AirAsia flight missing

[Gabriel]

Don't think af447 climbed to a stall, it just kept losing airspeed while the aircraft/pilots tried to maintain altitude until it stalled.

What are you talking about and what-s the difference?
AF climbed 1500ft at 7000fpm peak, losing a good bunch of its airpseed in the process and placing the plane above its ceiling, so the situation was not sustainable, and when the stall warning sounded, they pulled up again and climbed another 500ft before fully stalling.

[Gabriel]

somebody quickly explain (if you have a min) what might have happened here in non pilot terms if it was the same as the AF crash.
way i think of it with what i've read or seen)...Ice gets in the little pitot tubes (even tho these are heated) which give wrong airspeed indications...so plane is going faster than it thinks...pilot pulls nose up...but then it climbs too steep and then stalls....Is that right?
I'm not trying to speculate....im genuinely interested but the language you guys use is a bit perplexing to non pilots and i'd like to understand how this happens....if you dont wanna speculate on this one tell me in non pilot terms how it happened to the AF plane.

thanks

Well, you are quite close, except the highlighted part.
When the pitots fail (all three give different or unreliable readings) the plane doesn't "think that is going faster/slower than it is". It just "stops thinking".

It AF, the autopilot and autothrottles disconnected as soon as the airspeed became unreliable. After that, and for reason that nobody understands, the pilot pulled up. A lot. Very hard. Then you can continue with your description.

This accident might have been similar to AF. There are something that really look too similar (but note that many are just rumors or unverified information):

Airbus (A330 vs A320, but the systems design philosophy and "flight computer" logic are mostly identical).
Flying near strong storms in a latitude known as the "Inter Tropical Convection Zone" (ITCZ). A too steep climb (6000 fpm in this case, 7000 fpm in AF).
The flight climbs above its assigned altitude and losses a lot of speed.
The pilots are overwhelmed by the situation and don't even attempt a distress call.
The airplane crashes with high vertical speed and slow airspeed.

Again, much of the above is to be taken with a grain of salt (ok, better use the whole salt shaker). If you are going by the official information, all we know is that the plane crashed.

[Gabriel]

Not true, judging by flightaware data, on many short hops I've been on, those E175s easily climb at 3500fpm. Even a 747 going on a 6hr transatlantic flight can climb at 2500fpm initially.

Exactly, initially. Not at cruise altitudes.

[3WE]

Almost any aircraft can attain 6000fpm climb.... Until the speed bleeds off.

Yes!

The climb reported here is a little unusual for normal operations but with a good updraft and a little extra speed climbing at 6000 FPM is not super remarkable- but then again it likely indicates that something was wrong.

[3WE]

Any pilot can stall any airplane at any speed or altitude.

Yikes. I think your fundamentals are suffering from procedural overload. It's attitude, not altitude and the phrase does not include that a pilot can do it.

In fact if you limit a plane's nose-up elevator authority enough, a pilot may be unable to stall it at all sorts of speeds and attitudes and flight modes.

Right, Gabrel?

[Gabriel]

In fact if you limit a plane's nose-up elevator authority enough, a pilot may be unable to stall it at all sorts of speeds and attitudes and flight modes.

Right, Gabrel?

Not really. It might meet FAA's stall-resistant criteria, but that doesn't mean that it can't be stalled "at all sorts of speeds and attitudes and flight modes".

Take for example the Ercoupe. The plane was designed with limited elevator authority to avoid a stall and very limited rudder authority to avoid spin (that alone tells you that stall was not impossible, since no stall = no spin).

Now, take the Ercupe, firewall the throttle, nose over into a shallow dive, go to the red line, pull up to the load factor limit (or the end of the elevator authority) and the plane will not stall, but it will pitch way up. Go vertical and keep it there until you are completely out of speed. The only way out of that involves a stall.

On the other hand, take an airplane that, even with the autopilot off, will do by itself a lot of things that a human pilot would or should:
- Not roll 30° normally and, in worst case, not past 60° (even if the pilot hardly attempts to do so).
- Pull up to bleed off speed if the plane goes too fast.
- Never pull up past 2.5 Gs (even if that means ignoring the pilot's command).
- Not let the plane go past 30 deg ANU, even if that means ignoring the pilot's command.
- Pull up in turns to keep the nose (and altitude) from going down.
- Adjust pitch to keep vertical speed or pitch in changing airspeed.
- But if the speed goes too slow / AoA too high, then:

- Add power (even with the autothrottles off) to keep the vertical speed without losing any additional knot. If that doesn't work,
- Stop trying to keep the vertical speed or pitch and keep the AoA (or link the stick input to AoA instead of vertical speed / pitch as it was doing). If that doesn't work,
- Not let the AoA increase one single degree more past the "Alfa max", even if that means not only ignoring the pilot's command, but even pushing the elevator down to the stop when the pilot is pulling back to the stops on the sticks.

