The Great Ongoing Aviation Thread (general and other)

Miscellaneous and Personal Stuff I Must Share
4643

In other NTSB news, they recently released the docket and final report on an MD-80 failure to takeoff mishap.

Short version being the horizontal tail had been damaged and jammed by high wind overnight and the crew first figured this out when the airplane would not rotate, long after they were committed to a severe runway overrun.

The twist in the tale of the tail is that this was a known problem w MD-80s that Boeing had addressed by a pilot procedure change a year before. Which change this company had ignored, so these pilots had no idea about the problem and the way to detect it on preflight.

I flew the MD-80, DC-9, & 717 for about 6 years total, all long before the FAA/Boeing procedure change. They all have substantially the same tail subject to the same risk of damage in high winds. I’d never have known to check for that either. It’s odd how even after ~2,500 airframes flying tens of thousands of hours each for 3+ decades, even in the late twilight of the fleet’s collective life, there are still gotchas being uncovered.

4644

Why aren’t all airliners required to check for functional control surfaces? It’s such a basic part of a check list. And why couldn’t they trim it to climb out. That would be a different system.

4645

Did you read the report? Or any of the many pages of system and elevator damage description?

The design of the DC-9/B-717/MD-80 tail is very interesting.

The control columns are attached to small trailing edge tabs, left column to left elevator trailing tab, right column to right elevator trailing tab. The columns are ganged together by a frangible mechanism under the cockpit floor. So the columns normally move together until/unless you pull hard enough in different directions then they uncouple from each other and then are completely independent control paths to the tail tabs. I did that once inadvertently on the ground. Caused a cancellation and a 20-manhour repair. The Boss was not amused that day.

The elevators are attached to … nothing except their hinges at their leading edge. They simply free-float in the airstream. There is no direct cockpit control of the elevators at all.

To pitch up, you haul back on the control column which moves the tab(s) down which in turn flies the elevator in the opposite direction (up) which in turn pushes the tail in the opposite direction (down) which therefore pushes the nose in the opposite direction (up).

It’s not relevant to this accident but the DC-9/B-717/MD-80 rudder & ailerons are the same way. The wheel & pedals move tabs that should move the surfaces. In the presence of sufficient airflow and if those surfaces aren’t jammed somehow.

When the elevators are mechanically jammed in the down position, it doesn’t matter what you do with the control columns and tabs; the elevators aren’t moving. And the same would apply to a jammed aileron or rudder.

As to pitch trim …

In theory the electric stab trim might be able to generate enough tail downforce to overcome the jammed elevators generating tail upforce. In fact I bet it could. But probably not in the time available and also with no guarantee you could maintain control when airborne.

We had 4 criteria for abort above 80 kts and below V1: engine fire, engine failure, predictive windshear alert, and the catch-all, a belief that the airplane was incapable of flight. That list was a company mantra, but comes pretty much right out of the Boeing & McD-D books and FAA’s industry guidance. Good bet most operator’s mantra is similar.

So what are the things that constitute “incapable of flight?” That’s deliberately kept kinda vague and open-ended as all catch-all categories should be. But my personal list included things like collision, structural damage, all engines fail, big explosion, gear failure, and … you rotate and it doesn’t.

Which last discovery will always happen above V1 since V1 is always at/below VR and the airplane is still accelerating the whole time you’re pulling and it isn’t reacting and then you’re deciding what to do about that. So now you are committed to a reject that will almost certainly run off the runway end at some speed and perhaps major speed. But that’s a hell of a lot better than getting airborne and discovering at 300 feet and 180 knots that you’re not in control, then nosing or rolling in.

As to checks …

The standard pre-takeoff checks include running the controls to the stops. There are no surface position indicators for the tabs, much less the elevators / ailerons / rudder. So in an MD-80 that’s just proving the columns can move the tabs. But it’s actually not even proving that. It’s only proving that you can move the controls to the stops in each direction without obstruction, and that what you can feel of all the downstream mechanical stuff feels pretty much like it always does.

