So Boeing had a lot at stake in the Flight 427 investigation. There now had been two mysterious 737 crashes with similar circumstances. Boeing had to reassure its jittery customers that its plane did not have a fatal flaw.
Some airlines were so nervous about the crash that they instantly reacted to every theory that came up. When the NTSB raised the possibility that one of the plane’s thrust reversers had suddenly deployed in flight, a few carriers immediately called and said they wanted to disconnect them. Boeing officials sent telex messages to the airlines nearly every day for the first two weeks, telling them in one message, “We would like to stress that, contrary to some media reports, at this point the NTSB has not concluded there was a partial in-flight thrust reverser deployment. Boeing has no recommended operator action at this time. If the investigation shows any specific actions are recommended or required, operators will be notified.”
Boeing had a special team to deal with crashes—Air Safety Investigations, which was a select group of engineers run by a shrewd manager named John Purvis. The team’s mission was to help the NTSB and determine whether Boeing needed to fix any problems with the plane. For Flight 427, Purvis’s office was overseeing an unprecedented effort involving more than a hundred Boeing employees. Engineers who designed flight controls were studying the 737 rudder system for possible flaws. Aerodynamic experts were studying the numbers from the flight data recorder, trying to estimate when and how far the rudder had moved. The company’s structures experts were helping the NTSB make sense of the wreckage.
The crash was so important to Boeing that the chief engineer for the 737, Jean McGrew, had been appointed to work with Purvis and oversee the company’s response. McGrew said his mission for Flight 427 was simple: to help the safety board find the cause of the crash, and—if the 737 was to blame—fix it. It bothered him that some people thought Boeing might be trying to cover up flaws in the plane to save money. McGrew said he was open to any theory, even if it meant Boeing might be at fault.
From his boyhood in Missoula, Montana, McGrew seemed destined to be an aviation engineer. He delivered the Missoulian newspaper and used his profits to buy model airplanes. He loved the challenge of designing and building something that could actually fly. He started his career with the Douglas Aircraft Company as a young expert on flutter, the potentially catastrophic condition that can make an aircraft shake so violently it breaks apart. He liked flutter because it was complicated. Some engineers preferred a single, more simple discipline, but McGrew chose flutter because it involved virtually every aspect of aerospace engineering—structural mechanics, stress analysis, design, and aerodynamics. The more complicated, the better.
He was an early computer expert who introduced Douglas engineers to the power of the PC. He rented an Apple II and holed up in a Manhattan hotel room, crunching numbers to prove that their new MD-80 jet could safely land on LaGuardia Airport’s runway pier. He also was one of the first guys at the company to have a computerized Rolodex.
In 1989 he retired from McDonnell Douglas and jumped to Boeing, which was the aerospace equivalent of switching from Pepsi to Coke. He became the chief engineer for the 737, which meant he was the plane’s godfather, responsible for any design changes.
He was so smart that his mind seemed to leap ahead during meetings, solving problems that weren’t even on the agenda. His mind was like a mainframe. He studied a problem, processed the data, and reached a conclusion. One of his kids once gave him a gold Slinky on a plaque that said, RELAX, a reminder that he often got wrapped up in his job, but he rarely had time to play with it.
McGrew had a wry sense of humor and a love of tennis, boating, and spicy food. He solved problems with a cool, analytical approach and calculated everything to close tolerances. He knew precisely how long it took to get to the airport and always cut it close so he wouldn’t be sitting at the gate wasting time. He fit well into the Boeing engineer-dominated culture. He looked the part—a thin fifty-six-year-old man with wire-frame glasses and a pocket protector that he kept hidden inside his shirt pocket, as if he were slightly embarrassed about it. He loved computers, but he was not the kind of guy who would surf the Web or play games. To McGrew, computers were tools, not toys.
His decision about wearing a pocket protector was calculated with the same logic he used for everything else. He wanted pens and pencils within reach, but he didn’t want stains on his dress shirts. Never mind if the pocket protector looked nerdy. Data in, data out. A decision was made, his pockets were protected.
Boeing and USAir kept elaborate records on everything that was done to every plane, and Ship 513 was no exception. There was paperwork virtually every time a mechanic touched it, even for the routine transit check the night before the accident. The papers said the mechanics at the Windsor Locks, Connecticut, airport had examined the engines, checked the brakes and tires, switched on the exterior lights, and checked the galleys and lavatories.
The NTSB examined the maintenance records and found that 513 had suffered the usual ailments and hiccups of a seven-year-old plane—a blown tire, a fuel line problem, tiny cracks in a floor beam, a few spots of corrosion—but nothing extraordinary. Investigators scoured maintenance records and found a few “carryover” items, minor problems that did not need to be fixed right away. The engine thrust reversers were showing some wear. And the cabin floor was soft and spongy at Row 5 and was fixed with a temporary patch.
A year before the crash, the plane had undergone the rigorous USAir “Q check,” a three-week exam in which mechanics inspected every beam and panel. The plane had gotten the requisite minor checkups since then, and it complied with the latest safety requirements from the FAA. Ship 513 seemed to be in great shape.
Cox checked his records and found he had actually flown it several times, but not in the last several months. Each 737 had its own quirks. They felt different to pilots because of their age, their individual equipment, or because they occasionally got banged by a passenger walkway. Some were cranky and had a rougher ride than others. But 513 was known for a smooth and steady ride. When Cox spoke with a colleague who had flown it recently, the pilot said he was sorry 513 was gone.
“It was a good one,” he said.
When the systems group arrived at the Parker Hannifin plant in Irvine, Cox, the ALPA representative in the group, could feel tension in the air. To the Parker employees, this was like a visit from an IRS auditor. The NTSB group was there to see if the rudder power control unit—Parker’s bread and butter—had caused a crash that killed 132 people. Employees were jittery.
With the help of Parker technicians, the NTSB team pulled the silver-colored unit out of the crate and methodically took it apart, piece by piece, examining each one under a bright light. They checked one of the filters and found nothing out of the ordinary. A group member acted as the scribe to document the tests. He wrote, “No large particles.”
They checked the O-ring seals on the filters. The scribe wrote, “No deficiencies.” They opened one end of the power unit and peered inside with a borescope. “No evidence of impact marks or abnormal wear marks.” They opened another cover plate and peered inside at the fluid.