|
TABLE OF CONTENTS
(title page)
FACTUAL INFORMATION
ANALYSIS
CONCLUSIONS
SAFETY ACTION
APPENDICES
|
|
|
CASB Majority Report |
|
Sequence of Events
The Board was unable to determine the exact sequence of events which led to this tragic accident. The significant destruction of the aircraft at impact and during the post-crash fire, the limited flight data recorder information, and the lack of cockpit voice recorder information were all factors which prevented the determination of the exact causal sequence. Nevertheless, no pre-impact failures or malfunctions which could account for the accident were identified. Thus, the following scenarios were not considered consistent with the evidence gathered during the investigation: uncommanded deployment of a thrust reverser; pre-impact fire; pre-impact explosion; inappropriate aircraft configuration; hydraulic system failure; flight control malfunction: potable water system leakage; and physical failure of one or more engines.
Furthermore, the Board believes that there is sufficient evidence to conclude that ice contamination of the wing and the resulting degradation in aircraft performance was a significant factor.
There is significant evidence in the form of ice accretion calculations, pilot reports, and weather observations to suggest that, during the approach to land, ice accreted on the leading edge of the wing and that, while the aircraft was on the ground, additional roughening of the upper surface of the wings occurred because of the freezing precipitation and possibly frost. Since the aircraft was not de-iced, the contamination which accumulated during the approach and station stop remained on the aircraft for the take-off. The performance calculations, computer simulations, and flight simulator testing all demonstrated that the performance of the aircraft was consistent with the reduced aerodynamic efficiency and resultant high drag associated with wing ice contamination.
It is possible that other factors such as an engine compressor surge and the use of an inappropriate take-off reference speed contributed to this occurrence; however, their precise contribution could not be determined. The Board considers the following to be the probable sequence of events which occurred during the attempted take-off.
The take-off roll proceeded normally, and rotation was commenced at or about the speed calculated by the crew. The calculated rotation speed was at least four knots below that appropriate for the aircraft weight and may have been as much as nine knots below that appropriate for the aircraft weight. This lower rotation speed probably resulted in a delayed lift-off and extended take-off roll. Nevertheless, the aircraft lifted off and commenced climbing. The simulator tests did, however, demonstrate that the use of lower than normal take-off reference speeds reduced the chance of a successful take-off with ice-contaminated wings. Lower than normal take-off reference speeds would reduce the already limited speed margins above the stall.
At lift-off, rotation was probably continued towards the expected pitch attitude necessary to achieve a normal climb schedule. After lift-off, and, as the benefits of ground effect decreased, the aircraft's degraded aerodynamic characteristics would have become apparent to the crew. These degraded characteristics would initially have resulted in a lower than normal rate of climb for the pitch attitude set. In response, it is probable that the pitch attitude was increased to achieve the desired rate of climb. However, simultaneously, the drag effects of the contamination would have caused the rate of acceleration to decrease, followed rapidly by a decrease in airspeed. The extended position of the landing gear indicates that a normal climb rate was never acheived.
Further performance degradation may have occurred as a result of a compressor surge in the number four engine. Although there was no definitive evidence to indicate that the number four engine was not operating at high power at initial tree impact, this possibility could not be eliminated. Computer simulations demonstrated that lessor amounts of ice contamination were required to result in the observed performance degradation, if coupled with a loss of thrust in one engine.
Soon after the airspeed began to decrease, the aircraft stalled. Computer simulations and tests in the flight simulator demonstrated that, with ice contamination present, a stall would occur at normal climb-out pitch attitudes. The crew would have received very little warning of the impending stall: the stall occurred at a significantly higher than normal airspeed, and, because the angle of attack at which it occurred was lower than normal, it is probable that there was little or no advanced warning from the artificial stall warning.
The heading change to the right was typical of other jet transport aircraft stall accidents and thus could be directly attributable to the stall. It is also possible that the heading change reflects a loss of thrust involving the number four engine.
Once the stall had occurred, there was insufficient altitude available to effect a recovery. Furthermore, the change in aircraft pitch characteristics caused by the ice contamination could well have made aircraft pitch control more difficult. The normal nose-down pitching moments which occur at stall would likely have been changed to a nose-up pitching moment.
Previous stall accidents involving DC-8 aircraft have shown that compressor surging at the high angle of attack associated with stall is not uncommon. Thus it is also possible that the lower ground impact rpm of the number four engine reflects surging in the engine after the stall had occurred. The angle of attack at initial tree impact was determined to be about 21 degrees. Witness observations of the yellow/orange glow could have been the result of flame emanating from the engine which accompanied a compressor surge.
The full trailing-edge-up elevator position suggests that when impact with the terrain became imminent, the pilot applied full-aft control in an instinctive effort to avoid ground contact. Despite this effort, the aircraft struck trees, while in a severe stalled condition about 20 seconds after lift-off. Breakup of the aircraft commenced immediately, and, upon impact with the ground, an extensive fuel-fed fire commenced.
|
|