6 MARCH 2005, 12,000 metres above the Florida Keys. A deafening bang echoes through a packed Airbus A310 airliner. The carbon composite rudder on Air Transat flight 961 has snapped off, and is plummeting into the ocean below. The pilots manage to land safely at Varadero airport in Cuba, but it is the narrowest of escapes; as the rudder broke away, stresses on AT961’s composite tail fin may have come close to tearing that off too. Had that happened, the rest of the plane would have followed the tail fin into the ocean.
Eight months later, a hangar in Memphis, Tennessee. Engineers are carrying out routine maintenance on a FedEx-owned Airbus A300, and accidentally bash its carbon composite rudder. The engineers take a careful look to see if the blow has done any serious harm, and can hardly believe their eyes. Whatever harm they might have done was nothing compared to what had already happened inside the rudder. Hydraulic fluid from the plane’s control system had somehow seeped into the material and attacked the composite. Its sandwich structure was coming apart. Subsequent pressure tests showed that, in flight, the whole thing could have disintegrated at any moment.
These two incidents may not be unrelated, according to an investigation into the problems with the FedEx plane. Published in March, the US National Transportation Safety Board (NTSB) report notes the likely “applicability” of the FedEx rudder damage to the incident above the Florida Keys.
Advertisement
The FedEx case is significant because it means there is a potential threat from hydraulic fluid getting inside a composite structure. No one yet knows how common hydraulic fluid leakage is in the A300 fleet or how fast it damages composites, nor what is the risk of losing an entire rudder or tail fin. Even more importantly, the FedEx incident shows that routine inspection techniques don’t necessarily reveal problems with a composite structure. It has suddenly become clear that the use of composites, now a central part of aircraft design, creates a new and critical challenge for the industry.
“For the first time we are seeing applications of composites where, if they fail, the aircraft is dead,” says Philip Irving, an expert in aviation damage tolerance at Cranfield University in Bedfordshire. If there are doubts over the standard inspection routine, the question that has to be answered is a simple one. What is the industry going to do about it?
This dilemma has come at a critical time. Plane makers are replacing more and more metal in aircraft with composite parts, in everything from wings and fuselage to elevators and tail fins. The Airbus A380 superjumbo, for example, is 25 per cent composite by weight (see Diagram), and in 2008 Boeing hopes to launch its half-composite 787 Dreamliner.
There is certainly a good economic case for using composite materials. Although they might seem little more than high-tech mud and straw, composites made using layers of tough carbon fibres held together by polymer resins are as strong and stiff as metals but can weigh far less (see “Stopping the cracks”). Lighter planes boost profits by carrying more passengers for the same amount of fuel.
For plane owners there is a further economic advantage. When an aluminium alloy aircraft reaches 15 years old, its owner must start using expensive and time-consuming inspection tools such as eddy current generators to look for cracks caused by the repeated cycle of take-off, flight and landing. “Part of what we are selling with the Boeing 787 is an airplane that needs less maintenance,” says Al Miller, director of technology for the 787 programme at Boeing. “Airplanes sitting on the ground for a couple of days make you no money. You won’t have to do that with the 787 – almost all the scheduled inspections will be visual,” says Miller.
It is much the same story for the A380 superjumbo, says Simon Waite, a structures inspector at the European Aviation Safety Agency in Cologne, Germany. He says that when EASA certifies the Airbus A380 for flight, the regime of routine maintenance checks will be “80 to 90 per cent visual because that’s the cheaper, better option for operators”.
They are not alone in this view. Current thinking in the airline industry is that most problems inside a multilayered composite tail or flap would show up as visible marks, ripples or wrinkles on its surface. And even if they didn’t, the industry believes that damage is certain to be picked up by the other linchpin of aircraft maintenance, the acoustic tap test.
In many cases this is as simple as it sounds: an engineer taps a wing or section of fuselage with a coin or a small hammer and listens for any small change in pitch that signifies a void or crack. It’s a simple, cheap and quick process that can be carried out on the tarmac, without the need to dismantle the aircraft. Just as with visual examination, the tap test has proved its worth through decades of use.
However, as the NTSB report on the FedEx A300 suggests, when it comes to composite materials this may not be enough. The rudder from the FedEx jet is a type used on about 400 other Airbus jets. It comprises a hollow honeycomb core made of a Kevlar-like polymer, strengthened on its interior and exterior surfaces by carbon-fibre-reinforced plastic panels (see Diagram). The report says engineers found that one-third of a square metre of the rudder’s inner skin had separated from the honeycomb core, greatly weakening the structure.
Within the damaged composite were traces of Skydrol, a hydraulic fluid that had leaked from the rudder control system. Needless to say, it shouldn’t have been there; the danger that hydraulic fluids pose if they get into a composite structure – just as with water, oil, de-icing fluid and cleaning solvents – is well known. These fluids can attack bonding adhesives or the honeycomb material, or can freeze and expand at high altitude, forcing laminated layers and sandwich structures apart and weakening the material. “The composite in the FedEx rudder was supposed to be sealed,” says Mary Anne Greczyn of Airbus North America in Herndon, Virginia. The rudder is only designed to withstand Skydrol on its external surfaces, she says.
