No nothing wrong with it.
You need to distinguish between the ADR disagree due to the icing which was the event that lastet less than a minute and initiated this accident and the invalid airspeed due to the high AoA which occcured much later and was a consequence of it..

Airbus FCTM text.
 

Gentlemen, ladies,

Although the airplane was apparently never in an extreme attitude, the pilot somehow did not recognize the stall condition, or the position of the horiz. stabiliser. Could it be overreliance on Airbus' Flt Ctl Systems and mentally 'ruling out' basic indications?

In any case, the following statement, especially its last paragraph, in the Airbus Flight Crew Training Manual (OP-020, Abnormal Attitudes) reflects what I believe is a fundamental problem in the attitude of Airbus, and their design philosophy, and could be a contributing factor to the fate of AF 447.

"It is very unlikely that the aircraft will reach these attitudes, because fly-by-wire provides protection to ensure rapid reaction far in advance. This will minimize the effect and potential for such aerodynamic
upsets.
The effectiveness of fly-by-wire architecture, and the existence of control laws, eliminate the need for upset recovery maneuvers to be trained on protected Airbus aircraft.

Why did they not rely on the ground speed to regulate their thrust ?

sustaining 400kts ground speed @FL300 or 350kt ground speed @FL200 could have save them.

Why in case of IAS malfunction they do not use the ground speed (GPS AND inertial both provided it) to evaluate their air speed ?

peplum: The pitot-static system provides Air Data, relative to the medium which the aircraft flies in, unlike the other sources you listed. An analogy; think of a boat on a river - if the river is moving at 30kts and the boat is going against the flow at 30kts, then it's GPS and INS might show 0kts... but the boat would see a flow of 30kts at it's hull.

I've finally caught up on the day's messages, and I'm intrigued and saddened that with such safety-critical data sources (Air Data sensors), there is still little redundancy in design offered. I.e. Are there 3 different air data sensors designed by 3 different companies, based on 3 different sampling principles, and then 3 of each onboard?! Pilots, and computers/software, can only compute and respond effectively (without error) if their senses / sensors can guarantee dependable data.

This could well be a systemic disease inflicting mankind lately, take for example the design of Fukushima and other BWR's around the globe... listen to what ArnieG has to say about that...
The Implications of the Fukushima Accident on the World's Operating Reactors | Fairewinds Associates, Inc

We're not on a river but a jet in the atmosphere.

The wind could be evaluated at max +-75kts in this cas. Even if you add or remove 75kts you're always in the flight domain of the plane using my values !

Anyway think one moment. To fly a plane to must evalute the air speed. Because they were high altitude, and the 3 pitot was off, the only indicator effective at this moment was ground speed.
unless you prefer to open the window and reach out to assess !
I'm not Einstein but I hardly imagine none of the 3 pilots try to use this evident method to adjust thurst and sustain a correct FL.

As JCjeant said it, >300 posts over speculation, human-vs-computer battle.

But this accident is like any-other a combination of several factors and their consequences, I'll try to summarize those :

1°/ Pitot failures
- No more valid speed indication
- A/P A/T disconnection (corrective actions taken by crew : "I have controls")
-Switching to alternalte law (Identified by PM "Alternate law")

2°/ Climb to FL380
-The report says that after the initial right bank the pilot performed a corrective action with a left and up S/S deflection.
-Pitch gradually increase over 10°UP. Vertical speed goes to 7000fpm (corrective action : Pitch down)
-Bank values oscillating between 12°right and 10°left. (Corrected by the pilot, but maybe a hint of the flight enveloppe limitation and turbulences)

3°/ First stall warning and pilot corrections
-PF sets TO/GA, and Up-stick force (Seemingly an approach to stall procedure on a non-FBW aircraft : power and initially a small up-stick force as a reflex to maintain constant pitch)
-Consequence of the up-stick force on the A330:
-The pilot fly a G-acceleration, thus constant upward acceleration is commanded as long as the stick is in the 'up' sector
-Auto-trim sets THS up to 13°

4°/ Stall At one point between 02h10'51 and 02h11'40
AoA rises to 16° which is the basic stall AoA at low altitude, at FL380 stall will surely occur at a lower AoA.

5°/ Entering "deep-stall" Between 02h11'40 and 02h12'02 ---> Lasting 22sec only. At a maximum altitude of FL380 the A330 is now completely stalled and descending.
-AoA reaches 40°(Computers switches to 'Abnormal attitude law' (AoA>30°) and auto-trim is disable (Trim wheel is now stuck at 13° up even if it could have been moved by the PF, but with no force-feedback on the stick how to use the trim without QRH/FCOM references?)