As the NTSB report says, one thing that cannot be detected by the feel of the columns during the normal control check is jammed elevators; it’ll still feel exactly normal. Which is exactly why that procedure amendment was made a few years ago to detect this problem on the exterior preflight, since that’s the last chance to do so. And that was the amendment this company simply ignored.

The 737 is semi-similar in that, although unlike the MD-80 all the control surfaces are hydraulically driven, there are, like the MD-80, no cockpit indications of their actual motion. You’re just running the controls to the stops to feel for binding or anything else weird. With only your own subjective sense of what’s weird. And of course that lack of surface position indication was inherited from the 727 and 707 before.

Newer kinds of airplanes do have surface position indicators so during the check the pilots can see indications of actual surface motion and they can verify that each surfaces’ response is normal. But not on these antiques.

4646

That sounds like a stabulator tail. Which still begs the question, why aren’t airlines required to check the functionality of control surfaces. V1 seems like a bad time to sort things out.

4647

Again the point is that certification standards of the 1960s permitted that design. Modern certification standards do not. Airlines are generally permitted to use old airplanes as they were designed. Yes, some upgrades have been mandated.

But every single mishap should not result in a massive hardware retrofit project. That way lies madness.

4648

No, it’s a tail that uses servo tabs to move the elevator. If there is no airflow over the tail surfaces, they won’t move on the ground. So you can move the column back and forth in the cockpit, and even a perfectly functioning elevator won’t move. So you can’t wiggle the controls and verify that the control surfaces are moving, and moving in fhe correct direction.

The servo tabs would move, and you could watch to make sure they move correctly, but if the tail itself is jammed you wouldn’t see it. Not until the plane was going fast enough that the elevator should move - and doesn’t.

4649

Surely that’s the purpose of following the Boeing bulletin. To verify the elevators are not jammed.

4650

Yes, that is the purpose of the bulletin. But the content of the bulletin amounts to:

During exterior preflight, look up at the tail. If both elevators appear much trailing edge down they might be jammed. Stop right now and get them inspected by maintenance with a cherry picker before departing.

The bulletin has nothing to do with any control movement checks performed from the cockpit either at the gate or during taxi.

It was / is completely routine to taxi out behind an MD-80 and see one elevator full up and the other full down. Especially along the taxiway parallel to the runway where the airplane is likely to be experiencing a tailwind. They simply flop around in the breeze. Which is exactly why, when the “breeze” gets to be 50+ knots & gusty, all that violent flopping may eventually bend or break something internal and now you have a jam.

4651

I don’t disagree with anything you’ve said. But it’s still the duty of the crew to ascertain the integrity of the plane.

I’m surprised the bulletin didn’t go further. The elevator(s) can be tested on rollout before V1 by lifting the nose slightly to demonstrate they’re functional. Lifting the nose to the full travel of the strut maintains steering control.

The bulletin should also have required every plane be checked by a mechanical whenever winds exceeded 50 knots while the plane was on the ground.

4652

There’s a different bulletin for the maintenance folks that did just that. Although I don’t recall without checking exactly which wind speed it specified. When the DC9 was brand new 40 years ago there was a mandated check for winds over 65 kts. So the concern has been there all along, but 1960s engineering analysis wasn’t quite up to the task of modeling gusty flow around a building.

Again it took 2500 airframes about 50K hours each for this problem to manifest twice across the fleet. There isn’t a smoking gun here.

4653

True.

4654

In other unhappy news:

More details here:

A quick look at Google maps shows hilly forested terrain everywhere that’s not a road or the nearby town. Except for a couple of decent-sized farms & golf courses well off-axis to the runway. So if this was an engine failure, there are darn few places to conduct a forced landing, and definitely none from low altitude if taking off to the northwest.

The 90/270 turnback (if that is what they were doing) has a very low success rate. The articles are confusing and semi-contradictory on this point. It’ll be a couple weeks before NTSB gets around to releasing the prelim.

Even being god’s gift to aviation can’t save you from some circumstances nor from some mistakes in potentially salvageable circumstances. Don’t know which it is yet, and we may never know.

