Lyman . . .
 

Huh...?
Voice generated stall warning, "STALL, STALL, STALL..." intermittent up to 53 seconds.
My ears are not pots to piss in. I WOULD hear that, and I WOULD respect a stall warning, and I WOULD respond to it! To be sure, I WOULDN'T be pulling back on the stick and climbing 3000 feet, no matter what the Indicated Air Speed says. ...But maybe that's because I'm a captain with 20,000 hours.

Huh...? They didn't hear the stall warning? Didn't know that it MEANS stall?

These were 2 copilots who couldn't fly manual in IMC. :{

Why ? Why did he climb ? His first action should be level off. PNF doesn't know either - his comments seem to imply PF is not aware he is climbing.

And what could go wrong if you climb inexorably - Airbus is the first a/c without a service ceiling ?

He doesn't know his a/c flight modes / laws then.

Nose is already up - so you suggest he isn't looking at pitch either ?

Yep, because otherwise we'd have had a stall warning... oh, we did, for 50odd secs, but that's ok that's gone now because we've recovered from that stall, without ever putting the nose down...

But it doesn't - or at least not as I understand it - in particular if you get into alpha-prot mode, you have to act to get out of it. Leave it there and I think you'll end up in a phugoid around your max alt.

Maybe they also had no training (at all says the report), and no practice, in hand-flying at altitude. Maybe they have since had 447 scenario thrown at them in sim, before such training, and it didn't go so well... (or they have had the training and then the excercise and are thinking how they might have done untrained).

If every pilot had to be trained for every possible situation he could get into to be a qualified pilot he would be in his 40's to get his first job on his own. Pilots are usually people that learn the basics and understand the results of altitude, type of airplane and how thousands of situations can confront you.

A qualified pilot knows how to deal with never before encountered problems. We all did but some people here believe these FO's were not at fault because they were not taught high altitude full stalls in an airplane advertised as stall proof. Well they got into alternate law and couldn't deal with it because it did stall and they didn't know how to recover. Who is at fault for that? I don't blame AF or AB, I blame the pilots. If AF hired them knowing they were not able to handle a stall at all altitudes then maybe they get some blame too.

If these two were allowed to fly by themselves without the captain in the cockpit they should have been more qualified in my opinion.

I think (hope) there will be signficant work on that in the HF report. BEA has already expressed opinion and recommendation (see 4.1 of latest interim report)AF have already made some changes - 5.1.2 in the report refers.

GlueBall. Not like you to be coy. He had a STALL WARN at the drop, and the STALL WARN with STALL was concurrent with NO CUES. He'd been briefed STALL could be bogus with UAS, so let's dismount the high horse?

Immediately with the elevated noise in the cockpit, and what do YOU do?

You and others keep repeating data that is not untrue, but needs to be considered with other data. And you do it to frame the Pilots as stupid, newby, ne'er do wells. ENOUGH.

His screen was not recorded, remember? What did he see? What do you think he saw sufficient to condemn him? Was there no data to be recorded? OR was there data that was corrupted, and after boluxing up the deal, did not make it to the DFDR?

You are stuck on "54 SECONDS OF STALL WARN...bla..bla..bla.." Without completeing the picture with what I just posted.

UAS when this deal happened was a crapshoot. 31 crews landed safely, this one went in the drink.

31:1. You like those odds? I cannot believe an underwriter would indemnify anyone with the state of UAS and the sloppy workarounds given the crews forced to fly with it.

After 447, the AF pilots went on strike to cause the Line to get serious about r/r Pitots. Airbus was later forced to issue bulletins, one of which said: DO NOI RESELECT AutoPILOT, the a/c may climb without command.

Come to think of it, 447 did just that.

This: Your A/P disconnects. What do you do? the a/c is trending down, and rolling right. You don't know UAS is what caused the loss, it may be MET. You do nothing? Or do you risk losing something against which to frame PITCH and POWER, NOT KNOWING that is indicated?

