The pilots were greeted by a smattering of flashing warning lights, but they did not grasp the scope of the problem until one of the flight attendants made their way to the cockpit, according to report about the incident from French regulators. They brought a phone from one of the passengers, and on it was a photo of the damage to the engine, which was easily visible from passengers’ windows on the right side of the cabin.
The plane, an Airbus A380, which was supposed to be cruising comfortably at 37,000 feet, made an emergency landing in Canada two hours later, and no one was injured. But regulators warned the incident could have played out much differently if debris from the explosion had hit the aircraft instead of plunging to the ground.
The ordeal set French authorities on a years-long mission to find the lost engine pieces and pinpoint the root cause of the problem, requiring investigators to survey miles of terrain made perilous by deep, invisible cracks in Greenland’s ice sheet and the constant threat of polar bear attacks. The endeavor was also hampered by months of inhospitable storms, limited daylight and low visibility. Researchers ultimately found the key piece of debris — the engine’s fan — by accident, when a robot mapping glacial crevasses happened to roll over the spot where it was buried nearly two years after it had fallen from the sky, said Austin Lines, a US-based engineer who aided the recovery effort. It was packed in four meters (or about 12 feet) of snow and ice.
Studying the recovered debris showed the engine wasn’t damaged during maintenance, as investigators initially predicted. Rather, the problem seemed to be linked to weakness in the metal used to create the engine’s giant front fan — indicating what first appeared to be a freak accident may not be an isolated incident, according to a September report from France’s Bureau of Enquiry and Analysis for Civil Aviation Safety, or BEA, which led the investigation. Engine manufacturers have already worked to address the problem, but the BEA is now calling on regulators in the United States and Europe to take a closer look at how aircraft engines are designed, manufactured and certified for flight — hoping more careful scrutiny can root out such defects before they happen.
In-flight engine failures remain extremely rare, according to US and European authorities. But the unexpected conclusions from the BEA’s investigation highlight how an excruciating, 21-month search for a lost engine part was key to understanding how to prevent the same disaster from striking twice.
Searching Greenland’s tundra
The day after the 2017 Air France flight, BEA investigators and representatives from the plane’s and engine’s manufacturers, which included Airbus, General Electric and Pratt & Whitney, gathered at the Canadian airport to survey the plane’s damage.
Investigators pored over data in the aircraft’s flight data recorder — or “black box” — to hash out exactly when the explosion occurred and determined the debris likely landed about 60 miles from Narsarsuaq in the southwest of Greenland. Within days, helicopters were dispatched and investigators scoured the pure white landscape for signs of the large fan. But after one week and three unsuccessful search flights, the terrain was already buried in fresh layers of snow.
But both of those initial efforts failed, in part because the radars weren’t searching deep enough below the icy surface.
With another brutal winter looming, the search was once again put on pause.
Lines told CNN Business that at one point, investigators dropped a replica of the engine fan into the snow, just to make sure the radars they were using for the search could accurately detect the buried metal. But they couldn’t. And for months, the replica debris was lost, too.
France’s Onera research lab, which was behind the effort to use SAR radar on aircraft to locate the debris, also found the data it collected was too messy — or “noisy” in engineering terms — and the Onera team spent months developing new ways of analyzing the information before finally narrowing down the search field to a handful of possible locations, according to the BEA report.
“I don’t think anyone would care that much if a bunch of dudes went out with a robot and didn’t do much with it,” he joked.
In June last year, a five-person team, including Lines and a team of Icelandic mountain guides, flew by helicopter to the excavation site. A small dome-shaped tent built to withstand the severe winds sheltered them during their three-day-long excavation effort. At night, they slept with rifles next to their sleeping bags — a precaution for a polar bear attack.
Hidden crevasses posed the constant risk of the ground beneath the crew’s feet caving in, and they used metal rods to check the ice’s depth before trekking onto new territory. An unseen crevasse could have even been hidden underneath the dig site, so they wore harnesses with ropes attaching them to a nearby anchor point as they shoveled snow.
Lines, who had earlier helped dig the engine fan replica out of the ice sheet, was the only member of the five-team recovery crew that had been part of that effort and knew how grueling the work would be.
The first few meters of snow and ice above the engine fan shoveled out easily, but Lines used a chain saw to hack apart the thick, compacted layers of frost further down. The crew carved a ramp into the excavation site so that a sleigh operated by a pulley system could be used to shuttle about 20 metric tons of snow out of the pit.
“We had a lot of sunshine because the sun doesn’t really set [that time of year],” Lines said. “So we just worked through the night, and then went to bed for a few hours and then woke up and just started digging again.”
Finally, on day three, the tips of the engine’s fan blades came into view.
The battered piece of debris later proved crucial in understanding what actually went wrong on the 2017 Air France flight. Investigators determined that it wasn’t a maintenance issue, as previously thought. The engine actually failed because of a phenomenon called “cold dwell fatigue,” which caused the metal in the engine’s fan to fail far sooner than anticipated. Part of the problem could have stemmed from the fact that jet engine designers didn’t fully understand the limits and weaknesses of the type of titanium — called Ti-6-4 — that was used in this engine. The material is also extremely common across the aerospace industry.
In fact, according to the final BEA, report, “the mechanisms at the origin of the initiation of a cold dwell fatigue crack were still not completely understood at the time of the accident and are still not understood today.”