Hi,

Academic blah blah :)
I have not to learn you the fact that the real security issues will point to some actors
And the court will point exactly the same actors with some nuances ... that's a lawyers game to be performed .. for the spectators we are :)
Rarely a court of justice will point responsibilities to actors not already involved in the security issues
It's granted that the pilots will take the most of the weight .. and so by consequence .. AF for many negligence
Yourself have already the verdict !
All this will be confirmed in some loooong years ... when all the smoke is settled :)

But being on an aviation forum I find it waaaay more relevant and interesting to find out what went wrong and why than assigning blame to any of the involved parties. There are other professions who are paid to to the latter and I'm more than happy to leave that part (which I find not very pleasing) to them.
Therefore count me in for the academic blah blah as well.

@ takata
So far this thread plus the previous four total 9677 posts and counting. Apparently, questioning, probing and trying to understand technical and engineering "stuff", as you put it, regarding the aircraft and its FBW controls seems to be not the thing to do in your mind. The conclusion seemingly has been reached by you, ergonomics and interface issues were the cause, end of discussion. It will be interesting to see if in the final BEA report if they are in agreement with your conclusion. Hopefully, they will have a more "in depth" examination that goes beyond your present "bandwidth."

autotrim, stall warning, stable stall, et al...
 

Hard to keep up, and I am a very fast reader.

- Zorin asked about the autotrim, and others have tried to figure it out. From what I see of the tech manuals and inputs from other contributors, it looks like a basic way of providing the elevator the most authority and reducing the drag of an elevator that is "x" degrees deflected to maintain the current gee or AoA command.

Bear in mind that my primitive FBW system did not have an elevator at the rear of a horizontal stabilizer ( HS). The whole HS moved for pitch commands.

As with the 'bus, HAL "auto trimmed" to our gee command trim setting. Unlike the 'bus, we could trim for a gee, and not be limited to a one gee command.

- Gerard asked a great question, and it is one of mine.

What was the "stall warning" all about after the pilot pushed over a bit?

- Chris brought up the relatively benign deep stall characteristics. It's what got me joining this august grope of "experts", wannabe's and actual 'bus drivers. And others.

All who have flown a delta wing raise your hands! Concorde counts. Mirage variants count. The one I flew as a yute was so deceptive that it was scary - F102. Very slight "buzz", but no real buffet or burble. Great directional stability and lateral stability. But the altimeter was pegged at 10,000 feet per minute going down.

The modern commercial heavies have really great aero characteristics that can make a stall insidious. And the stall is not like the Airbus manuals depict on the lift versus AoA curve. There is no sharp break in the curve at "x" AoA. It's a very gentle curve and one can fly at fairly extreme AoA's without the sharp pitch excursion we all saw when checking out in a Chipmunk or Cessna or T-28 or......

The problem occurs when the jet reaches a certain AoA and c.g. and speed combination that prevents normal pitch down commands from being effective. It's that pitch moment graph I posted years ago ( heh heh).

The test pilot maneuver resulting in a deep stall that I posted was not the classic entry to a deep stall in the Viper. The classic entry was a fairly steep pitch, low gee, rapidly slowing speed, and running outta air molecules for that HS or THS to use to get the nose back down. Does that sound relevant here?

I gotta admit, that from this old fossil's FBW experience and perspective, that there are too many autopilot functions embedded in all the 'laws". The jet seems perfectly capable of flying to the basic limits that all the heavies, if any, can match. And my problem is the "basic" limits seem to take short shrift behind roll angles, autotrim, mach warnings, etc. etc. ad nauseum. Then I read here that one reversion law commands the motors!! BEAM ME UP!! If not in some autopilot mode, let the plane fly. Sheesh.

I thank all here for allowing a "lite" pilot with some FBW experience back when the Earth was still cooling to participate. I am now SLF, and I wanna feel comfortable about the jets I ride in.

If you were not such a :mad:, you would understand that anybody, including pilots, may be induced in error, even by serious Air France or Airbus shortcomings. So first step is to understand what goes wrong.
Stop trolling, please.

Honestly, my picture of the accident includes everything I read here. I don't reject anything, you may notice my input has to do with proposing some things, some perhaps a bit bizarre, but what about this accident is NOT unusual? I do intend to embrace the Truth when it arrives, my perspective may not project that, but questions cause thought, unless rejected, and I think that is too bad.

