Air France 447 - On topic only!

[EconomyClass]

Altitude Loss

I remember Eastern Airliens Flight 401 crashing in the Everglades. I believe it was the deadliest crash in the USA at the time. I remember being incredulous that the whole staff of the cabin were all trying to solve an indicator light for a landing gear, bumped the plane off autopilot, failed to notice the descent of the plane, only realizing the crash was about to happen when they saw the ground. This can happen? Remedial steps were taken afterward, but I suspect it just means that a different scenario might have to happen to bring about the same problem. If AF447 was descending at the rate suggested by data, then the cockpit voice recorder should contain at least one voice saying something about the altimeter dropping like a stone. That's why I wonder if the voice recorder contents are being withheld till the final report is issued, just to postpone the fury.

http://en.wikipedia.org/wiki/Eastern...nes_Flight_401

[Gabriel]

Knew it didn't add up:

http://www.flightglobal.com/articles...itude-law.html

The abnormal attitude law is a subset of alternate law on the aircraft and is triggered when the angle of attack exceeds 30° or when certain other inertial parameters - pitch and roll - become greater than threshold levels.

Alternate law allowed AF447's horizontal stabiliser to trim automatically 13° nose-up as the aircraft initially climbed above its assigned cruising altitude of 35,000ft.

The stabiliser remained in this nose-up trim position for the remainder of the flight, meaning that the aircraft would have had a tendency to pitch up under high engine thrust.

Crucially the abnormal attitude law - if adopted - would have inhibited the auto-trim function, requiring the crew to re-trim the aircraft manually.

After stalling, the A330's angle of attack stayed above 35°. But while this exceeded the threshold for the abnormal attitude law, the flight control computers had already rejected all three air data reference units and all air data parameters owing to discrepancy in the airspeed measurements.

I see a flaw here. With all likelihood the THS would have already been at 13º ANU by when the plane reached 30º of AoA (and hence when the abnormal alternate law would have been adopted if the ADRs hadn't been rejected), so even if Abnormal Alternate Law had been adopted at that point it would have made no difference in the THS getting to 13º and staying there for the remainder of the flight.

My opinion is based on two sources: My interpretation of the BEA report (that I quoted below) and a simple analysis (added below between brackets):

At 2 h 10 min 51, the stall warning was triggered again. The thrust levers were positioned in the TO/GA detent and the PF maintained nose-up inputs. The recorded angle of attack, of around 6 degrees at the triggering of the stall warning, continued to increase. The trimmable

horizontal stabilizer (THS) passed from 3 to 13 degrees nose-up in about 1 minute [starting from when? probably since the pilot first pulled up at about 2:10:05] and remained in the latter position until the end of the flight.

Around fifteen seconds later, the speed displayed on the ISIS increased sharply towards 185 kt; it was then consistent with the other recorded speed. The PF continued to make nose-up inputs. The airplane’s altitude reached its maximum of about 38,000 ft [an airplane doesn't keep climbing with more than 30º of AoA, so the AoA was much lower here, around the onset of the stall], its pitch attitude and angle of attack being 16 degrees [here, the AoA was 16º, but the THS was already at 13º].

Finally:

Failure to realise a need for manual re-trim was central to the loss of an Airbus A320 over the Mediterranean Sea about six months before the AF447 crash.

Maybe, but if the pilot was pulling nose-up, why would he apply nose-down trim even if he knew that autotrim was off and manual trim was required instead?

[Gabriel]

(2) In approach to stall situations (stickshaker) in RVSM, I want pilots to ALWAYS apply corrective pitch reduction and then carefully add thrust if needed. Releasing backpressure on the stick will not reduce pitch on an Airbus in Normal or Alternate Law.

Oh yes it would. Why not?
Normal and alternate law are load factor demand. If you are pulling back you are commanding more than 1G. If you release the back pressure the plane will strive for 1G, and that requires less lift and hence less AoA (as long you are not around the stall zone).

(3) For full stall, I want them to know how to determine the nature of the stall and have full situational awareness. Now that, I realize, is a bit optimistic, but it is what I think we should always strive for. For instance, if the THS is at its upper limit, there isn't any chance of stall recovery

Are you sure of that? Why design the THS with enough travel to reach a position where there can exist no other state tahn stall, even with full nose-down elevator?

