In the meantime, Flight 427’s power control unit continued to pass every test. Samples of hydraulic fluid from the PCU, along with samples from other USAir, Southwest Airlines, and United Airlines planes, were sent out for testing. The safety board even spiked one sample with bleach to make sure technicians from Monsanto did not tamper with the results. The Flight 427 samples passed, although several of them were unusually dirty.
Technicians at Boeing tried a test using a typical rudder valve and hydraulic fluid so gummed up with particulates that it looked like Dijon mustard. The contamination was much worse than anything the investigators had found on the plane, but they wanted to see how much the valve could endure before it malfunctioned. The fluid was so thick that it kept destroying Boeing’s hydraulic pumps. But in test after test, for thirty hours, the equivalent of stomping on the rudder pedal five thousand times, the valve worked fine.
The bird theory was still alive, mostly because Boeing was so insistent about it. Members of the structures group were back at the Pittsburgh hangar, where the windows had been covered with black paper. They shut off the lights, donned strange-looking orange sunglasses, and shone a black light along the wreckage. The sunglasses were supposed to make it easier to see the faint glow from bird remains. It was a bizarre scene: people wearing weird sunglasses walking slowly around the wreckage, waving a black light the size of a baseball bat.
On the number three leading edge slat, they saw a faint glow. They brought the light closer. It wasn’t bright, but it was definitely stronger than the surrounding area. They switched on the hangar lights and examined the spot carefully. It was a small clump of material that looked like dirt or mud. It wasn’t bird proof by any means, but it was worth testing.
It was time to call Roxie Laybourne, the world’s premier expert on feathers.
Laybourne, who was eighty-four years old, was a legend among crash investigators. She was a tiny woman with short gray hair and thick eyeglasses who worked at the Smithsonian Institution’s National Museum of Natural History. She wore a white lab coat and Reeboks, and she walked hunched over because of back and neck problems. She had begun studying feathers in 1960 after a Lockheed Electra collided with a flock of starlings in Boston, killing sixty people. Suddenly the government cared a lot about birds.
Identifying them was an odd but necessary task in aviation. Engine and airplane manufacturers needed to know how well their products could take a direct hit from a laughing gull, a Cuban whistling duck, or—their biggest fear—a Canada goose. Once they got Laybourne’s reports, they knew how much damage different birds could cause, which enabled them to improve their products to better withstand a bird strike.
Examining feathers under her microscope, Laybourne had discovered unique patterns in the stringy fragments of feathers called downy barbules. Some barbules had little triangles, others had rings. She found similarities between birds from the same family, even those that lived thousands of miles apart. Pigeons and doves were easiest to identify because the nodes on their barbules looked like crocuses. The hardest were songbirds because their nodes all looked alike.
With tiny feather fragments—some no bigger than bread crumbs—Laybourne solved thousands of bird strike cases. In one, she used a fragment of down from a pilot’s shoulder patch to identify the herring gull that broke through the canopy of a Harrier military jet. She also used her skills to help Customs identify birds smuggled into the United States, and she once helped the FBI catch a murder suspect.
Laybourne’s office at the museum was a reflection of her unusual occupation. Her bookshelf included Birds of Nepal and Raptors. On her door were a poster from the Israeli Air Force warning pilots to watch out for birds and a Far Side cartoon that showed Santa Claus and his reindeer splattered on the nose of a jumbo jet. She was philosophical about the conflict in the skies. “As long as you have man and birds flying,” she said, “you have the potential for problems.”
Cindy Keegan, the head of the airplane structures group for the Flight 427 investigation, brought the suspicious clump to the museum. But when Laybourne saw the tiny brown sample, she doubted she could be much help. It was smaller than a dime, a mixture of sand and dirt but nothing that even remotely looked like a feather. She eyeballed it for a moment and then examined it under the magnifying glass she wore around her neck. Nothing. She pulled off the cover on the microscope, put a sample on a slide, and took a closer look. Still nothing.
The next day, two other bird strike experts happened to stop by her office. One, Major Ron Merritt, headed the Air Force’s famous BASH Team. The other, Eugene LeBoeuf, was the FAA’s chief bird scientist. After examining the sample, they agreed that it was only dirt and vegetation.
Laybourne phoned the NTSB. “There’s no bird material here,” she said.
11. BACKDRIVE
The data recorder on Flight 427 was like an eyewitness to the crash with one eye closed. It took only thirteen measurements of the flight, which provided an incomplete picture of what happened. It saw the plane yaw to the left and then roll out of the sky, but it didn’t see why. That meant the NTSB and Boeing had to make educated guesses about what had happened. Investigators were pretty sure the sudden yaw was caused by the rudder. No other flight control on a 737 could cause that kind of movement. But they did not know whether the rudder moved because the pilots stomped on the pedals or because something went wrong in the hydraulic system.
That had become the central question in the investigation: Was the crash caused by man or machine? As other theories were ruled out, Boeing and ALPA began to differ about why the rudder had moved. Boeing kept raising the possibility that the pilots had mistakenly stomped on the pedal, whereas ALPA kept suggesting the rudder system had malfunctioned.
The small number of measurements on flight data recorders was a function of money. If it were up to the NTSB, every airplane in the United States would be required to have 300-parameter boxes immediately. But the airlines balked at that request because it cost tens of thousands of dollars to equip each plane with the more advanced recorders. It wasn’t that the boxes were any more expensive. The recorder on Flight 427 could easily have handled dozens of additional parameters. It was the multitude of new sensors and wiring that ran up the cost. Because the 737 was a relatively basic airplane with steel cables that moved the flight controls, it was costly to install sensors. A newer, fly-by-wire plane could be equipped with a sophisticated flight recorder more cheaply because the sensors and computers were already there.
Boeing assigned some of its smartest aerodynamic engineers to perform a kinematic analysis of the crash. Their goal was to take the undisputed facts from the thirteen measurements in the flight recorder and come up with estimates for things the box did not record, such as movement of the wheel and the rudder. The job was even tougher because the plane had been bounced around by the wake from the Delta 727. There was no doubt that the wake had jostled the USAir plane, but it was up to the engineers to figure out which movements were caused by the wake and which were caused by the ailerons and rudder.