When investigators found the wreckage of USAir Flight 427, the biggest piece they saw was the rudder. Now they're trying to determine if it was also the culprit.
Most clues point to the rudder. Pilots and engineers testifying at a public hearing this week said a sudden turn of the rudder probably caused the plane to suddenly roll to the left and plunge to the ground.
The wreckage of the Boeing 737-300 showed no evidence of a bird collision or engine reversal. There were no other airplanes in the immediate area and the FBI found no signs of a bomb. The rudder seems the most likely cause.
But no one is sure how it could suddenly turn in mid-air. Despite thousands of tests on the wreckage, investigators are still puzzled.
Boeing engineer Paul Cline has run about 5,000 tests on rudder system parts since the Sept. 8 crash. But he told the National Transportation Safety Board hearing Wednesday that he found no evidence that the components were to blame.
"This bothers me as much as anyone else," he said. "I've spent many hours awake, thinking of this."
All 132 people on board were killed when the plane plunged nose-first into a ravine.
The rudder is the biggest flight control on the 737, moving the plane sideways like a rudder on a fishing boat. Pilots say it's unlikely that the Flight 427 crew was using it as they cruised above Aliquippa, Pa.
Pilots move the rudder with their feet. Pedals in the cockpit are connected to cables that run through the belly of the airplane. Those cables are hooked to a power control unit, often called the PCU, that converts the cable movement into hydraulic pressure. That hydraulic muscle turns the 20-foot rudder.
The rudder is not connected to the autopilot system, but it is partially controlled by the yaw damper, a device that automatically makes dozens of small adjustments to the rudder every minute to keep the flight straight and smooth.
Much of the testimony this week has focused on the PCU and the possibility that it malfunctioned. If the PCU jammed, the rudder might have steadily shifted to one side, causing the plane to plunge before the pilots could regain control.
Another scenario: The yaw damper might have malfunctioned. If it stopped getting return signals that the rudder was moving properly, the yaw damper might have continued to push the rudder to one side.
So far, there has been virtually no physical evidence to prove that anything malfunctioned.
Engineers found "very high levels" of particulates in the hydraulic fluid, some four-times larger than a 25-micron filter is supposed to allow. But those particles were still extremely small, comparable to "specks of dust in a sunbeam," according to Bernie Turner, a Boeing flight-controls engineer.
He said a micron is "an awfully teeny thing."
Engineers found no scars near the PCU's valves to indicate they jammed. When engineers mixed a special batch of hydraulic fluid with an extremely high concentration of particles, the PCU still worked fine.
But there have been at least 16 reports of other rudder problems on Boeing 737s since February 1991.
In July 1992 in Chicago, a United Airlines pilot discovered he could not fully depress his rudder pedal. When engineers examined the PCU, they found hydraulic leaks and components that didn't fit together tightly enough. Tests found the rudder might actually go the opposite direction from the pedals.
In April, a Continental Airlines plane cruising at 37,000 feet over Honduras suddenly rolled to the left. The crew then made a precautionary landing. Last Thursday, the safety board got reports of sudden rudder shifts on two Air France 737s.
Investigators also are exploring whether there's a connection between the Aliquippa accident and the 1991 crash of a 737 in Colorado Springs, Colo., one of only a handful of crashes the safety board has never solved. The rudder was a suspected cause in that crash, but the safety board said it lacked enough evidence to single out the rudder as a primary cause.
The Federal Aviation Administration then proposed a rule that would have required more frequent tests and possible replacement of a shaft leading to the standby PCU, which acts as a backup if something goes wrong with the primary one. But Donald Riggin, manager of the Seattle aircraft certification office for the FAA, said his agency withdrew the proposal after officials became convinced that they "did not have an unsafe condition."
In response to the Chicago incident, the FAA last February issued a separate rule requiring airlines to test the PCUs every 750 flight hours and replace them with newer models within five years. Airlines balked at the rule, saying it was too costly, but the FAA said it was necessary.
The USAir jet that crashed had been inspected as part of that program, but it still had an older model PCU, Riggin said.
He said the five-year deadline to replace the older PCUs was necessary to minimize the economic impact on the airlines and assure that enough PCUs could be manufactured. "We were confident in (the testing) to pick up problems."
Why did Flight 427 fall from the sky?
The investigation into the mysterious crash of USAir Flight 427 in Pittsburgh is now centered on the rudder. Investigators suspect that the rudder suddenly shifted left, causing the Boeing 737-300 to roll and then plunge 6,000 feet to the ground. They have conducted tests on the rudder's power control unit, which uses hydraulic pressure to make the rudder turn right and left.
Rudder: Vertical hinged section on tail that can be turned left or right to control plane's direction.
Rudder pedals: Pilots push them back and forth to move the rudder from side to side.
Yaw damper: Uses gyroscopes to detect plane's position, moves rudder slightly to counteract tail's tendency to oscillate (waver right and left).
Cables: They move back and forth when the pedals are pressed.
Power control unit: The cables then move the rudder's power control unit, a hydraulic device that actually moves the rudder (enlarged below).
One theory under discussion is the possibility of contamination in the hydraulic fluid. Filters are supposed to keep the hydraulic fluid pure, with particles no bigger that 25 microns. But tests found particles as large as 100 microns in some parts of the unit _ much smaller than a human hair, but possibly big enough to cause problems.
Clues in the tail
Investigators are trying to determine if the servo got jammed, causing the rudder to turn completely to one side. Valves in the servo move hydraulic fluid, providing power to move the rudder.
The Servo is the brain of the control unit, converting the cable movement into hydraulic pressure.