The Black Box
That much was evident from the flight data recorder. But neither it, nor the cockpit voice recorder, explained the origin of the yaw, nor why the pilots had been unable to stop it. "We don't know the extent of the rudder movement" due to limitations in the plane's recorder, says Curtis Meisenheimer, an aerospace analyst with Frost & Sullivan, a consulting firm in Calif., who adds that recorders installed in new planes do record the rudder position.

Frozen solid?
Still, the record of the plane's flight naturally focused attention on the rudder, and that's where the recent tests, performed at Boeing facilities, come in. It seems that under certain circumstances, the complex rudder control device, which functions something like a car's power steering control, can seize up and push the rudder all the way to one side. Under these conditions, to be specific:

	The rudder power unit cools down to about -40 degrees (Celsius or Fahrenheit -- the two scales meet at this temperature). That's a normal temperature in the stratosphere, where the USAir plane had been cruising before it began descending.
	The hydraulic fluid in the rudder system was slightly contaminated (the testers simulated this by taking fluid from an actual jet).
	The fluid was abnormally hot -- as it would be if the hydraulic pump was damaged.

These were the conditions that caused the rudder power unit to jam, which could have started the sequence of events that caused the Pittsburgh and Colorado Springs crashes. But just because this could explain the crash does not mean it does explain it.

Counterarguments?
Although it's not meaningful to argue that the series of events was so improbable as to be impossible (737 accidents are very uncommon, so whatever causes them is just as uncommon), there are some reasons to doubt the "frozen-controller" explanation.

	The control unit would normally stay warm because the yaw damper (defined) is usually constantly adjusting the plane's flight.
	There is no sign -- yet -- that the hydraulic fluid in the Pittsburgh plane was actually hot enough to cause the rudder problem.

Unfortunately, the jammed control unit in the lab did not carry marks indicating that it had been jammed. The hydraulic pump from the Pittsburgh crash was destroyed, so it's impossible to tell whether it might have had a role in overheating the hydraulic fluid.

For the past two years, with one 737 crash investigation (Colorado Springs) unsolved and another (Pittsburgh) proving highly enigmatic, Safety Board officials are worried about a possible pattern in these yaw-and-roll situations. Thus, even before the Pittsburgh investigation is completed, and even before the revelation about the rudder control unit, the Board issued a request for changes in the plane's rudder controls.

There are 1 2 3 4 5 6 7 8 9 10 11 12 documents. (Glossary | Bibliography)