6°/ Stall warning 'bugs' begin
-The AoA is so high that Pitot airflow might stall, showing speed below 60kts
-Stall Horn STOPS (IMHO one of the key info that kept the crew in a wrong situation awareness)

7°/ More corrective actions from PF at ~02h12'17
-Pitch down input (correct action to get out of the stall)
-As airfow gets back in the pitots 'speed became valid' and Stall Horn rings again.
(IMHO this is the other key info given to the pilots: the PF brain might have taken it as a logical computer :
Alarm rings ---> Erroneous Action ---> What was my last action? ---> Pitch down input <---CONCLUSION---> Pitch-down input = Error)



As the aircraft is still in a stall with a more than 10.000ft/min rate of descent, the BEA did not communicate on the period between FL350 and FL100 but a major brainstorm/communication or even actions might have taken place in the flight deck.



8°/Approaching FL100
-Less than one minute before impact
-The other pilot takes over (or maybe the captain from behind)
-Only info is 'AoA remained over 35°'
(We do not have any info on the last minute inputs, possibly a desperate maneuver)


I hope this will help picture out the overall situation, so people don't only focus only on one computer or one pilot input.

False. IAS is OFF but GS is OK.
Air speed can be sufficiently evaluated from ground speed to adjust thrust and do not stall.

Using computers to help high workload
 

Sure you can create models to handle any situation, but the risks are naturally increasing with each missing parameter. It's no fuzzy logic, although that would be a good concept to start with :8

I read many inputs on random newspaper forums, asking why cannot GPS-data be used calculate speed if pitots freeze... Beyond the obvious of not understanding the difference between air speed and ground speed, I came to think about something:

Would it be out of question to create an automation graceful degradation, which:

a) would use a pattern of ground speed vs IAS to judge faulty airspeed readings. Give an advisory to pilots before actually making the call to consider it faulty (less voting of three pitot readings).

b) in case of actually loosing reliable IAS and forced switching to alternate law or what ever mode, would use GPS for displaying pilots an educated guess of airspeed, and would use GPS for calculating altitude and a safe attitude and power setting, which would keep the plane in the air in those conditions. The pilot would then have the option of flying the plane manually or rely on an rude "autopilot" using only pre-calculated values to keep the plane in a safe as possible envelope with data available.

c) use GPS to judge whether the plane is stalling or IAS too far out to be considered reliable (re: no stall warning below 60kts).

Will GS appear on the PFD?
Then they will have to check on the ND or FMS, okay it is just a few inches away but then the aircraft was on full auto and released to full manual the workload must have significantly increase in the first seconds, and one might consider the 7000ft up vertical speed to illustrate the amount of workload and a related lack of visual scanning

You're right. I'm very surprised to not read in the official media the air speed evaluation from the ground speed. I'm quite sure that it's too complicated to recognize the pilots were too clueless and accustomed to use their computers and to consider this simple tips that could save their lives.

On A320 it does. GS and TAS.

They were 3 pilots. None of them had this simple idea to read the ground speed. Damn, the wind would be at max 75kts (may be more we don't really care).

Open the window and evalute the wind !
No, just read the ground speed.

You're at FL300, sustain as much you can 400 kts GS
You're closer of FL200, aim 350 kts GS.

Automatic pilots makes intelligent people too stupid when deconnected.

And please indicate air incidence should be considered main indicators.

Genghis,
"these chaps were experienced and current pilots unlikely to make such an elementary error. "

Only one of them made the mistake, the other two did not know what he was doing. PF used up 27'000 feet and PNF then had a go with less than 10'000 feet left. You are there to monitor the other guy but cannot see what he is doing, so you assume he is getting it right, at least to start with? In my book that gives you a single point of failure and that is a bit naughty.

I grossly over simplified, but sometimes 'simple' provides us the best analogy - fluid mechanics aside. As the last resort option (beyond having desirable sensor redundancy etc), GPS or INS could have been the only, albeit inaccurate substitute for air data available during the reported pitot icing. Can a 'bus pilot tell us what the safe operating airspeed margin 447 might have had at FL370?

Temporary pitot failure may have been a factor, but it's not a good enough reason in itself for an otherwise fully intact a/c to fall from the sky. They recognized UAS, there's a procedure for it, why didn't they apply it? Or at least common sense? The plane was flying well before the pitots failed, it would very likely have continued to do so if they had maintained pitch and thrust. Instead for reasons yet unclear they performed a manoeuvre leading into a stall. Then they either failed to interpret the situation or didn't / couldn't apply appropriate measures to recover from it. Again for reasons unknown yet.
There's still far too much open questions and to most of them 'pitots' doesn't seem to be a good answer.

What are you looking for, Vs @FL370 ?

chaz88z

Your 2nd item omits to say that there were two stall warnings after the PF corrected the bank to the right. Subsequently the aircraft climbed at a high rate. The high climb rate must mean that the stall warnings were not valid; could they be triggered by invalid airspeed rather than excess angle of attack?