ENOUGH.

Infrequent flyer, I've posted this too many times, you know what I think,. and rehashing it solves nothing.

I believe, Heartr and Soul, that taken with some objectivity, there is reasonable explanation for the mistakes that were made.

Except: With UAS, you get AL2. ROLL is direct, PITCH is not, it is commanded through a filter, and Elevators are subject to AUTOTRIM.

To me, what's left of NORMAL LAW amounts to a half assed autopilot, not FBW, so why would you want partial autopilot sharing command of your controls?

seriously

I am not a pilot but I have read this whole thread and would like to get this clarified. The captain was saying 'No, no, no, don't climb'. Doesn't that indicate that he is aware of the stall? And after the computer voice is saying 'pull up, pull up', the captain replies 'Go on: pull'. Is he saying that because he realizes that it is impossible to recover from a stall at below 4000 ft?

Also, why doesn't the captain resume control? Is he just standing there the whole 3.5 minutes?

No, he got there in the last minute and couldn't figure out what they had done. He needed to be there in the beginning to know what was happening.
He is the only one I feel had nothing to do with the crash. He was taking his required rest. The two FO's got it into the stall situation it was in. It was too late for him to fix it.

Maybe I wasn't clear enough; because that doesn't answer my question: Doesn't the fact that he told the pilot not to climb, point to that he was aware of them being in a stall?

Also the claim you make that he had to be there to know what was happening doesn't really jive what I have read in this thread. Shouldn't he have been able to tell they were in a stall judging from the loss of altitude? And why didn't he resume the control of the airplane? Why did he just stand there?

Correct, it is surprising he did not push to get back on his seat and have control on the sidestick.
Actually he came back 2 min 40 sec before the end.

AP remained disengaged, but could we know if they tried to re-engage one ?

What do you mean by "but could we know if they tried to re-engage one"?

I don't think any captain with less than one minute of observing what is happening out of sleep could figure out why they are descending 10,000 fpm with the nose up. We normally have competent copilots so don't think this can happen. I always trusted mine. Mine had lots of experience too. These two didn't. We didn't hire people with under 5,000 hrs. It makes a big difference when you hire unqualified people.

As the crossbars were visible by intermitence, maybe the crew may have been tempted to re-engage one AP. Would the FDR register an action on either AP pushbutton ?

That's a misconception. The protections were "advertised" as being able to prevent stall, this is true - but the "FBW Airbus can't stall" myth came out of press misunderstandings and public misinterpretations. Other modern myths include things like "Apple computers don't crash and can't get viruses" and "ABS in modern cars means braking distances are shorter".

The common thread running through all these myths is that they contain a grain of truth, but have been transmogrified through retelling until they actually become untrue in themselves. Apple computers do crash (and how - believe me!), and in some situations ABS can extend overall stopping distance. People who work with these machines are supposed to know this, but they can't stop the myths from gaining traction.

In the case of the Airbus myth, it is the Normal Law protections that make it almost impossible to stall the thing as long as all systems are functional, but with the failure or disengagement of systems that those protections rely on the aircraft becomes no more difficult to stall than any other airliner in the sky. If the PF was not trained to know that, then it is a massive dereliction of duty on the part of those who trained him. If the PF was trained to know that and in a moment of panic - or under pressure - forgot, then it was a tragic mistake - and one that many more experienced pilots, including senior and decorated Captains, have also made in the past.

He left the cockpit 9 minutes earlier (if I recall correctly). Do you think he had fallen asleep by then?

And what other cause can there be descending 10,000 fpm with the nose up other than a stall?

My neighbor flies All Airbuses and claims alpha protection would make the Hudson River Sully landing easy. Now you are saying that is not true????

The Hudson landing was accomplished with Normal Law active due to the decisive action of the crew in getting the APU going. Had they not done so and relied instead on the RAT, then the ditching would have been in Direct Law, the protections would not have been active, and it would likely have eroded the safety margin somewhat.

