Statement by Jim Hall, Chairman
National Transportation Safety Board
on the Investigation of the Crash of Alaska Airlines flight 261
February 8, 2000
Good afternoon and thank you for coming.
As you know, the National Transportation Safety Board has been investigating the crash of Alaska Airlines flight 261, an MD-83, since the early hours after the accident. My colleague, Member John Hammerschmidt, was on-scene at Port Hueneme, California to brief you last week on that investigation.
Our investigators are still on scene in California, and will be there for some time to come. I want to again thank the United States Navy for their work on the mapping and recovery effort, and the U.S. Coast Guard for their work in the early days of the investigation. I also want to thank the state and local authorities who have provided us with so much help during this investigation.
In addition, there are a lot of people whose contributions are less apparent to the general public. I want to acknowledge the work of the Alaska Airlines Care Team, the Disaster Mortuary Operational Response Team (D-MORT) of the Department of Health and Human Services, the American Red Cross and the Salvation Army.
Flight 261 was an MD-83 aircraft, N963AS, purchased new by Alaska Airlines in 1992. It had accumulated 26,584 hours of service and 14,315 flights.
I would like to start off today by giving you preliminary information from flight 261's Flight Data Recorder. I will from time to time also reference information off the Cockpit Voice Recorder. To assist you and the general public, my prepared remarks will be placed on our Internet site, www.ntsb.gov, immediately following this press conference.
You see on the table next to me the two recorders recovered by the Navy from about 700 feet below the surface of the Pacific Ocean. Dr. Bernard Loeb, Director of Aviation Safety, is behind the CVR, and Dr. Vernon Ellingstad, Director of the Office of Research and Engineering, is behind the FDR.
The FDR is a Sundstrand UFDR that contains 48 parameters of information, including stabilizer trim position and elevator position. The 25-hour magnetic tape captures the entire 2 hours and 43 minutes of the accident flight, as well as information from previous flights.
After takeoff from Puerto Vallarta, the autopilot was engaged at approximately 7,500 feet of altitude. About 13 minutes later, the autopilot was disengaged when the aircraft was at 29,000 feet. During the following three minutes, the aircraft climbed to 31,000 feet.
You will remember that when we released CVR information last week, we told you that when the approximate 31-minute recording began, the pilots were discussing an existing problem with the aircraft's stabilizer trim. According to the FDR, the crew had flown for 1 hour and 53 minutes with the autopilot disengaged.
It was at this point that the first event of note occurred. This chart represents the last 14 minutes of the flight, using transponder radar data depicting the aircraft's altitude against a time scale.
Approximately 12 minutes before the end of the FDR recording, the data indicate the aircraft was cruising in straight and level flight at an altitude of 31,000 feet and airspeed of 301 knots calibrated air speed, with the autopilot engaged.
Simultaneous with autopilot disengagement, the stabilizer trim moved to the apparent full nose-down trim position in about 6 seconds, and remained that way until the final upset. As recorded on the CVR, the crew commented that they were not able to maintain vertical control and actions to overcome the problem were discussed. The airplane began to descend at an average descent rate of 7,000 feet per minute, more than 3 times the typical rate of descent from cruise flight. During this sequence the speed brakes were deployed. After about a minute, the aircraft regained what could be characterized as controlled flight. At this point, flight 261 was at about 24,300 feet.
For the next 9 minutes or so, the aircraft was in controlled flight, descending from 24,000 to about 18,000 feet. Toward the end of this period, the crew extended the slats and then extended the flaps for a period of little more than 30 seconds, and the CVR reflects comments that the aircraft is controllable in this configuration. The crew then retracted the slats and flaps. CVR and FDR data show that the airplane remained in control at this time.
Things then began to happen very quickly. The aircraft was at about 18,000 feet, airspeed 270 knots, pitch attitude 2.7 degrees nose up. The stabilizer was in the full nose down trim position, and the elevator was deflected more than 12 degrees in the nose up position. This elevator deflection was approximately 50 percent of full travel.
