and cannot conduct it away,” McEnteggart states.
As internal damage increases during the test, so does the heating. This usually means that a composite must be cycled slowly, in order to avoid excessive temperature rise. But SSHC monitors the temperature rise, and allows the test to be carried out at the maximum possible rate.
“We can increase the frequency up to the point where it sees the specimen start to heat up – and then it stops,” says McEnteggart. “We've done some benchmark tests, which are nearly 30% faster.”