This thread exemplifies what I think is a solid plus in analysis. There is no boardroom with a coffee maker and sandwiches which instill a faux civil atmosphere. It is a rare pilot who can be scrupulously objective. There are plenty here, and I admire all the people who post here.

As of now, it looks rather bad for the pilots, all three. I cannot figure out why in the world BEA would accede to pressure to release such an unsatisfying Note. It has made the adversarial nature of the discussion ever more polarized, not that it has any relevance on their Report.

Hi Turbine,
Certainly not. Discussing systems and engineering "stuff" is very fine with my bandwith allocation. What I'm very less interested is about such fallacious and circular argumentations based on nothing but greviances, which discards investigation work in order to insert some supposed hidden informations, revealing the cover up of imaginatory supposed "facts".

I'll give you an example:
Bearfoil's point is that the aircraft attitude at AP disconnection was in some abnormal attitude (pitched down below -13°) and slow, but fast - such fact, of course is hidden by the BEA - then UAS was immediately due to "turbulent airflow" while she zoomed 3000 ft due to crazy THS systems triming her to the max because of some nose down roll tendency, but Pilots fought the roll and nose up pitch, but were few seconds later back to 3NU which was now increasing to 13NU while applying full back stick during half a minute, and so on...

Here, let me tell you that it is in fact a huge waste of bandwith, in my book.

Waste of time?
 

I have followed this thread since Day 1 and contributed occasionally and asked questions when my personal knowledge ran out. It seems a great pity that we have to persistently wade our way through piles of cr*p to get to the 'good stuff', and most recently I have to take issue with bearfoil. With the exception of little green men from the planet Zog, it seems that any kooky idea with a shred of Airbus-itus (definition: anything built by Airbus must be implicitly designed to screw with the crew in some nefarious and more convoluted manner than it did in the previous post) must be somehow to blame, and blow me down with a feather, the crew must have been overtaken by HAL as it went increasingly 'nuts' (obviously hell bent on dashing the airplane into the ocean despite the heroic efforts of the crew).

The last post I bothered to read by bear involved some absentmindedly forgotten-by-BEA nose-down pitch (of -9 degrees!) and overspeed (I think it was), that the PF fought against resulting in the zoom-climb of doom....blah, blah, blah.

Look I'm all for theories, but let's face it - this imaginary world belongs firmly over in Rumors and News. Please take this lala-land and post it where it belongs. Please.

I have learned so much over here in the relative sanity of the Tech Log - massive :ok: to mm43, gums, Chris Scott, Turbine D, takata, henra, PJ2, DW, '33' and others. The sanity you have brought here is greatly appreciated.

I'm not sure what this post itself contributes other than allowing me to vent some... :{

@ Zorin 75:
 

Just before!

In NORMAL law αsw = 23° (fixed value for all flight phases)
In ALT & DIRECT Law αsw = function of M and S/F config.
S/F is not applicable, in this flight phase αsw = function of M only.

αsw---- MACH


-Inhibited on GND (if not in test)

Initial STALL STALL STALL
 

Well, it seems I'm the one making wishful thinking while an endless Bearboiling of circular points seems to make your own days. Obviously, what makes you think that she was near STALL at handoff, when AP kicked off, is that the STALL WARNINGS sounded twice, then stopped. On the other hand, in many UAS case, without changing anything about the flight parameters, this happened exactly the same.

From here come your insistence that she could not have been in "controled flight", no matter if everything else is pointing that she was, except a starting roll to the right.

On the other hand, I have already underlined many other cases of UAS events recorded that are obviously pointing at exactly the same "suspect" stall warnings when the FLIGHT LAW is switched from NORMAL to ALTERNATE. For me, this is certainly a major issue (if confirmed) of the Warning System because I really think that it could have played a major role during the following sequence leading to the total loss of control by the crew.

In fact, there is a VERY simple BASIC rule in the system :

* NORMAL LAW -> COMPLETE SAFE FLIGHT ENVELOPE USED BY ELECTRONIC FLIGHT CONTROL SYSTEM
* ALTERNATE LAW -> REDUCED SAFE FLIGHT ENVELOPE USED BY ELECTRONIC FLIGHT CONTROL SYSTEM

Hence, any switch from NORMAL to ALTERNATE LAW should be immediately followed by a reduction of the safe flight envelope at both ends : in clear, the safe margin should be expanded concerning both the overspeed warning and the stall warning (nonetheless, the aircraft real "flight envelope" is still the same, but its defined "safe limits" are changed due to some certification concerns about its systems state and possible manual flight).

