by Bill Adair
The government issued its first requirement for planes to have recorders in 1941, but the order was rescinded because of maintenance problems and poor reliability with the early boxes. The pilots union, ALPA, fought against having them on commercial planes because of fears that the recorders would be used as mechanical spies. But finally the union relented, partly because a recorder cleared an ALPA pilot who was falsely accused of flying too low. In 1957 the government issued another mandate that planes be equipped with recorders measuring airspeed, altitude, heading, and vertical acceleration. The boxes were primitive—a stylus moved up and down, scratching continuous lines on a strip of foil—but many of them survived crashes and provided valuable clues.
Today, modern recorders store their data on a durable computer chip that can take hundreds of measurements. It records basic parameters such as airspeed and altitude, and it shows what the pilots were doing—whether they were pushing on the rudder pedals or turning the wheel. That information can be especially valuable because it answers the man-or-machine riddle of many crashes.
The recorder on the USAir plane had only thirteen parameters—altitude, airspeed, heading, pitch (whether the nose was pointing up or down), roll (whether the wings were level or rolling down to the left or right), and engine power. It had only two measurements that told what the pilots were doing in the cockpit. One showed when they were pushing the button to talk with air traffic controllers, which allowed investigators to synchronize the flight data recorder with tapes from the CVR and the Pittsburgh control tower. The other showed whether they were pulling or pushing on the control column, the “stick” that made the plane climb or descend. The recorder did not measure what was happening with the rudder or whether the pilots were pushing on the rudder pedals. Haueter’s investigators would have to figure that out by themselves.
The labs of the NTSB are messy places. In the metals lab, twisted pieces of airplane wreckage are spread on a countertop like body parts awaiting an autopsy. In the flight recorder lab, mangled orange boxes are piled on a table, many still caked with mud. On a nearby wall is a bank of gadgets that look like a dozen VCRS—computers used to download the information from flight recorders. Another computer can convert the data into a color animation, to show a plane crash like a Saturday morning cartoon.
Technicians in the lab could see that the data recorder from the USAir plane was badly damaged. Dirt and yellow insulation from the plane had gotten inside the box when the 737 struck the hill. The steel cocoon that protected the computer chip had broken away from its mounts and smashed the circuit boards inside the recorder. But the cocoon had done its job. The data were fine. The technicians transferred the data into a computer, converted the raw numbers into rows and columns that were easier to read, and zapped it all by modem to the Pittsburgh command center at the Holiday Inn.
The command center had become a chaotic place. The phones rang constantly with calls from witnesses and others with theories about the crash. A swarm of people converged on Haueter every time he walked into the room, bombarding him with questions about computers, meeting times, phone calls, logistical arrangements. He wondered if he would ever get a chance to actually investigate the crash.
The first person at the Holiday Inn to see the data from Flight 427 was John Clark, a white-haired NTSB engineer. He sat cross-legged on the floor in a corner of the room, studying the results on his laptop computer. The numbers showed the plane was descending from an altitude of 5,984 feet when the left wing dipped. The wing stayed down for about fourteen seconds, then started to level off, then rolled down again. The nose had been up slightly when the wing dipped, but the nose quickly plunged toward the ground. The vertical Gs—the forces of gravity on the plane—told a frightening story. As the plane spiraled toward the rocky hill, the centrifugal force on the passengers and crew reached nearly 4 Gs. That meant a two-hundred-pound person like Emmett would have felt like he weighed eight hundred pounds. The numbers showed how quickly the pilots lost control. Just twenty-eight seconds elapsed from the first hint of trouble until impact.
As Clark looked down the column for the plane’s heading, he saw something unusual—an abrupt change, which meant the big 737 had suddenly yawed to the left like a car beginning to skid sideways on a wet road. Other measurements showed that a split second later the left wing rolled toward the ground and the plane plunged nose down. Clark knew many things could make a plane yaw and roll like that, but the most likely was a sudden move by the rudder.
