The devil is in the deadly details

Could you solve the puzzles in time?

Careful pilots use checklists. One item on all checklists calls for the controls to be free. After studying two accidents, one in a new production twin on a first flight and one in an experimental jet, because the ailerons were reversed, I paid extra attention to controls free and correct. I looked at the ailerons when I deflected them, every time, and made sure they moved correctly. I knew full well that they worked fine on the previous flight, and that they had not been subject to any maintenance, but I just wanted to make sure.

Now I’ll ask you the same question that I recently asked myself. Do you ever look back at the elevator as it moves and verify that it is both free and correct?

You can probably guess why I asked that question and your guess is probably correct. I’ll tell you why anyway.

An airline pilot did a beautiful two-year restoration of a 1947 Piper PA-12 Super Cruiser, N3280M. I feel a kinship because my first airplane was also a 1947 PA-12, N3389M, built not long after 80M.

PA-12 crash
The controls were free, but were they correct?

The restoration included all new fabric and replacement of the flight control cables and electrical wiring.

Ready for the first flight, the pilot got takeoff clearance from Rwy 27L at Orlando Sanford International airport to remain in the traffic pattern.

According to a witness, the airplane accelerated normally for takeoff, pitched up, and continued to pitch up into a full stall, rolled to the right and nosed in on the right side of 27L. The inevitable cell phone video was consistent with this description. The airplane was consumed by post-impact fire.

In determining that most everything was functioning properly before the cash, the NTSB found one anomaly. Manipulation of the elevator control cables revealed that a nose-up stick input resulted in a nose-down deflection on the elevator and vice-versa. Further examination revealed that the elevator control cables were improperly rigged, such that they were attached to the incorrect (opposite) locations on the upper and lower elevator control horn.

When I was considering the reversed aileron cases I wondered if, after missing that during the pre-flight and checklist work, a pilot would have had any chance of recognizing this after liftoff and compensating for it. Given the split-second nature of the event I thought the chances of surviving it were between slim and none, and much closer to none.

I went through the same mental exercise on the PA-12 accident where there was, to me, a bit more of a puzzle.

A normal takeoff in a PA-12 would involve raising the tail after a little speed has been gained, then running level while accelerating to flying speed, and then lifting off with a little back pressure on the stick. That would not have been possible with the elevator control hooked up backwards. Pushing the stick forward to raise the tail would only serve to keep the tailwheel on the runway.

Cessna control check
A lot of pilots check the ailerons, but how about the elevator?

I wondered if the fact that this pilot flew airline jets could have been a factor. I have seen jet pilots accelerate to V1 without their hand(s) on the control wheel and then place both hands on the wheel and rotate at Vr. For an airplane that is not at a positive angle of attack while on the ground, that is an accepted way to take off if there is no crosswind.

In a PA-12, with hands off the stick, the airplane would roll in a three point attitude until it reached flying speed, or, the tail might come up just a bit as it accelerated. That would depend on where the trim was set. Then, when it flew, still hands off, it would seek its trim speed. If the nose came up naturally as the airplane accelerated, and the pilot tried to lower the nose, the reversed elevator controls would mean the pilot would actually be raising the nose by pushing forward on the stick.

I don’t think any pilot would be able to figure this out before the airplane crashed. If the pilot knew the elevator was backwards to begin with, he might have a chance, but if he knew that in advance, he’d hardly be trying to fly the airplane.

That’s just one reason why the preflight check of the controls is important. It’s also important to move all controls through the full range, for two reasons. You want to make sure they are free through the full range but you also want to make sure your seat position allows you to move every control to the stop. Many pilots seem reluctant to aggressively use the controls, and it’s not really good to fight the controls, but you want to be able to go to the stop if necessary.

I owned one airplane where this was quite pertinent. A Cherokee Six doesn’t have an overly large vertical fin and rudder and the lower part of the surface is mostly out of the air flow. The few times I got ice on my Six, the only ice on the vertical was on the top half.

