Pelican's Perch #27:
AS261, the Media, and Pitch Control
The recent crash of Alaska Flight 261 near Los Angeles triggered a torrent of TV and newspaper reports that displayed profound ignorance of pitch control systems not only among reporters (where it might be expected) but also in the pilot community. After taking a few potshots at the media, AVweb's John Deakin describes the three basic types of pitch control systems plus a bunch of variations, and talks about what can go wrong and how pilots should react.
The recent loss of Alaska Airlines 261 was like an unexpected hard punch in the gut for me, and for everyone in the airline business, because we know that Alaska is one of the very finest airlines, with excellent attitudes prevailing in the cockpits and cabins, good equipment, good maintenance, good training, and highly skilled pilots who operate in somewhat more difficult weather and terrain conditions than most.
I can personally attest to this. I often have the privilege of riding in their jump seats as I commute between my home in Seattle and my crew base in Los Angeles. I have come to know some of the crews, and I can only say Alaska Air is a place I'd like to work. Most of us know that if this tragedy can happen at Alaska, it can happen anywhere.
The loss to family and friends is immeasurably greater, and most of the people in the industry are acutely aware of that, too.
Along with the rest of the world, I watched TV and newspapers a lot over the days following the crash, and as always, I was struck by the abominable job they do with "breaking news." With rare exceptions, there seems to be a compulsive need to fill the airwaves with senseless chatter, and fill column inches with meaningless prattle. The rare times they do bother to get genuine experts like Barry Schiff or John Nance on TV, they seem to go after them for the taped "sound bites" while not giving them much chance to make any intelligent comments. At the same time, they'll give endless minutes of airtime to some "talking head" that knows nothing except how to look his or her best on camera. A three-second sound bite is not enough for a real expert to say much. Astonishingly, network newscasts presented some general aviation pilots as "experts" when it was clear they had not the slightest idea how any jet transport systems work. Their calm, assured comments were flat wrong, and seriously so. I guess to the TV networks, "a pilot is a pilot, and airplanes are all alike."
While I'm slamming the media, let me mention another pet peeve. I feel like taking a roundhouse swing at my TV set with a baseball bat every time some fool with a mike shoves it in the face of a distraught relative, and asks the inevitable inane question "How do you feel about losing your whole family?" Have these morons no decency? Are the people running the TV news shows really that stupid and heartless? No one I've ever talked to will admit such "coverage" is a good thing, or worthwhile. This is carrying "if it bleeds, it leads" much too far!
On the other hand, there was some tasteful coverage on the Seattle stations of the many services and memorial gatherings that took place, although I thought those should have been more private, too. Seattle was hard hit by this one, as many of the passengers were from that area.
But enough of my soapbox.
The other thing that struck me was the profound ignorance of pitch control systems, not only in the media (where I expect it), but also among the aviation fraternity. I thought it might be useful to do a column on this very complex subject, if only to point out how many different systems and variations there really are ... and probably reveal my own ignorance in the process!
No, I am NOT trying to solve the AS261 crash that is the job of the NTSB, who seems to be doing this one right, in my opinion. However, I hope that what follows will help you to better understand the investigation into the horizontal stabilizer.
Just like the feathers on an arrow, the basic tail feathers on an airplane help to keep the airplane going forward in a straight line. But there's a lot more to the story when we start talking about controlling and trimming pitch.
C.G. vs. C.L.
Briefly, most pilots know that the Center of Gravity (C.G.) is that point on or in the airplane where it "balances" in all three axes. Put an eyebolt there, and hang it from the roof of the hangar, and it should remain in any attitude you place it. This is the point where all the "down" forces (gravity) are centered in level flight. On most common airplanes in level flight, it's somewhere above or below roughly the one-third point on the wing's chord line. The Center of Lift (C.L.) is exactly the same idea, it's the point in or on the airplane where all the "upwards" lift forces are centered, a "balance of lift forces." For fun, let's mentally put another eyebolt at this point. (Don't confuse "C.L." with "CL" which is used for "Coefficient of Lift," another matter entirely.)
New pilots (and a few old ones!) are always amazed that not only do the C.G. and C.L. NOT meet at the same point; we don't even want them to. Instead, the designers arrange the location and shape of the wing and the weight distribution of the airplane so that the C.L. is behind the C.G. This means that if you hang the airplane from the fictitious eyebolt at the center of lift, the airplane will promptly point its nose at the floor.
We avoid this distressing condition in flight by putting a horizontal tail surface on the airplane with some special characteristics to hold the tail down. This is nothing more than a small wing, mounted upside down. The curved part (if any) will be on the bottom, and the flat (or less curved) part will be on top. This is to counterbalance the nose-down situation described above.
We don't want too much of a good thing, because the more down force that tail has to produce, the more lift the wing has to produce to support the total weight. Picture a 4,000-pound airplane, with 100 pounds of "download" on the tail, and you'll see that the rest of the airplane must produce 4,100 pounds of lift. The drag from producing that extra lift can be considerable. Some cargo airlines take the extra trouble to load cargo so the aircraft is very close to the aft C.G. limit. This reduces the nose-down pitching moment, which reduces the "down force" needed to counteract it, all of which reduces the drag and can make a very noticeable difference in fuel burned.
A Balancing Act
But the most important result from all this is stability. Any increase in airspeed will produce more lift at the wing and more "down force" at the tail, making the nose come up, which will tend to reduce the speed again. Decrease the airspeed and you'll get less lift on the wing, less "down force" on the tail, and the nose will want to pitch down to regain the speed.
If the horizontal tail should happen to fall off in flight, the normal airplane will violently pitch nose down to approximately vertical, and stay there to impact. The good news is that tails rarely fall off.
If the designer didn't make use of this balance of forces, a little backpressure might pull the nose up, then the natural forces would make the nose come up even more, and the pilot would be in a constant battle to keep the nose pointed where he wanted the airplane to go. Some of the very early aircraft had this problem, and were quite difficult to handle. This is also why an aircraft gets a good deal more pitch-sensitive with an aft C.G., when the C.G. moves back closer to the C.L. and very little or no down force is required at the tail.
To this point, virtually all aircraft are the same, with exceptions being aircraft like the Beech Starship, the Wright Brothers "Flyer," Mignet's "Flying Flea" and other oddball types. Let's not go there, this time.
Three Different Trim Systems ... and Some Variations
However, even on "mainstream" airplanes, the devices to accomplish this basic purpose can be very different from each other, with some really strange ones used over the years.
Stabilizer and Elevator
Perhaps the first among what I'll call "conventional" systems is the simple fixed horizontal stabilizer with a movable elevator, as found on the Caproni N'20 or the deHaviland DH4 pictured here. Note the primitive external cables, simple control horns and complete lack of any device for trimming. Also see the Curtis Robin with internal cabling, and an external pushrod and bellcrank system to move the elevators. If you preflighted one of these, you got to see it all!
(My thanks to the Seattle Museum of Flight at Boeing Field for allowing me close access to the museum aircraft pictured here.)
Why were there no trim tabs? Most of these aircraft were very slow, so they had a very limited range of speed between stalling speed and redline speed, even in a dive. The pilots usually sat in the rear cockpit, well away from the C.G., but the airplanes were designed for an average pilot's weight back there, and any fuel, payload, or passenger seating was roughly on the C.G., so that anything from "no passenger" to a real heavyweight didn't matter very much, at least for C.G. purposes. These aircraft also didn't fly straight and level much, so no one cared if the pilot had to hold a little pressure on the stick to correct the balance during the rare straight and level portion of the flight.
Stretching the Point
But pilots are a lazy bunch, and often clever, so I'll bet it didn't take too long before some of them figured out how to use a few strips of rubber from a discarded inner tube tied between the stick and some handy point in the cockpit to relieve the load, and sliding the rubber bands up and down the stick to get it "just right." How do I know this? Because I've used this little trick for aileron and rudder trim!
Later, some designers would use a variation of this for trimming, building in some sort of bungee system connected to the cables, with an adjustable lever in the cockpit to vary the tension on the cables. The bungee created a new "center point" while still allowing elevator inputs.
It didn't take too long for the early designers to respond to the whines of the poor, overworked pilots, and one method of alleviating pressure on the stick was to add a small fixed tab at the back of the elevator, as on the 1928 Boeing P-12. By flying the aircraft, then bending the tab a little, it didn't take long to set the aircraft up for some specific pilot weight, configuration and speed. We see these little tabs on rudders and ailerons even today, correcting for minor flaws in the shape or balance of forces in roll and yaw.
Adjusting these fixed tabs is non-intuitive, for they must be bent in what most think is the wrong direction. This Boeing P-12 obviously had a nose up tendency that distressed the pilot assigned to it, forcing him to push the stick forward, holding some "down" pressure on the elevator. The little tab is bent UP, creating a small local amount of camber or curve at the trailing edge, and this creates a tiny bit of "lift" at that point, tugging at the trailing edge of the elevator (down in this case.) The advantage here is that this trimming force occurs without placing any stress on the control system. This can be important, because the control stick is a long lever, which means it is very powerful, and the elevator takes a lot of force to move it against the airflow. But the little bell cranks and lever arms in the system are not very long at all, which means they must be very, very strong.
Fixed tabs work great for fixed conditions, but as airplanes got faster, and as they could be loaded at different C.G. locations, something more was needed. Two common devices came into being: the adjustable trim tab, and the movable stabilizer.
A good example of the adjustable trim tab can be seen on the Grumman-American Cheetah. This one is set for considerable nose up trim, perhaps after the last landing. With this system, if the stabilizer is not at the correct angle for the loading, the elevator will be unfaired, creating drag, and if the pilot has to trim to hold it there, the trim tab will be unfaired (with the elevator) as well, creating more drag. This is not a problem on slower aircraft, but it becomes important at higher speeds.
The earliest example of an adjustable stabilizer that I know of is on the Taylor (later Piper) Cub, and it remained a favorite method on Pipers for many years, as on the Piper Tri-Pacer pictured here. The leading edge is simply cranked up and down by a system of cables and pulleys. By setting the stabilizer angle to remove all stick pressure, all variations in speed and C.G. can be corrected, and the elevator will be faired with the stabilizer in normal hands-off flight, reducing drag. This didn't help the early Pipers much, for they were no speed demons, but this is exactly the same arrangement found on almost all modern jet transports! Another neat feature of the movable stabilizer is that full elevator authority is always available when trimmed.
For example, without the moving stabilizer, and with an elevator that moves 10 degrees up and down, if the pilot must hold five degrees of elevator position for level flight, then there is only five more degrees of elevator available in that direction. With the movable stabilizer, the elevator should always be faired when properly trimmed, so full travel is available in either direction.
Another variation is the all-moving horizontal tail. The entire surface is hinged at roughly the center point. When the pilot moves the stick or yoke, the entire horizontal surface ("stabilator," as in "stabilizer+elevator") moves as one unit as on the Cessna Cardinal or the Piper PA-28R-200. But variations in C.G. and speed still affect the airplane; so again, some form of trim is needed. On this Cardinal, there will be some heavy-duty cables moving the stabilator in direct response to the pilot's input on the yoke, and smaller cables from the trim tab control in the cockpit to the trim tabs on the back of the stabilator.
Note the very interesting "slots" in the leading edge of the Cardinal stabilator. Cessna added these when they discovered that the early Cardinal stabilators had the nasty habit of stalling abruptly during the landing roll, causing the nosewheel to crunch abruptly and unceremoniously onto the runway. To remedy this stabilator stall, the designers were forced to add the slots to enhance airflow over the "top of the wing" (the bottom of the tail). These slots were an afterthought, added well after the Cardinal was first introduced, and retrofitted to the pre-existing fleet.
When you see slots like this on a wing, the air enters below the leading edge, and exits on top of the wing. Slots like this are often used on a wing just forward of the ailerons to improve roll rates. You may also see them used more extensively on STOL ("Short TakeOff and Landing") aircraft like the Helio Courier, or the Dornier DO-28, where they are installed for the length of the leading edge of the wing.
Bungees, trim tabs, and adjustable-incidence stabilizers these are the three simple, basic pitch control systems.
Cheating with Tabs
Control system designers often "cheat" a little, too. They will often add a little tab that looks just like a trim tab, except there is no control for it in the cockpit. This tab will be connected directly to the main surface (stabilizer or stabilator) by a linkage that will move it with or against the motion of the elevator (or stabilator). It may move more in one direction than the other. If the tab moves opposite to the surface it's hooked to, it assists the pilot by lowering the force needed to move the surface, and is called a "servo tab." If it moves with the surface, it adds to the force required (and improves the effectiveness), and is called an "anti-servo tab." Both will sometimes be called "geared tabs." These will be added and adjusted as needed to produce the desirable stick forces and control harmony. In some cases, they will be separate tabs, as on this Goodyear FG-1D Corsair. Here the two inboard tabs are trim tabs, and the two outboard tabs are servo tabs, according to Chris Avery, who flies one of the few remaining ones in existence.
Another interesting tab is the "spring tab" or "flying tab." The yoke is directly hooked to a cartridge with a heavy-duty spring in it, and that cartridge is hooked to the tab. Initial motion of the yoke will move only the spring tab, which will "fly" the elevator (or stabilator) into a new position. Once the spring cartridge runs out of travel, it will then move the elevator directly, as on the C-46. Or there may be no springs at all as on the DC-9, which runs all direct pitch control from the yoke by tabs alone. This is a very effective device, making relatively light controls, while providing immense control power.
Various combinations of all these may be found on some airplanes, and they are not limited to just pitch control. All of them have been used for roll and yaw, as well. On the C-46 you will find both spring tabs and conventional trim tabs on the elevators and the rudder. The first picture here shows the left elevator with the trim tab neutral, and with someone in the cockpit pulling back on the yoke, deflecting the spring tab. Note the control blocks are in place, preventing elevator motion, but there is a considerable amount of yoke movement to produce this tab deflection. The next picture shows the opposite yoke deflection for airplane nose down. Old-time check pilots liked to stump new hires with "Which tabs are which?" The ancient acronym is "OTIS" (as in elevators) for "Outer Trim, Inner Spring," but of course this may not apply to other aircraft! The outer trim tabs are also unusual on this airplane, being set so that the zero point has one side up 15 degrees, the other side down 15 degrees. This is a quick and dirty fix to control flutter. Wartime needs did not allow elegant fixes, and this artifact remains to this day.
The 1952 Martin 404 really gets carried away the deceptively simple-looking rudder tab on this one is a combination of servo, trim, and spring tab! Setting it up is a nightmare, and the designer is rumored to have spent the rest of his days cutting out paper dolls in the loony bin.
Using all these tricks and more, some very large airplanes have been flown only with cable systems. Most notable in this regard was the Hughes HR-1, the largest airplane ever built. (Howard Hughes hated the widely-used nickname "Spruce Goose.")
Big Airplanes, Weak Pilots
But larger airplanes get really heavy on the controls, and even the strongest pilots need some help. Most large aircraft today use various versions of hydraulic "boost" where the first motion of the cable system moves a hydraulic valve to simply assist the cable motion or as on the 747, where all surfaces are entirely hydraulic, with no connection whatsoever between the controls and the surfaces.
Hydraulic controls add a whole new layer of problems, and the solutions to those add more complexity, and more chances for failures. Fully hydraulic control surfaces are "irreversible," so the pilot has no feeling for speed, as the controls always move with complete ease. Certification requires that controls should get stiffer with increasing speed, so entirely separate "artificial feel" systems must be developed, added and tested, with possible failures and procedures for dealing with those failures.
As jet aircraft entered service, some additional problems were encountered, and solved in ingenious ways. All the jet transports operate over a very large range of speeds (100 to 500 knots). Most will double their empty weight when loaded, and some will nearly triple it. Designers must provide for much greater pitch control. For example, it is necessary to correct for the C.G. shift caused by flight attendants walking up and down the length of the airplane. Of course, some FAs have more effect than others. I hasten to add that the Alaska Airlines FAs are all slim, trim, smart and beautiful, so this doesn't apply at that airline. (I don't want hot coffee dumped in my lap the next time I ride on Alaska!)
On jets, the differences between IAS and TAS are far greater, Mach effects come into play, the aircraft must operate over a large range of temperatures (-54C to +54C in the 747, for example) and pressures (from 15 PSI at sea level to 2-3 PSI at altitude). It's a whole different world, and the design challenges are formidable. For one thing, the 747 changes its length by a couple inches with the extremes in temperatures. Cables can get very sloppy when it is cold, without special devices to keep them taut all the time.
The simplest and most common pitch system on jets is probably the trimmable stabilizer and moving elevator, just like the old Piper Cub. Due to much larger aircraft, much faster speeds, and much greater control forces, the stabilizer's leading edge is moved by heavy-duty electric motors (DC-9) or by hydraulic motors (most Boeings.) In all cases I know of, these motors drive a jackscrew arrangement, where one end of a long and very strong threaded rod is attached to the leading edge of the horizontal stabilizer, while the aft end of the horizontal stabilizer is hinged to allow roughly 15 or 20 degrees of movement. This jackscrew is exactly like a huge bolt, and the bottom end runs in a coupling exactly like a huge nut. In some cases the jackscrew itself spins inside a fixed "nut," while in others the jackscrew is stationary and the "nut" turns. Either way, the jackscrew mechanism runs the leading edge of the stabilizer up and down as needed. Generally, the pilots will have electrical toggle switches under their thumbs to run the stabilizer trim up and down at a high rate for maneuvering at lower altitudes and speeds. Most pilots turn on the autopilot for high-altitude, high-speed flight, as these airplanes are fairly miserable to hand-fly in that regime. Sometimes there is another control lever to run the stab trim, perhaps at a reduced rate, in case the electrically powered toggle switch fails. The autoflight systems will trim at the slowest trim rate.
Two Types of Brakes
Once all this machinery is set in motion to move the stabilizer, inertia comes into play and there would be considerable "coasting" if the designers didn't install some sort of system to "brake" the stab trim to a stop when the pilot (or autopilot) releases the trim switch. For electric motors, the circuit will often power the windings in both directions at once, bringing everything to a quick stop. For hydraulic systems, the usual control valves will do the trick with closed valves blocking all fluid flow. The result is trim motion only when wanted, and "instant stop" when done. This braking system is usually engaged all the time when trim is not being used, and must be released prior to trimming. When you see dual switches under the pilot's thumb, one is usually to release this brake; the other is to run the trim. As a safety measure, both must be moved to get the desired trim change.
The Jackscrew and Trim Motor Stall
By its mechanical design, a jackscrew is almost "irreversible." Just as you cannot normally turn a nut by pushing on the bolt (unless the threads are extremely coarse), the immense forces on the stabilizer cannot move the jackscrew arrangement.
Early jackscrew systems on Boeing 707s and Douglas DC-8s were badly flawed, as designers did not realize how much force might be needed in one unusual case. In one early accident, one of these aircraft flew into a massive updraft, pitched sharply down (weathervane effect) while gaining altitude in the updraft. Instinctively, the pilots responded to the altitude gain and began trimming nose down (leading edge of the stabilizer full up), arriving at that unfortunate stabilizer position just as the airplane ran out of updraft and began to respond to the nose-down attitude and trim with a vengeance! The airplane nose pitched down hard and the speed built up very rapidly. The pilots promptly pulled back on the yoke to get the nose up, and when they ran out of elevator travel, they began frantically trimming as well. But the rapidly increasing speed placed such a load on the out-of-position horizontal stabilizer that the trim motors could not move it from the full-stop position. The elevators were not sufficient to overcome the badly mis-set stabilizer, and the airplane was lost.
The solution to this was bigger, more powerful trim motors on later airplanes, so that no combination of factors will "stall" the stabilizer trim drive motor. Also, trim limits for the thumb switches were reduced, with full trim available only by "manual" handles, or the like. New prevention and recovery techniques were also developed notably, leaving the autopilot on (but with altitude hold off), and also flying by straight pitch attitude without chasing the airspeed or altitude with control inputs or power/thrust changes. Since these procedures were discovered and implemented, no further upset accidents have occurred unless there were other factors involved. Northwest's legendary Paul Soderlind was an early pioneer in the research into high-altitude upsets, and developed the techniques we still use today.
There were also malfunctions in the electrical control systems that ran the trim motors, so "stabilizer brakes" were added to some. If the trim is in motion, and the yoke is moved in the opposite direction, the trim is blocked from further movement by various means, physical or electrical. On the 727, it was a heavy-duty pin that acts just like the "Park" position on an automobile's automatic transmission. When activated, it would slam into the mechanism with a huge "klunk," bringing it to an instant and violent stop. I hated testing that puppy, it just wasn't a good sound, and it had to be hard on that pin!
Several different failures can cause unwanted trim, so most airplanes have some sort of "trim-in-motion" warning. Boeing aircraft tend to have little wheels that go clickety-clack when the trim is in motion. The 727 has one that is very large and noisy, but with the added benefit of the pilot being able to actually crank that wheel to move the jackscrew directly (slowly and painfully). It takes a lot of turns and a strong arm, but it works. The 747 has a similar wheel, but it is driven by very light-duty cables off the jackscrew assembly, and serves as an indicator only. Early Douglas DC-8s had no warning at all, until one was nearly lost at JFK and only quick action by the crew saved the airplane. The trim ran away to full nose-up, but the quick-thinking captain rolled into a bank steep enough to help counter it until they could sort it out. Ever since then, Douglas products have a loud, very intrusive horn, or even a voice warning that seems to yell at the pilots, often just when someone keys a mike to talk to ATC. You can probably tell which warning I prefer!
Move the Tail
Finally, some airplanes have "all moving" tail surfaces, like the F-4. These are most often seen on transonic and supersonic airplanes, due to the very strong control forces needed in those regimes, and some very bizarre effects from shock waves that require major inputs. These are universally driven by hydraulics, and the systems must be very carefully designed to permit moving the "center" point for trim, and for synthetic "feel" to the pilot in all speed ranges from stall to maximum speeds.
Some Unusual Variations
Concorde (please don't put "the" in front of "Concorde") has a most interesting pitch system, as there is no tail as such. Most of the trailing edge of the delta-winged bird is taken up with moving surfaces called "elevons" (elevators+ailerons) most of which serve multiple purposes. There are no flaps at all every landing is a "zero-flapper." According to James Bedforth, a captain on that magnificent airplane, this control system is one of the most daunting systems to learn during the transition training. Not only are there multiple functions of the surfaces, it is the world's first "fly by wire" system (with cable backup controlling the hydraulic actuators). Trimming simply resets the center point of control travel, and there is a system to provide artificial feel to the pilots. Further pitch control comes from moving fuel around among 13 tanks to improve the C.G. as needed for the various stages of flight. The fuel system is the second most daunting system to learn, according to James.
Lockheed, never content with "simple" when "complex" will do, put another oddball on the L-1011. This looks like a moving stabilator with trim tabs or flying tabs, but the hydraulics actually drive the stabilator, with the "tabs" being moved by linkages to assist! There is no direct connection between the "tabs" and the control yoke, according to Chuck Tully, a longtime captain on this machine.
Lockheed also did a complex design on the Connies, where pitch control is by a conventional hydraulically boosted elevator with the usual tabs, on a fixed stabilizer. But with the hydraulic system out, elevator forces on this very large old aircraft would be so high that pilots could not control pitch. So Lockheed put a shift mechanism in to change the leverage when hydraulics fail. This reduced the available elevator travel by 60%, but reduced the force needed by the same amount. You can bet that system gets checked on every flight!
One of the most bizarre pitch control systems was installed on the unfortunate Grumman F10F "Jaguar," a little-known and even less-respected effort by a manufacturer of many fine airplanes. Famed test pilot Corky Meyer was the only person to ever fly this abomination, and he did a fascinating article in the April 2000 issue of Flight Journal. (Highest praise for this publication, it never fails to delight me.)
Picture a long, skinny torpedo-like thingie mounted right up on the very tip of the vertical stabilizer, with a stabilizer fixed to that. That's funny enough, but now picture that whole torpedo as being fully free to tilt up and down, no connection to anything, except at the hinge point. Finally, a very small controllable wing was mounted to the very front of that, and this was hooked to the stick for pilot control. I get the distinct impression that Corky didn't like it much. I get the feeling that somewhere early in the design, someone made a bet that he could do something very different. He did, but I doubt he tried to claim any winnings on this turkey.
But enough of the mechanics; let's move on to some failure modes.
A "runaway" stabilizer by itself is extremely dangerous because it can be so insidious at first. If the pilot is maneuvering, it will manifest itself by increasing pressure in one direction, but it moves fairly quickly, and by the time the pilot realizes that his trim inputs (thumb switch) are not having any effect, the stab trim is well on its way to a very dangerous condition. It takes very quick thinking to realize "hey, we've got a runaway," then reach over and flip the cutout switches to stop the stab motion. Yes, the "trim-in-motion" warnings will sound, but since they are constantly sounding during all normal operations, pilots tend to blank them out. For this reason, most jets have some sort of automatic "stabilizer brake" installed. The pilot's yoke moving in opposition to the trim usually actuates this device. In other words, if the trim is running away and moving to produce "airplane nose down," the normal reaction of the pilot will be to pull the yoke back, and that will "brake" the trim motion fully automatically and well before it gets out of hand. It may cut off electrics, or with a hydraulic trim it may electrically close a valve, or otherwise halt the trimming.
With the Autoflight On
Perhaps worst of all is the runaway stab trim when the autopilot is flying the airplane, because the autopilot itself will counter the initial trim change with simple yoke movement, without the pilot even being aware of it. Yes, most airplanes have warnings that will sense a mis-trimmed condition, and even sound an alarm or light a warning light, but by the time action is taken, considerable mis-trim may exist.
Punching off the autopilot is a very natural thing to do, but it's usually the wrong thing to do. That may produce a very abrupt pitch change that could be very hazardous, either to passengers or to the structure of the airplane. Far better to disable the trim, take some time, analyze the problem, get the trim back in line (if possible) first, either with an alternate stab trim system (if installed) or by changing the speed of the airplane to match the trim, while either leaving the autopilot on or holding the yoke VERY firmly as the autopilot is disconnected.
Unwanted Nose-Up Trim
Now, let us assume that for some reason that trim signal gets "stuck" and starts running the leading edge of the stabilizer DOWN, pitching the airplane's nose UP. Initially, the pilot (or autopilot) can handle it by feeding in elevator against it but by the time the cutouts are moved, there will be considerable nose-up trim, and the pilots will be holding a lot of forward force on the yoke. If there is no backup, or if the stab becomes "jammed" at that point, it will be a real chore to hold the nose down. With two (or more) people, they might take turns, or brace a knee against the yoke.
Solution? Slow the airplane down until the airplane is at the speed where that trim would be needed. At that speed, the pilots have plenty of control. When it comes time for landing, putting the flaps down will cause the nose to pitch down, and the pilots may have to hold considerable back pressure, or restrict flap extension, or even make a no-flap landing.
Unwanted Nose-Down Trim
Much more dangerous is the runaway trim in the nose-down direction. Now the pilot will have to exert a constant heavy pulling force, and this gets VERY tiring VERY quickly. The only solution may be to speed up, but if you are already at cruise, that will have little or no benefit. Any attempt to slow down will produce greater and greater nose down pitching forces, and the pilots will have to exert more and more "pull" to counter it. On the airplanes like the DC-9 with limited elevator authority (tabs only), this will get very hairy very quickly. As speed drops (or the aircraft descends and TAS drops), the pilots will eventually run out of back yoke, and after that the nose will begin to drop inexorably. That should increase the airspeed, which would help, but the natural reaction to increasing airspeed would be to pull the power off, making the situation worse.
As control is lost, the airplane will go into an outside loop, airspeed will quickly go past redline, and at some point, structural breakup will probably begin. This is almost certainly what happened to Alaska 261, and this may have been exacerbated by some part of the tail breaking away, further reducing control.
These pitch control systems have proven so reliable for so long on so many airplanes that most airlines have gotten away from training for "runaway trim" and "jammed stab." We used to do it all the time, every six months. We'd end up making an ILS with both pilots holding very heavy pressure to control pitch. That's realistic, but miserable, and few of us thought it a worthwhile exercise after doing it once or twice. Or the instructor would fail it with a nose-up runaway, so that changing to landing configuration would make it a non-event. I'm personally glad I don't have to do it every six months, but the unfortunate part of doing away with it in practice is that many pilots have now forgotten the basic principles behind dealing with pitch trim failures, and that's bad.
One of the basic principles with this kind of failure is to DISABLE the stab trim, and LEAVE IT THAT WAY. There are exceptions to every rule, including this one, but this is one system where you're almost always better off just leaving it alone once the immediate failure is contained. In fact, that's often true of most emergencies, and is precisely the reason we are not supposed to do much trouble-shooting on any systems.
Be careful up there!