The Mystery of Flight 427

by Bill Adair

  She would reply, “I don’t care.” She was afraid the investigation was going to destroy little bits and pieces of the man she loved until there was nothing left.

  People from the NTSB and Boeing called him at home, day or night. Everything was an emergency, even when it really wasn’t. It reminded Dedik of the story of the little boy who cried wolf. These guys—and they were nearly all guys—were always crying wolf. Worse, they treated her like she was Haueter’s secretary. It infuriated her. “I don’t care about the office, I don’t care about 427, I don’t care about anything,” she told him one night. “The victims are dead. There is nothing you can do about it. You know what? It’s not going to make any difference whether you solve this today or tomorrow. There is nothing that is so important that you have to deal with it right now.”

  She had an important job, too, keeping the world safe from a nuclear disaster. She worked hard and traveled the globe to negotiate with other countries and make sure they were complying with international agreements. But she didn’t obsess about her job, and she knew how to draw a line between work and the rest of her life. When she left the office, she left her job behind. That was one of the things that she had liked about Haueter when they met—he had a life outside his job. He was different from the other Washington men. But Flight 427 had changed him.

  Some nights when he arrived home, she would tell him not to say a word about the investigation. Other nights she would give him ten minutes to talk and then make him promise not to bring it up again.

  She asked him once, “Is it so awful to have an investigation unsolved? Does that mean you’re a failure?”

  “Yes,” he said.

  Some people in aviation, and a few at the board, thought that too much emphasis was being put on coming up with the probable cause. To them, the NTSB frequently got tied in knots trying to find the precise cause of a crash and did not focus enough on preventing future accidents. The reformers, led by retired NTSB aviation safety chief C. O. Miller, wanted the board to issue “findings” after each crash, with the emphasis on preventing accidents and reducing hazards.

  But to Haueter and most others at the NTSB, it was important to name the probable cause. They thought that the public needed answers about why a plane crashed. It gave everyone a sense of closure, and it gave people confidence that it was safe to fly. If the board didn’t name a probable cause in USAir 427, people would wonder for years whether the 737 was safe.

  Haueter was a long way from solving the mystery, however. It was nine months after the crash, and the investigation had stalled. They’d done hundreds of tests on the rudder system and come up empty. Many other theories had fizzled. Several colleagues at the safety board thought Haueter should admit he was stumped and give up.

  With no new leads, Haueter went back to old theories that had been ruled out. It was a sign of how desperate he had become. Leads that had been dismissed nine months earlier were alive again. He was willing to consider anything, even if it did not involve the rudder.

  Brenner and operations group chairman Chuck Leonard conducted a second interview with Fred Piccirilli, a witness on the soccer field who thought he saw smoke coming from the plane before it hit the ground. The wreckage showed no sign of an in-flight fire, but Piccirilli had credibility because he was a USAir maintenance employee. He had come up to Hall at the hearing in Pittsburgh and told him about the smoke, so Hall asked the investigators to talk to him one more time. In the second interview, he again said the smoke was “orangish-reddish-brownish” and was coming from a spot in front of the right wing. He saw no fire but said the smoke remained in the air after impact and dissipated slowly. His account hadn’t changed much from the first time. The problem was, there was not a single piece of evidence that backed up what he said.

  The bird theory had also been born again. Never mind that no one had found a single feather in the wreckage or that Roxie Laybourne had conclusively ruled out the suspicious clump. The Boeing investigators had persisted about the theory, saying it was possible a bird had broken through the nose of the plane and hit a rudder cable. They wanted to do one final check of the wreckage with a black light to look for bird remains. This would be their last chance because the NTSB was about to release control of the wreckage to AAU, USAir’s insurance company, which planned to put the pieces in crates and move them to a warehouse.

  Supplee, the USAir mechanic, had been summoned back to the Pittsburgh hangar to help conduct the final bird check. He thought the whole exercise was a waste of time. They had already ruled out birds, yet Boeing persisted with any theory, no matter how absurd, that would clear the airplane.