That said, I'd take issue with use of the term "easy" regardless of the context - no water ditching could be considered that based on the things I've read and learned over the years.

Indeed, what do we know of PF's screens? If a/p was tried, (it had been tried, with uncommanded climb as a result before), and he was confused by his FD, well, who knows? For many reasons, including comment on the second STALL ("pas valide?"), no conclusions re: the right side actions are legit, at this point.

Hi,

Lyman
May I ask .. what do we know of the NPF screens ?
Recorded ,remember ?
What did he see ? what do you think he saw was sufficient for not bother to take the command of the plane .. and let the PF to condemn him and al ?

bubbers44
The captain was not out of sleep .. this is a urban legend
Please check the time chronology .. BEA report N°3 and make some maths to know time gap of captain leave flight deck and captain enter again flight deck
How he can be already sleeping in this amount of time ?
I know that during the two last WW there was soldiers who marched asleep and falling on the ground like a ripe apple ..
If the captain was so tired .. I ask questions about his day of rest at Rio :eek:

Ok, even though my neighbor gave the Hudson river ditching credit to the autopilot I didn't agree with him because Sully had to manage his speed to stop his descent rate prior to contact with the water. Landing wings level is quite easy so I give all the credit to Sully for making the best out of a bad situation. I always flew Boeing aircraft by choice so know very little about what it is like to fly the AB. I retired before I had to be concerned with it.

That's a shame bubba, to have finished your career without experiencing the beauty of flying a FBW Airbus. A truly remarkable airplane.

bubbers
 

Agreed...whilst I agree his handover prior to leaving the flight deck perhaps wasn't the best I think it's a tough ask to expect him to walk onto the flight deck having perhaps been asleep seconds earlier and immediately assimilate all that had gone before (and yes, it is certainly possible to be asleep within 9 minutes of leaving the Flight Deck; been there, seen it, done.... ZZZZZZ).

Glueball,

Interestingly, at a time of very high stress, you probably wouldn't.

Anyway, the captains role should not weigh in this accident. He was on his break and it is considered legal to leave the two FOs alone. Period.

At a big ME airline, they stow the crew bunks in the furthermost corner of the aircraft, between zillions of passengers and trolleys. Captains would never ever make it to up front again in a panic situation, and it's considered legal by all authorities.

If anyone thinks, the captain could have easily shifted his break, then consider:

- There is not only the very one zone of increased danger on such flights, maybe the other one is 3 hours down the road.
- Each crew member has to have 3 hours of uninterrupted break to have it legally considered.
- Generally the first and the last hour is when the captain is in his seat, legally I think it is only above 20000ft when FOs may occupy the left seat.

Now fit this into the AF flight and you will see, that he was almost constrained to be on his brake at the time.

To me this accident shows that a crew composition of 3 should contain 2 captains. Not only or necessarily because of more experience, but moreover because of the lack of leadership training and inhibit of take-over capabilities. This was very much displayed by the PNF-FO.

blubber, off topic, the Hudson rider was not in normal law because of RAT but because both engines still delivered residual thrust and all hydraulics kept going. I agree that it is much easier to settle down an Airbus in normal law than a conventional aircraft (you just pull - aircraft does the rest, that's where the myth is coming from).

I think even in these days when training emphasises CRM we are still bound by training scenarios from the piston engine era. All type rated pilots today are proficient in dealing with an engine fire/failure at V1. The chance of actually encountering this with a modern jet aircraft is very small. Apart from seeing a chart labelled 'flight with unreliable airspeed'. I have never been exposed to any simulator training similar to the AF scenario. Should there be a major re-think of initial and re-current type rating requirements?

Slightly off-topic, I know, but:
"beauty" - I don't think so. Clever it may be, especially the general ergonomics, comfort & system design, but as far as getting a kick out of flying them; well, it's rather like eating a toffee with the wrapper still on.

