by Bill Adair
Laynor didn’t argue much against Phillips’s memo, he just sat on it. The memo stayed on his desk, gathering dust. Phillips, normally one of the NTSB’s most cautious, gentle investigators, complained loudly to everyone that the fixes were crucial. These were not a bunch of crackpot recommendations, he said. There were persuasive arguments for each of them.
Haueter agreed. He felt so strongly about the safety fixes that he went over Laynor’s head to Hall. But Hall was unwilling to get involved. If Laynor wasn’t ready for the recommendations, no one was. The 737 safety fixes would have to wait.
Laynor was still interested in theories about the wake turbulence from the Delta plane. He said it was too much of a coincidence that Flight 427 had flown through the precise spot where the Delta plane had been seventy seconds earlier. He figured the wake had to play some role in the pilots’ loss of control. He kept pushing for a flight test with a 737 and a 727 to show whether the wake was strong enough to flip a plane.
To Haueter, it was ridiculous to think that a wake could flip a 737. He repeatedly said that planes would be falling out of the sky every day if that were true. But he knew that he had to test Laynor’s hypothesis. If he didn’t, the wake turbulence theory would haunt them for years.
The FAA had its own 727 that could be used for the tests, but Haueter found it surprisingly hard to get a 737. They were the best-selling jets in the history of aviation, so he figured that some owner somewhere would be willing to lease one for a month. But he kept getting turned down. The 737s owned by the major airlines were all booked, and they couldn’t spare one for a month, even though the NTSB was going to pay. Leasing companies had some available, but they didn’t want to help. They were afraid the plane might be damaged by the test and did not want the publicity linking their plane to one that killed 132 people. Haueter offered to paint it white so that no one would know the owner, but he still couldn’t get any takers.
Finally, he went to Hall and said he needed help. The test was worthwhile, Haueter said, but no one wanted to participate. Could Hall exercise some leverage and find a plane?
Hall got on the phone with FAA administrator David Hinson, and they jointly called USAir chairman Seth Schofield. Suddenly USAir changed its tune. The airline would loan a 737.
Everyone agreed that the $1 million cost of the tests would be split by the NTSB, the FAA, Boeing, and USAir. ALPA decided it could not afford to contribute, which Haueter thought was a bit cheap. Here’s a bunch of guys making $160,000 a year, but they couldn’t afford to support the tests?
16. THUMPS
Jim Cash had the creepiest job in aviation. As the NTSB’s expert on cockpit tapes, he listened to pilots die every day. He estimated that he had heard about three hundred die horrible deaths, plus several hundred more who had been fortunate enough to survive. It was fitting that Cash was shy and soft-spoken. He heard pilots hollering and screaming every day, but when he talked, his voice was so quiet that you had to strain to hear him. He was forty-three, with rosy cheeks and a boyish face. There were no aviators in the Cash family when he grew up in upstate New York—his dad had a blue-collar job at a power plant, his mom was a housewife—but Cash built model airplanes and got interested in flying. He was awarded an ROTC scholarship to Syracuse University and studied electrical engineering. It wasn’t that he really wanted to be an engineer, it was a matter of survival. He hated writing and liberal arts and figured his best hope for graduating was to be an engineer. He joined the air force and flew F-4 fighter planes, but you would never know it by talking with him. He had none of the ego or bluster of a fighter pilot. He was in the air force for eight years but chose to leave when he was due for a desk job.
He couldn’t find anything he liked at an aerospace company and stumbled across an opening in the NTSB sound lab. He had no experience listening to cockpit tapes (air force planes usually do not have them because the military does not want the tapes to fall into enemy hands), but his experience as a pilot and his skills as an electrical engineer made him ideal for the job. Like Haueter and Phillips and seemingly everyone else at the safety board, Cash loved to build and fix things. He built a pool in his backyard and spent four years restoring a ’72 Porsche 914.