Eeek! - What is this ridiculous fascination with minimising height loss everyone is quoting (modern pilots) as having during cruise ... ???

At 35,000 goddam feet... really... who is dreaming up these latest training schemes ?

Peplum and others:

You simply cannot use ground speed.

The wind could easily be 150kts or more in any direction ( I don't kow where you get the idea that 75 is a maximum Peplum, it just isn't). So, if you maintain 400lts GS as you suggest you might be doing 250 or 550 or anything in between. i.e. you have no idea. At near to service ceiling the safe speed range can be as little as 30 kts wide and yet you think it is fine to use a method that has a 300 kt (or more range of error). I can only assume you are not professional pilots to have made such a basic error.

I've read the stall alert is invalid when AoA is too high bevcause AIR speed measure cannot be intrepated as valid.

Why did they sustain a such AoA ? Do the plane displays (sufficuently) it on instruments ?

"...on protected Airbus aircraft"

Once in alternate law, the aircraft was not protected anymore and the reaction to a stall warning is like in any other airplane.
Ofcourse a correct pilot action is needed.

So calling this a fundamental problem in the attitude of Airbus and their design philosophy seems incorrect.
Alternate law and stall recovery should be included in the training of the pilots.

Where do you take the 30kts from?
WRT to ground speed the margin is roughly 100kts GS. (~M0,68 - ~M0,86)
The mistake you made is the same one sees over and over again. the 30kts range is IAS. For inertia of the airframe, it's GS that counts. IAS is aerodynamics, Gs is mechanics.
You can't have it both ways, sorry.

GS 400 kts would have been btw much better in any case than GS 110kts (that's what they had roughly). GS should have given them a clue how the thing was plowing through the air. Even if not exact enough for perfect flight, it might have saved the day.

I do not pretend to be a pro. I'm not. That's not a pilot question. In life, whatever the job you have. When one parameter is vital for you and you don't have it anymore you have to evaluate it.

My thinking is when you do not have anymore reliable IAS, how do you evalute it ? Which clue do you use to evalute it. If not GS, which other do you propose ?

What did source of information they can use to evalute IAS : Ground speed and that's all.

Of course the range of possible speed is narrow at FL350. So, why to pitch-up ?
One wind can be fastet then 75kts, so descend... or praise for it to be less than 75kts but you have to set-up a strategy to replace the automatic pilot and pitot/mach meter did.

Why did they expect climbing to FL380 ? To my knowledge atlimeter was OK. So they should know what they did climbing so high.

Other pilots have reported to live comparable IAS disfunction on airbus and do escape.

Sorry but I do not understand why they do no exploit G.S.

No, in this situation pitch attitude and power is the primary reference.

I wonder if this will be the deciding factor for Airbus to implement back-driven sidesticks?

Edit: added "power" to statement above; thought it was obvious.

I would say both. pitching 0° at 0 kts means nothing.

Henra

At the limit of the envelope the margin can be as little as 30kts TAS from Vls to Mmo. MAch and GS have nothing whatsoever to do with it. The aircraft stalled. This is an aerodynamic event. How have you calculated your 'safe' mach numbers because they bear no relation to reality mach .68 could well be below minimum at high level and high weight.

I fail to see what the aircrafts inertia has to do with it. It may have a slight bearing on rate of stall recovery but the important info for stall recovery are aerodynamic. IAS, TAS AoA.

I say agian GS has nothing to do iwth it and only an idiot would use it to handle the aircraft when he has attitude and thrust available to him.

GPS Useless: Fly Attitude
 

The aerodynamics of the plane is a perfectly good airspeed indicator.

For a given weight and cg, the pitch angle (nose up/down), as shown on the Attitude Indicator, will indicate airspeed. Pitch down, and the plane speeds up; pitch up and the plane slows down. It's there staring at the pilot throughout every flight.

Oh, that happens to be the basic of the Unreliable Airspeed Indicator training.

Of course, some 35 years ago, the FAA began preaching that pitch controls altitude and power controls speed, and too many bought into it, but that's for another thread..

It counts as soon as you hit the ground. But to keep a plane in the air airspeed is all that matters.

No, GS is pretty useless. What would have worked is pitch and thrust.

The reference was at FL350 @ ~200t and for the Mach assuming still air to show how wide the difference is regarding TAS.

at FL380 it was probably indeed down to 30kts but going there was not a good idea in the first place. The idea of peplum to go to a lower altitude where the margin is wider and to go to a reasonable GS is better than what happened here, even though not perfect. It is not a good strategy to fly the plane by GS but it could be used to check plausibility of one's own assumptions about what is going on. GS can't be trumped and if GS says 110kts, you get a clear pitcure about what's wrong.