I based it at an airport with a single north-south runway so there were notable crosswinds in the windy seasons and more than once there was not enough rudder authority to push the nose straight down the runway before touchdown in a strong crosswind. In other words, I got to the rudder stop before I solved the problem. The solution was to go around and fly to another airport with a runway more nearly aligned with the wind.

MD-83 crash Michigan
The pilot caught the error in five seconds, but the MD-83 still went off the end of the runway.

Then there was the MD-83 takeoff abort with the University of Michigan basketball team on board. The final report is far from complete, but preliminary information indicates that this crew might have been fooled by the fickle finger of fate.

The crew apparently did the controls check and nothing seemed abnormal so they started the takeoff. Initial reports said the takeoff was aborted because of wind gusts reaching well over 50 knots but this didn’t have much to do with what happened.

When the crew did the elevator control check, the left elevator moved but the right one did not because it was jammed in the nose down position because of a mechanical problem. When the airplane reached 152 knots, which was Vr, the pilot flying applied back pressure to rotate the airplane but the pitch attitude did not change. The pilot realized that something was wrong in five seconds, by which time the airspeed had reached 166 knots. The takeoff was rejected and by the time the acceleration had turned into deceleration the airspeed had reached 173 knots and the airplane went off the end of the runway.

The crew and passengers were okay but the airplane was substantially damaged.

It will be interesting to see how the NTSB treats this in the final report. You can’t see the elevator surfaces from the cockpit of an MD-83 and from preliminary information it appears that the controls in the cockpit moved normally. In at least some military operations an outside observer visually checks the control surface movements as the pilot does the pre-flight but I don’t think that has ever been done in airline operations.

This crew didn’t know anything was wrong until the airplane would not rotate. Should they have caught that in less than five seconds? Maybe yes and maybe no and a first thought might have been related to a possible effect of the extremely strong crosswind. The takeoff was on Rwy 23L and the wind was from 260 at 35 with gusts to 50. That would get at least some of your attention.

If they had rejected the takeoff in, say, three instead of five seconds after it wouldn’t rotate, the airplane might have still gone off the end and I think it’s going to be hard for the NTSB to find fault with the crew on this one.

The maintenance on the airplane could be a whole ‘nother subject. The airplane did function normally on the previous flight and the last maintenance was done before that flight but something had to have been amiss for a while.

Gulfstream IV gust lock
A quick check of the controls would have caught the engaged gust lock in a Gulfstream IV that crashed in Massachusetts.

The crew and passengers in a Gulfstream G-IV did not fare so well after a rejected takeoff. This one was almost hard to believe.

In the investigation, the NTSB learned that this crew did not do the controls free check before 98-percent of the preceding 175 takeoffs. That was critical before the accident flight because the crew was not too good on the rest of the checklist and did not release the gust (controls) lock.

When the airplane reached Vr and the pilot not flying called for rotation, nothing happened. The controls were locked. Eleven seconds elapsed between the call for rotation and the first sign of brakes application. Four more seconds passed before the power was reduced. By that time the groundspeed was up to 162 knots and only 1,373 feet of runway remained. The airplane went off the end at a good clip and hit things before ending up in a shallow ravine where it blew up.

I have known a lot of pilots who don’t make full use of checklists, especially in familiar airplanes. This accident illustrates the danger of operating in that manner.

Moving on, I’ve never seen a checklist item about checking the elevator trim as free and correct but trim was at least on my last minute killer items checkist: Instruments-Controls-Fuel-Trim.

How many pilots flying with electric trim ever go to the trouble of ensuring that the electric trim operates correctly?

Apparently a Baron pilot did not do this when the airplane was returned to service after maintenance and the electric trim had been hooked up incorrectly and worked in reverse. The pilot took off, never figured this out, lost control of the airplane, and crashed.

Why did he lose control? One thing that was not covered in the old Part 23 certification rules was a requirement that an airplane be controllable with the trim run all the way in either direction. There are rules on maximum required force for all imaginable configuration changes and the limit is 75 pounds with two hands on the wheel and 50 with one hand. That is quite a push or pull but it did not apply to full trim. Maybe that has changed in the new Part 23, maybe it hasn’t, but for a number of pilots that would be like shutting the barn door after the donkey is out.

