The Loss of Alaska Airlines Flight 261 - January 31, 2000

(For those of you who may have logged onto this site on Tuesday, I apologize for being slow to post this update - the concentration of myself and so many at ABC on properly covering this accident has consumed every waking moment.)

As many of you know, being a pilot for Alaska Airlines myself (currently on personal leave), this is a very difficult accident to deal with in purely emotional terms. I must be considered biased because of my connection with the airline, but I also know as a matter of virtual certainty that the two pilots did everything humanly possible to bring their passengers and themselves back safely despite the control problem(s) they were battling, and the reasons why recovery became aerodynamically impossible must be fully discovered.

As I have said in person and on the air countless times, it is in the interests of everyone to maximize the flow of information and explanation about what happens in the aviation business, good or bad, because our society depends on the aviation system to such an extraordinary degree. What follows does not seek to solve this accident, or prejudge what happened. What I intend to do is explain some of the key systems that are more than likely in the direct causal chain of this accident so that you can better understand the unfolding investigation and the facts as they become available in piecemeal fashion. First, on Monday afternoon, an MD-83 piloted by two very seasoned Alaska Airlines pilots reported to Los Angeles Center a problem with the aircraft's "Stabilizer Trim" system. The flight was en route from Puerto Vallarta, Mexico, to San Francisco International Airport, and was passing the L.A. area at Flight Level 310 (31,000 feet) at the time. A request was made to divert to Los Angeles International (LAX) which was less than fifty miles from their position, and with the help of Los Angeles Center and SOCAL (Southern California Approach and Departure Control) the flight was given several descent clearances and vectors to LAX, and at one point - apparently at the request of the crew - given radar vectors that kept them over "the Bay" or over water as the crew worked with whatever problem they were having. Within approximately seventeen minutes of the onset of the problem, Flight 261 was observed by several other flight crews and a park ranger on the Channel Islands diving into the Pacific Ocean, out of control, and at times inverted. Subsequent search and rescue efforts have proven fruitless, leaving the airline and the NTSB to accept the terrible reality that the 83 passengers (which included over thirty members of the Alaska Airlines 'family' riding on passes) and five crew members are dead.

There are more facts known, of course, but this is the basic profile: The pilots reported having difficulty maintaining altitude because of a problem with the Stabilizer Trim (See explanation of the trim system below). Reportedly, they had descended involuntarily from 310 to 240 (24,000 feet), and at some point one of the pilots reported to air traffic control that they had the aircraft under control, but the other pilot says "no we don't." They were given further descent clearance as they requested, and stated that they needed to configure the aircraft. When the flight was in the vicinity of seventeen thousand feet and not far from Santa Barbara, it departed controlled flight and slammed into the water.

What Happened?
What Happened? First, please understand that, just as I've said on the air and with every major accident, the responsibility for determining what happened rests with the NTSB (National Transportation Safety Board). However, there are some clues that can be responsibly discussed, and an initial direction that the NTSB will follow.

Specifically, the crew themselves reported a Flight Control Problem. I put that in caps because it denotes a major aircraft system.

First, let's discuss the physiology of the Stabilizer and Trim system.
The Primary Flight Controls consist of the Ailerons (moveable panels on the back of the wings which control roll left and right), Rudder (The moveable vertical panel on the rear of the vertical tail fin which controls yaw left and right), and the elevators (the moveable panels on the rear of the Horizontal Stabilizer which controls the nose-up and nose-down pitch of the aircraft). When we use the term "Horizontal Stabilizer," we're referring to the smaller "wing" on the top of the tail at the aft end of the aircraft (also called a "T-tail" for obvious reasons). While the Elevators (the moveable panels on the rear of the Horizontal Stabilizer) are the primary control for pitching the aircraft up and down (climb and descent), canting the Horizontal Stabilizer up and down is an important function of controlling any swept-wing aircraft through its impressive range of flying speed. Why? Because as you accelerate such an aircraft from, say, 180 knots of airspeed up to speeds of 300-400 knots, the lift produced by the wings becomes correspondingly greater, and the nose of the plane wants to rise. In the cockpit, the pilots must do one of two things to compensate for that tendency if they want to maintain the same altitude: They must exert continuous forward pressure on the control yoke to oppose the tendency of the airplane to pitch up, or, they must trim the aircraft's Horizontal Stabilizer using the so-called "Pitch Trim" system to neutralize the nose-up tendency. They do this by canting the Horizontal Stabilizer (the entire surface) up a bit (with respect to the front of the stabilizer). This exerts more upward pressure on the tail, which in turn exerts more downward pressure on the nose, and causes the nose to pitch down slightly. If done correctly, the resulting control forces in the cockpit are completely neutral. When that's the case, the pilots can take their hands off the yoke and the aircraft will maintain its altitude. Slow down, you need to trim nose up. Speed up, you need to trim nose down. It's that direct a relationship.