Cash was renowned for finding clues in the tapes. He knew the distinctive clicks of a flap lever and the grind of the landing gear. By studying the whines of the engines, he could tell if they were operating at full power. His office at the NTSB was a darkened room that was stacked high with audio equipment—amplifiers, tape players, audio mixers, and graphic equalizers. But the place where he worked his magic was a powerful Unix computer that ran a program called WAVES. It allowed him to translate sounds into squiggly lines that showed frequency, volume, and energy. When he wanted to identify a strange click or rattle, Cash compared that fingerprint with one from a known sound. He spent lots of time recording sounds on airplanes on the ground so he could match them with the mysterious noises. He was like a police fingerprint expert, always looking for a match.
It had been ten months since the crash, and the NTSB still had not identified the thumps heard on the 427 cockpit tape. Cash had taken the tape to the FBI’S sound laboratory in Quantico, Virginia, to see if they came from gunshots or an explosion, but the FBI found that the thumps did not have the unique signatures of sounds from a gun or a bomb. Since then, Cash had spent hours studying the thumps on his computer. He discovered that there actually were three distinct sounds, the first coming just as Emmett completed the sentence “Oh, yeah, I see zuh Jetstream.” But Cash had been unable to match the signature with any of the sounds from the stomping-slamming doors test conducted at National Airport a few days after the crash.
If the thumps had been heard at any other time on the cockpit tape, Cash would not have worried about them. Tapes often had sounds that couldn’t be identified. Human beings and airplanes made so many noises that it would be difficult to identify them all. But Cash thought it was important to find the source of the thumps because they occurred at the most crucial time, right before the plane rolled left and twisted toward the ground.
It was fortunate that he had four channels of sounds on the tape. Most 737 tapes had only three—one for each pilot and one from the area microphone in the ceiling. But someone on Ship 513 had mistakenly left the jump seat mike on, which meant there was a fourth source of sounds.
As he studied the squiggly lines on his computer one day, Cash noticed that the thumps were picked up by the microphones in the ceiling and the jump seat but not by the headset mikes that Emmett and Germano were wearing. That meant the sound was probably traveling through the frame of the airplane rather than through the air, which suggested it came from outside the cockpit. He also noticed an important difference in the sounds: the ceiling mike picked up the first thump 9/100 of a second before the jump seat mike did. He thought that was an important clue.
Late one summer night, Cash and a few helpers arrived at the USAir terminal at National Airport. They boarded a 737 parked at a gate and switched on the plane’s power. As the voice recorder was running, an FAA employee crawled through the baggage compartment and pounded on different spots with a rubber mallet.
They used two walkie-talkies. The FAA employee announced his location right before he pounded, so they could log where the sounds came from. When Cash got back to his office and examined the sounds in WAVES, he checked the timing of each to see which ones had the same 9/100 of a second interval. He found a match when the FAA employee pounded on the fuselage about twelve to sixteen feet behind the cockpit mike. That meant the sound came from Row 1 or Row 2 in first class.
He was halfway there. He knew the location of the thumps. Now he just had to figure out what caused them.
Every time a 737 burped, Vikki Anderson’s phone rang. It didn’t have to be anything serious. The usual hiccups of a plane—minor autopilot problems, rudder anomalies, or bumps from wake turbulence—were now being reported by 737 pilots as potentially serious incidents that might be rel
ated to the USAir crash. Many pilots had become jittery about the 737.
As the FAA’s lead investigator on the crash, Anderson provided technical help to the NTSB, to explain how the FAA certified the 737 and how it inspected USAir. She also had to help the FAA decide if there was any need to take action because of the crash. She had been a Braniff flight attendant for twenty-three years and had joined the FAA because she had grown weary of the vicious cycles in the airline business, when Braniff would go bankrupt and then reemerge to fly again. She talked about the airline like it was a movie with sequels, Braniff I, Braniff II, et cetera.