@Zorin_75:
No contradiction:
For flying it's IAS that counts.
Inertia (i.e. Decelleration/accceleration) happens based on TAS. (Rho/2 * V_TAS ^2).

Power + Attitude = Performance, on a C152 or A330.

ok pitch -10 Toga and you'll soon see if that all that counts :D

Well done Zorin. Its pitch and power that would save them.

Pitch alone tells you nothing by the way, Graybeard, since a moderate pitch with a high angle of descent=high AoA as in this case.

I hope the non pilots can now accept that GS without air data is just meaningless for aerodynamics, absolutely meaningless.

busTRE:
A moderate pitch with high angle of descent means not enough power to hold altitude. Use power as needed to control altitude.

This assumes a non-stalled wing, of course, which such events begin with.

Zorin_75: From an engineering perspective; invalid (faulty) air data from non- design-redundant sensors appear to be the weakest link imho, triggering a cascade of errors thereafter.

Band-aid, extrapolated solutions just don't fly with me. If someone is going to give a pilot or a computer Air Data, then sure as aunt betty make it dependable, fail-safe Air Data. It's really just that simple. Alternatively, a crew of two pilots, and MacGyver might suffice.

I suggested this on the initial AF447 thread, and didn't get much response, but I don't see why it would not work:

As long as you have a valid IAS, you systematically compare it to GS using GPS and compute wind speed and direction from that. You store this value. If the IAS fails, you use this stored value to correct the GPS-derived GS to get an estimated AS.

Except in the case of severe wind shear, why would this not work?

You could present the estimated AS as a range, with some degree of uncertainty, incorporating previous variations in wind speed, if any, or just an uncertainty factor (perhaps growing over time) to account for possible changes in wind speed.

Whilst i can see the argument for pitch & power as overriding targets, as a non pilot, surely there is more to it than that? For example, during the previous hr or so of the flight (at FL350) there must have been an "average" values for pitch, power and GS, that have been clearly demonstrated to "fly" the airframe in a sensible trim.
So surely, in the case of upset, these values would have been a nice place to start trying to aim for, no matter what your AS instruments are saying to you?? Clearly, under no conditions would a groundspeed of 110kts be acceptable (unless you are flying into the MOTHER of all headwinds, which the rest of the flight would have demonstrated to be false??)


The big question, is what parameters and causal factors lead the 3 pilots to incorrectly determine the necessary actions to recover?

Like all things in this world, everything is a compromise, i'm sure we could build an "uncrashable" aircraft, but no one would be able to afford to buy or fly it!

The most important thing, is that if clear and accurate action/response recomendations are forthcoming, that these are precisely implemented.

Unfortunately, there are no absolute answers in risk management, it's a "make the best job you can with the resources availible" problem.

spagiola: I think that would work (in a brief sensor fault condition), but estimated values degrade with elapsed time and environmental transcients as you say - wouldn't it just be better to be able to depend on reliable air data?

THS Problem on a Falcon 7X
 

Another leak (not verified): http://jacno.com/prov/images/extrait-fdr-cvr.png

See posts #83 and #84 on this page (in french): Forums Aviation Civile &bull; Afficher le sujet - Airbus Rio-Paris : les dessous cachés des enquêtes


A THS related incident involving a Falcon 7X and the subsequent release of an EAD http://jacno.com/prov/EASA_EAD_2011-0102-E_1.pdf and the grounding of the world's Falcon 7X fleet.

Are both AF447 and this Falcon 7X sharing some software or hardware components ?

Comments ?
Regards.

Don't get me wrong, I'm all for reliable sensor data, of course.
Still, such a failure shouldn't (and doesn't need to) lead to losing control of an aircraft. Any accident is a chain of events with often many contributing factors. So far, large pieces of the puzzle are still missing.

I am curious as to what caused the initial upset ( with a climb rate in excess of 7000 fpm) according to flights website. The statemet that the Captain briefed the crew about turbulence ahead is a little ambiguous. Did he mean immediately ahead ? or just a reminder that it was forecast on the flight ? That question is relevent ? Altough they did make an initial turn of 12 degrees just prior to the upset. I guess they were avoiding weather .

Also , could there be a link to this and the QF 330 incident where they unexpectantly had an upset ? Cant recall all the details of that incident , apart from unexpected , climb or desent .

davionics, #353

Absolutely. There's no engineering reason why it can't be fixed either.
They have had years to deal with it as well. I know pilots are a hairy
chested breed who don't like to make a fuss and think that they can
handle any situation, but my argument is that they should never be
expected to handle some of them. I think the regulators will need to
have much sharper teeth in future, both to overcome industry inertia
and airline beancounter interests.

Doesn't the A380 have a laser based or alternative air speed sensor
as a option ?...