Tests of the FedEx rudder in a pressurisation chamber showed that separation between layers got worse every time the rudder was pressurised and depressurised, just as would happen on each flight cycle. What’s more, none of the current inspection techniques – visual inspection and tap tests – would necessarily have revealed the hidden damage, the NTSB report says.
Looking for signs
David Maass, a former composites engineer with helicopter maker Sikorsky who now runs composites consultancy Flightware in Guilford, Connecticut, believes that in the light of this information, plane makers and regulators could cause disaster if they press ahead with inspection regimes based on visual checks and tap tests of easily accessible areas. “Trying to determine visually if a composite laminate has internal damage is like trying to diagnose a brain tumour by looking at someone’s face,” he says. James Williams, a mechanical engineer and composites expert at MIT, agrees, describing the use of visual testing techniques on composites as “lamentably naive”.
So what is the solution? That is – quite literally – the million-dollar question. While there are plenty of other ways for engineers to scan composite structures, such as ultrasonic and thermal sensors, most have not been adapted for widespread use. While they can spot problems in composite parts taken back to the lab, they have not been turned into reliable, quick and easy-to-use gadgets that any engineer can pick up and use on a plane on the tarmac – the kind of tool airlines need if it is to be useful. “We need the industry to develop new technology that can detect critical faults easily,” says Nick Stoss, director of air investigations at the Canadian Transportation Safety Board, which is still looking into the flight AT961 incident.
Such gadgetry may not be far off, however. Jean-Pierre Monchalin and his colleagues at the National Research Council Canada in Boucherville, Quebec, are patenting a laser-based technique that could expose hidden damage inside composites. Their system uses pulses of laser light to heat a patch on the surface of the material. The resulting thermal expansion creates vibrations that move through the material, and any breaks or delaminations in the composite structure alter the vibrational frequencies. For example, the larger an area of separation between composite layers, the more it will lower the frequency. Any change is picked up using light from a second laser, with the help of an interferometer.
Monchalin calls this technique a “laser tap test”, and has shown it can detect damage to composites that is invisible at the surface. Even better, the laser pulses can be sent along optical fibres to reach high-up or awkward locations, so the system should allow engineers to inspect a composite airframe without removing parts.
For all the problems, however, there is no doubt that Boeing and Airbus will supply safe aircraft when they launch the 787 and A380. Both companies say their pre-manufacturing checks, computer modelling and accelerated ageing tests, combined with visual testing in service, will ensure their planes are safe. Boeing plans to test a complete plane through a simulated three aircraft lifetimes: that’s 165,000 flights. It will undergo pressurisation cycles, torsion, twist and bending. And 1 metre-long guillotine blades will be fired at the fuselage to test the composite’s resilience to bird strike, for instance.
Meanwhile, Roland Thevenin, composite expert at Airbus, says his company is designing sensors to monitor the A380’s structure in flight. Until that system enters operation, he says, visual checks will be key: “If damage is not visible, the integrity of the composite structures is not affected.”
The industry has a vast body of knowledge on how to monitor and predict the behaviour of metal planes. However, Irving says, composite materials are at the same stage metal was 60 years ago. “We have done a lot of work in the lab but haven’t really used it in anger. Engineers feel it ought to be OK, but there’s always this nagging doubt that something you hadn’t thought of could happen.”
Stopping the cracks
Composites are immensely popular engineering materials because they are up to 40 per cent lighter than aluminium yet 20 per cent stronger. However, great care must be taken when using them because composites fatigue and fail in a different way from metals.
Aviation composites comprise a polymer resin matrix that contains layers of strong reinforcing carbon or glass fibres. Layer after layer of these reinforced matrices are glued together and shaped in a curing oven to provide preciselythe strength each component needs.
In a metal, although a fatigue crack can appear at a very low stress level – about 10 per cent of the metal’s breaking stress – such cracks normally grow only slowly, allowing ample time to spot them before catastrophic failure. To start a crack in a composite, on the other hand, you need to apply somewhere between 50 to 70 per cent of its breaking stress. But while it is much more difficult to create the crack in the first place, it will then reach a critical size much more quickly.
This risk of sudden failure means that aircraft using composites are supposed to operate at stress levels far lower than any critical level – even when, for instance, they experience high loads created by atmospheric turbulence or aircraft wake. “But things will still fail, so regulators need to establish an inspection interval and a detection regime that will always catch them,” says Phil Irving at Cranfield University in the UK.
New clues?
The damaged FedEx rudder has rekindled memories of an accident in 2001 that killed 275 people. Flight 587, an American Airlines Airbus A300, lost its composite tail fin shortly after take-off and crashed in the New York borough of Queens. In 2004 an inquiry blamed the pilot for aggressively swinging the rudder from side-to-side to evade the turbulent wake of an aircraft ahead of it, which produced side loads on the tail fin 1.9 times its load limit. The fin snapped off.
In March this year the US National Transportation Safety Board said that flight AT961 experienced similar forces when its rudder broke off in flight: “These high stresses may have been dangerously close in magnitude to those that caused the in-flight separation of the tail fin during the accident involving flight 587.” Yet as Robert Spragg, a lawyer representing some of the flight 587 victims’ families, points out: “With the aircraft (AT961) at cruise, where would side loads have come from?”
In April, US congressman Anthony Weiner called for a fresh look at flight 587 in light of the new findings. But NTSB spokesman Keith Holloway says the board doesn’t see the need to reopen the investigation “at this point”.