He walked over to Haueter. The data were still rough, he said, but the big shift in heading was significant. “There is something going on here with the yaw,” he told Haueter. “It looks like this airplane had some type of rudder event.”
It was an encouraging lead. But NTSB investigators have an old saying: Never believe anything you hear in the first forty-eight hours. Early clues can be overrun by new evidence. The first few theories about a crash—known affectionately as the causes du jour—often don’t pan out.
Still, Haueter felt confident that he would be able to solve the mystery. It was only Day 1. They had good data and a clear CVR. They were making progress.
Before the crash, the hill off Green Garden Road had been a peaceful retreat from city life. It was a thirty-minute drive from downtown Pittsburgh, with thick woods separating it from the steady traffic on Route 60. Children often picked wildflowers in the meadows. Deer wandered through the woods and drank from a creek at the bottom of the hill. The only signs of urban life were the USAir jets that flew overhead, making their final approach to the Pittsburgh airport.
Flight 427 crashed on secluded land owned by George and Mildred Pecoraro, who had lived there for nearly thirty years. They had been displaced before. They lost access to the land when Route 60 was built in the early 1970s, but they moved back after they bought right-of-way and built a dirt road in 1981. They lived in a two-story house at the top of the hill, about one-fourth of a mile from where the 737 crashed. The night of the tragedy, they ended up in a nearby Hampton Inn and weren’t fazed by the fact that they were assigned to Room 427. Mildred said they weren’t superstitious.
Thousands of fragments from the big plane were blown hundreds of yards away, into fields owned by George David, a police officer in nearby Aliquippa. David grew hay on his 61 acres and loved the solitude of the place. The deer were so friendly they would eat apples right out of his hand. But now, the day after the crash, his peaceful hill looked like the site of a military invasion. Yellow and red police tape was strung around the trees. Helicopters pounded overhead as trucks from the National Guard, the Salvation Army, USAir, and Allegheny County brought supplies and volunteers. Tents were set up along the dirt road as field offices for the Beaver County coroner.
Down the hill, the Green Garden Plaza shopping center had become the nerve center for the crash, with TV satellite trucks parked bumper to bumper and more than two hundred reporters crowding around the command center. The Green Garden merchants all pitched in. The Hills department store gave blankets and other supplies. The Chevy dealer became a temporary headquarters for the Hopewell government. Stress debriefing was available at the New York Pizza Shop.
When Haueter saw the body parts scattered around the hill, he decided to treat the site as a biohazard area. Investigators had rarely worried about diseases before, but he had just taken the government’s training on biohazards and felt there was enough danger from blood and fluids to justify employing the full OSHA protections. The investigators would have to wear plastic suits.
Several local officials disagreed. Beaver County coroner Wayne Tatalovich told Haueter the plastic suits weren’t necessary. Sure, there were lots of fragmented bodies, Tatalovich said, but there wasn’t much blood. He felt the site would not be any more dangerous than a morgue. They argued for a while, but Tatalovich wouldn’t budge. He said Haueter’s team could wear the suits, but the coroners would not. Haueter warned everyone at the first meeting, “This is a biohazard zone. All safety board e
mployees will have to respect that. I can’t force anyone else to wear them, but it’s a good idea.”
When the USAir plane plowed into the gravel road at 300 miles per hour, it shattered like a crystal vase thrown on a concrete driveway. The 109-foot plane splintered into hundreds of thousands of tiny fragments, many no larger than a plane ticket. The USAir logo was usually found on hundreds of items in the plane, but one of the airline’s mechanics noticed an odd pattern to the logos he found in the wreckage. They said “US” or “USA” or “Air,” but he could not find any logos that were intact. Everywhere he looked, USAir had been torn apart.
The site was littered with seat cushions, hundreds of shoes, and thousands of Business Week with the headline THE GLOBAL INVESTOR on the cover. The magazines were everywhere.