Especially in this case where the plane remained in Alternate Law (load factor demand and active autotrim), by pushing the stick fully forward you are commanding the lowest load factorwithin the Alternate Law envelope. I don't know what that lowest load factor is, but is probably zero or near zero. The plane was falling at about 1G. So pushing the stick will have the effect of the computers trying to reduce the AoA. That means nose down elevator and nose down aoutotrim to cancel the force on the elvator.

...there may be a bug in the system, or this might just be the auto trim attempting to trim the downforce out of a very steep FPV as commanded by the sidestick inputs. It is designed to do this. Even while falling out at 10,000+ ft/min, it appears that the commanded PFV was something like a 10° climb gradient.

No Airbus control law is FPV demand. The sidestick commands either load factor (in nornal or altenate law) or elvator deflection (and hence AoA) in direct law.

But if logic instead allowed the ADR's to continue providing AoA data to the EFCS, the flight would have entered Abnormal Attitude Law, and the autotrim would have frozen, and the PFD would have advised the pilots to trim manually. That's how I'm seeing it as this point.

Re: My previous post.

More important in my view, the Stall Warning could have kept working.

Apparently from the BEA report, at one point the stall warning stopped (because the AoA was considered below the false 60kts speed, but the pilots surely didn't kneo it was for that). At that point the pliot applied nose-down command (why? I don't know), and then....

.... the stall warning sounded again!!!!! (because the false airspeed went above 60kts again, again surely th pilots didn't know it was for this).

Pilot's view:

I'm pulling up.
The stall warning stops.
I push down.
The stall warning starts again.

What would be your next step if you were the pilot?

Of course, that was very well into the event already. The pilot had already failed the UAS procedure (or even keeping the plane more or less stabilized), the stall prevention (apporach to stall), and the initial stall recovery.

At that point, I'm afraid, a pilot would not be able to recover even if that would mean just pressing the "save me" button.

[Gabriel]

Apples to oranges.

What he was getting at was that the 447 crew was out of the loop due to design.

Not to mention day versus night and he did have a panel to work with. He sure had talent and ... a pile of luck and fate was stacked with him that day.

[off topic]

In my opinion, Sully made just one thing especially right that day. One single major thing that saved the lives of all the persons in that plane, and maybe of many more on ground:

Having the clearness of mind and the balls to decide to try the Hudson (when 3 airports that, maybe, we could try to reach).

That is a hard thing for a pilot. To accept that the plane WILL be lost, that you will not land in a runway, especially when it is not clear that you won't reach a runway. The call of the airports nearby is very strong and very hard to disobey. He did it, and kudos for that.

The rest was a piece of cake: he only had to land in the longest and widest runway ever, no need to touch down in the touch down zone (you could be as long or short as you want), no need to touch with the centerline between the wheels (the river was a dozen of planes wide). I mean, really, what was the alternative? To aim the nose directly into the river?

And he didn't do all the things perfect either:

The landing could have been smoother (the vertical speed at touchdown was quite higher that the average landing) He flared too high too slow and the alpha-floor protection prevented him from keeping pulling up.

He said "brace for impact" over the PA. He could have advised about the water landing so the passengers could retrieve the life vests from under the seats and the FA were prepared for a water evacuation.

He didn't use the "ditch" switch that closes several valves related to pressurization in an attempt to prevent water from entering the cabin. That was at the end of a long checklist and they didn't have time to run it up to the end. But they didn't go to the ditching part (config, speed, etc) and still made the landing. He knew about the ditch switch and it could have occurred to him to use it.

[/off topic]

[Evan]

I see a flaw here. With all likelihood the THS would have already been at 13º ANU by when the plane reached 30º of AoA (and hence when the abnormal alternate law would have been adopted if the ADRs hadn't been rejected), so even if Abnormal Alternate Law had been adopted at that point it would have made no difference in the THS getting to 13º and staying there for the remainder of the flight.

The loss of auto pitch trim in ABNORMAL ATTITUDE LAW should have resulted in a prominent USE MAN PITCH TRIM message on the FMA (FPD)*. I think this could potentially make a big difference:

(*There is nothing in the FCOM or FCTM that specifically states this, but they do state that this message appears when auto pitch trim is lost.)