1) In our case, the high speed limit warning would be reduced from Mach 0.86 to Mach 0.82. Also, the low speed warning (alpha-prot) could be rised (I did not find the relevant data about it, then, if wrong, see point 2).

Nonetheless, neither overspeed nor stall protections could be applied because of the current declared airspeed monitoring by EFCS : ALT2 is immediately applied, no protection could kick without a valid airspeed ; After this point, the PFD function displayed (overspeed and stall warning speed limits) will disapear from both PFD.

Next, following an UAS confirmed in ALTERNATE LAW, the calculation mode of this stall warning limit will change again from one function using an AOA corrected by a Mach value supplied by those ADRs to another one using a default Mach parameter. This would change the granularity of the AOA measured and subsequently will increase the margin for the stall warnings.

My explanation (so far) is that, when UAS is declared, there is a small time window where the limitation of the safe flight envelope is applied to the stall warning limit which last during the resolution of an UAS monitoring : it takes 10 seconds for the system in order to confirm ALT2 (after that, the system could not subsequently revert to NORMAL LAW without a full ground reset).

So in fact, the aircraft, during those 10 seconds window, without changing any flight parameter, will fly at speed closer -or below- the threshold of the STALL WARNING speed. But, following the UAS confirmation, the new function will apply another AOA warning limit. Then, this would stop the WARNINGS... if the reduced threshold has been overshot during those 10 seconds. So my bet is that the STALL WARNINGS stopped at 0210:14 and would have sounded during the 0210:05-0210:14 time window.

2) If alpha-prot calculated speed do not rise when switching to ALTERNATE, another simple explanation is that a drop of the polled speed, due to unreliable airspeed situation, is not filtered by the stall warnings during the same defined 10 seconds time window. Hence, if the polled (unreliable) speed goes below alpha-prot speed, this alarm is sounding. Once UAS is confirmed, as precedently, the new UAS STALL WARNING function is declared and the alarm is stopped.

Those 2 points are based on the history of published UAS events :
Here are the tables of the 36 UAS events declared before the publication of 2nd BEA report; one may see that 12 cases of "suspect" STALL WARNINGS were reported (33%). Moreover, those stall warnings do not let any trace outside the CVR or crew reports (they are not part of the maintenance post flight reports), and this survey was mostly incomplete about many cases listed. So it may be assumed that more cases of "suspect stall warnings" were not recorded.
http://takata1940.free.fr/list.jpg

http://takata1940.free.fr/stall.jpg

Dozy, multiple quotes:
Originally Posted by Machinbird http://images.ibsrv.net/ibsrv/res/sr...s/viewpost.gif
JD-EE, we were temporarily lead astray by forum members who apparently believed that autotrim would not resume control once you made a manual trim input.

That would be me, having misread the documentation I had.

Quote:
I strongly suspect there is some misinformation adrift in the Airbus community that needs correction. I wonder how BEA will address that?
Again, it was just me - the "Airbus community" were the ones who kindly corrected me via PM.

Dozy, you can't take all the responsibility, there were others as well putting out this misinformation.

To be fair with Dozy, this "misinformation" comes from Airbus itself, as the manual states in one page that the use of manual trim will "freeze" autotrim, and somewhere else that the PRIMs do not fault when manual imputs are applied, but that the PRIMS will re-synchronize with the manual trim imputs, which have the priority over autotrim.

Takata,
Quote:
Originally Posted by Machinbird
The problem with the trim is that it moved to a high aircraft nose up setting without crew awareness.
Takata
Yes, assuming he knows he made the stick inputs.
Quote:
Originally Posted by Machinbird
It appears that a FBW aircraft requires the pilot to know exactly what mode the aircraft is operating in or else the question arises, "What's it doing now?"

TakataTakata, suppose we sat a pilot down and hit him with a fire hose stream. Do you think he can remember his birthday while this is going on?

Perhaps a couple of recent accidents make me think that way.