  When he walked into the hangar, Supplee noticed that the musty smell of chlorine and jet fuel had faded since his last visit six months earlier. The windows had again been covered so his team could work in darkness. They donned the ridiculous-looking yellow-orange glasses that supposedly made it easier to see the glow from bird remains. They had two black lights, so they split into teams and started on different sides of the wreckage. Supplee and a forensic expert from the Armed Forces Institute of Pathology went to the place where the leading edge of the wings had been partially reassembled and waved the light over it. Nothing glowed. They moved to other piles of wreckage. Supplee picked up each piece, and the guy from AFIP waved the light over it. Still no glow. They checked hundreds of pieces and did not see anything glow.

  Suddenly they heard a shout from across the hangar, where a Boeing team member had a second black light.

  “Guys! Look at this!”

  Everyone hurried over. They could see a piece of wreckage glowing like a Jimi Hendrix poster in a dorm room. If this was a bird, it must have been a big one. How could they have missed it before? Then they removed their glasses and shone a flashlight on the wreckage. It wasn’t a bird. It was the fluorescent paint they had used to draw grid lines at the crash site. They must have painted the wreckage by mistake.

  They laughed and went back to the tedious job of inspecting each piece. A little while later, the AFIP expert noticed a slight glow on one piece. “Might have something here,” he said. It was shaped like an X, about one and a half inches long and half an inch wide. Supplee could see it clearly through his glasses. They took a Q-tip and a solution and swabbed the area gently to get a sample. When they added a solution to test for organic materials, it glowed slightly.

  But lab tests were negative. Once again, birds had been ruled out.

  Some days for Haueter went like this:

  9 A.M. Boeing calls and whines about the investigation.

  10 A.M. ALPA calls and whines.

  11 A.M. It is USAir’s turn, followed by a second Boeing whining session after lunch. Haueter wanted to shout: “Give me a break!”

  He believed in the party system, but on many days the parties behaved like children. In the absence of solid answers, they had retreated to positions that protected their own interests. Boeing and Parker Hannifin, the manufacturer of the PCU, saw no evidence that the rudder system had malfunctioned. They still wanted Haueter to spend more time scrutinizing the pilots. But ALPA and USAir saw no evidence that the pilots had made a mistake. They wanted Haueter to conduct more tests on the rudder system.

  Haueter was frustrated by the lack of progress. Many clues pointed to the rudder system, but he and Phillips could not find any proof that it had failed. The power control unit not only passed every test, it seemed to have the strength of a superhero. So, in May 1995, Haueter decided they should step back and take a broader look at the plane, instead of just focusing on the rudder. Maybe they would discover some wild idea that would lead them to the solution.

  He, Phillips, and Vikki Anderson, the lead investigator for the FAA, flew to Greensboro, North Carolina, and rented a car to drive to Winston-Salem, where USAir had a big maintenance hangar. The tan-colored hangar was big enough to hold three 737s for inspections that USAir called Q checks, the major overhauls where planes were stripped to their frames so mechanics could r
eplace anything that did not meet FAA standards. A 737 got a Q check every 11,000 flight hours, or roughly once every three to four years. It was a great place to look for inspiration because everything in the planes was exposed. The 737s looked like they were naked.

  A steady rain fell outside the hangar as they wandered around one of the stripped planes, looking at the maze of cables and wires that were normally hidden by the floorboards and the aluminum skin. They scrutinized the wing, trying to see anything that might have caused the smoke the witness had reported, and crawled into one of the baggage compartments. Phillips was struck by how much older planes smelled like coffee. They carried so many thousands of gallons of coffee over the years that it seemed to permeate everything. It was a familiar, comfortable smell.

  They sat in the cramped baggage compartment and talked about the possibility that an obese passenger fell through the floor and landed on the rudder cable—an idea that had been dubbed “the fat guy theory.” Would the fat guy have pulled on the cable enough to move the rudder? Probably not, but it might be worth testing, Phillips said. They looked at the plane’s tail and discussed what might have happened if a bird had struck the vertical fin. Could that have turned the rudder?