She started with the FAA as a cabin safety specialist, analyzing flight attendant training and evacuation plans, and then moved to the agency’s accident investigation office nine months before the Hopewell crash. This was her first major accident. As she was flying to Pittsburgh the morning after the crash, she wondered if she could handle the gruesome job. She wasn’t sure she could deal with body parts lying everywhere. She had done well in the months since then because she was forthright about her feelings. While many men in the investigation bottled up their emotions, Anderson was honest about what she was going through. She was a fun person to work for, easygoing, thorough, and well organized.
Only a handful of women were part of the tight fraternity of crash investigators, but Anderson had worked hard to learn the ropes. She read every crash book she could find, and she wasn’t bashful about asking questions. She would freely admit when she was baffled by an engineer’s mumbo jumbo. She kept the FAA team organized and made sure meetings didn’t drag on too long. A tiny woman with big brown eyes and a warm sense of humor, she could liven up the most dreary meeting. When they discussed the gurgling sound that Andrew McKenna had heard in Seat 1A, Anderson suggested it was someone sitting in 3A gargling. When she visited the Winston-Salem hangar, she said she was open to all theories about the crash, “even the one about Russian death rays.”
The FAA had declared the 737 safe. An independent team of FAA engineers and inspectors had conducted a Critical Design Review of the plane for six months and found no major problem that could be linked to the accidents in Hopewell and Colorado Springs. The team listed twenty-seven recommendations to improve the plane, everything from new screens in the wheel wells to a consistent definition of the word “jam,” but it found no serious problems with the plane.
Yet, despite the FAA’s assurances about the plane’s safety, pilots remained nervous. In cockpits and airport hallways, 737 pilots eagerly swapped gossip about recent incidents and discussed what they would do if their planes suddenly began to roll. At least once a month, Anderson would get a call from the FAA security guard saying there was a pilot at the front door who wanted to talk to someone about the crash. Sometimes the pilots had theories, other times they just wanted to find out the latest from the investigation. One young pilot—Anderson doesn’t remember his name or his airline—was surprisingly nervous about the 737. He needed reassurance that the plane he flew was not going to fall out of the sky. Talking to him was eye-opening. It made her realize that some pilots were truly scared.
Anytime a pilot reported a problem that sounded the least bit serious, the FAA and the NTSB quickly sent a team to check it out. A July 1995 incident in Richmond, Virginia, was typical. Anderson’s boss, Bud Donner, interrupted her during a meeting and said, “Vikki, I’m going to need you in a little bit. We’ve got another little adventure.”
A USAir plane carrying forty-eight passengers was making a left turn toward the runway when it suddenly rolled to the right. The captain shut off the autopilot, but the roll continued. He quickly turned the wheel and pushed the left rudder pedal to regain control of the plane. The rolling stopped, the plane leveled off, and they landed safely.
Anderson and Donner discussed the event and decided it sounded serious enough to send people to Richmond. FAA investigator Jeff Rich quickly rented a car and raced down I-95 to the airport. The NTSB also dispatched someone.
The plane was still parked at the gate when Rich arrived. For three days, he and other investigators crowded into the cockpit and ran tests on the plane’s navigational computer, programming the computer so it thought the plane was flying and then checking to see how it might have sent a faulty signal. In the meantime, the plane’s flight recorder was sent back to Washington to be analyzed in the safety board’s lab. Ultimately, the roll was blamed on a faulty autopilot and on the pilots themselves, who did not realize the autopilot was steering the plane.
Two weeks later, Anderson’s home phone rang while she was in the shower bathing her cocker spaniel, Pepe Lopez (a dog her kids named after a brand of tequila). An Aviateca 737 that originated in Miami had crashed into the side of a volcano in El Salvador, killing all sixty-five people on board. The plane was making a final approach to San Salvador during bad weather. Early reports suggested that there weren’t many similarities with Flight 427, but no one could be sure. Clues to the Hopewell crash might be lying on the side of the volcano. Anderson packed her bags and flew to El Salvador.