Quote:
Originally Posted by Machinbird
In the case of the AF447 crew, they had no attention to spare to consider what the trim might be doing

Takata, have you ever been in a really stressful situation while flying? From your comments, it appears that you have not. That is fortunate for you, if so, but it gives you a blind spot with respect to the effects of in flight stress. You seem to be assuming that the initial pitch up to FL375 was deliberate.
Quote:
Originally Posted by Machinbird
thus the nose up demands caused the trim to run silently to a high setting

Nothing down low when setting up for a landing, a whole lot is wrong when it happens at cruising altitudes. The problem is the silent motion of the trim. They needed to know it was moving so that they could monitor it.
Quote:
Originally Posted by Machinbird
without the knowledge or anticipation of the crew.

Yes, I think that they didn't want all that nose up initially. Later, after the stall, who knows?
Quote:
Originally Posted by Machinbird
Once at a high setting, the trim acted to stabilize the aircraft in the deep stall that they eventually achieved.

Just an opinion at this point, but the addition of TOGA thrust coinciding with the second stall warning is too reminiscent of an airline approach to stall procedure as recently taught, except pitch control was improper/clueless as it had been since PF said"I have the controls."
Once in the stall, it was imperative that stall recognition occur, but it seems, it didn't. Without AOA indicatiors, without stall warning, without comprehending that the decreasing altitude was real, there was one final item that could have explained the situation and that was the THS trim position.
Quote:
Originally Posted by Machinbird
When the Captain arrived on the flight deck, he had to puzzle out what he was seeing. From the jump seat, the indications were essentially hidden by the wheels themselves.

There is no surety that the Captain would have been able to puzzle out the situation in time if he had seen the THS trim, but if he had noted its very abnormal position, there is a logical thought process that should follow that would lead to stall recognition. For him to check the THS position, he would almost certainly have to form the thought to check it and then lean over for a peek.
Quote:
Originally Posted by Machinbird
The THS autotrim system has virtually eliminated the problem of trim-runaway, but it has replaced that problem with a new set of problems.

We seem to agree that they likely never tumbled to the fact they were in a stall. The recent Perpignan trim related accident combined with the AF447 accident seem to indicate that there are issues in trim position awareness in Airbus land.
My apologies for a rather cursory reply, unfortunately I have limited time during the week.

CG position ?
 

TD, I have partially addressed your question here.
As the take off CG was very much forward (23.3%), and despite the aft fuel transfer, I figure the ideal target of around 38% for fuel saving could have been approached only toward the late stage of the cruise phase, not after 3 hours of level off.

To have a CG more fwd than aft should help for the handling, and stall recovery as well ...


I don't demand much, just criticize what, IMO, deserves such.
You, obviously, do not tolerate such critics to be made. Fine with me.

As you're not willing to publicly discuss Habsheim, despite my invitation, why can't you quit mentioning it ... ?

Hi Machinbird,
Well, I would say that what the PF really wanted during his initial imput is quite irrelevant concerning the "THS issue". You should read again the BEA note as you seem to be the one assuming that the THS was trimmed during the initial climb to FL375 (please, don't trust everything Bearfoil is posting!). You'll discover that everything started a while later: this THS was trimmed from 3NU to 13NU between 0210:51 and 0211:50, therefore, until this point, everything seems about right about its behavior:

"0210:51, the stall warning was triggered again. The thrust levers were positioned in the TO/GA detent and the PF maintained nose-up inputs. [...] Around fifteen seconds later, [...] The PF continued to make nose-up inputs."

What would make you believe then that the PF maintained (at this stage) all those nose-up imputs without knowing it?

In fact, stall warnings seems to have sounded correctly at this point (see above) while the PF was trimming his THS near its max by applying sustained NU orders! Hence, this THS was fully trimmed during the stall sequence, not before it!
Without puting the sequence in the correct order, this will be fairly useless to discuss its usefulness as to alert you that you are going to stall, when you are fully stalled, but still trimming it the other way!

Beside, I also believe that there is a great deal of chance that the initial PF nose-up order, and following climb, was not voluntary but rather the result of an overcontrol of the roll tendency. Nonetheless, this is absolutely not the issue discussed about this THS setting. When THS goes there, it is quite hard to believe that the PF wanted something else than those sustained nose-up orders... Consequently, what more could he have learned from his THS setting that he already didn't knew while pulling up?

Your concern about it is only understandable with hindsight. Should they have tried to recover from a stall, at some point, this certainly would not help... but then, he put it here at the first place when it was certainly not the right thing to do at all...
My point stop here on this subject.