Passengers’ belongings were scattered through the woods and on the road. It was as if the crash had taken a snapshot of each person’s life, revealing that person through his or her possessions. There were plaid boxer shorts and another pair with red diamonds; sweatshirts from Purdue University and the New York Renaissance Festival; T-shirts for Hooters, Soldier Field, the Chicago Bears, Harley-Davidson, and Bugs Bunny. There were lots of mangled and burned books: Forrest Gump, the Pocket Prayer Book, Rush Limbaugh’s The Way Things Ought to Be, The Chamber, by John Grisham, a management training manual called Tiring Up Commitments During Organizational Change, and a copy of the Bible. Investigators also found lots of everyday stuff: a garage door opener, family snapshots, a Swiss Army knife, pocket calculators, Kodak film, a rosary, and a teddy bear.
While many of the local volunteers were vomiting in the woods or sobbing at the horror, Cox remained unfazed. He had been to many crash sites as an investigator for the pilots union, and he knew the smells and didn’t dwell on the body parts. He was there to look at wreckage, not people. He had been appointed the ALPA representative on the systems group, which was shaping up as the most important group in Haueter’s investigation. The systems members would examine the plane’s flight controls and hydraulics to see if they had caused the crash.
Each morning Cox joined the other members at Green Garden Plaza to be sealed inside his rubber suit. It was like getting dressed for surgery. They had latex gloves, boots, and surgical masks. The boots and gloves had to be taped to the suits, which made the outfits unbearably hot. With everyone wearing an identical white rubber suit, it was also difficult to tell people apart. They eventually wrote their names across their backs, as if they were wearing football jerseys. Everyone involved in the investigation also had to wear a colored bracelet to get access to the site. To foil trespassers, the color changed every day. Several people had been arrested trying to sneak onto the hill to take pictures of the wreckage and the body parts.
Cox’s first assignment was to pick through the flattened wreckage of the cockpit. Getting to it was difficult because of the trees and hilly terrain, so the systems group enlisted the help of the Allegheny County Delta Team, a paramilitary group of public works employees who responded to the crash like they were invading Kuwait. “You want a road? We’ll build you a road,” said one Delta member cheerfully. A few hours later, there was a gravel road straight to the cockpit.
Cox found that picking through the wreckage wasn’t easy. The investigators had to use a pulley on a big metal frame to lift the largest pieces. Some were buried several feet underground. Others had been flattened like aluminum cans. Much of the wreckage was buried beneath a thick layer of wire that looked like burned spaghetti. Their first priority was to see what they could learn from the gauges and switches. Cox was especially interested in finding bulbs from the cockpit warning lights. Lightbulb filaments stretch when they get hot, so the investigators could tell if a warning light had been on by measuring the filament. But Cox discovered that every light had been shattered.
The softest things in Ship 513 had survived with the least damage—seat cushions, handbags, and hundreds of shoes looked fine. But the rest of the plane was torn apart and hard to identify.
“That looks like junk,” said one of the Delta Team members, pointing to some twisted metal.
“It’s a nose-gear strut,” said Cox.
The gauges provided a few clues. The captain’s airspeed indicator was covered with mud, but when Cox cleaned it off, he saw the needle had stopped at 264 knots—the plane’s speed when it hit. A needle on the hydraulic pressure gauge indicated that the B system was at 3,100 pounds, which told Cox that it had full power when the plane crashed. The plane’s hydraulics are crucial because they move the landing gear and flight controls such as the rudder, elevator, ailerons, and spoilers. The needle for the second hydraulic system—the A system, which moved the landing gear—was missing.
Elsewhere on the hill, other teams were finding more clues. The 737’s engines were badly damaged, but the members of the power plant group could see that the fan blades were bent opposite to the way they rotated. That meant the engines were running at impact, which ruled out the possibility that engine failure had caused the crash. Everyone looked for parts from another plane, on the theory that the big 737 might have collided with a Cessna or a Piper. But so far, none had been found.
Cox was perfect for the systems group. He was a 737 pilot who knew every inch of the cockpit and, like Haueter, he loved dissecting the mechanical and electrical systems that made the plane fly. He was not a do-it-yourselfer as Haueter was—Cox was away from home too much to have time to build things—but they shared a fascination with solving mechanical mysteries.