Maybe, but if the pilot was pulling nose-up, why would he apply nose-down trim even if he knew that autotrim was off and manual trim was required instead?

The BEA report is frustratingly vague:

At around 2 h 11 min 40... The PF made an input on the sidestick to the left and nose-up stops, which lasted about 30 seconds.

Meaning what? That the pitch command was cancelled at about 2 h 12 min 10?

At 2 h 12 min 02...At that moment, the thrust levers were in the IDLE detent and the engines’ N1’s were at 55%. Around fifteen seconds later, the PF made pitch-down inputs.

At around 2 h 12 min 17, a pitch down command. Still in excess of 10,000 ft to recover. We know both pilots were looking at the PFD's, because they both report no valid indications. If the message USE MAN PITCH TRIM was displayed, I think it is fair to assume they would have seen it. If having ABNORMAL ATTITUDE LAW in effect would have produced this message, then it could have made all the difference. But were they ever trained on how to use the man pitch trim during stall recovery?

[Gabriel]

I think I worded this poorly. Releasing the stick will not reduce pitch, it will maintain the last pitch command due to neutral static stability, but reducing back-pressure, i.e. moving the stick back towards neutral, will decrease to pitch rate and lower the AoA. Do I have that right?

Since the stick is self-centering, what is the difference between "releasing the stick" and "moving the stick back to neutral"?

I've addressed this a couple of post above.

[Evan]

Oh yes it would. Why not?
Normal and alternate law are load factor demand. If you are pulling back you are commanding more than 1G. If you release the back pressure the plane will strive for 1G, and that requires less lift and hence less AoA (as long you are not around the stall zone).

Check my recent post where I corrected this. I meant "releasing back pressure" as in "releasing the stick" not "releasing back pressure" as in "reducing back pressure". (see the attached diagram).

Are you sure of that? Why design the THS with enough travel to reach a position where there can exist no other state tahn stall, even with full nose-down elevator?

No, I'm not sure of this. Perhaps the stabilizer downforce is less than I expect at full deflection. The stabilizer travel is designed to trim a steep FPA for neutral static stability.

No Airbus control law is FPV demand. The sidestick commands either load factor (in nornal or altenate law) or elvator deflection (and hence AoA) in direct law.

Again, see the attached diagram. Neutral sidestick maintains PFV due to auto pitch trim (neutral static stability).

[Evan]

Since the stick is self-centering, what is the difference between "releasing the stick" and "moving the stick back to neutral"?

"moving the stick back towards neutral, i.e lowering the pitch rate command. I expect that this would lessen the amount of pitch commanded, but when the stick is then released, the last recorded value is maintained. I am actually asking if this is the case here.

[3WE]

They lost the engine and not just lift but the slats on that side.

So with 20/20 hindsight they now have a better slat design but you still need to fly the airplane.

You infer that AA 191 pilots failed to fly the plane while distracted with a number of warnings?

[Gabriel]

"moving the stick back towards neutral, i.e lowering the pitch rate command. I expect that this would lessen the amount of pitch commanded, but when the stick is then released, the last recorded value is maintained. I am actually asking if this is the case here.

Check your diagram... closer.
I have to go now but I'll come on this later.

[3WE]

...The rest was a piece of cake...

Gabriel: Careful there, my friend. I think we are witnessing a classic case of mis-translation between Spanish and Englais.

What Sully did was NOT a "piece of cake"! The probability of a survivable ditching is $hitty at best.

I think what you are trying to say is that what he did was "straight forward" and that he followed procedures that almost all pilots could and would have done (not just right out of the A320 owners manual, but a 757 manual, or even a Cessna 150 manual).

Also, given that ditching is so often fatal- I have to wonder if he didn't do something "magical" (and perhaps lucky, or perhaps a special skill he had read about) during the initial contact with the water?

[Evan]

Check your diagram... closer.
I have to go now but I'll come on this later.