Hi Takata,
Just a brief note, my assumption is that the trim initially traveled slowly nose up and began to really move as the aircraft approached the stall. I do not attribute the nose up attitude to the trim as some are wont to do, it was merely the result of more nose up input than nose down input from the SS, averaged over time-that and what must have been a completely fouled up scan.

Is there any sort of pitch indication adjust function on the PFD? The PF's control inputs act like he had an erroneous level flight pitch setting. I am strictly a steam gauge guy so I've never had the opportunity to see how the other side currently lives.

Yes it did, but look at it from the standpoint of the PF. He got a stall warning and applied what to him seemed like an approach to stall recovery, and the stall warning eventually extinguished (because he was deeply stalled). The aircraft attitude changed very little. Why should he think he was stalled. He just didn't have airspeed indications, and now the aircraft is acting really strange, dropping wings and things like that, and the altimeter is doing the digital equivalent of unwinding.:eek:

associated actions
 

Hi Machinbird,
I agree & in 2009, there was no stall warning procedure in the QRH.
The guidance was buried in FCOM, Supplementary Techniques, Flight Controls which describes the initial actions which the PF took.

"An aural "STALL, STALL, STALL" warning sounds at low speeds. Upon hearing it, the pilot must return to the normal operating speeds by taking conventional actions with the controls:
Thrust Levers...TOGA
At the same time:
Pitch Attitude...Reduce
Bank Angle...Roll Wings Level"

Unfortunately he then lost the reference power setting he had before the UAS event. I would guess that he recalled the power and pitch attitudes he learned during his conversion course for unreliable speeds and remembered the TOGA + 15 degs pitch ... (but that doesn't work at FL 375.)
The next bit reads "Thrust/Pitch .... CL/5degs Above FL 100"
but by then they were so very deep into the stall they would have needed at least 10 degs nose down to accelerate.

Why do you label these stall warnings 'suspect'? BEA has investigated several of these cases and writes in its Interim Report No.2: I think there are number of other statements in your post that you may wish to rethink. Alpha-prot is the AoA at which Normal Law changes from an Nz law to an alpha-law. In this connection you are several times referring to speed where the reference should really be to AoA (*). Alpha-prot is only relevant in Normal Law and I doubt it is even calculated in Alternate Law since it has no significance in that law. In Alternate(1) there is a protection called 'Low Speed Stability' that is driven by IAS instead of AoA, and changes alternate law to direct law. It is active from about 5 kt up to about 10 kt above the stall warning speed (*), depending on weight and slats/flaps configuration. Low Speed Stability is lost in Alternate (2) Law.

P.S.:: (*) For given configuration, weight, c.g., and altitude, AoA is a function of Mach and load factor. V-alphaprot and V-alphamax shown on the speed scale of the PFD are calculated for 1 g from alpha-prot and alpha-max, respectively. V-S/W is calculated for the actual "gee" and the stall warning threshold AoA.

Please, what is this transition from g-control to pitch-control you are talking about and where is it described in the FCOM? Does it apply to A-330?

CONF iture

Thanks for the CG explanation. That was something I was trying to understand, and I missed it in your previous posts.

Hi,

takata
takata I will tell you what goes wrong and will be wrong in the next decades
During the next ten to twenty years .. forecasts show that companies will grow very quickly (especially those of the Persian Gulf region)
The demand for pilots and other personnel will be huge
The consequences of these expansions will be reduced formation and training while the aircraft are becoming more complicated
Ironic indeed ....
The result of all this is easy to predict
The planes will start dropping like flies killed by insecticide
This is what's wrong (and already present)

Serious .. Air France ? ... I disagree (see audit results and stats)

So...

Have some airlines really been training NU inputs and HITHRUST for stall recoveries when this seems to produce further (and usually unwanted) NU trimming?

Has there really been a history of excessive ND Stall recoveries creating ground proximity incidents - forcing this re-evaluatiobn of traditional stall recovery methods?

Or... has this type of recovery procedure stemmed from 'The Devil makes work for Idle Hands'..

e.g there has been no specific events creating a need for high power, NU recovereies, other than that modern fan-engined aircraft are (nominally) capable of such recoveries.

One area of confusion in my mind Chris S may be able to clear up. What I think I see here and from BEA is that the THS did not move from its cruise angle until the top of the zoom climb? Given that it is supposed that PF held a nose-up demand most of the way up, why not? There must have been a good 40 seconds or more of 'demand'.

BOAC: More 'imaginative thinking?'....

