How will 787's new materials fare in a crash landing?
Boeing's new 787, the first airliner built largely from carbon fiber infused with epoxy resin, won't behave like a traditional metal airplane in a crash landing. Engineers have spent years modeling and testing the differences, and say they are confident the design is safe.
Seattle Times aerospace reporter
When a Boeing 777 lost power and crashed short of the runway at London Heathrow Airport in 2008, the landing gear collapsed and a strut pierced the passenger floor. Yet apart from one broken leg, there were no serious injuries.
When an Airbus A340 landing in bad weather skidded off a Toronto runway in 2005, it broke into pieces and caught fire. But in the minutes before flames engulfed the jet, all 309 people aboard evacuated safely.
Though such accidents don't always end so well — last month 158 people died when a Boeing 737 slid off a runway in India — today's metal airliners are designed to be survivable in a crash landing.
How will Boeing's new 787 Dreamliner, the first airliner with a body built largely from carbon fiber infused with epoxy resin, fare in such a crash?
The new material is tough. But hit hard enough, it breaks rather than bends. And in a fire, the epoxy resin burns.
One early Boeing computer simulation was disturbing.
In 2005, as design of the Dreamliner advanced, a previously undisclosed Boeing analysis showed a crash that is survivable in a largely metal 777 would be deadly in a 787: The impact would shatter the bottom of the 787 fuselage and deliver a jolt severe enough to kill all the passengers.
A Boeing engineering manager called the outcome a "potential showstopper" for the Dreamliner.
Boeing says a key design change and subsequent physical tests prove the final Dreamliner design is now as safe as a metal airplane.
And while a few critics remain concerned, the Federal Aviation Administration (FAA) is close to certifying the jet as safe to fly passengers.
When an auto company develops a new car, it must run more than a dozen full-scale crash scenarios, witnessed by safety officials. Every test destroys a car.
But running full-scale tests of big jets crashing is considered impractical as well as too expensive. As a plane heads toward a crash on land or water, there are too many possible impact variations to test every scenario.
So today's metal airplanes have been certified largely using computer simulations. Manufacturers validate their virtual results with smaller-scale physical tests — flexing the wings, stressing fuselage panels to breaking point.
The FAA and Boeing agreed in advance exactly what testing was needed to prove the 787's safety.
The 2005 Boeing document that laid out the deadly Dreamliner crash scenario was an early mathematical analysis, prepared for the 787 program by structural-dynamics experts in the company's Phantom Works research unit.
A computer-generated drawing from the internal report shows that in a simulated crash, the 777's metal lower fuselage crumples, as one would expect. But the rest of the airframe, including the floor of the passenger cabin, is intact.
Lower fuselage shatters
In the composite-plastic 787, by contrast, the lower fuselage is shattered, with multiple holes. And the passenger floor has broken away from the fuselage and collapsed, leaving passengers with little chance of reaching an exit.
Worse, the Boeing study projected that the impact on passengers would be much more severe in a 787.
The highest survivable impact in a crash landing is considered to be about 20 g, meaning a nearly instantaneous deceleration equal to 20 times the acceleration caused by gravity.
The study projected that at a vertical descent rate of 22 feet per second (about 15 miles per hour), the average peak impact on a passenger's spine would be 15g in the 777.
In the 787, though, that impact would be a deadly 25g, the study concluded.
In March 2005, Phantom Works project manager Vince Weldon sent an e-mail to Boeing's chief technology officer, Jim Jamieson, flagging the simulation as "very dire."
An aeronautical engineer, Weldon worked 46 years in aerospace, half of those at Boeing. At Phantom Works, he assessed the use of advanced composites for future airplanes, though he had no direct role on the 787 program.
Weldon's concerns were examined by a panel of Boeing technical experts chosen from outside the 787 program. Its review endorsed the jet's composite-material design.
"He raised questions. They were investigated," said Boeing spokeswoman Lori Gunter. "We did not proceed with the design until we were sure it was safe."
In 2006, Weldon was fired by Boeing after an allegation that he used a racist remark about a superior in the course of pushing his concerns internally. Weldon, 72, denies that and says the accusation was a way to discredit and get rid of him.
Boeing made structural changes after the 2005 analysis that dramatically improved the jet's crash safety, said Mark Jenks, a vice president on the 787 program.
It redesigned rows of short wedge-shaped support posts beneath the cargo floor so they progressively collapse on impact, absorbing energy and reducing the impact felt in the passenger cabin.
Paolo Feraboli, an assistant professor at the University of Washington and director of its Lamborghini Lab for studying advanced composite structures, who worked for Boeing on the 787 program, said the support posts "fail in a very progressive, very stable, very energy-absorbing fashion."
With this change, Boeing's computer model now projected a much better impact result.
But unlike homogeneous metals, multilayered composites are very difficult to simulate accurately on a computer, said Feraboli.
"We don't currently have the knowledge and the computational power to do a prediction based on purely mathematical models," he said.
So to convince the FAA that its computer model matches real-world results, Boeing did some physical tests not required on previous metal planes.
In 2007, Boeing performed a key "vertical drop test" of a partial fuselage at its Apache helicopter-manufacturing facility in Mesa, Ariz.
An 8-foot-long section of the fuselage's bottom half, with full luggage containers beneath the passenger floor, was dropped from 15 feet onto a thick steel plate. It hit at an impact speed of 30 feet per second, around 20 miles per hour.
That's about 10 times the typical vertical descent rate when a big jet lands, and three times the rate its landing gear is required to withstand.