Hi,

So .. no problems .. all were killed .. but with respect of the law..
Dura lex sed lex

OK, now either this is getting lost in translation or I'm seriously concerned. If this neighbour of yours flies the FBW Airbus and is not aware that the autopilot/FMC and the FBW architecture/FCU are completely separate entities, then I'd suggest remedial training!

So - last time, because this is offtopic. Hudson landing was FMC/AP off (i.e manual control), Normal Law (i.e hard protections active).

I dunno, looks like it can be fun to me!



[EDIT : The video is there as a bit of light-hearted fun, and as a demonstration of two of the Normal Law protections - it is certainly not meant to be considered a serious part of the discussion! ]

Well, I dunno, but it looks like a piece of juck to me !! More like an arcade game, NOT something to EVER leave the ground. Guess I'm just too old.

Hi,

Notice:
Mostly time ..the right seat pilot keep his hand on the thrust levers :8

Rather narrow in scope, this video shows the potential problems of a philosophy that has as its goal to prevent pilots from exceeding the a/c limits.

There is nothing fundamentally wrong with such an approach, but the devil is in the intent.

Intention drives every human endeavour, and goals have brackets.

First of all, it presumes to fill a void in aviation that wants expansion. What are the statistics of Loss of Control accidents and parenthetically, CFIT. The answer is there are none that cannot be massaged.......

A premier example of CFIT is the introduction flight of the very aircraft that "introduces" a solution to CFIT. Habsheim.

The pilot exceeded the limits of the a/c, and it crashed. Embarrassing.

The upshot is there are still professional and wellfounded discussions surrounding the philosophy, and "statistics", of course, are utilized parochially to push one or the other agenda.

Clearly, AF 447 exemplifies problems this philosophy has yet to wring out of its expression in commercial flight. Criticism comes from her own pilot crews, and those of the main competitor.

There are first hand criticisms of real time events in the record.

"Je ne comprends riens". "What is wrong?" "What do we do"?

"Disregard the STALL". "Do NOT disregard the STALL". "Do NOT re-select the autopilot". "The aircraft may climb unexpectedly"

These are incendiaries, to be sure. The important part of the conversation is the conclusion that there is no conclusion. This is troubling at every level.

AF 447 is too full of failures to isolate the philosophy of the design as cause. Did it contribute? That is the question.

And that there is such a question remaining is troubling. Doubt is no friend of safe travel.

Quoting the "Habsheim Clown" in this context is enough to get you discredited wih the few remaining people here that try to come up with valid (and occasionally valuable) comments. Keep it up.

From the video (and many earlier posts) Seems like the design enable mediocre pilots the ability to fly with ease if there are not too many demands on them. Let's visit the bank limiter. Apparently it is there so a person with little ability/situational awareness will not exceed what the software engineer figured out what the limits shall be. Can anyone imagine why you would ever need to exceed them? What about to avoid a mid air? Not that it would require a 90 degree bank but it is nice to know it is there if needed, but wait, you have to be able to actually fly in that regime in order for it to be effective. What? No prior experience/training in those 1000 hours of experience?
Next lets discuss the system that prevents a stall. Why is it there? Maybe so inexperienced pilots can cope with most situations. We need an aircraft for the masses, check this out....ERCO Ercoupe - Wikipedia, the free encyclopedia

I am not actually belittling the airplane or design. All if I am saying that is if we keep designing airplane so foolproof we need to stop producing fools that fly them with little or no practical experience.

Deep Stall/ Stall Recovery
 

Any of you on the forum with test flight experience, particularly with fast jets may well wish to consider why large aircraft cannot be equipped with stall recovery parachute systems.

Or, you know, giving good pilots the knowledge that, more than 99% of the time, there's something covering their back if they make a mistake, or knowing that they can push the limits of the airframe safely if they need to make a positive evasive maneouvre.

The text in bold makes me sad. How many times can I say that the limits were set by the aerodynamicists and aero engineers - and discussed with the engineering pilots (in consultation with line pilots) - before the software engineers got anywhere near it until it sinks in?