Where the heck do you SEE that written?

So where are the "held nose up inputs"??? It's just NOT there. Why invent this?

The comment regarding the THS comes later. Read the note.

Takata and others, Zorin_75 asked a very pertinent question about the persistent fixation on the autotrim. Why IS everybody so fixated on it?

Surely nobody here thinks that turning autotrim off or simply coupling the elevator to the stick would have made a difference, I hope. The elevator jack screw runs too slowly for the pilot to use it for primary control. So the dual action, a quick action partial solution and a chugging along slower action is derived as the best way to handle it all. But the real problem is the full to the stops NU input that persisted until the plane was in an attitude from which recovery was problematic.

Hadn't it gotten into a condition such that the elevator authority was sharply limited by it's being fully stalled? By the time they made the final ND input there was no way to pull out of their stall. (I mentioned the thought of breaking out of it with unequal thrust to introduce an asymmetry to break out of the stable stall. And, yes, I can see no sane pilot wanting to do that. I vaguely wonder if it could have led to breaking the stall before it hit.)

The plane was crashed, essentially, by the persistent NU command not the THS. I've talked about training to use the trim wheels. But on thinking it over and too vaguely remembering the stall characteristics of the aircraft vis a vis position of the elevator I suspect all the elevator would do is slightly modify the NU forces on the aircraft. Are you worrying about something that appears to make a huge difference yet makes no real difference at all when the AOA of the aircraft is some 60 degrees?

Just askin....

- Hmm I thought that is what I posted, never mind! Specsavers?

The 'Nose-up demand' certainly does not fit with my 'understanding' of events (which again you MIGHT have noticed (Specsavers?)) but does appear frequently here as a mantra so I was looking to see if it was possible, actually. I'm sure there are those here who think PF made the a/c climb, or am I imagining that? Anyone own up or can we eliminate that 'theory??

JD-EE - indeed , and I have discussed this with my tame AB expert - we both agree that after 40 + years of 'doing aviation' neither of us had ever contemplated a situation where a tailplane could be so stalled that a nose down elevator command could be a nose up.

JD-EE Great minds think alike, wish mine were one. There is very little to go on from the investigators, and that partially explains the breadth of comment here re: the most important phase of flight. Induction of Fail.

Too many shiny objects to harrumph about, and the meat of the matter in abject darkness. For ten seconds before A/P drop, and ten after, 447's fate was being sealed.

Either one believes the PF (and auto before him), created the stage for an absurd and fatal climb, or one believes they did not, and the hand of Satan slapped them silly at 2:10:00.

In between, (and in the belly of the only important data, that which is missing), this flight was lost.

Who is to say the airframe did NOT actually STALL after the very first WRN? Was the fatal parabola initiated at the beginning of ACARS, or some time before?

Maybe it's me, but I sense that most here assign controlled flight to this a/c until the apogee of the zoom. By definition, one thinks, LOC happened at the "the aircraft began to climb...." Whether or not the PF "caused" the event, is not as important as what caused the event that "caused the event".

Snapshot thinking? a/p dropped for a reason, so in strict terms, can one absolve the a/c?

Of course not. This was a continuum, and I truly believe that when the data is complete, and in the fullness of time, there will be still some adversarial and parochial gripes to be had.

In the meantime, Shadow is most intriguing. With the chronic and slow loss of "Read Speed", wouldn't the a/c Trim Power UP and NU also?

Why hasn't anyone addressed what the a/c might do with the need to "Maintain Speed and Altitude" when in autoflight? Because BEA have ignored it? It is "off limits"?

It occurs to this nomex clad cattle prod wielding ex-pilot that some one should at least attempt it?

Could the a/c have been at max power and NU somewhat at drop? Enough nose up to chirp the STALL and Cricket at 'one' NU left input?

The precise data in this bracket of unaddressed time is seductive.....

BOAC;
While it may be an impression on the part of some, the reference for any of this is not here but the BEA Update, and it does not state that NU SS was held most of the way up. It states that after the plane started to climb the PF made ND control inputs and alternately left hand right roll inputs.

