SPRINGFIELD, Va. -- Federal investigators probing a 1994 USAir crash that killed 132 cast doubt on the safety of the world's most widely used commercial aircraft yesterday, suggesting that the Boeing 737 might have a fatal design flaw.

Safety experts also questioned whether design changes already ordered by the Federal Aviation Administration have worked, noting that some 737s with the modifications have still reported unexpected problems in flight.

The investigation has "raised questions in many minds about the design and operation of the 737's rudder system," said Jim Hall, chairman of the National Transportation Safety Board at a hearing here on the crash. "We will determine if more needs to be done."

The NTSB is expected to conclude its 4 1/2-year-old investigation -- the longest in its history -- today, when it determines the likely reason USAir Flight 427 flipped on its back and dived to the ground near Pittsburgh, killing all on board.

It also will consider 10 recommendations to the FAA, and reportedly among them is a mandate that sweeping, expensive modifications be made to most 737s.

Staff presentations at the board hearing centered on evidence that a flawed design in the 737 can cause the planes' rudder to flip unexpectedly to one side, sending the aircraft into a sudden, uncontrollable downward spiral.

Staff members also rebutted Boeing Co.'s claim that the crash was caused by pilot error.

Using graphic flight simulations, investigators showed that the plane's pilot would have to have steered the plane in the wrong direction, without the co-pilot interfering, and held it there until it crashed, a scenario one NTSB official called "implausible."

Investigators also questioned Boeing's claim that a similar crash of a United 737 in Colorado Springs, Colo., in 1991 was caused by a freak "wind rotor." All 25 aboard died.

Such a rotor would have been more than 30 times more powerful than any ever recorded in Colorado Springs, and would have to have followed the plane's flight pattern almost perfectly to cause a crash, they said.

"If the rotor's not there, you're back to the very unlikely scenario that the pilots did it themselves," said NTSB member George W. Black. "Or, there was some kind of mechanical malfunction."

Flaw would jolt industry

A finding that Flight 427 crashed because of a faulty rudder design could send a shudder throughout the passenger airline industry, where the Boeing 737 is the most common aircraft, with more than 3,000 in use worldwide.

The aircraft, first approved for flight in 1967, has flown some 91 million flight hours, carrying a number of passengers equivalent to the entire population of the globe.

While the 737 enjoys one of the best safety records in the industry -- one crash for each 1 million flights -- several crashes and other mishaps have been attributed to mysterious movements of the planes' rudders, the vertical flap on the tail that controls lateral movement.

Flight 427 experienced a radical left swing of its rudder, known as a "hardover," causing it to flip on its back and spiral to the ground.

NTSB investigators focused yesterday on a device in the plane's tail, a servo valve, which regulates the flow of hydraulic fluid into the system that controls rudder movement.

In some extreme instances, those valves have been shown to jam in a way that causes the rudder to move in the direction opposite to the pilot's commands.

The board's crash simulations suggested that the pilot of Flight 427 turned the plane to the right and pressed hard on its right rudder pedal, a procedure that would correct movement to the left in normal conditions. But the plane banked left, flipped over and dived to the ground about 25 seconds after the unexpected movement began.

The FAA ordered newly calibrated servo valves for all older 737s in 1997, but even some planes with the new equipment have reported unexpected, uncommanded rudder movements.

A MetroJet 737 with the new part made an emergency landing at Baltimore-Washington International Airport last month after an unexpected rudder movement caused it to turn to one side.

Bernard Loeb, director of the NTSB's office of aviation safety, said yesterday that the plane's manufacturers have known of potential problems with the rudder servo valves since the early 1960s. Technicians first redesigned the component after one malfunctioned during tests when the aircraft was being developed, he said.

Details of the 500-page report on the Flight 427 crash will not be made public until after its expected approval today. If the board recommends design changes to the 737's rudder system and the FAA approves them, the changes could cost more than $100 million.

What board may do

Staff members hinted yesterday that the board will recommend new procedures for landing 737s, including a speed increase that could offset the effects of a rudder hardover.

The board is also expected to seek the installation of flight data recorders that capture more information for crash investigators, including records of how flight controls in the cockpit are moved and what effect they have on control surfaces such as the rudder, ailerons and elevator flaps on the tail.

Yesterday's board meeting was attended by more than 100 relatives of Flight 427's victims, a reminder that Boeing could have more at stake than simply mechanical changes in its aircraft.

Should the NTSB find today that the plane crashed because of a design flaw, the company could be exposed to lawsuits from family members.

Robert Johnson, whose mother perished aboard Flight 427, attended with six family members and a lawyer.

"It's not even about closure at this point," said Johnson, 32. "It's about people taking responsibility and making sure this never happens again."

Pub Date: 3/24/99