A 90 degree bank would endanger the structural integrity of the airframe, especially if pulling G's in the pitch axis. The protections in the A320 allow the pilot to order a full right bank up to 67 degrees and will keep the G loading below 2.5 if pulling, as specified by the aero engineers who designed the thing - if you need more than that for an escape maneouvre then you shouldn't be on the flight deck in the first place.


That's the point, the protections allow you to do it safely - whether you're the best pilot in the world and know the precise airframe limits or not - because pilots like Gordon Corps tested the thing to the aerodynamic limits and discovered what was safe, and included those limits in the design. In Normal Law you can pull full back and bank and the aircraft will give you the maximum possible response without fear of doing damage.

As I said before, even the best pilots have bad days, just ask KLM. The point is that if you exceed the safe AoA in Normal Law then the aircraft will have your back.

Cutting back on training was never the intent of any FBW airliner design no matter which side of the Atlantic they are built. That particular SNAFU belongs squarely at the door of airline management.

Before Landing Check List,

I doubt that Airbus attempted to design an aircraft for the lowest common denominator. As someone else stated, airline management has mistakenly decided that the protections/etc are a substitute for stick and rudder skills.

In my short six years flying narrowbody Airbii, I've realized that the airplane demands a masters degree level of understanding in order to deal with any situation other than normal. In A447's case, the pilots were faced with the need to properly evaluate the situation, to understand what changed in the 330's flying characteristics depending on the exact failure/control law downgrade, to hand fly an aircraft with different flight characteristics than they had likely EVER experienced, etc.

Airline management, especially in my US LCC experience, attempts to short cut training to save money. For example, my carrier only trained UAS after this accident. IOW, the manufacturer is not responsible for the cost cutting in training.

I'd still rather fly a Douglas.

Hurrah - he gets it!

Really? The fact that a planeload of people got out of the Hudson River cold, wet, but very much alive would tend to go against that suggestion, especially given that that A320 involved was not much more than a sophisticated glider for the few minutes it was in the air, and those few minutes were all the crew had to troubleshoot and land it.

I mean sure, ECAM output can look very dense and not especially human-friendly, but it's no different than Boeing or MD's equivalent, and you don't need to process pages of ECAM to keep the thing in the air - that's your colleague's lookout, and he or she will tell you what you need to know.

Well, seeing as they apparently hadn't manually handled *any* jet at high altitude, that's going to be a problem no matter what they were flying.

While it's very easy to say this from my little office/den, and I acknowledge that - the handling characteristics don't change a great deal between Normal and Alternate Law - pitch handling is practically identical, and roll is slightly more sensitive in the latter.

To the best of our knowledge, what was presented at the start of the sequence was this:

FMS/AP Disconnect - So you're going to have to handle her manually. Be prepared, but first of all do nothing with the controls for a few seconds to see what she's doing by herself. Any inputs you make must be slow and gradual.

Unreliable Airspeed indications - OK, so you don't know how fast you're going via the usual channels, and you're in moderate turbulence, so just keep her straight and level using small and gradual inputs if you have to. Use pitch and power to keep her stable.

Alternate Law 2 - So now you have to fly the aircraft manually with no hard protections, so be even more careful with your inputs. She wants to fly, so keep pitch and power steady (bring power up a little if you have to - if, say, you pulled the power back to transit turbulence earlier on).

If any of that seems more difficult than it would be in any other airliner, for what it's worth I can't see it.

As an SLF well-versed on some of what that company did in the '60s and '70s I can assure you I wouldn't!

