Gulf Air Crash Probe Spotlights
|After a missed approach, the A320 had begun a 360-deg. orbit at 500-600 ft. when it plunged into the sea.|
Although "no single factor was responsible for the accident," according to the newly completed final report, investigators stressed the need for optimal crew resource and cockpit workload management. The airline implemented a number of safety initiatives following the accident.
On Aug. 23, 2000, a Gulf Air A320 operating the Cairo-Bahrain route as Flight 072 plunged into the sea about 2 naut. mi. northeast of Bahrain International Airport, killing all 143 passengers and crewmembers ( AW&ST Sept. 11, 2000, p. 57; Sept. 4, 2000, p. 77). The crash occurred at 7:30 p.m. local time, about 1 hr. after sunset.
The flight crew had disconnected the autopilot to conduct a VOR/DME approach. It was aborted about 0.9 naut. mi. from the Runway 12 threshold owing to excessive height and speed--at 584 ft., nearly twice as high as the aircraft should have been, and at about 180 kt. instead of the prescribed 136-kt. target speed. "We are not going to make it," the captain said. Two minutes before the captain decided to abort the approach, the aircraft was flying at 330 kt., at 2,300-ft. altitude.
Gulf Air's updated A320 flight manual now specifies that a 250-kt. speed limit must be observed below FL100 (10,000 ft.), exclusive of specific air traffic control requests.
After informing ATC, the crew did not perform the published missed approach but instead initiated a 360-deg. left turn, at 500-600 ft., in preparation for another landing attempt. Such a low-altitude, 360-deg. orbit is a nonstandard maneuver, investigators stressed. Moreover, the go-around was made with deviations from prescribed flight parameters and involved considerable variations in bank angle, g-force and altitude, they added. During the left turn, the aircraft's altitude varied between 965 and 332 ft., and bank angle reached 36 deg.
INVESTIGATORS ALSO NOTED that the standard go-around procedures were not applied, including rotation to 15-deg. pitch-up. The crew was trying to make the left turn to re-intercept the approach path, but the aircraft instead crossed the extended runway centerline at an angle of about 90 deg., causing the pilot to say, "We overshot it." Pilots re-creating the accident in the simulator found that only moderate bank angles were required to properly complete the circle to a landing, though the approach was not well-stabilized due to the short distance from the re-intercept point to the runway. Flight tests showed that it would be hard to confuse any other local object with the well-lighted runway. The aircraft continued the left turn and came back over the runway itself.
More surprisingly, during the final phase of the flight, the pilots did not respond to the 9-sec. "whoop, whoop, pull-up" warnings of the ground proximity warning system (GPWS), indicating that they were not perceiving a serious threat or relying on information provided by flight displays.
Earlier in the approach, the crew had ensured that the flight directors were turned off. Now Gulf Air is implementing an Airbus modification that automatically restores the flight director bars at go-around initiation.
The Bahraini investigation team, backed by the U.S. Naval Aerospace Medical Research Laboratory (NAMRL), determined that the pilots' misperception of the aircraft's pitch orientation, exacerbated by "multiple cockpit distractions," seriously contributed to the mishap. France's BEA accident investigation bureau, the U.S. NTSB and Airbus also participated in the inquiry.
Information provided by the A320's digital flight data recorder confirmed that the flight crew suffered from spatial disorientation during the go-around. The final plunge was caused by the pilot-in-command pushing the side stick for 11 sec., reducing the load factor to about 0.5g and producing a 15-deg. nose-down pitch attitude and rapid acceleration as "Toga" (takeoff/go-around) full power was applied.
Investigators pointed out that, in total darkness, somatogravic illusion could have caused the captain to perceive falsely that the aircraft was "pitching up." He responded by making a "nose-down" input, and the aircraft descended and thereafter flew into the sea at 280 kt., in approximately 6.5-deg. nose-down pitch. Somatogravic illusion is where acceleration creates a false sense of pitching up, and it can affect both crewmembers simultaneously. The effect has been known for decades and is one that pilots are warned about when they are instructed to rely on their instruments instead of body perceptions.
NAMRL experts stressed that "the mishap represents a tragic but scientifically interesting accident in which a series of events led to a physiologically normal misperception of pitch orientation by the pilots." The Navy experts noted that the A320 pilots did not perceive the true attitude of the aircraft owing to the compelling nature of false but concordant information complicated by a high workload and task saturation.
HOWEVER, NAMRL'S calculations showed that the error in the crew's body-perceived pitch would be about 7 deg.--i.e., they would perceive a 12-deg. attitude when in reality it was 5 deg. NAMRL said the illusion was "evidenced by the fact that the pilot did not follow the GPWS procedure (. . .) since he believed he was approximately level."
But the pilot's actual pitch command was much greater than this error--it was 15-deg. nose-down when it should have been at least 5 deg. nose-up, or three times greater than the somatogravic error.
In addition to disorientation, a well-established factor that contributed to the catastrophic sequence of events, the investigation team's findings indicated that the flight crew's proficiency, as well as the airline's screening for new and upgrade pilots, was questionable.
THE CAPTAIN, who may have bordered on overconfidence, had logged 2,402 hr. as flight engineer and 4,416 hr. of pilot time, including 1,083 hr. on the A320 and 86 hr. as an A320 captain. His most recent line check, a few weeks before the crash, indicated that all competency elements were satisfactorily completed. Four months earlier, however, he almost flunked his captain upgrade checkride by doing poorly in a rejected takeoff and engine failure after V 1 decision speed.
According to colleagues interviewed by investigators, the first officer was disciplined but shy and may have been too reserved to dispute the captain's handling of the approach. He became a first officer four months before the accident, had failed his initial A320 proficiency check in 1999 and had logged no more than 608 flight hours.
Gulf Air now requires screening by an accredited aviation psychology organization for new pilots and first officers upgrading to captain.
Michael A. Dornheim contributed to this report from Los Angeles.
© September 2, 2002 The McGraw-Hill Companies Inc.
from this link