We see a this best in the Excel graph which A33Zab provided, (post #691) in the previous thread - He has shown, and I think accurately so, that the THS does not begin to move from its 3° position until the aircraft is almost at the apogee. The SS was not NU most of the way up but was NU briefly which initiated the pitch-up then ND which began a slight level-off though the aircraft kept climbing. The stall warning occurred @02:10:51 according to A33Zab's graph and a continuous NU SS and application of TOGA thrust is coincidental with that event, and the THS begins to roll towards NU. The aircraft actually stalls at about 02:11 as the descent begins, while the SS is continuously held in the NU position with the pitch attitude at 16deg and thrust set at TOGA. According to the graph, the THS reaches 13NU at around 35,000ft in the descent at about the same time the captain enters the cockpit.

I suspect that while the THS position moves to continuously neutralize elevator forces just as any trim does, the movement is a "follow-up", the rate of change being governed by the pitch basic control loop which has a number of discrete inputs but primarily the sidesticks and the accelerometers. A long discussion has already taken place on Nz law and trimming to 1gee. Manual control of the THS can be taken any time by rolling the trim wheel forward or back. The following is a schematic of the THS general arrangement:


http://www.smugmug.com/photos/i-Ndwn...NdwnjLk-XL.jpg

Again, the BEA Update states that the PF made "a left nose-up input". It isn't a "theory". Back-stick was later held during the stall, and for 30 seconds during the descent. Also, ND SS was briefly held after the initial pitch-up and briefly again during the descent. Knowledge of all other times will have to wait for the next Interim Report.

BOAC, I "think" what I meant was that a ND input would still come out NU and an NU input would also be an NU input. Whether they reversed as gums has said can happen, I don't know. I don't THINK the charts showed a full reversal even to the level that ND was more NU than an NU input. All I remember is that it looked like both resulted in NU type motion.

And I think rudderrat hit the nail on the head for why the PF pulled nose up even though he was above FL100. The big (well trained) print buried in the FCOM called for TOGA and NU. (If it really is presented as quoted or as I remember seeing it in the past with smaller print for the above FL100 part it was presented poorly. It should break right at the beginning for above FL100 and below FL100 so the decision is made early and then procedure can be followed logically and linearly. (And thank you thank you for that post rudderrat.)

bearfoil, logically the airspeed loss cannot be slow. If it was the heaters would be effective. For the icing to happen it would have to take place very quickly in a fashion to overrun the heater's capacity on an instantaneous or nearly instantaneous basis.

It took awhile for that coin to drop, too. The original statement bothered me. I just figured out why.

PJ2, would the original NU command be a persistent, albeit moderate, command or would be a short one to establish the increased AoA and then returned to neutral? I got the impression from choices of wording that the original NU command was transient.

JD-EE, transient and then neutral, then with slight ND, then NU more continuously, is my impression.

PJ - that fits with my understanding, so where the *** does 7000fpm come from? That is one hell of a r o c for a 'transient nose-up followed by nose down'. To take your words "but was NU briefly which initiated the pitch-up" - I am still having difficulty in envisaging 7000fpm in a 200t a/c at FL350 from that - are you content? I am, by the way, quite aware of what BEA are saying about when the THS moved and I have assumed it was to compensate for the increasing NU demand on the elevator from PF's 'stall recovery'.

Hi JD-EE,
Respectfully, the procedure recalled by rudderrat is to apply Nose-Down (reduce pitch attitude) at the same time as TOGA... while, in our case, the pilot apply Nose-Up (increase pitch attitude) at the same time as TOGA... Then, this is quite hard to conclude that he was following this procedure!

This is what (below) he is talking about; you may also notice the relevant part about the aural stall warning sounding at "altitude":
"... it warms that the aicraft is approaching the angle-of-attack for the onset of buffet. To recover, the pilot must relax the back pressure on the sidestick..."

http://takata1940.free.fr/A330_ALT%280%29.jpg
http://takata1940.free.fr/A330_ALT%281%29.jpg

takata, that is correct if you are not reading in a hurry.

This is what rudderrat quoted:

So this is what the pilot was doing. He never got down to the other part.
And note that this is from the FCOM not the QRH. Rudderrat remarked that in June 2009 the QRH had nothing to say about stall....

Is what you quoted from 2009 or a current manual. It was noted in some incarnation of this thread that the procedures had been altered and something had been added to the QRH regarding stalls.

Assuming the PF was sane and coherent, (I do), NU would be input to counterract ND, or a descent of some sort. After his correction, neutral stick. If the slight rotation he commanded (NU) then appeared too much, he would input ND (he did). Wait, that did not work, she still climbs. A bit more ND, and what is that pesky Roll? More Roll? compensated. More Roll still? and continued Climb? Increasing climb rate?? More ND and More.