What can go wrong?
Among the malfunctions that airline pilots train to handle on each type of aircraft, malfunctions of the Pitch Trim System have specific procedures. The most benign problem is an inoperative trim system that fails in flight. Essentially, the aircraft will continue to be completely controllable, and even "in trim" whenever it's flying at the same speed as when the trim failed. The most serious problem is an uncommanded movement of the pitch trim system, a so-called "runaway trim" situation. Although extremely rare (I've personally experienced one runaway in 13-thousand hours of flying, and that was in an Air Force C-141), a runaway pitch trim can get very serious very fast if not recognized immediately by the pilots, because the trim system can end up canting the Horizontal Stabilizer to an extreme nose-down position, or an extreme nose-up position. If such a thing should happen in flight, the pilots will have to pull or push the control yoke in the opposite direction of the trim to maintain level flight. In other words, if the pitch trim goes nose-down, the pilots must pull on the control yoke to deflect the elevators up to neutralize the Horizontal Stabilizer's attempt to put the aircraft into a nose-down dive. Since as much as a hundred pounds of physical force must be applied and maintained by the pilots (in the absence of returning the trim to the right position), you can see how such a condition would be a major emergency requiring the immediate coordinated action of both pilots, and a tremendous amount of physical effort.

I know this sounds terrifying, but let me assure you of two things: First, it is an extremely rare situation because of numerous safeguards I'm going to describe; and second, the few times it has happened (prior to this week, at least), the flight crew has been able to discover the problem and stop the movement of the pitch trim system significantly before the stabilizer ran to an extreme position.

Safeguards:
Immediate detection of any uncommanded movement of the pitch trim is paramount, so the first safety measure is to make sure the trim system cannot move - in other words, the Horizontal Stabilizer cannot be repositioned - without some sort of significant warning noise in the cockpit. In the MD-83, as with all earlier models of the DC-9, there is a loud tone heard every second when the pitch trim is in motion. You literally can't miss it. There is also a warning light.

In Boeing's 737, the earlier 727, and other models, there is a physical wheel on each side of the center console in the cockpit (between the pilots) which actually rotates in one direction or another, physically cabled to the pitch trim "jackscrew" in the tail. The trim system can't move without moving the trim wheels which make a very loud and clacking sound (and the pilots can even physically grasp and hold the wheels to stop a runaway).

Detection alone, however, is not enough. The manufacturers also build in automatic brakes on the pitch trim system which will not allow it to move without being commanded to move (and thus will not allow the Horizontal Stabilizer to be canted up or down). In the Boeings, the very act of opposing the motion of a runaway pitch trim with the control yoke causes the pitch trim brakes to slam into place. In the MD-83, the pitch trim brake is always on unless positive electrical power is applied to remove it while trimming.

Finally, each type of aircraft has several more safeguards built in to prevent any single failure from sending the trim system into unwanted motion. Therefore, to get a runaway pitch trim that actually repositions the Horizontal Stabilizer to a significant degree, (a) The primary switches or control circuits have to fail to start the motion (such as a short in the pitch trim switches on the control yokes); (b) the pilots have to be unaware of the motion, which means the aural and visual warnings will have to be rendered ineffective; (c) the brakes must fail; (d) the immediate emergency action all pilots are trailed to take to instantly disconnect the trim system (deprive it of power to move) has to fail. Only if all the conditions exist can the trim system motor the Horizontal Stabilizer into an extreme position requiring the pilots to battle it for control by using heavy control forces in the opposite direction.

