Cheers.

I thought that was the case, but wanted to make sure.

So - with that checked, your assertion about the discrepancy between the selected VS and actual VS during the descent phase of the sequence is correct (i.e. the FDs could well have been commanding a high nose-up pitch angle). However, during the descent phase with the aircraft stalled, the PF would have had a job keeping the nose consistently above 0 degrees, and chasing a positive FD command then would have been practically impossible.

What's interesting to me is what happened in the climb phase, and for me it's vital to keep these two phases distinct in terms of analysing FD behaviour and correllating it with the PF's actions. In both cases of the FDs' reappearance during the climb phase the discrepancy between the pitch angle, selected V/S and actual V/S should never have indicated an instruction to put the nose up to 15 degrees, but that's what happened.

Picking this apart is a complex endeavour!

Hi DozyWannabe,

Please see post #113 by A33Zab.
I can't explain the initial climb without flight directors (perhaps work load, surprise, CBs, Lightening flashes, smell of ozone, noise on the cockpit from hail, AP disconnect wailer, ECAM dings, ALT Alert warning, "Stall Stall" SV, and calls from the cabin crew, etc. played some part.)
However A33Zab's picture seems to correlate quite well a bit later.

OG:

This is somewhat of a conundrum in the sense that when you reach -10 degrees, as you say a long time before actual recovery, to hold that pitch attitude do you...

1. release the SS, i.e. FCS positions the elevator to hold what?

Still stalled, It's likely the elevator would go right back to full ANU to hold ~1G and the existing flight path which is quite a bit steeper downhill than 10 degrees at this point. Rule out SS release?

2. or is the SS manipulated, i.e. what is necessary, considering FCS elevator response, to hold a constant pitch attitude while stalled? If not continuing full forward, how do you do this? Some occasional aft SS necessary? (rhetorical :})

The FCS has a pretty big say in what you actually get here in practicality, although the theoretical side is certainly sound and I understand that's as far as you intended to take it or that anyone could take it. :ok:

In practicality, I wouldn't come off the forward SS stop (with whatever extended nose down pitch rate was actually available) until I knew that I was 'flying' again. This approach, of course, may change some of the values. :)

I hope we're looking at the same post - I'm seeing one with this image, which deals almost exclusively with the climb phase...

In fact I've (roughly) highlighted the areas where the timeline says the FD was available and in which mode:

http://i1088.photobucket.com/albums/...ned/VS1_FD.png

In the first case, I'm not sure what ALT CRZ* might have presented, but in the second case, even with +6000fpm selected, the FDs engage with an actual VS in excess of that. In that case, would the FD have commanded nose down?

You definitely would have to apply forward SS until starting to pull up. How much SS I don't know. This post has a graph that shows the normal acceleration NZw.
Another post shows it for another trajectory.

N.B. In these graphs NZw is the acceleration normal to the flight path, whereas it is usually measured normal to the airplane's longitudinal axis.

Perhaps I should have written "assumed to be flown at a constant pitch attitude"
Even in the stall the aircraft always responded to elevator is a conventional manner, and I didn't really anticipate any difficulty in using the stick to maintain constant pitch. So I didn't give the practical piloting technique a lot of thought :)

But as you say, I didn't intend it to be taken too far - I was just curious to know whether there was, in fact, any chance of recovery.

It doesn't really matter. What many of us are saying is that FD 're-enable' should either be inhibited without a deliberate selection by crew OR, less desirable, it should re-engage in 'acquire selected altitude' mode, although this latter could cause problems if slow and below that altitude. Whoever decided it should just 'accept' the existing VS? Did anyone ask a pilot?

Presumably if the A/P was not inhibited also it could just re-engage and rear up to pitch for 6000fpm up? How sensible. Now, wait a minute, that rings a few bells..................:ugh:

OK465;
In the sim exercises, for recovery the SS was held full nose-down from the beginning of the stall warning at about FL360 until the wing was unstalled at about FL250, about 40 seconds total time, with a maximum achievable ND pitch of about -12 deg with an average of -10deg. The thrust levers were in the CLB detent and the THS was initially at 13.6deg and was returning to the normal cruise setting.