BEA tells the Truth, and I see it this way. By BEA's data alone, where is PF screwing anyone's pooch on the way up? On the way down, with gobs of AoA, the Pitch can reverse with a Stalled Tail Plane. Remember this is not so much a Tail as a variable incidence wing.

JD-EE. Chronic and slow enough to allow the TailPlane to trim for it. And engines to develop max thrust (or close to it).

BOAC - we don't know the details of the SS movement so it isn't possible to correlate such movement with rate of climb. The aircraft would have had a nominal pitch attitude of around 2.8deg roughly - SS movement aft of 3 - 4cm (previously said 2 - 3cm), as measured at the top of the stick would produce an enormous but clearly brief rate of climb of the kind we see here - it would be the equivalent I suspect, of about a six-inch rearward movement of the control column on the B737, just to try to equate a sense of the large changes involved in pitch and climb.

"Two-hundred tonnes" I believe is immaterial here as the wings produce commensurate lift for the design. It is a matter of mass, momentum and available energy (far more than 200T worth) from the wings at the initial CAS, which, given the eventual pitch attitude of 15deg clearly would be quickly unsustainable but in the short term, achievable. The BEA Update states that the VSI reduced to 700fpm and it is easy to understand/imagine the ballistic trajectory which resulted prior to the start of the descent.

Aside from my original thoughts of responding to the UAS drill instead of "doing nothing" while getting out the QRH for pitch-and-power settings, this almost looks like this was a reversion to original training where the approach to the stall is taught in transition or initial courses at lower altitudes in which the goals have traditionally been minimum loss of altitude, (much discussed in earlier threads) and "powering out of the stall" using TOGA thrust. The initial pitch attitude just may be some over-controlling which resulted in checking the stick forward a bit before responding again to the stall warning with back-stick, (driving the THS up, as we see). As I mentioned a number of times before, once the aircraft departed level flight, the "knowns and cues" for stable, level flight were gone and situational awareness, (what pitch? what power?) became problematic. The simulator exercise required a lot of nose down and I don't recall seeing what the THS actually did during the exercise and didn't look at the THS indiction. The descent rate was 7500fpm and the stick had to be held full forward. This was a stall from a pitch-up and the PFD looked like this in the recovery:


http://www.smugmug.com/photos/i-6svs...6svsRpq-X2.jpg

Obviously, it isn't what the pilot was doing as he applied sustained nose-up, no? Otherwise, every single procedure calling for TOGA would fit with your definition of what he was doing. It would better fit with WINDSHEAR procedure in this case.
As the BEA did not provide any hint about any manual change of N1 rate during the first part of the climb (between 0210:05 and 0210:50), we may also think that the PF realised he climbed to 37,500 ft without adding any thrust from 35,000 ft and was trying to compensate for it. Maximum N1 is about 115%, but I don't know if such level could be reached in current conditions... at 35,000 ft, he was flying at about 95% N1.
This FCOM quote is dated March 2003 (Rev 18). Most people did not look at the FCOM 3 which is providing "supplementary techniques" (3.04.27).

PJ2 "Once the aircraft 'departed' level flight". Other than level is not a 'departure', yes? Departure means a loss of aerodynamic flight, no?

Per BEA, how is the a/c known to be at S/L flight at dropout? PF obviously corrected for a ND or descent (or overspeed) of some sort, right? Assuming his NU input was a stupid or inadvertent blunder is not supported by the Data, surely? He could have pulled NU to correct an overspeed whilst NU already, yes? This would fit with such an unusual roc so quickly. Would the a/p have trimmed the NU to control speed with added power it had applied to "correct" a "slow" IAS due particle ICE plugging?

I do not mean to nitpick here, but your commentary is the Gold Standard here, and am I missing your drift? Maneuvering to retreat from STALL WARNING is not a STALL recovery, that is why the book allows for back pressure. I cannot find where it allows NU? Only back pressure, to minimize altitude loss? In fact, it advises lowering Pitch at this point, which is not the same as ND. This is parsable, but I read it as "non-negative" 'Target Pitch', to maintain altitude. I find nowhere a direction for PITCH UP.

Situational Awareness is the key, of course. Strictly speaking, loss of autopilot, for whatever reason, is a loss of SA machine-wise? Whether for UAS or inability to maintain programmed flight envelope limits?