All five persons on this cargo DC-8 died in the crash right after takeoff. The aircraft may have had some icing on the wings caused by very cold fuel that would not have been present during preflight inspections. But the aircraft should have been flyable. So what caused the crash? Alcohol, cultural norms, and Crew Resource Management.
— James Albright
The captain appeared to be intoxicated and that was noticed by several people, including the first officer. But nobody stopped the flight. The crew was Japanese and the captain was American. Does that matter? I think it did.
- Date: 13 Jan 1977
- Time: 0635
- Type: McDonnell Douglas DC-8-62AF
- Operator: Japan Air Lines
- Registration: JA8054
- Fatalities: 3 of 3 crew, 2 of 2 passengers
- Aircraft Fate: Destroyed
- Phase: Initial Climb
- Airports: (Departure) Anchorage International Airport, AK (ANC, PANC), USA;
- Airports: (Destination) Tokyo-Haneda Airport (HND/RJTT), Japan
- The aircraft arrived at Anchorage at 0503. The incoming flightcrew reported that the only weather they encountered en route was a layer of fog on the final approach at 800 feet and that they did not encounter any precipitation or icing.
- Two contract mechanics stated that there was ice on the inlet guide vanes, the engine cowlings, and the engine bullet noses, but no ice was reported on the airfoil surfaces. The JAL personnel stated that they did not see any ice on the aircraft. One contract mechanic advised the JAL representative that the engine anti-icing system should be used by the next crew to clear the ice in the engine inlets.
- The outbound flightcrew was wakened about 0330, left the hotel by taxi about 0430, and arrived at the JAL dispatch office about 0500. The taxicab driver who brought the outbound crew to the airport stated that he became concerned by the captain's actions in the taxi and called his dispatcher to report his impressions.
- About 0450 the taxi dispatcher called the operations agent for the contract maintenance company and reported that one of her drivers had taken an "intoxicated" JAL captain to the airport. The operations agent stated that "...it seemed logical that JAL would detect anything unusual and act accordingly." He further stated that at 0620, he notified his line manager of the conversation with the taxi dispatcher and that "I felt that if the captain was intoxicated JAL OPS . . . or his first officer would have stopped the flight immediately." The JAL dispatch personnel and the inbound JAL crew stated that they noted nothing unusual about the outbound crew.
- The outbound crew consisted of an American captain and a Japanese first officer and flight engineer. They went to the aircraft about 0515 and boarded the aircraft with the two cattle handlers. The driver of the crew car, a friend of the captain, stated that "...he was in good condition as far as way's I've seen him sometimes and I made that that statement before I ever heard any rumors that he was supposedly drunk or had been partying or whatever."
- A review of the cockpit voice recorder indicated . . . The taxi checklist was completed and the takeoff data, the flap settings, and the trim settings were again reviewed. The captain, in response to the challenge "anti-ice, de-ice, and rain removal," said, "Ok, we will use engine anti-ice." The de-ice system was reported "off" by the flight engineer.
- The captain taxied the aircraft southeast on the ramp, past the terminal toward runway 24L. He stopped on the ramp after being instructed to hold short of runway 24L. After several communications with the controller, the aircraft taxied onto runway 24R, and reported "...ready for takeoff." The tower advised the captain that he was on runway 24R which the captain contradicted. The controller then issued taxi instructions to get the aircraft to runway 24L. The captain made a 180° turn on runway 24R before he finally taxied to the taxiway which leads to the approach end of runway 24L. The crew again reported that they were ready for takeoff at 0633:37.
- Takeoff was initiated and at 0634:32 the captain called "maximum power." At 0634:50 the captain announced, "I have" and at 0634:52, "80" (knots) was called by the copilot. At 0635:10, "Vee one" was called by the copilot and at 0635:16 rotation was called and acknowledged by the captain. At 0635:19.5 the captain called "Ten degrees" and at 0635:21.4 the first officer called V2. At 0635:26.2 a sound similar to aircraft buffet was recorded. This sound became more frequent and continued until the sounds of impact. At 0635:32 the first officer called "Gear up" and at 0635:33 the flight engineer said "Too much speed (steep)." At 0635:38 the engineer called "stall" simultaneously the stick shaker sounded and continued until 0635:39.3, when impact was recorded.
- A witness near the departure end of the runway saw the aircraft climb to an estimated altitude of about 100 feet above the ground, veer to the left, and then slide "...out of the air."
Source: NTSB AAR-78-7, ¶1.1
- The initial blood alcohol level of the captain was 298 mgs percent and a vitreous alcohol level of 310 mgs percent recorded in test conducted 12 hours after the accident by the Alaska Medical Laboratory.
- A blood alcohol level of 100 mgs percent is considered to be legally intoxicating for drivers in the state of Alaska. The National Safety Council Committee on Alcohol and Drugs has determined that a blood alcohol level of 180 to 200 mgs percent would result in mental confusion, disorientation, dizziness, exaggerated emotional state (fear, anger, grief, etc.), disturbance of sensation (diplopia, etc.) impaired perception of color, form, motion, or dimensions, decreased pain sense, impaired balance, muscular incoordination, staggering gait, and slurred speech.