At this time, the flaps began to extend to 11 degrees. Approximately 3 seconds after the start of the flap movement, the slats began to deploy.
Beginning at about 4 seconds after the beginning of the flap/slat deployment, the pitch attitude data show the airplane pitching nose-down at a maximum rate of 26 degrees per second, and reaching a nose-down pitch attitude of 59 degrees in approximately 3 seconds. The nose-down pitch rate decreases over the next 2.5 seconds as the maximum nose down attitude of 70 degrees was reached. During this rapid nose-down pitching motion, the aircraft experienced a negative 3G vertical acceleration. Although the pitch attitude began moving rapidly in the nose-up direction at that time, the aircraft did not reach level flight thereafter.
A 60-degree per second roll rate to the left began as the pitch attitude approached the maximum nose-down value. The remaining roll values from the recorder are consistent with the airplane rolling into an inverted position.
This final descent from 17,900 feet lasted just over a minute. The aircraft wreckage is located at Latitude 34 03.5 North, Longitude 119 20.8 West.
That is all we can currently report to you now from the two flight recorders.
I'd like to turn to an examination of some preliminary radar data. Dr. Ellingstad is placing before you a chart showing the route of flight 261 over the last 6½ minutes of its flight. The route of flight is derived from transponder information, and the last transponder beacon recorded an altitude of 1,600 feet. The open circles and open triangles on the chart indicate primary radar hits, which are possible reflections of radar signals off of objects. Some of these targets were recorded up to 2½ minutes after the last transponder beacon from flight 261.
You see where we have marked the area of the loud noise we reported to you last week that was heard on the CVR. This basically corresponds to the beginning of the final descent of the aircraft that I've just explained. These primary radar hits might be indicative - and I emphasize MIGHT be indicative - of something coming off flight 261 near this point. I will note that the path of the primary targets is consistent with the direction of the recorded winds at the time of the accident. Using these radar data, we have instructed assets of the U.S. Navy to search an area of the ocean where we believe something that would have departed the aircraft at that point could have landed, about 4 miles from the main wreckage site. This is of course a difficult task, complicated by the fact that we do not know whether something did actually separate from the aircraft.
As to other activities in this investigation, the Navy has completed mapping the accident area with underwater side scanning sonar and video. This information will help guide us in developing a salvage plan. Last night, the Navy found and recovered about an 8-foot section of what we believe to be the left horizontal stabilizer and some portion of the center horizontal stabilizer.
The Ventura County Medical Examiner's Office has told us that the remains of three of the persons aboard flight 261 have been identified and their families notified. Additional identifications and notifications are expected later today.
Our Maintenance Records Group convened in Seattle on February 2 and has interviewed 15 Alaska Airlines maintenance and dispatch personnel in Seattle and Los Angeles. The last heavy maintenance check was a "C" check on January 13, 1999 (C checks are conducted about every 15 months).
This aircraft had two previous stabilizer trim maintenance write-ups, both of them in 1999. In October, the trim system checked out OK and the aircraft was returned to service. In November, the alternate trim switch was replaced.
There were two widely publicized incidents last week involving other aircraft that reported stabilizer problems. Last Friday, we reported our findings on the February 2 American Airlines incident in Phoenix.
On February 4, Alaska Airlines flight 1613 departed Reno, Nevada for Seattle. Shortly after takeoff, the crew reported primary and alternate trim systems were inoperative. The crew declared an emergency and returned to Reno.
A Safety Board investigator traveled to Reno to oversee the Board's investigation. In the post-incident examination of the trim systems, the primary trim system operated initially normally, but the alternate trim function was intermittent. Both the primary and alternate trim motors were replaced, and at that point, the alternate system operated normally, but the primary system was inoperative. Several components associated with the trim system have been removed and are being shipped to our laboratory for analysis. We are also in possession of the flight data recorder and are reading it out in our laboratory. The cockpit voice recorder was written over by subsequent events during the flight so was not useful for our investigation.
This investigation has gone very well in the week since the tragic loss of flight 261. There is much left to be done, and we'll share further information with you in the future as appropriate. I'll be happy to take some of your questions.