With the FPV symbol available, the FPA could be observed just above initially at -40deg, (pitch -10deg).

It began to move, initially very slowly up, about 15 seconds after full ND SS;

- at 29 seconds after full ND SS, the FPA had moved from -40deg, (pitch at -11) to -25deg, (same pitch);

- in the next 5 seconds it moved from -25 to -15, (FL257);

- at 38 seconds after full ND SS the FPA was -9deg, (pitch -5deg) with the wing unstalled and the CAS at 255kts.
Absolutely correct!

Presumably - given the presence of pilot engineers in the people required to sign such behaviours off - yes.

Which is exactly why, unlike the FD, the AP *is* inhibited under such circumstances.

Hi Dozy;

Re the discussion with RRR on the FDs...just to clarify, when they re-engage, they synchronize with the current V/S or FPA aircraft parameter, and not with anything that is selected.

For reference - what RRR said was:

We really need to be sure about this, because it will have had a massive impact on what the FD may or may not have been displaying.

One thing I haven't seen mentioned on here (only reading the latest thread), is that it took 40 seconds to call the captain after autopilot kicks off, and another minute for him to arrive. I'm not a pilot, but it seems it should be policy for the captain to be called immediately in that circumstance.

We are sure...:)

The select VS window on the FCU will NOT be open (i.e. the display will be blank) until the FD's return. There is no active mode until FD's are available.

The display window will then 'open' and will, on the FCU panel, display the existing VS at FD recovery, as will the FMA, i.e. +1400 VS. The FCU value & the FMA value cannot be different.

edit: BTW, in ALT2 the A/P is inhibited permanently for the remainder of the flight.

Dozy;
I clarified the FD matter in post #119 with help from various documents at my disposal. While complicated to explain, the FD operation is fairly straightforward after training and some experience with the system.

It works extremely well, and there is a very good reason why the UAS memorized drill requires that they be turned off.

I suspect you're well aware that building in an "auto-shutoff", like the oft-cited issues with the stall warning NCD matter and the movement of the THS, may create new, separate problems with their own extremely-rare-though-possible risk exposures, and isn't as straightforward a solution as it may seem. It was stated many threads ago that this is a performance accident not a technical one. The issue here is performance, not system characteristics in extremely rare circumstances, for no designer can possibly expect to anticipate and then counter all possibilities and permutations of degradation and system behaviour. As I have observed many times, this is, in essence, sadly, tragically, not a complicated accident.

OK465;
Precisely.
Which clarifies BOAC's concern that the AP would engage at "+6000fpm"; It wouldn't, in ALT2B.

The AP would engage in Normal Law but, under normal protections/limits it would also ensure a smooth reduction from an obviously-too-high VS to a stable level-off or gentle climb.

In any case, (because of the Madras accident), as speed reduces, VS reverts from VS to OPEN CLIMB and the thrust reverts to SPEED.

PJ - out of interest, how did you get the sim to an AoA of 40degrees? Starting recovery 'at the stall warning' would not achieve that?

When and How long are PRECISELY known, where have you been ?
Also FD commands have been recalculated by Airbus, or call it BEA if you like.

Everything is presented on a video animation we will not see.
The animation presented during the press conference has no interest.

I think the AoA was closer to 30deg, (-40 FPA, -10 Pitch), but I held the stick fully back until the stall warning and kept it back. The sim pitched down to about -2 or 3 degrees, even with power on and the THS moving towards -13.6deg. The stick was then placed fully-forward and held until it was clear the wing was flying again.