- Call it Airbus...
And you get a plane that is as unstallable as it gets.

That is, until an icecube hits the pitot and the plane quits and says "your plane, but now treat me like a Piper Cub -one without airsped indicator" (or even worse, "I'll keep keeping the pitch for you, so I'm not stable in speed or AoA, but I will not protect you from slow speed anymore") and you have a plane that is AT LEAST as stallable as any flown by a pilot that may be not very current with the stalls and might be thinking that he is still flying an unstallable plane.

Oh, I forgot, add a bit of "what is it doing now" and call it Airbus (again).

[Evan]

- Call it Airbus...
And you get a plane that is as unstallable as it gets.

That is, until an icecube hits the pitot and the plane quits and says "your plane, but now treat me like a Piper Cub -one without airsped indicator" (or even worse, "I'll keep keeping the pitch for you, so I'm not stable in speed or AoA, but I will not protect you from slow speed anymore") and you have a plane that is AT LEAST as stallable as any flown by a pilot that may be not very current with the stalls and might be thinking that he is still flying an unstallable plane.

One of the principal reasons you can't build an unstallable plane is the method by which computer systems can ascertain situational awareness. Pilots have their eyeballs, their central nervous system and their equilibrium; computers have ring-laser gyroscopes, radio signals and air data sensors. To make envelope protection possible it is necessary to have at least two sources of the same information and a way to compare the two for discrepencies. If they disagree, both must be considered invalid because it is impossible to know which to trust. To make these systems practical (fault-tolerant) a third source is needed, allowing for the system to continue functioning with one failure. In this configuration the sources each 'vote' and either the majority wins or, if all three vote differently, the system shuts down and the pilots take over (fault-passive).

There are very few scenarios where all three sources will be unreliable. A double or triple gyroscope failure has never happened to the best of my knowledge. The only IR occurances were actually ADIRU failures (both on Airbus and Boeing planes). Air data occurances are rare but not un-heard of because of common air data sensors exposed to a common environmental phenomena such as icing. AF447 showed us (though the industry was alreadly aware of it) that this could occur at any phase of flight. In a double data source failure scenario, the system is legally blind, envelope protections cannot function and the plane is eminently stallable. Thus the pilots must be careful not to stall it (fault-passive).

But, because of this 'voting' process, even in normal law a stall-resistant plane is stallable due to another rare but not unheard of phenomena: bad voting. It is possible for two or even all data sources to be unreliable and yet still in agreement. In this case the system will continue to function assuming there is no failure, and thus malfunction. This could occur due to things like equally blocked pitots or frozen AoA sensors or a glitch in the comparator logic. In this scenario, any plane is stallable in normal law. It is therefore imperative that the pilots always monitor both their primary instruments and their autoflight systems and know instinctively how to react when they behave erratically. (fault-passive) It is their job to do this. It is their job to know how to do this.

And so it is plain to see that any airplane is stallable at any speed or altitude (yes altitude because we were talking about altitude excursions and safe vs unsafe altitude).

If you understand one thing about the A320, understand this: it IT IS NOT UNFAULTABLE, IT IS FAULT-TOLERANT ON A GOOD DAY AND FAULT-PASSIVE ON A BAD ONE. This also applies to Boeing, Bombardier, Embraer and every other airframe manufacturer in the sky.

FAULT PASSIVE only applies if the crew is alert and prepared to react to fault situations. That requires supplemental training, CRM and situational awareness. In complex modern aircraft, situational awareness requires procedural discipline.

If they are not trained, prepared and alert then the aircraft is fault-intolerant and not airworthy.

Oh, I forgot, add a bit of "what is it doing now" and call it Airbus (again).

Or call it 777 or 787 or any commercial aircraft designed since 1990.

[elaw]

Or call it 777 or 787 or any commercial aircraft designed since 1990.

Or any aircraft at all, really, regardless of when it was made. It all depends on the pilot's understanding of the aircraft's systems.

In a practical sense, a 172 is much more "stallable" when piloted by someone who for some reason thinks you can decrease AoA by pulling back on the yoke than when piloted by someone who thinks the opposite.

Yes, strictly speaking the "stallability" of the aircraft itself has not changed, but the likelihood of a stall occurring when the plane+pilot system is in operation is greatly increased.

[Gabriel]

One of the principal reasons you can't build an unstallable plane [.......]

So many things to say...

I'm sure you will believe me that my answer can (and normally would) be much longer than your post. But I don't feel like doing right now. Perhaps later, perhaps in another thread. But I will leave 2 comments:

- I don't agree with a LOT of things that you say, both on "how things are" and even more on "how things can or can't possibly be".