4655

That’s rough. That’s one of those ones that really forces you to realize that if it could happen to him, it could happen to me. None of this avoidance thought like “I’m safer than that guy.”

4656

A comment I often make about single and light twin takeoffs is that before brake release you need to have consciously decided that you’re willing to crash in order to avoid being killed. And put some actual thought into where you’re going to cause that crash. Because if you need to do it, there won’t be time to figure it out from scratch.

A challenge in any two-pilot flight is a difference of opinion on how willing you are, at the moment of truth, to aim the airplane at something that’s going to destroy the airplane and probably injure both of you.

Discipline, panic, survival instinct, wishful thinking, startle, expectation bias, etc., all come into play with no time to spare and no practical ability to talk once instinct takes over. Dying in a fight over the controls has a long pedigree in aviation, even if the documentary evidence of same is sparse except on modern big airplanes with high tech DFDRs.

4657

Excellent advice. I’d add that if you regularly fly out of a certain field, take the time to figure out what your altitude is for a safe return, and below that figure out in advance where you’d go in the case of a powe failure. "If I lost the engine now, I’d go for that field. " “If I lost it now, I’d put it down on that road.” Cement the idea in your head that turning around and trying to land back on the runway is almost never a good idea.

Another good idea: Get up to a nice high practice altitude, set the plane up for best climb, then yank the power and see how much altitude you lose when you try to A)get the nose down and set up best glide, B) do a 270, then a 90 degree turn to put you on the reciprocal heading as if you are landing on the runway. I’ll bet you’ll be shocked by how much altitude you lose if you haven’t done that before. Once you load up the wings in a descending turn, the altitude will drop off very quickly.

4658

Your second paragraph is really a gem. Although it’s usually 90 first, then 270. :slight_smile:

Depending on details of aircraft, configuration, etc. I’ve heard rules of thumb at 800 or 1000 ft. Nowhere close to the 200’ to 400’ where people tend to try it.

4659

My Grumman AA1 was worse than that. Stubby little wings with a high wing loading for a light aircraft. At my home field they used to let us do ‘overhead break’ landings. If I crossed center field at 1500’ and pulled the power, a 270 for landing would put me pretty much on the numbers. In that plane, you couldn’t make the field if you lost the engine in the pattern unless you were flying really tight, and almost never if you had to follow someone on final. So overhead break landings were considered ‘safer’, but I doubt it. A standard, stabilized approach is probably better for a recreational pilot. I did them because they were fun. I loved that place (Namao AFB, RIP).

I knew a guy there who flew F-104 Starfighters. He said that some joker had inserted an ‘engine out landing procedure’ in the manual:

In case of engine failure:

  1. Reach under seat. You will find a brick.
  2. Open canopy.
  3. Throw out brick
  4. Follow brick to ground.

The other F-104 cold war joke he liked to tell:

Q: How do you get your own F-104?
A: Buy a plot of land in Germany and wait.

4660

My dad, Air Force radar operator, used to say that the glide path on an F-104 was “point of impact equals engine out location unless there was a headwind, in which case it was slightly behind the engine out location”.

4661

Yeah, the F-104 was not great at gliding. In reality it had a glide ratio of about 5:1,

The Starfighter really was a great airplane. I believe it still holds multiple records. And it flew fine when used for its intended purpose - high altitude intercept of enemy bombers. Unfortunately, that mission went away when ICBMs came along, so the F-104 was repurposed as a low level fighter/interceptor. And that’s where it developed its reputation as a ‘widowmaker’.

Especially bad for the role were its downward-ejecting seats, done that way to prevent the pilot from being cut in half by the tall tail had they ejected the normal way. If you could roll on your back, you could do that and eject. But if the airplane was out of control and at low altitude, it would ruin your whole day.

4662

Which unfortunately almost nobody does. One of the ways to mitigate it is to make a short field takeoff and corkscrew up around the airport pattern until you get to a useful altitude. Something I use to do at night because you can’t even see where you’re going to crash.

the hardest thing to do is to push the nose down immediately in order to stabilize airspeed. You’re already committed to descending so plan A is to do it with as much control as possible.