In this image, the stall warning was constantly sounding. The image of one of the recoveries may help envision the attitude, rate of descent and recovering speed even while still fully stalled:

http://www.smugmug.com/photos/i-PrSV...-PrSVWNM-M.jpg



I posted a schematic a while back and it may help as a refresher:

http://www.smugmug.com/photos/i-9NjS...-9NjS58r-L.jpg

If this video exist .. maybe this will be seen on the judicial process ...
That can be required by some parties

Dozy, recalculation of F/D commands is covered in section 1.16.5.4, graph is on page 96 of the English edition. Values on graph are shown relative to aircraft symbol on ADI, positive is nose-up, negative nose-down.

Since the stakes are pretty high, I bet video will leak, as will CVR track, eventually.

There seems to be some departure from 'actual' parameters in that sim (as expected), as you describe a pitch down below the horizon with full THS and power. 447 managed to hold a high nose attitude in that situation. How much stick effort was required to hold the nose down against the thrust couple? Full forward, half?

I should clarify, to calm DW's fevoured brow, that I do KNOW the A/P would not 're-engage' - I was attempting to transfer the apparent FD 'logic' to the A/P as an example, in my opinion, of misguided design. After all, we do have enough examples of the AB A/P ACTUALLY causing a dramatic 'uncommanded' climb.

JD96
 

Call him/her back if you must and only when it's safe to do so, but:

1. You should have two type rated pilots at the controls who are capable of handling the aircraft, with or without the autopilot and with or without the captain being on the flight deck.

2. Calling the captain (via "phone" or similar) would take one of the pilots "out of the loop", albeit briefly, at a time when both pilots should be firmly focused on flying the aircraft.

3. In any event dependant on crew rest facilities, sleep patterns and umpteen other things you may not get the captain back onto the flight deck in a matter of seconds, it will possibly/probably take several minutes, which is the reason for point 1.


AF may have failed on at least one of those points.......

@PJ2
I share that concern. PF achieved a higher pitch attitude and stalled at lower speeds, mainly due to the fact that some ND SS and or reducing th NU SS input led to some g- reduction below 1.0 g without effectively reducing the pitch rate (see the Nz graphs of the respective time frame). TOGA application at that point (2nd StallWarning) and SS NU input forced a further increase of pitch without solving the speed and AOA problem. Thus when the stall finallly occured, pitch was a lot higher and speed was slower than what you describe in your sim exercise. With the nose lower like that gods g (gravity for the nonflyers) works a lot earlier than it would have for AF447, and the speed would be back in flying parameters a lot quicker.
Presently on leave on my iphone, therefore no further comments on the report.

Looking back at one of HN39's graphics on post stall recovery, I was reminded of an earlier debate over "super-cooled" water. The matter has not been cleared up by the BEA as their conclusions include:- The equatorial areas of the Atlantic and the western Pacific are the spawning grounds for sometimes large tropical revolving storms. The mechanisms associated with these maritime environments and their affect on the upper atmosphere is still not fully understood.

Enough of that pseudo transparency ...
Video animation of selected recorded data is basic stuff - Where is it ?
Full data are not available to all in the judicial process - Why is that ?

How naive you can be ...
There was no such thing as a CVR track leak, there was intentional release of deformed data in order to format the opinion :

According to Amedeo from Le Figaro, the AF447 captain had stated :
On ne va pas se laisser emmerder par des cunimbs

But now the last CVR transcript from the BEA is :
On n’a pas été emmerdé par les cunimb hein

Completely different meaning.
I can explain if you need help.

What connection between Le Figaro and Airbus ... I let you guess.

It appears also that Fabrice Amedeo is the same guy that took part in this other piece of propaganda where the Airbus Chief Pilot himself was generously interviewed from the impressive A380 flight deck.



2 options for the families of the AF447 pilots :
Either they accept what's on the menu right now and try their best to resume 'normal' life.
Either they fight the system in research of the full story ...

Yes indeed ...
But all this is weird for me (made sens .. or not)
Indeed this sentence (cunimb) is called when the plane is near Natal (the captain announces that he sees Natal just before pronouncing the sentence about cunimbs)
It does cunimbs they met on "en route" .. Rio-Natal ?
Curious to see a meteo graph on this route .... (I see nothing special in the TEMSI appendix 12)
Or the sens is ... "it was no cunimbs between Rio-Natal" ?
It's usual to have cunimbs between Rio and Natal ?