Cox was a meticulous guy who kept his life and cockpit carefully organized. His bookshelf was a reflection of his personality: All the Tom Clancy hardbacks were on one shelf, all the books about flying on another. Paperbacks were together, separated from the hardbacks. An errant copy of his wife’s Martha Stewart Weddings put in an appearance on the maritime shelf, but it didn’t stay long. Cox kept their finances on their home computer, and he maintained precise records about where their money went. When his wife, Jean, came back from shopping, she had to separate the expenses into categories such as Household and Gifts.
Whether he was flying a difficult approach into O’Hare on a stormy night, driving 100 miles per hour in his sports car, or just balancing his checkbook, Cox was in control. On his business card he was “Captain John Cox.” He had 8,000 hours flying 737s, and he loved the plane. He had flown lots of others in his twenty-four-year career, but he had always preferred the 737. He liked its smooth landings and the solid way it handled a crosswind. “The airplane tends to make you look good,” he said. He loved the challenge of mastering a machine, maneuvering the fifty-ton bird through winds and clouds and heavy rain and still managing to touch down so gently that the passengers in back could barely feel it.
Cox also craved speed. The speedometer on his fire-engine-red Acura NSX went to 180 miles per hour. He had gotten the needle up to 125. He referred to the $80,000 sports car as “the toy,” but he treated it with reverence. When he parked at a store or restaurant, he put it in a remote corner of the lot, parked at an angle across two spaces so no one would nick his doors. He kept a cloth cover on the car, even when it was inside his garage. When he removed the cover, he folded it up carefully, one side at a time, as if he was folding a flag.
The son of a Birmingham banker, he grew up in a family that had no connection to aviation. But he got interested in airplanes as a toddler. One of his first words, uttered when he was two, was “Constellation,” the big plane that he watched taking off from the Birmingham airport. He got his private pilot’s license at age seventeen, flew charters and corporate planes at eighteen, and then joined Piedmont Airlines at twenty-six. He became a USAir pilot when the two airlines merged in 1989.
Cox was trim, with the graying hair, silver moustache, and tanned good looks that seemed standard issue for an airline pilot. He was one of ALPA’s technical experts, a rare pilot who understood the complex engineering of the planes he flew. Even his doodling was intricate. The margins of h
is notepads were filled with complex geometric figures that looked like M. C. Escher drawings.
His union had a Jekyll-and-Hyde reputation in aviation safety. C. O. Miller, a former NTSB official, often said that ALPA had done more to promote safety in the skies than any group except the federal government. The union had helped design the national air traffic control network and the instrument landing systems that guide planes toward a runway. It played a big role in changing cockpit design to reduce pilot mistakes (the easy-to-read T design on the instrument panel is one of its legacies), and it was a strong proponent of crew resource management, which has improved communication in the cockpit.
But so far as some critics were concerned, ALPA had a reputation as a union that used safety as an excuse to get more money and generous work rules. Najeeb Halaby, a well-respected FAA administrator from the early 1960s, once said there were two ALPAS—the one that made substantial contributions to safety and the one that masked its economic demands “under the guise of safety.” The problem, Halaby said, was that he could never be sure which ALPA he was dealing with.
Among accident investigators, ALPA had a reputation for sometimes making excuses for pilots when there was overwhelming evidence that they had screwed up. The union would claim the pilots were influenced by the design of the plane or try to blame air traffic controllers or some mechanical problem. In some crashes, it was obvious that the pilots had made a stupid mistake—they simply forgot to set the flaps for takeoff or they flew into a bad storm. But ALPA would throw up smoke screens and make excuses.
Cox, however, was respected at the NTSB because he was not a strident unionist. He was regarded as one of the Young Turks at ALPA who were more like accident investigators than defenders of the pilot brotherhood. He had taken the highly regarded accident investigation course at the University of Southern California and followed its open-minded approach. “The evidence leads you where it leads you,” he often said. If that meant a pilot was at fault, so be it.