Is it all not progress? When I did my A320 course in Toulouse in 1988 ( course 4 I think ) not many people had mobile phones. I used to fly to Lagos frequently before that and there was no telephone service that worked in Nigeria. All of a sudden mobile phones arrived there and the problem solved. Should the Nigerians have been taught to use the old system first or was it cheating to use mobiles? We have millions of motorists using GPS systems and a good majority couldn't read a map I would guess. It's the same with ships that still seem to run aground/sink etc. There have been many aircraft accidents recently like THY into Amsterdam where they actually unbelieveably managed to stall on the approach. Technology will continue to make a pilots' job easier and it can't be stopped. Does anyone really think that Airbus are going to give up the sidestick and non-moving thrust levers. It's progress. The horrible truth about AF447 is that they didn't avoid the thunderstorms and everything that happenned after was a result of this.
All the talk about heavy crew, captain going on his/her rest at the wrong time is irrelevent as the remaining 2 pilots in control should have been just as competent as the Captain. The fact that they were not would seem to point to Air France training issues. I guess what I am getting at is that technology is trying to make the pilots job easier and that is not likely to change because in theory it makes the whole flying operation safer and that's what the passengers and everyone else wants. Gone are the days of us old Jet Provost pilots, now it's all autopilots, flight management systems, TCAS, ACARS and the rest of it. I don't think pilotless airliners are anywhere in the near future (passengers wouldn't fly on them) but look at the success of the Drones against Al Quaida.
I flew the Air France route more times than I can remember and it was always dodgy around the ITCZ but we always deviated the cbs by as much as was necessary. The reactions of the AF pilots was amazing by initially not avoiding the weather and then allowing the most junior of the 3 pilots to a) fly into the weather and b) allow him to control the situation catastrophically whilst the other co-pilot appeared to let it happen without questioning it until it was too late. The poor old Captain appeared when it was far too late, hardly his fault and whether it was a bad decision to have taken his rest at that time I would dispute as his copilots should have been just as competent as him. That's the way it was in the airline I flew for. Anyways that's my view as one of the first A320 pilots and a great believer in the product and subsequent variants ie 340, 330, 380 and future products.

The aircraft was designed to protect itself against just such clowns. I agree, but he was not a clown, nor was he anyone's concierge. He was a line pilot, I know little about him, and of course it shows that any aircraft will foul up.

The knock on Airbus has always been that the aircraft is aloof, and can be mystifying, to even qualified pilots.

I'd have thought that the builder of the aircraft would have exercised better judgment in selecting a pilot to show off.

Look, the aircraft crashed, and it must have been embarrassing all around.

Irony, or maybe comeuppance?

Off to a bad start, then.

Great hoorah here about "Interface". What a disastrous word. It is generally applied to more sophisticated models than pilot/plane.

It is a machine. It is not complex, and it is not someone's poodle. The part about flying this machine with some special deference is pathetic. No quarrel with the platform.

I admire the spunk it must have taken to continue with the idea. Most designers and builders after such a kerfuffle would get into building tractors, or summat.

Habsheim is a mythic, Chris, an epic, even. Beowulf, or the Odyssey. Acceptance of its majesty brings one peace.

Wait. Achilles, then.

You've Got It Right!
 

(My bolding above.)

Absolutely correct! But there is a bit more. Someone, maybe Dozy, asked me earlier what an experienced pilot would do when confronted with the AF447 situation. My response was essentially Dozy's answer above. But, it seems to me that the big thing missing in training of pilots flying highly automated aircraft is the absolute need for the crew to constantly be monitoring just what the autopilot and the flight control system are doing. You need to be fully aware of the situation every second. You might be a bit relaxed in daylight and VMC, but at night in foul weather the PF needs to have his eyes glued to the attitude indicator with occasional glances at other instruments such as altimeter, airspeed indicator, and engine instruments. If he does that, he will instantly know the situation when something goes amiss and the autopilot and/or autothrottle click off. As long as attitude doesn't change too drastically, airspeed won't either. Then he simply needs to ride it out while discussing things with the other pilot. If power was set properly to begin with, the natural stability of the aircraft will keep everything else in line. As soon as he changes anything, he upsets the apple cart. I think complacency has caused pilots to assume that the autopilot and flight control system will correct for all problems. As we can see from this most regrettable incident, they won't.