I know this is complicated stuff, but the bottom line is simply that multiple, major failures have to occur before a true in flight control emergency is present.

Now, in the case of Flight 261, the pilots reported they were having trouble with this system, and had lost altitude before getting it marginally under control. That indicates some sort of nose-down trim had been applied. Later, they reported that the pitch trim was "frozen" (or so we are told at this early juncture). That makes the situation even worse, because if the Horizontal Stabilizer is now in the wrong position, a frozen system means the crew cannot motor the Stabilizer back to safer setting.

How can this happen in the first place? First, the system can fail internally - a remote possibility. That means that the actual mechanism in the tail that cranks the Horizontal Stabilizer's front end up and down is either jammed or stripped and won't move, and that also means that neither of the electric trim motors will budge it. In that case, the flight crew is stuck with whatever setting they've got, and may have to land the aircraft in a backbreaking effort to maintain opposite forces on the control yoke.

Second, the system can temporarily fail. What does that mean? That means that because of the airspeed of the aircraft and the extreme "air loads" on the tail (due to the opposite struggle of the elevators opposing the stabilizer, a nose-up versus nose-down tug of war), the motors on the pitch trim system may be unable to develop sufficient torque to reposition the Stabilizer, which, after all, is a very large surface (a small wing, in fact). If this happens, the only method of decreasing those tremendous airloads is to slow the aircraft, which is what the aircrew reported they were doing. The question is, can you slow down enough to regain the ability to move the pitch trim before you run out of elevator control, which decreases with decreasing airspeed.

In addition, any of us would want to do a "controllability check" while still at fifteen to twenty thousand feet to find out how slow we could safely fly the airplane for a landing approach. We would never want to do such experimenting close to the ground where there's no vertical room for recovery if we go too far in the test. Without doing a controllability check, we would be gambling unacceptably by guessing at what airspeed would be safe. During a controllability check, we would carefully extend the flaps, and perhaps the landing gear, and check the flying characteristics of the airplane in order to decide precisely how to safely fly back to an airport runway. It was during this phase of flight that the pilots of Alaska 261 apparently lost the ability to maintain controlled flight. Obviously, and tragically, nothing they did from 17,000 to impact enabled them to regain control and stop the dive.

Conclusions:
They have to wait, of course - conclusions, that is. What we have right now is just the following:
-The Pitch Trim system of the MD-83 being used as Alaska 261 was implicated by the pilots themselves as having caused a controllability problem.
-The pilots were unable to maintain controlled flight, which indicates a major malfunction or series of malfunctions in the flight control system.
-The pilots were correctly engaged in the process of configuring and otherwise probing the envelope of safe flight (airspeeds, configurations, etc.) With which to safely approach Los Angeles International when the aircraft departed controlled flight.
-The aircraft, with all aboard, were lost.

The cockpit voice recorder has been recovered as of Thursday morning and will probably be a significant help in knowing the precise nature of the pilot's struggle (what they were trying to solve to the best of their knowledge and discussion). The flight data recorder will be found soon, despite the apparent disconnection of its "pinger" (the small appendage which creates a sonar-readable sound for recovery purposes when under water). The FDR should show the NTSB the precise position of the Horizontal Stabilizer at all times during the flight, which will be invaluable information in reconstructing the sequence of events which led to the emergency. However, the NTSB will need to recover the entirety of the vertical tail structure, the Horizontal Stabilizer and the entire stabilizer trim system in order to know precisely what was wrong, and how the problem started.

In the meantime, it is appropriate, I think, to recall even in the depths of sympathy and shock over this loss, that airline flying continues to be the safest form of transportation, and that while galvanizing in every way, this tragedy is an anomaly, not a warning of deterioration. As aviation does so uniquely and well, we will suck this accident investigation dry of methods to guarantee there will never be a repeat.

And finally, please say a prayer for those lost, and for the grieving families of those left behind. There but for the grace of God, go any of us.

Back to top or previous page