Cessna 172 trim wheel
When in doubt, check the manual trim wheel – don’t try to override the autopilot.

There was a Cessna 182 accident, in England I think, that involved the autopilot and the electric trim and something the pilot did incorrectly before takeoff. Because of the pilot’s error, the takeoff was made with full nose-up trim. The nose pitched up and the airplane stalled and crashed. The pilot had apparently not made a last check of the trim.

There have also been problems with pilots who don’t fully understand autopilots and electric trim. If you try to override an autopilot, the trim will sense a need and trim against your force and it will keep trimming until you quit pulling, for example. Again, if the trim goes to the stop the force might be too strong to overcome.

Years ago, an airline pilot sensed that the altitude to capture on a descent had been incorrectly entered and sought to manually level at the correct altitude while the autopilot was still flying. As best I recall, control wasn’t lost but when the crew realized that something was wrong and took the correct action of punching off the autopilot, the airplane was way out of trim and went whoop-de-do, giving some passengers a bad time.

The startle value of something like that can’t be underestimated. A pilot expects an airplane to behave in a certain way and when it doesn’t, it’s a complete surprise. Perhaps there is an identifiable pilot trait that would suggest quick recognition of such but avoidance is the best way.

I was talking recently with a person I taught to fly 50 years ago, when he was a teenager. He hasn’t flown since but has remained interested and he asked me: What’s with all these old guys being careless and crashing old airplanes? He had read in his local news about a pilot in his 80s who had trouble getting both engines going on an old 310, who exhibited some confusion before departure, and who crashed soon after takeoff, all according to news reports. There was also word that a witness saw smoke streaming from one of the engines.

There are a lot of reasons an airplane engine can be hard to start. Some of those reasons should preclude taking that engine for an airplane ride and some are innocuous. The pre-flight run-up might give a further clue but I know of several cases where pilots, one an old friend, came a cropper because they took off in a twin after one engine showed obvious signs of distress during the run-up. Maybe this is a case of a pilot trying something in a twin he wouldn’t even think of in a single.

Takeoff by Cessna
If it doesn’t feel right on the takeoff roll, don’t rotate.

The takeoff itself has to be considered a part of the pre-flight preparation because that is what it is, right up to the moment when you lift off.

How so?

It’s an excellent practice to take stock of everything after full (or takeoff) power is applied and the roll begins. I always liked to include the exhaust gas temperature reading(s) in the scan because I knew where it was supposed to be and if it wasn’t there I needed to do something different. My P210’s fuel control system had no temperature compensation (I think it came in a Cracker Jack box) and in warmer temperatures the mixture actually needed to be leaned on takeoff lest it be too rich for proper operation. I knew where to position the mixture control for a given outside air temperature and the EGT check was to make sure it was in the right place.

Everything else aside, on takeoff I always asked myself if the engine was running strong and smooth after power was applied. If it didn’t sound good and feel good, I didn’t want to go flying. I would hasten to add that in 20,000 hours I didn’t abort a lot of takeoffs but the ones that I did abort needed to be aborted.

I guess the moral to the story is that there are an almost unlimited number of gotchas in flying and the best avoidance plan is found in carefully checking and considering everything. Maybe twice.


  • Not too long ago an F-15 based in Europe, I believe, crashed and burned attempting takeoff after depot maintenance activity. Turns out the elevator controls were cross-connected. The video was horrific. The pilot was fighting the a/c and didn’t abort in time to avoid the smashup.

    Man, you would think the controls in a sophisticated a/c like an F-15 would be foolproof but clearly not.

    Thanks for your articles. I try to learn something from every one and I usually find several nuggets in each article.

