ROANOKE TIMES
Copyright (c) 1996, Roanoke Times
DATE: Wednesday, July 3, 1996 TAG: 9607030064
SECTION: NATIONAL/INTERNATIONAL PAGE: A-11 EDITION: METRO
DATELINE: WASHINGTON
SOURCE: The New York Times
Air-safety investigators struggling with two unsolved crashes of Boeing 737s said Tuesday that an incident involving another 737 last month could provide important clues.
In the incident last month, the plane unexpectedly rolled to one side after a movement by the rudder that the pilot did not ask for - the same sequence that investigators believe happened in September 1994 to USAir Flight 427, which crashed on its approach to Pittsburgh, killing all 132 people aboard.
Among those killed in the USAir crash were Virginia residents Dave Lamanca of Roanoke, Richard Talbot of Blacksburg, Dewitt Worrell of Lexington and Steve W. Wyant of Charlottesville.
The same sequence may have also happened in March 1991 to a United Airlines 737 that crashed as it was approaching Colorado Springs, killing all 25 aboard.
Like the two planes that crashed, the aircraft in last month's incident, a 25-year-old Boeing 737 belonging to a small new carrier, Eastwind Airlines, carried a primitive flight-data recorder that did not capture many details of the incident. The plane was bought from USAir and is still maintained by that airline.
Since the Pittsburgh crash, the National Transportation Safety Board has been studying all ``uncommanded roll'' incidents in 737s, in which one wing dips unexpectedly. Investigators have looked at more than 50 such incidents, but this is the first since the Pittsburgh crash that they cannot explain.
Late on June 9, in calm, clear weather as the plane was approaching Richmond, Va., on a flight from the airline's headquarters in Trenton, N.J., the pilot felt what he described as a ``bump'' on the rudder, and then a more substantial swing as the plane rolled to the right.
The captain told the investigators that he ``stood on the opposite rudder pedal'' to try to correct the roll, and extended the ailerons - the flight-control panels on the wing surfaces - nearly all the way. The plane landed without injury to the 48 passengers and crew of five.
In the Pittsburgh case, the flight-data recorder and cockpit voice recorder do not make clear what happened, but engineers say the plane's trajectory could be explained by a sudden swing of the rudder all the way to one side.
Engineers still do not know what could have made the rudder swing that way, but they believe that the problem may be centered on the standby rudder actuator, which is the only part of the rudder assembly on the Eastwind plane that was not replaced as they dealt with a series of problems in May and June.
On new planes, flight-data recorders generally track more than 100 different kinds of data, but the one on the Eastwind plane tracks only 11, as did the Pittsburgh plane.
In a letter to the Federal Aviation Administration, James Hall, the chairman of the safety board, said, ``Under slightly different circumstances, the Eastwind incident could have become a third fatal B-737 upset accident for which there was inadequate flight-data recorder information to determine the cause.''
The letter, sent Monday and made public Tuesday, noted that the board had asked on Feb. 22, 1995, that such recorders be installed by the end of last year.
The safety board had asked for immediate action on early 737s and improvements over a longer period on other old planes.
In a statement Tuesday afternoon, the agency said it agreed with the intent of the safety board's recommendations on flight-data recorders and was still working on a proposal.
Airlines have complained that wiring their planes to capture additional data, like the position of the rudder and the position of the rudder-control pedals in the cockpit, would mean taking them out of revenue-producing service for long periods.
The accidents are particularly troubling because more than 2,400 Boeing 737s are in use.
According to the safety board, the Eastwind plane had three earlier problems with its rudder. Shortly before the June 10 incident, most of the critical parts of the rudder-control system had been replaced, except for a part called the standby rudder actuator.
This is normally used only in backup mode and is part of the yaw damper, an automatic system that swings the rudder up to 3 degrees in either direction to keep the plane pointed in the direction it is traveling.
The yaw damper may have been ``misrigged,'' Hall's letter said, and could have pushed the rudder up to 4.5 degrees.
Standby rudder actuators have been known to stick if metal parts inside rub together and metal from the softer surface accumulates. But Boeing Co. and others have insisted until now that such a failure cannot cause the rudder to bend by more than 3 degrees.
In this case, apparently because of a maintenance error, instead of having 3 degrees of movement to either side, it was 1.5 degrees to one side and 4.5 to the other. This is the first time investigators have found a mechanism for a swing of more than 3 degrees, but both planes evidently experienced an even larger rudder movement.
LENGTH: Medium: 93 linesby CNB