  USAir mechanics had tagged along to answer questions. “Do you have any ideas?” Anderson asked them. “Is there anything, no matter how far out, that you have noticed about the plane?” They had a few suggestions, but they were all theories that the NTSB had considered before. Anderson went up to the cockpit and sat in one of the seats. She stepped on the rudder pedals, turned the wheel, and moved the control column up and back. She was struck that it was such a simple airplane, about as sophisticated as an old VW Beetle.

  She and Phillips went to the wheel wells, where the hydraulic lines converged and where the landing gear was stored in flight. A USAir mechanic who accompanied them was concerned about the vulnerability of the hydraulic lines there. If a tire blew out, it could do serious damage, he said. Boeing had once had screens that protected these fragile components, but the company had stopped using them.

  “I’ve always wondered about things being in such a small place,” the mechanic said.

  “I have, too,” Phillips said. But he had found no evidence of such an explosion in the wreckage.

  They spent lots of time around the tail, looking at the power control unit and how it fit inside the vertical fin. Could a bird have broken through the tail and hit the cables or rods that led to the PCU? Haueter doubted it. The angle of the vertical fin would make it hard for a bird to break through. It would be deflected before it could pierce the skin.

  Rain was pouring down as they had lunch at a pub near the hangar and then took a commuter flight back to Washington. It had been a good chance to see the innards of the plane, but they felt no closer to knowing what had happened.

  Boeing had a theory about the pilots’ feet. The company wanted to test the damaged pilot seat tracks to determine if they indicated where Emmett and Germano were sitting. The pilots might have been so far from the rudder pedals that they had to stretch to use them. Maybe that caused one of the pilots to push too long and hard on one of the pedals.

  The seat tracks, metal strips that allowed the pilots to slide their seats forward and backward, were brought to the NTSB’s metals lab in Washington. Metallurgists inspected the tracks, but they were so mangled that the experts could not find proof of whether the seats were too far back. Besides, the position of the seat was not important, since each pilot had a crank that adjusted the rudder pedals forward and backward, so even when a pilot had the seat back as far as it would go, the pedals could be in the proper place. At six-three, Emmett was so tall that it was logical that he would have his seat back.

  Cox felt that Boeing was going to ridiculous extremes to blame the pilots. “Boeing is desperately trying to do anything they can to clear their airplane,” he said. He was resolute that the pilots were not at fault. “I still think it’s a systems problem with the airplane.”

  A few days later Cox and the other investigators flew to the NASA-Ames Research Center near San Francisco to take a ride in one of the world’s most advanced flight simulators. The idea—also suggested by Boeing—was to see if Germano and Emmett had been so startled by the wake turbulence that they could not recover the plane.

  The NASA vertical motion simulator, or VMS, was part of SimLab, the world’s most advanced laboratory for studying pilots. The lab gave out patches with the simulator’s logo, which had four stars and arrows pointing in every direction. It looked like a recipe for motion sickness.

  The VMS resembled a wandering elevator inside a cavernous ten-story building. Like Boeing’s M-Cab simulator, it could be programmed to perform like many different aircraft—fighter planes, big transport jets, even the space shuttle. It felt more realistic than other simulators because the cab could go up and down sixty feet, compared with just a few feet on M-Cab. That gave pilots a better illusion of the ups and downs in a plane. For the Flight 427 test, airsickness bags had been taped to the walls, just in case.

  Malcolm Cohen, a NASA expert on the inner ear, had been invited to determine if the pilots had lost their bearings when the plane started to flip. He rode twelve times—several with his eyes closed—and was surprised at how smooth the ride was. He said it was so smooth that the pilots should not have been disoriented. Cox had watched the simulator go back and forth a few times before his turn and said it would be “an E-ticket ride,” like the most thrilling ones at Disney World. He then walked across the ramp, buckled himself into the left pilot’s seat, and put on the headphones. The safety board had brought along the cockpit tape so riders could listen to the crew.