- "And you get a plane that is as unstallable as it gets" is different for "unstallable".

[3WE]

No, planes cannot really be made "stall proof" by limiting the elevator authority. It might meet FAA's stall-resistant criteria, but that doesn't mean that it can't be stalled "at all sorts of speeds and attitudes and flight modes".

Take for example the Ercoupe. The plane was designed with limited elevator authority to avoid a stall and very limited rudder authority to avoid spin (that alone tells you that stall was not impossible, since no stall = no spin).

Now, take the Ercoupe, firewall the throttle, nose over into a shallow dive, go to the red line, pull up to the load factor limit (or the end of the elevator authority) and the plane will not stall, but it will pitch way up. Go vertical and keep it there until you are completely out of speed. The only way out of that involves a stall...

I think we are having a slight language barrier and bickering over imprecise, gray-area wording...

UNLESS you do what you describe with the Ercoupe, it's going to be difficult to stall: Normal cruise, STEEP turns, turning from downwind to base where you pick up a tail wind, a LONG approach where you forget to monitor the airspeed on short final, pulling up relentlessly...

...for a very wide range of conditions, you do not have the control authority to simply haul back and stall the plane and you probably cant. You have to dive it to the redline and then pull up...a somewhat special set of circumstances.

...or you have to drop it from a balloon, or zoom off of a cliff at 20 MPH, or drop a heavy weight from the cockpit that's tied onto the tail to make the CG go aft, or a crazy wind shear, and then it will stall. But again, these are somewhat special circumstances.

Indeed a plane can be stalled at almost any airpseed and attitude- although sometimes it requires some weird stuff to happen.

But any pilot cannot stall any airplane at any altitude or attitude...The Ercoupe and it's limited elevator authority was the perfect example!

[elaw]

Or in different words... any plane can be stalled under almost any condition by taking action that is intentional and falls outside the limits of what most people would consider "normal flight" (excluding aerobatics).

But stalls resulting from mental lapses or poor technique during normal flight operations are much less likely in some planes than in others. If you get a little slow on approach in an Ercoupe, the plane's behavior will be much more benign than if you do the same in an Extra 300.

[Gabriel]

I think we are having a slight language barrier and bickering over imprecise, gray-area wording...

UNLESS you do what you describe with the Ercoupe, it's going to be difficult to stall: Normal cruise, STEEP turns, turning from downwind to base where you pick up a tail wind, a LONG approach where you forget to monitor the airspeed on short final, pulling up relentlessly...

Yep, I agreee. It requires something really "special" to stall an Ercoupe.

What I meant is that you cannot make an airplane "100% unstallable no matter what" (i.e. no grey areas left) just by restricting the elevator authority (and, I dare to say, by adding any number of additional features).

An A320 is closer to "unstallable" that the Ercoupe (now we are back in the grey area) because it will not even let you do the extreme zoom climb in the first place (well, most of the times while in direct law anyway).

[Gabriel]

Or in different words... any plane can be stalled under almost any condition by taking action that is intentional and falls outside the limits of what most people would consider "normal flight" (excluding aerobatics).

But stalls resulting from mental lapses or poor technique during normal flight operations are much less likely in some planes than in others. If you get a little slow on approach in an Ercoupe, the plane's behavior will be much more benign than if you do the same in an Extra 300.

Yes, but try to escape from a CFIT with the CG towards the forward limit and you might be missing that elevator authority that was removed to prevent that you accidentally stall when the CG is towards the aft limit.

There is no free lunch (until you put a computer in the middle, and even then, when TSHTF...)

[Evan]

So many things to say...

I'm sure you will believe me that my answer can (and normally would) be much longer than your post. But I don't feel like doing right now. Perhaps later, perhaps in another thread.

Yes, make a new thread. I'd love to hear what you have to say to that.

And 3WE can we PLEASE leave SR-71's and Ercoupes out of the AirAisa thread. When I say 'any plane' I am of course saying this in the context of commercial aviation. There is no large passenger aircraft that cannot be stalled under the right circumstances, even the A320 in normal law.

[3WE]

And 3WE can we PLEASE leave SR-71's and Ercoupes out of the AirAisa thread. When I say 'any plane'...

I did not bring Ercoupes nor SR-71's to this discussion.

And "any pilot can stall any plane at any altitude" remains as another inaccurate statement.

Finally, FWIW and until I hear otherwise, I am NOT going to believe that yet another set of highly-trained airline pilots intentionally but obliviously made strong and sustained control inputs that are consistent with the memory checklist for a great way to stall a plane...

...why would they violate something so basic?...

...I blame something mechanical, weather or computer related...

Yes, I've been wrong about this before.