Videos of the drop test show the fuselage section slamming into the ground and completely flattening along the bottom — evidently fractured and broken, since the plastic doesn't bend. Beneath the passenger floor, small bits of the fuselage support structure fly off.
Floor stays solid
But in contrast to the 2005 computer model, everything above the cargo floor appears solid. The crucial passenger cabin floor and its supports remain intact.
And Boeing said sensors at the passenger seat showed the impact forces were survivable.
"The integrity of the floor area and overall extent of the damage were all within the bounds we expected and required," said Jenks.
He wouldn't disclose the impact forces recorded at the passenger seats. But he said the results validate the 787 design.
"This structure is as good as the 777," said Jenks. "That's what the model is showing when we finalized the design and then ran this test."
On the sidelines of an industry conference in Seattle last month, a leading outside expert with decades of experience in aerospace composites said Boeing is aware its software models for simulating the behavior of composites aren't as good as those for metal.
To compensate, he said, Boeing's engineers built the most conservative assumptions into their design.
The outcome is a structure that is over-engineered and heavier than Boeing would like, the outside expert said. But he believes it should be safe.
However, one Boeing engineer who worked on the 787 design said he's still uneasy.
He said the energy-absorbing stanchions installed beneath the cargo floor will protect the 787 passenger cabin only from the vertical impact of a crash landing. But how will the composite structure perform if the plane hits an obstacle as it skids along the ground?
"There are other crash scenarios in which the stanchions can't help," the engineer said.
Fire, smoke, toxicity
The drop-test outcome raises one more issue: performance in a post-crash fire.
While the aluminum of a metal plane crumples on impact, composites tend to fracture or shatter. In a crash like that of the A340 in Toronto, would the 787 fuselage keep fire, smoke and toxic fumes from penetrating to the interior and overcoming passengers?
The good news is fire tests conducted by the FAA in November 2007 show that plastic composites like those used in the 787 stand up to fire much better than metal.
But one critique suggests that may be irrelevant if fire and fumes can enter through holes in a shattered hull.
A fuel-fed fire can melt through an aluminum panel in about a minute. With an added layer of thermal insulation inside the fuselage wall, the fire barrier holds up a further 4 minutes. That required 5-minute total provides passengers time to get out.
But the type of composite plastic on the Dreamliner will resist burn-through — and provide protection from the fire — for much longer than 5 minutes, even without insulation.
While the epoxy resin in the composite material ignites and burns, the mat of carbon-fiber layers chars like wood to create a protective barrier that holds back the fire.
Ali Bahrami, head of the FAA's Seattle office dealing with commercial-airplane certification, said the agency's tests showed the carbon-fiber composite not only resisted burn-through impressively but also prevented toxic gases from penetrating inside.
"Composite structure is performing better than metal and insulation together," said Bahrami. "... With composites you provide (travelers) a longer time to get out."
Boeing ran a series of its own lab tests, applying an external fire to a panel of the 787's composite material, with similar results.
"Because there are things that are new about this, we've gone way, way beyond what might have been basic requirements," said Jenks. "I'm personally extremely confident and comfortable."
Composite-materials expert Derek Yates is not convinced.
In an unpublished paper, Yates dismissed the FAA fire tests because they were done on an intact fuselage, which is not typical in a real crash.
Yates, 74 and retired, worked for Lockheed on the Trident missile, which had the first primary aerospace structure made from composite plastic. From 1997 to 2000, he consulted for Boeing.
Yates' strong views on the composites fire threat stem from work he did in the 1970s with NASA that resulted in FAA rules effectively banning the use of epoxy-based composites from aircraft interiors because of the fire hazard.
All airliners entering service since 1990 comply with that ban.
Yates is concerned that a 787 fuselage's underside will shatter in a crash — just as depicted in the 2005 crash simulation.
One worry is that as the plastic fuselage slides over the ground after the initial impact, the broken underside of the hull could rip open, creating holes through which toxic fumes and smoke from burning composite material might pour in.
Yates asked the FAA during the public-comment phase of the 787 certification process to require a fire test with a full-scale ruptured fuselage rather than an intact panel. His idea is to test if toxic fumes from burning composites will be substantial and penetrate any rupture.
FAA spurns request
The FAA rejected the request, Bahrami said, because in a post-crash conflagration, "the fuel fire is by far the biggest problem" — not the burning of the composites in the fuselage skin.
Boeing executive Jenks said that after the drop test, the bottom of the composite plastic fuselage was "crumpled, creased, cracked and fractured," but with only "small holes" in the skin.
"There weren't big gaping holes," Jenks said. "I don't think there's any reason to believe there'll be more holes per se in a composite fuselage."
Dan Mooney, Boeing vice president of development for the 787-8, said the composite fuselage doesn't break like glass.
"It doesn't shatter and disperse in lots of pieces," Mooney said. "It tends to hang together by the fibers."
Mooney said Boeing's tests show the burning of the plastic resin won't add significant risk.
In February, Boeing issued official guidelines telling airport firefighters they can use standard techniques to put out a 787 fire, adding that "from a toxicity perspective, the composite structure ... poses no greater hazard than an aluminum fuselage."
Boeing has completed all its 787 fire testing and submitted the results to the FAA for certification.
"We believe we're done," said Mooney.
FAA spokesman Allen Kenitzer said the agency's review of Boeing's data on a 787 crash impact and fire is almost complete. Certification is expected this fall.
"We don't anticipate any problems or unique difficulties with the remaining work," Kenitzer said.
Someday, a 787 will suffer a heavy crash landing. Only then will all of Boeing's analysis and the FAA scrutiny be put to a real-world test.
Dominic Gates: 206-464-2963 or firstname.lastname@example.org