- Of the 13 persons interviewed regarding the captain's activities before reporting to the airport, 5 close acquaintances said that he showed no signs of drinking or that he had not had a drink in their presence. Six persons who were not closely acquainted with the captain stated that he had been drinking or showed signs of being under the influence of alcohol with the 12 hours before the scheduled flight.
Source: NTSB AAR-78-7, ¶1.13
- The timing of the callouts at 80 kns, V1, VR, and V2 were compared to a CVR tape of another JAL DC-8 which took off under similar conditions. . . the time between the call for rotation and V2 was about 1.5 seconds longer in the transcript from JA 8054, when compared to the other DC-8 takeoff tape.
- Aircraft rotation appeared to have been within 1 second of the call, and the aircraft lifted off, as defined by the dip in the altitude trace, within 3 seconds after the rotation began.
- The initial rate of climb after liftoff appeared to be higher than that normally achieved by other DC-8's in similar conditions.
- JA 8054 reached a maximum recorded airspeed of about 164 kns; the airspeed began to decrease when the sound of buffet was recorded on the CVR. A maximum altitude of 284 ft (160 ft above the runway) was reached and, as the sound of buffet increased, the heading trace indicated a turn to the left; the vertical acceleration trace indicated a decrease in vertical loading; and the airspeed continued to decrease. The airspeed trace became erratic as though there was a disturbance in the airflow sensed by the alternate pitot static system. During the 3 to 5 seconds before impact, buffet sound increased, the vertical load factor increased rapidly, and the rate of descent decreased suddenly.
- Computer simulations . . . indicated that the aircraft must have been rotated to an excessive pitch angle just before it reached V2 in order to have produced the FDR recorded data. Finally, both studies concluded that the aircraft stalled just after V2 was called and that the stall continued and deepened to an angle of attack of at least 18° until impact.
- Conditions were favorable for the accretion of rime icing from the time the aircraft approached Anchorage until the crash.
- The temperature of the fuel remaining in the wings tanks was calculated to be about -8.3°F and the temperature of the fuel added at Anchorage was about 32°F. Calculations show that the temperature of the fuel in the wing tanks after refueling from 20 to 25.5°F. Because of the fog at 20°F and the below freezing temperature of the skin above the fuel tanks, the supercooled water droplets in the fog could have accumulated on the wing and formed rime ice. Although no such icing was reported by either a crewmember of a ground crewman, there was sufficient time between the preflight check and the takeoff for enough ice to form to degrade the takeoff performance of JA 8054.
Source: NTSB AAR-78-7, ¶1.16]
- The performance studies indicated that normally the stall warning system should have activated when the aircraft stalled at, or just after reaching, V2 instead of about 1 second before impact. Although all the system components were not recovered, and, therefore, could not be examined, the CVR indicated that the system was checked during the pretakeoff checks and the crew was apparently satisfied with the test. There were several reasons why the system may not have functioned, including changes in performance cause by airfoil ice, ice on the transducer, or improper calibration. The Safety Board was unable to determine why the stall warning system did not activate earlier in the accident sequence.
- Based on the performance evaluation and the computer simulations, the Safety Board believes that the subsequent slow acceleration resulted from rotation to about 15° after liftoff, a higher than normal pitch attitude. As aircraft performance deteriorated, the situation was probably worsened when the angle of attack was increased to about 18°.
- Icing of the leading edge or the upper surface of the wing would have lowered the angle of attack at which the aircraft would have stalled. The net effect, assuming that all other inputs remained the same, would have been a reduction of the angle of attack required to stall the wing.
- In view of the above, the Safety Board concludes that the recorded aircraft performance resulted from the pilot's control inputs aggravated by airframe icing.
- The captain's physical and mental states were such that he could not effectively control the aircraft. The amount of alcohol in his system would have severely hampered his reactions, coordination, and reasoning ability.
- It is extremely difficult for crewmembers to challenge a captain even the when captain offers a threat to the safety of the flight The concept of command authority and its inviolate nature, except in the case of incapacitation, has become a practice without exception. As a result, second-in-command pilots react indifferently in circumstances where they should be more assertive.
Source: NTSB AAR-78-7, ¶2
The National Transportation Safety Board determines that the probably cause of the accident was a stall that resulted from the pilot's control inputs aggravated by airframe icing while the pilot was under the influence of alcohol. Contributing to the cause of this accident was the failure of the other flightcrew members to prevent the captain from attempting the flight.
Source: NTSB AAR-78-7, ¶3.2.
NTSB Aircraft Accident Report, AAR-78-7, Japan Air Lines, Ltd., McDonnell-Douglas, DC-8-62F, JA8054, Anchorage Alaska, January 13, 1977.