Guy Norris/LOS ANGELES

The Airline Pilots Association (ALPA) and Boeing have submitted opposing findings to the US National Transportation Safety Board (NTSB) over the probable factors involved in the crash of a USAir Boeing 737-300 on 8 September, 1994.

The aircraft rolled rapidly to the left after encountering wake vortices from a preceding 727 and within seconds stalled at 4,300ft (1,300m), according to the NTSB. The 737 remained in a stalled condition for 14s while it pitched down and rolled left until it hit the ground, the Board says. Despite a three-year investigation, no firm evidence was uncovered to point to a definitive cause. Apparent similarities to the earlier unexplained crash of a United Airlines 737 near Colorado Springs led to intensive investigations of the 737's flight control system, particularly the rudder and its power-control unit (PCU).

ALPA says that evidence points to the crew fighting to overcome the rudder. The 737 "-has a history of uncommanded yaw. The current system does not meet current FAR [airworthiness] requirements", it adds.

Boeing says that after exhaustive tests on the rudder and PCU covering 75,000 engineering man hours, "-there is no certain proof of aircraft-caused full rudder deflection during the accident. The previously unknown failure conditions discovered in the 737 rudder PCU have been shown to not be applicable to Flight 427 or any other conditions experienced in commercial service". It adds that, although no certain proof of crew error exists, "-a plausible explanation for a crew-generated left rudder input must be considered. In Boeing's view, there is insufficient evidence to reach a conclusion as to the probable cause of the rudder deflection".

Boeing's investigations included a test flight in which the USAir departure from controlled flight was re-enacted. A full rudder input was made "-at the speeds we think 427 was flying at" says Ken Higgins, vice-president of flight operations, who adds that the crew successfully recovered. "The aircraft was fully controllable, but you cannot simulate surprise and human factors," he cautions.

Boeing's findings include the recommendation that flightcrews need to be specifically trained to handle "large upsets". It adds that 737 yaw-damper reliability enhancements are needed to "-reduce potential aircraft contributions to upsets". It concedes that "-737 flight-control anomaly procedures could be improved".

Enhancements to the control system are under way as part of a $140 million programme, with incorporation of an improved PCU on the production line from July 1997 and retrofits "now beginning". Improved yaw-damper couplers and pressure-reduction enhancements will begin on the production line in July 1998, with retrofit starting shortly after. Final changes to introduce more-capable flight-data recorders will be implemented in late 1998. The PCU retrofit must be completed by August 1999 and the yaw-damper and pressure-reduction changes by August 2000. The NTSB is to review both sets of findings before issuing its final report in 1998.

Source: Flight International