Her first view of the crash came from the TV in her hotel room, which showed grisly pictures that would never be shown in the United States. The bodies were relatively intact, which told her this was a much different accident from Pittsburgh. The plane had probably struck the volcano with a glancing blow. The next day she got a Jeep ride as far up the 7,000-foot volcano as possible, but she had to climb the rest of the way in her blue FAA jumpsuit, wearing a 30-pound backpack loaded with tools, water, cameras, and biohazard gear. She had thought she was in great shape, a three-times-a-week runner, but the volcano was so steep that she had to crawl the last 50 feet on her hands and knees. When she reached the 6,400-foot level, where the plane had crashed, she lay on the ground gasping for breath, then looked up and saw an eighty-year-old Salvadoran woman grinning at her.
Anderson figured the woman had probably passed her going up the mountain. Maybe Anderson wasn’t in such good shape after all.
Anderson and investigators from Boeing, the NTSB, and the Salvadoran government climbed through the jungle and inspected the wreckage. The voice recorder was found and was sent back to the NTSB lab, but the plane’s data recorder was missing. Anderson figured a villager who lived near the volcano had probably stolen it, thinking it was some kind of safe box with money inside. That kind of looting was common at crashes in Third World countries. Investigators could often tell they were nearing a crash site because they started seeing people sitting outside their homes in airplane seats.
After analyzing the wreckage and the voice recorder and interviewing an air traffic controller, the investigators decided that the pilots were at fault. They had approached the airport from the wrong direction and had flown into the side of the volcano. The Richmond and El Salvador accidents ended up on a list with more than fifty other 737 incidents that were explained by mechanical malfunction, pilot mistakes, wake turbulence, or bird ingestion. They did not answer the riddle of Flight 427.
17. THE ANNIVERSARY
The Flight 427 Air Disaster Support League was making progress in its effort to improve treatment of families after a crash. On June 20, 1995, members of the league and three other crash groups met with Chairman Hall and Secretary of Transportation Federico Pena at the NTSB offices in Washington. The families from the other crashes had the same complaints as those from Flight 427: It took too long for the airline to figure out who had died, the coordinators for the airlines were inadequately trained, and the companies did a poor job of returning personal belongings.
Aides to Pena and Hall scribbled notes as the families listed their complaints. One of Pena’s aides kept reminding him to leave for another appointment, but he stayed for an additional forty-five minutes. He and Hall came away convinced that the government needed to help. They directed staff members to draw up legislation creating a new government office to assist families after a crash.
That summer, the Flight 427 Air Disaster Support League was organizing two da
ys of events to mark the first anniversary of the crash. The group planned to have four memorial services, including one at the crash site on the night of the anniversary, and a luncheon for the families the following day.
Brett had not been active in the league. He agreed with its goals, but he wanted to keep his distance. When he and Joan’s brother Dan Lahart arrived in Pittsburgh for the anniversary, they stayed at the same hotel as other league members did, but they planned to skip the meetings and all but one of the memorial services. When they ate dinner at the hotel restaurant the night before the anniversary, they sat in a dark corner so they could have privacy.
Brett was hardly thrilled to be there, but he wanted to pay tribute to Joan and attend a memorial service at the place where she had died. He had endured the most painful days in the past year—his birthday, Christmas, her birthday—and hoped they wouldn’t hurt as much in the future. He had not dated anyone since he lost Joan and was not sure he ever could. “If I meet someone in the future that I want to marry, she will know deep down inside that this would not be happening, we would not be having children, had I not lost the first love of my life,” he said.
Brett and Dan visited the coroner’s office in the morning, to check whether any of the unclaimed shoes or jewelry belonged to Joan. None did. Then they drove to Sewickley Cemetery to see the monument to the passengers. It was a beautiful day, just like the day Joan died. Workers were planting 132 tulips that would honor the victims. Brett and Dan got choked up. They chatted with a woman whose husband had been on the plane. She said she often visited the memorial in the early morning and found the cemetery very peaceful.