I know that, in Normal and Alternate Law, the g-load commanded is proportionate to stick deflection, so reducing the deflection would reduce the g-load request. I also know that the Airbus has neutral static stability in Normal and Alternate Law, so upon releasing the stick to neutral, the auto pitch trim will maintain FPA until a downward pitch command (-g) is introduced. That is how I arrive at my understanding of sidestick-to-pitch behavior. But correct me if I'm mistaken...

[guamainiac]

When you come over that section of the Bronx at 3000 and have no power you have the canyons of the city, the river and then the docks and cliffs of the Jersey shore. In the distance to the west you will see Teterboro and that may have been an option but I personally doubt he would have made it.

Had he tried he would have been faced with the cliffs all with business and apartments and then a bit of the swampy flats.

In reality he had only the river as a sure bet, Newark and it's big runways also out in the marshlands would have been a few miles out of reach.

To stretch the glide with that much fuel and weight would have been insane.

Teterboro and Newark beckon like the sirens that lure you to the rocks.

[MCM]

Evan, you're using a simplistic version of "in the loop". It is far more in depth than just knowing when something is happening.

Its about the whole design of the machine, the whole operating concept.

Do you see that the very fact there is so much discussion on the nuances of the different flight control laws is a demonstration of that at work?

I'm not saying its a bad thing the way Airbus design aircraft - I never have. But I am saying that their design does make it harder for pilots to be completely in tune with what the aircraft is doing at all times. Impossible? No. Harder, yes. When you have a failure, to have to question what your control input is going to do, is a great demonstration of that.

Does it save more lives than it costs though? Absolutely.

[Gabriel]

Gabriel: Careful there. What Sully did was NOT a "piece of cake"! The probability of a survivable ditching is $hitty at best.

I think what you are trying to say is that what he did was "straight forward"

Last time off topic, on this branch at least, I promise:

I meant "piece of cake". Once the decision to "be in the Hudson" was made, the rest that he had to do (glide the plane at any 2reasonable" speed -it was not necessary to hold best glide speed or to hold any one speed for the matter-, while aiming at no specific touchdown point, flare and touchdown at a survivable vertical speed) was very easy to do (a piece of cake). He could not be too high, too low, too long or too short. If he was too fast then the flare would have been longer and floaty but that wasn't a problem. To flare the plane to prevent digging the nose in the water is easy, it's just what he does in every landing. I mean, really, what could he have done worse as to worsen the outcome? At the end, the good result of the landing was a combination of an average "stick and rudder" skill and the chance factors that the river was there and that it was a pretty calm river and not the middle of the ocean in the middle of a storm with 10ft swells at 90° from a 40kts crosswind.

When you come over that section of the Bronx at 3000 and have no power you have the canyons of the city, the river and then the docks and cliffs of the Jersey shore. In the distance to the west you will see Teterboro and that may have been an option but I personally doubt he would have made it.

Had he tried he would have been faced with the cliffs all with business and apartments and then a bit of the swampy flats.

In reality he had only the river as a sure bet, Newark and it's big runways also out in the marshlands would have been a few miles out of reach.

To stretch the glide with that much fuel and weight would have been insane.

Teterboro and Newark beckon like the sirens that lure you to the rocks.

In fact, the airport that was closer was La Guardia.

During the accident investigation the question whether they would have made La Guardia or not was raised. So they tried in the simulator several. And they succeeded. Sometimes. When the pilots did all perfect. A slight delay, a sharper or shallower turn, and they could not make it.

This shows that La Guardia within reach. Barely, but within reach. That also means that it was close enough to produce a strong temptation give it a try.

Now, of course that Sully didn't have this information. He didn't have the option to say "Ok pause, let's try it in the simulator, no that turn was too tight, let's try again. Ok this bank works. Let's do it". He had to assess the situation and make a decision. Fast. I've heard him tell that the Hudson was "the only surface long enough, wide enough and smooth enough that he knew they could reach".

"Knew" is the keyword here. It happens many times that a pilot tries to reach a place that he thinks that he can reach and finds himself trying to stretch a glide that violates the only laws that no pilot can violate no matter how hard he tries: the laws of Physics.

That's why, if Sully thought "Hey, I think we can probably make La Guardia. I know that we can make the Hudson", and he opted of the Hudson, I applaud him. And I'd do it even if the simulator had shown that La Guardia was easily within reach with margin.