Rudder Reversals May Have Caused the Crash of 2 Boeing 737s

From Associated Press

Government investigators used eerie computer simulations and dry investigative analysis Tuesday to argue that unexpected reversals in the rudders on three Boeing 737s caused a pair of airline crashes and nearly caused a third.

The findings were presented during a National Transportation Safety Board hearing into the cause of the September 1994 crash of USAir flight 427 outside Pittsburgh that killed all 132 on board. Not only may they explain that crash, but also the 1991 crash of United Airlines flight 585 outside Colorado Springs, Colo., that killed all 25 aboard, and a 1996 incident in which the pilots of an Eastwind Airlines 737 briefly lost control of their aircraft as they approached Richmond, Va.

The findings also raised questions about whether further changes are needed in the 3,100 737s flying worldwide. The twin-engine plane is the most common jetliner in the world, with 800 in the air at any time.


The safety board was expected to settle on a probable cause for the USAir crash--and reportedly propose a redesign of the 737’s rudder system--at the end of the hearing, which was extended into Wednesday because of the lengthy presentations.

A source close to the board said members will also decide to revise their admittedly vague opinion about the United crash, which they said in 1992 was caused either by a rogue wind or a loss of a “lateral or directional control system.” The accident is one of only four without a definitive cause in the board’s 31-year history.

The rudder panel runs up the tail of an airplane and sweeps the nose left and right. In a reversal, a jam or misalignment in the controlling linkage could cause it to go in the direction opposite of that which the pilots intend when they press the rudder pedals.

Under those circumstances, a pilot trying to recover from a minor control problem--as was the case for the crashes and the near-miss--could actually make the situation worse.

“A rudder reversal scenario will match all three events,” Dennis Crider, chairman of the NTSB’s Aircraft Performance Group, told board members.

He and other staff members illustrated their point by playing a series of computer simulations showing the USAir and United planes flipping over and flying into the ground, and the Eastwind plane wobbling as the pilots fought to regain control.

A second series of tapes depicted a possible view from behind the pilots as they struggled to regain control.

In the case of the two crashes, the ground was seen rushing toward the cockpit window as robotic depictions of the pilots struggled to regain control of the airplane.

Boeing Co., which makes the 737, withheld extensive comment but released a statement: “When the facts show us an opportunity to make (the 737) even safer, we will.” However, in earlier meetings with reporters, it dismissed a finding of rudder reversal, saying its own investigation of the three incidents has found no supporting physical evidence.

In a submission to the board, the company mentioned the possibility of pilot error in the Pittsburgh crash, a powerful wind in the United Airlines crash and a misrigged yaw damper--another element of the rudder system--in the Eastwind incident.

In their presentations, the NTSB staff members rebutted those explanations point by point, in one instance saying that the wind Boeing blames for the Colorado Springs crash would have had to have been 36 times stronger than any similar “wind rotor” ever recorded in that Rocky Mountain city.

Since none of the planes had modern, highly informative flight data recorders, the board staff was forced to make assumptions in developing its hypotheses. That approach prompted doubts for one board member.

“When I see all these assumptions, I get a little twitchy,” said Robert T. Francis II, the board’s vice chairman.

The USAir crash occurred Sept. 8, 1994, as the 737 approached Pittsburgh after a flight from Chicago. Despite clear skies, the plane suddenly spiraled 6,000 feet into the ground, killing everyone aboard.

Some investigators suggest a part in the rudder linkage--a hydraulic valve known as the power control unit--may have jammed, causing a rudder reversal. If the first officer pressed the right pedal to counteract a leftward movement of the plane, he would have held a flip-flopped rudder in the wrong position.

The Federal Aviation Administration, which certified the plane for commercial use, argues that the precise cause will never be known. Instead, it has focused on making the plane safer.

Working with Boeing, the agency has redesigned the power-control unit to prevent rudder reversals and also created a limiting device to reduce the amount of rudder that can be applied in high-speed situations. Both have also developed new training to help pilots recover from surprise changes in course.

About 100 family members attended the hearing, held in a hotel ballroom outside Washington to accommodate the crowd.

After watching one of the crash simulations, Angelina Hall, a Las Vegas woman whose mother died in the crash, said: “I’m sorry my mother had to endure what she did.”