Oops, that's harsh....

Hi DozyWannabe,
Please see page 98 of the final report.
http://www.blackholes.org.uk/PP/FDs.jpg
Apart from the first 20 seconds or so when PF pulled the nose up whilst trying to keep the wings level in ALT LAW, subsequently the FDs were reasonably closely followed.

From time 02:10:50 until 02:11:40, whilst the stall warning is sounding, the FDs are commanding nose up from 7.5 degs up to 20 degs.
It would be far less confusing if the FDs had remained out of view until reselected by the pilots, or if they reappear automatically then park at zero pitch and wings level command until a pilot manually "sync"ed them to a mode they want.
If the FDs had remained locked onto ALT FL350, then they would have commanded a nose down.

I fail to see the logic of this design.

What Owain has theorized tells me that "doubtful" is a dubious vedict. He pointed out that from the data available, THS and elevators still had lift and authority. That supports the idea that recovery is possible, but what isn't certain is the altitude penalty paid to recover. (Though some estimates can be made). As note previously, they became, once stalled, two test pilots not well briefed on what they were testing ... :( and didn't know they were now in that role. :{
Confuses me as well. Makes sense to me that FDs are to be selected OFF when UAS is encountered. Back to flying attitude and power ... old school flying.
Arrrgggh!

For PJ2, since you've spent some time actually training AB3xx pilots:
You seem skeptical that the instrument scan of the two pilots on the flight deck at the time, heavily dependent upon the FD. Do you find very many pilots who go through your training sessions to be what I might call "FD" dependent? Do the training scenarios give you the latitude to find out where their scans are strong, and where they are weak?

It would appear that if one pilot is too heavily dependent on the FD, (PF in this case), that increases the problem the other pilot has in helping the PF get back into the game. (If I am making comments to him based on a scan of nose and wings, and he is following his FD, maybe we have a chance for confusion?)
If that "FD boresight" were in progress (guessing) then my understanding of some "why did they do that" is somewhat improved.
But only partly.
PNF appears to have maintained something of an instrument scan, in re altitude at least, as he was working (in vain) to correct PF's various errors. (Note: seems that he'd have helped more by making sure the memory item of disconnecting FD was accomplished).
Seems that the former is the more desirable.

Hello Lonewolf_50;

I'm on the road at the moment and need some time to compose a response, as well as respond to RetiredF4's observations.

The flight tests did not go beyond approach to stall

Certification requirements have varied over the years. One always needs to check the certification basis for a given aircraft and, even then, it is not guaranteed that a variation wasn't agreed upon during the program.

As I recall, the 744 may have traded on the basis of the earlier models.

As to what the OEM ought to investigate, over and above the Standards, might be another matter.

Makes me wonder why I did 3 QTG's recently that encompass 'the stall', as validated by IAS, buffet onset, stall AOA, and VSI break? QTG's come from flight test data.

There is a big difference between 'the stall' and a 'developed' (better word, aggravated) longer term condition.

as JT says. :ok:

Wondering why we are talking about "pitch rate" command versus "pitch" command. Can't command a rate directly, nor pitch itself, so to get a "pitch" value you must hold the stick fore or aft to get there, then adjust because the 'fuser is still trying to command a gee value - nominal one gee corrected for pitch attitude. Am I reading the manual wrong?

That being said, what in the hell are the pilots doing trying to follow the FD's? That's one theory out there on the streets, from what I have read.

I have always felt that a constant nose down attitude and some power to gradually fly out of the stall that the jet experienced would have been possible. 5 degrees, 10 degrees, no matter. Very glad a few have tried it in the sim or calculated the required maneuver. The jet does not have a pitch coefficient moment as we had in the Viper at 40 or 50 degrees AoA that prevents a nose down pitch moment using elevators and the THS.