  • Every time a do my walk around on a kitfox IV, when checking the port side I check that the control surfaces move in the correct motion (both ailerons move opposite to each other, elevator moves together) then when I get to the starboard side I leave the door open so that i can see that when I move control surfaces again i make sure the control column inside moves the right direction for the movement I am putting on the controls. I have been doing this since day not because I was taught as a student but because before I learned how to fly I helped rebuild and build several planes and this just seemed like a good idea to me

  • I love your articles. As a relatively new pilot they are very helpful. I was with a friend once who saw me checking the movements and she suggested I check the ailerons when they are in a neutral position as well as when full elevator pressure is applied in both directions so Ill know they’ll work at all pitch’s. I honestly don’t know if it does any good but it makes me feel better about my plane and I darn sure don’t ever want to be “one those pilots” Richard talks about :-). Thank you for keeping the lessons simple to learn!

  • Another example of elevator trim wired backwards, “Thursday, December 12, 1957, U.S. Air Force B-52D Stratofortress, No. 56-597, from the 92nd Bombardment Wing, was taking off from Runway 5 at Fairchild Air Force Base (AFB)…” “According to eyewitnesses, as the aircraft left the runway, it made an abnormally steep climb to an altitude of approximately 2,000 feet. Fire started coming from the jet engines and pieces of metal began flying off the engine cowlings and mounts. The aircraft stalled, executed an abrupt, right wingover and appeared to level off. But, at an altitude of approximately 500 feet, the plane nose dived and crashed in a stubbled wheat field one mile west of the base.”
    “It was caused by faulty wiring in an electric motor controlling the horizontal stabilizer which controls the climb, making the aircraft to do the opposite of what the pilot intended. It was the first time faulty wiring had caused a B-52 to crash ”

  • M. Long,

    If you look carefully at the J-3, PA11, and PA12 aircraft there is a jack-screw that adjusts the angle of incidence of the horizontal stabilizer on those model aircraft thereby affecting the trim of the aircraft. If that were the case with this accident the pilot would have noted heavy stick forces that would be attributed to an out of trim condition. More than likely that would have been a survivable flight to a full stop landing. However, the opposite is true with a reversed control cable connection condition which resulted in this unfortunate fatal accident. Sadly, no matter how much time you have accrued in aviation, all of that time is behind you. What is most important is the next hour you are flying and how thoroughly you pre-flight can and shall determine the outcome as it did in this accident analysis. The first control movement check should be on the ground during the pre-flight walk around. The second check for smooth control movement in the run-up phase of the pre-take-off checklist.

  • This story is from Kathmandu, Nepal. Many years earlier of course. A part of the checklist begins with a ‘walk-around’. But I would say many airline pilots while doing the walk-around miss out few items or avoid doing it as many people including passengers are watching him/her. He or she feels kind of intimidated. Such actions may not have happy endings.
    This pilot was seen standing under the wing of his Piltaus Porter that was ready for departure to Lukla. His did his walk-around, did his checks whatever that was, waved his hand to ground crew, taxied and took-off. After airborne, the aircraft nosed dived and crashed. Fatal.
    Synopsis: External Flight Control locks were not removed. The pilot missed it and so did the ground crew. Among the deceased were wife and two daughters of Sir Edmund Hilary who was at the receiving end in Lukla.
    When walk-around and system checks are done four five times a day, complacency creeps in. Unfortunately, we have never devised a checklist to root out complacency. Sometimes we have to punch ourselves a few times.

  • I was not used to a rudder lock till I got a different airplane. I am worried I might forget to remove it so I tied a rope from it to the wheel chocks. This reminder works for me.

  • As I was having a Super Cub rebuilt, I noticed that my mechanic had rigged the elevator trim backwards. When I brought it to his attention, quite an argument followed. Though he was a pilot with a high performance bush airplane, as well as being a mechanic, he simply couldn’t get it through his head that I was right,

    Finally, I rolled the elevator trim to its nose down setting (the trim’s forward end pointed up, obviously confusing him), and pointed out that such a setting would force the elevator to its full down position. He finally saw the light.

    My wife and I decided to perform the final wing rigging . . .

Leave a Reply

Your email address will not be published. Required fields are marked *