  The cab moved down to one side of the big tower, as if the NASA people were getting ready to fire it from a slingshot. The ride began. It was much smoother than M-Cab, without the jerks of the mechanical stops. Again and again, Cox rode the simulator, concentrating on a different element each time. He watched the instrument panel on some rides and looked out the windows on others. The simulator’s windows showed a computerized display of the ground and sky that changed as the plane moved.

  He had seen a similar display in Boeing’s M-Cab, but he noticed something new this time. He was surprised at how quickly the plane went nose down and the windscreen filled with the ground. It seemed like the pilots were helpless to do anything. To Cox, it was more evidence that the crew was fighting to survive, but the plane was not responding.

  Haueter hardly ever got scared when he flew, but he did get the jitters during a flight in a USAir plane in the summer of 1995. The plane was just like Ship 513, a 737–300. It was approaching Washington National Airport when it hit turbulence. Startled, Haueter sat up straight and quickly looked around to find the horizon through the windows, to see if the wing had rolled down too far. What was going on? Had the plane had a rudder hardover? Unfortunately, the people beside him had pulled down the window shade. He looked out the windows on the other side, but he couldn’t tell how far the plane had rolled. A quick thought flashed through his mind: Wouldn’t it be ironic if the chief investigator for 737 crashes was killed in one?

  The plane leveled off, and Haueter breathed a sigh of relief—but he vowed to get a window seat in the future.

  A big reason for his jitters was that he felt the 737 needed immediate safety fixes. They had not come up with the probable cause, but he and Phillips felt they had found enough problems with the rudder system to ask the FAA to mandate some improvements. That was typical in a high-profile investigation. The NTSB often made safety recommendations long before it determined the probable cause. Phillips had written a fifty-page memo calling for eighteen safety improvements, most involving the rudder. The memo said the FAA should immediately require Boeing to devise a procedure for pilots to handle a rudder hardover. Phillips was concerned that pilots would be caught by surprise if they had an incident and wouldn’t know how to respond. The memo also called for long-term design changes to the PCU to preven
t a hardover.

  In response to Boeing’s inability to track rudder problems with its databases, Phillips also called for a joint program between government and industry to keep a database on maintenance and operational problems. He had been surprised that a company as sophisticated as Boeing could not easily list 737 rudder incidents.

  Haueter thought the memo made a good case for the changes. He also wanted to get a jump on the FAA, which was writing its own safety study on the 737 and relying heavily on what Phillips had learned. Haueter did not want the FAA to get credit for the NTSB’s work. “These guys are going to beat us to the punch with our data,” Haueter told Bud Laynor, the NTSB’s deputy chief of aviation safety.

  Laynor was a navy-trained pilot and engineer who had designed flight control systems on airplanes and NASA spaceships. At sixty-one years of age, he was in great physical shape, with short brown hair cut like he was still at the Patuxent River Naval Air Station and a thin, leathery face with creases in the cheeks. Laynor was the most respected technical expert at the board, so his approval for Phillips’s memo was crucial. Without it, the memo would go nowhere.

  When Phillips wrote it in the spring of 1995, he figured it would go through the usual editing by NTSB managers and would emerge essentially intact. But after a few weeks he was surprised that nothing had happened. The memo seemed to be stuck on Laynor’s desk. When he saw Laynor in the hallway one day, Phillips mentioned the memo. “What about those recs?” Phillips asked.

  “I’ll get to them,” Laynor said.

  But when he did get to them, Laynor decided the memo was premature. The 737 had more than 60 million flight hours, and not even one crash had been linked to the alleged rudder problem. The NTSB had no evidence that anything in the PCU had malfunctioned. Without proof, he felt it was premature to say there was a problem with it. Besides, the valve had a built-in backup. If one slide jammed, there was a second slide to oppose the jam so the rudder could still move or return to neutral.