If there is one thing about our primitive FBW system, it was a force transducer versus a position transducer for pitch and roll commands. So to hold a pitch attitude other than what Otto was trying to achieve to meet the gee command ( trim gee not a default of one gee), we had to exert "x" pounds. Releasing the stick commanded Otto to resume the trim command we could set. The 'bus does not allow the pilot to command a trim gee. So the pilot has to hold the stick fore or aft to get away from that one gee default trim condition. This feature was rather intuitive, despite lack of movement. To fight nose up trim we had to push forward, and vice versus.

This is still a great thread for discussion, and I am honored to be able to bitch and moan and contribute amongst all the commercial pilots here.

Lonewolf_50;In that context, "We've losts the speeds" doesn't appear to have had any relevance to "IAS doutuese", and as that item wasn't sought in the QRH (at least no mention made), the PNF became concerned with -Unfortunately, the whole episode can be placed into similar categories -

• Lack of training in high altitude UAS procedures
• Lack of training in high altitude handling
• Lack of CRM training.

Having said that, I and others have noted that a deviation would have resolved any issues, and we would have all continued on blissfully unaware that this accident was still waiting to happen.http://images.ibsrv.net/ibsrv/res/sr...ns/mpangel.gif

Until they passed 10,000 ft. there appears to have been no comment on their direction of travel... DOWN.

This was indicated ( and recorded). But it was not NOTICED by the pilots.

The presentation of this information might be improved.

( Seeing the hands of my old "Sensitive Altimeter" going round would have caught MY eye...)

( The lack of recognition of " STALL" has been discussed.)

Get the feeling that BEA doesn't want to dwell to much on areas which could be the basis of claims against Airbus.

Final Report section 1.18.1.1

JAR 25 Change 13

My emphasis. BEA were just stating the facts

Where does this come from? The BEA pulled Airbus up on the certification of the pitot tubes and the SB rather than AD approach with which they were to be replaced. Airbus was served with notice of intended action for manslaughter along with AF.

The BEA are *not* interested in protecting Airbus (nor should they be) and never have been.

OK. Doze, I’ve been a little too cryptic.

The stall recovery issue
A number of respected posters (such as PJ2 #127, #145, #165 and Owain Glyndwyr #148, #149 and there are others) have made a case for the possibility that the aircraft could have been recovered after the stall. BEA is in effect saying that it is only interested in what happened until departure from the flight envelope. It’s saying that while maybe or maybe not the aircraft would have been recoverable subsequently, the issue is, in effect irrelevant, because it didn’t cause the departure from the flight envelope.

What I want to say was the failure to attempt the stall recovery was another hole in the cheese, and logically shouldn’t be dismissed so easily by the BEA. Stall recovery surely is another part of the backup offered by having human pilots as the final resource. And just maybe, just possibly, the reason why the BEA is not interested is because it raises the question of the inverted stall warning. To have included a recommendation about the inverted stall warning could place Airbus Industrie under GREATER legal pressure.

The non-recognition of altitude loss
Similarly a couple of posters have expressed surprise that the BEA’s report does not particularly pick up the crew’s evident ignorance of the descent down to 10,000. Some have made the suggestion that they only noticed when one digit came off the digital display. Again, I am wondering, ever so gently, whether the failure to highlight this disadvantage of a digital altimeter versus an analogue one, was also influenced by a BEA desire to keep FURTHER pressure off Airbus Industrie.

Caveat
There are a lot of pages to read, and I haven’t had the time to absorb everything, but think the above thoughts, which seem to share something with JCJeant, might be worth injecting.

The BEA does address the issue with the stall warning inhibition.

The presentation also stated that while recovery at a later stage *may* have been possible if the crew had advanced handling training, the fact that none of them did rendered a recovery very unlikely.

Quite some time ago I stated a belief that the inhibition of the stall warning below a certain speed was very unlikely to be Airbus-specific. This is something I still believe quite strongly.

The recommendation is stated on page 214 of the Final Report: