Deadly plane failure raises questions about safety of widely used engine

Southwest Airlines Emergency Landing
National Transportation Safety Board investigators examine damage to the engine of the Southwest Airlines plane that made an emergency landing at Philadelphia International Airport.

A commercial airplane’s engine failure over Pennsylvania on Tuesday bears similarities to problems on another plane less than two years ago, but federal investigators declined to say there was cause to worry about a class of engine shared by both aircraft and used by hundreds of airlines around the world.

“We are very concerned about this particular event,” said Robert Sumwalt, chairman of the National Transportation Safety Board. “Engine failures like this should not occur, obviously. To be able to extrapolate that to the entire 737 fleet … if we feel this is a deeper issue, we have the ability to issue an urgent safety recommendation.”

The NTSB was not taking that step Wednesday, he said in Philadelphia, adding that it needed more information. The planes, both Boeing 737s flown by Southwest Airlines, had fan blades that broke inside the engine. NTSB investigators have found evidence of fatigue cracks in both broken blades. Sumwalt did not rule out a manufacturing problem as a cause.

On Wednesday evening, however, the FAA said it “will issue an Airworthiness Directive (AD) within the next two weeks that will require inspections of certain CFM56-7B engines. The directive will require an ultrasonic inspection of fan blades when they reach a certain number of takeoffs and landings. Any blades that fail the inspection will have to be replaced.”

Southwest Flight 1380, flying from LaGuardia Airport in New York City to Love Field in Dallas on Tuesday morning, made an emergency landing in Philadelphia after its left engine experienced an uncontained failure that appeared to be caused by a fan blade’s breaking. The debris that blew out of the engine struck the fuselage, breaking a window and leading to the death of Jennifer Riordan of Albuquerque, N.M, who was sitting in the 14th row, Sumwalt said.

Camera icon TOM GRALISH / Staff Photographer
National Transportation Safety Board Chairman Robert Sumwalt at Philadelphia International Airport.

Riordan died from “blunt impact trauma of the head, neck, and torso,” according to the Philadelphia Medical Examiner’s Office. She was wearing her seat belt at the time of the engine failure, Sumwalt said, but witnesses said the force of the plane’s pressurized air escaping the cabin pulled her partially out of the window before fellow passengers were able to pull her back in. The flight carried five crew members and 144 passengers.

The titanium blade is one of 24 attached to a hub at the front of the engine. It broke in two places, Sumwalt said, at mid-span and near its connection to the hub, but it appears the break near the hub happened first. The fatigue crack investigators found on the broken blade was inside of it and not visible to the naked eye, he said. The NTSB will perform metallurgical tests on the area, he said.

On Wednesday, investigators were still collecting pieces of metal that surrounded the engine from as far as 65 miles away from Philadelphia. The cowl around the engine includes a containment ring — a band of durable material such as Kevlar, the material used to make bulletproof vests, said David Brookstein, a Temple University professor of mechanical engineering.

“It’s the bulletproof vest of the engine,” he said.

Trying to contain something moving as fast as a turbine blade that’s broken loose is difficult for almost any material, experts said.

“We’re pushing the limits of technology to contain those fan blades,” said John Goglia,  a former NTSB member. “We don’t have a material strong enough to contain it all [at] that high speed.”

Less than a year ago, both the engine’s manufacturer, CFM International, and the Federal Aviation Administration expressed concern about fatigue on blades in the CFM56 engine series. They focused on the benefits of a more robust inspection protocol, but limited the recommendations to planes with 15,000 cycles, or about four years, of service without time in a maintenance shop.

A spokeswoman for Southwest said Wednesday that a maintenance recommendation from the manufacturer of the plane’s engine that addressed problems with fan blades would not have applied to the engine involved in Tuesday’s incident.

“The engine was not subject to the bulletin released last year,” said Brandy King, the Southwest spokeswoman.

The NTSB would not confirm that in a news conference Wednesday. The agency will track the history of each of the 24 fan blades in the engine, Sumwalt said.

King did not specify why the bulletin didn’t apply. She declined to provide information about the engine’s maintenance history.

A review of FAA repair data records on the plane involved Tuesday showed no previous significant engine problems.

Southwest said Tuesday that it would subject all its engines to an accelerated maintenance regimen, including ultrasonic inspection of fan blades, that should be completed in 30 days

CFM, a partnership between General Electric and the French company Safran Aircraft Engines, was working with the NTSB on the investigation, a spokeswoman said Wednesday. She and other experts described the CFM56 series as workhorses, with a good record and used by 300 airlines. The initial bulletin advising airlines to give special attention to fan blades came from CFM, but the company would not say whether another would be forthcoming after Tuesday’s accident.

“I can’t rule it in or rule it out in general,” said CFM spokeswoman Jamie Jewell.

The FAA declined to say Wednesday whether the engine failure over Pennsylvania would prompt any additional actions or changes to its proposed directive.

In August of last year, the FAA proposed a directive to mandate the testing the CFM bulletin recommended, an ultrasonic inspection that would reveal flaws in metal parts. That proposal is still going through the FAA’s approval process.

Ultrasonic inspections operate on the same principle as an ultrasound, said R. John Hansman Jr., a professor of aeronautics and astronautics at Massachusetts Institute of Technology.

“It’s the same general idea, except now you put it on the blade, and instead of looking for baby pictures you’re looking for holes or cracks,” he said.

Routine maintenance of blades, which can be swapped out of an engine, includes removing them every 3,000 hours of flight and looking for nicks, Sumwalt said. Those nicks could be caused by flying debris, experts said, and could increase stress on the metal, making it more likely that fatigue cracks could develop.

The CFM bulletin, and a subsequent FAA proposal, was prompted by an August 2016 incident similar to the one that happened Tuesday. An engine identical to the one that failed Tuesday lost a fan blade, causing an uncontained engine failure that threw debris into the fuselage, though in that case none entered the passenger cabin. The damage depressurized the cabin and forced the plane to make an emergency landing in Pensacola, Fla. That incident remains under investigation by the NTSB, but in both Tuesday’s accident and the one in 2016, the NTSB observed evidence of metal fatigue where the fan blade broke.

Fan blades, typically made of titanium, bear a heavy load on planes like a 737.

“When you design the airplane or the engine, this is probably the highest-stressed piece in the airplane,” Hansman said.

When a jet turbine is spinning, the tips move nearly at the speed of sound. The blades draw 2,500 pounds of air per second into the engine, according to a 2013 review in Progress in Aerospace Sciences.

They also must withstand significant heat variances. The air coming in is as cold as minus-75 degrees, while the air being vented from the rear of a jet turbine is hot. Depending on wind currents, that hot air can be blown back toward the engine’s interior, increasing temperature changes that cause metal to expand and contract.

Fatigue in fan blades is often caused by the blade’s vibrating over a long period, eventually causing little cracks to develop. But such cracks are more likely to be caused by a nick, as could happen if the blade is struck by flying debris, or by defects in the metal, experts said.

Though experts didn’t rule it out, several said it was unlikely there was a widespread problem with manufacturing or materials on CFM56 engines. The scrutiny on the fan blades is intense, they said.

“What it says to me is we’re starting to discover some events with the engine that were not obvious before,” Goglia said.

He described aviation as a relatively young technology. Knowledge has grown substantially, and the maintenance demands of aircraft are well understood, but the two incidents involving fan blades could hint at an area where engineers need to be more conservative in their estimates of how often parts need to be maintained or replaced, Goglia said.

The NTSB investigation into the 2016 engine failure is still open, he said, because the investigators have yet to figure out why it happened. Tuesday’s event might shed more light on what caused the blades to break, and how to prevent that from happening again.

“They don’t have the cause,” he said. “They have all the circumstances around it, but they can’t pinpoint what happened. Maybe the missing link could be here.”

In the Philadelphia situation, the NTSB has a year to 15 months of work ahead.

Interviews with the pilots and crew were underway, Sumwalt said, and investigators were reviewing the recordings of the flight crew during the incident. They will look at the broken window to determine what caused that damage, he said. And the NTSB has removed the side wall in Row 14 to see how the plane was dented.

Staff writer Michael Boren contributed to this article.