Machinbird

Originally Posted by OK 465
So when the FD's returned, there was no visual indication of VSW, i.e. no real good visual indication of stall proximity or the inadvisability of pulling up again at this point.
If you will notice, no where during the pull up was there ever a mention of actual altitude by the crew, only that weird "go down" recommendation from PM.

It would seem that this crew was already so stressed, that neither of them actually read the altitude numbers, and they only noted the scrolling of the altitude tic marks on the PFD.

What could cause this level of stress? Other than it was a black and turbulent night, and neither of the pilots had experience with UAS or Alt2 at altitude (which is enough to make most pilots edgy), it would appear that the initial overcontrol of roll caused an adrenaline flood when the aircraft refused to calmly follow PF's piloting efforts.

Something really got to these guys. What else is unusual about the first 30 seconds after the AP dropped that would set them on edge? How many of the 30+ flights that didn't lose control experienced roll oscillation?

Yep!

OK465

Clandestino:

You are correct that the CAS (which BTW is referred to in the report as "calculated airspeed' not calibrated nor IAS, the term you use) is actually recovered at 2:10:34 at 223 knots.

However, the FD's become available at 2:10:47, and as you correctly state, CAS 'has already been recovered' and is at 216k with a theta of 5.6 degrees at 2:10:49 at ~FL375, eventually topping out at an 'apex' of FL379.

The stall warning occurs 4 seconds after the FD's return (2:10:51) at a '6 degree' AOA (FD bars steady not flashing).

We may be splitting hairs here, I'm an old guy and get distracted easily.

Entirely correct as far as vertical direction of flight. (I'm old and occasionally get tired also. )

I think there's a considerable effort going on to make it 'explicable', and that effort unfortunately may indeed fail.

(I understand that pilots need to know the state of their airplane to control it, and those that do will perform well no matter. As you say, sometimes they don't, but as you also say (if I may paraphrase without going back and directly quoting), "pilots 'learn' their aircraft from reading their manuals".)

CONF iture

Typical Clandestino stuff here :
I write how illusionary it is to pretend to maintain indefinitely a 5 deg pitch in altitude but you find 5 reasons to state I’m wrong but now you feel the need to precise it is only for the necessary time to get the QRH out …
How long do you think it takes at CLB thrust from FL350 200T before the AoA matches the 5 deg given to the pitch ?
I have no precise idea just a guess it would not take much but I’m certainly not ready to test and validate or not your theory.

No pprune in that, and not only Airbus.
And on the contrary, to not apply that procedure for low level has proved to be deadly, but then the procedure was not published yet …

HazelNuts39

HazelNuts39

If you allow me to 'split hairs' on terminology: The airspeed indicated on the display (corrected for instrument error, if any) is commonly referred to as IAS. If the ADIRU applies appropriate corrections for errors in pitot and static pressures, IAS is equal to CAS. If the pitots are blocked or at large AoA the indicated airspeed is not equal to CAS. The report refers to the captain's IAS as 'Vitesse conventionelle' or 'calculated airspeed' or 'CAS' when it means IAS.

Regarding the ACARS messages FLAG ON CAPT(F/O) PFD SPD LIM, BEA#1 says: "This message indicates the unavailability of the FMGEC's characteristic speed calculation function." There is no reference for Takata's assertion that it would be latched. IMO it could well be a temporary condition similar to the unavailability of the FD's - the SPD flag that occurs later is not latched (see IR#3 page 44). When the FD's return at 02:10:47, the F/O's airpeed (ISIS) is still 121 kt. The stall warning speed of 207 kt would have been off-scale on the F/O's PFD.

VGCM66

I am guessing no subtitles yet on this French reconstruction anywhere...

?

CONF iture

Interesting graph HN39
So in theory it takes about 3 minutes before both angles match.
Alpha SW is below Alpha Max.
Alpha SW is not even 1 deg above the 5 deg AoA.
Where would be Alpha Stall ?
What altitude would be reached when both angles match ?

BOAC

We are heading for 1300 posts now, and the oozlum burd is circling once again. SURELY wheter or not 5 degrees is 'right' or 'wrong' is immaterial in this accident? Had the crew maintained 5 degrees and crashed I would be interested!

OK465

1301 BOAC, go play with NG guys...

Quote:

There is no reference for Takata's assertion that it would be latched. IMO it could well be a temporary condition similar to the unavailability of the FD's - the SPD flag that occurs later is not latched…

HN39:

This is baffling to me. There's also no reference contradicting Takata. Having done some UAS work for sim lesson plans before I retired…

I can get a ‘latched’ SPD LIM flag with NO characteristic speed info on BOTH PFD’s by simply inputting a speed discrepancy of 50 knots, failing a single ADR in use (either 1 or 2, or 3 if switched), and then reviving the ADR and removing the speed discrepancy.

I will be in a ‘latched’ alternate law of some flavor (amber hash marks), both FD’s are back, but the absence of characteristic speed info is permanent, however the SPD LIM flag remains also???

I think I have ALT2”B” OCD.

HazelNuts39

True, but do you need three minutes to get the QRH?
The stall is not clearly defined. I would put it somewhere between 9.5 and 10.5 degrees.
37353 ft after three minutes.

CONF iture

No it should not, but don't forget you're here in the theory territory where everything is by the book, total control of the pitch, no turbulence, above average crew who will get the right data from the right page with no delay. In practice many distractions may interfere ...
It is bringing complexity where none is needed.

I restate my position :
IMO, the procedure for UAS in cruise phase should clearly mention to adopt the usual parameters for such phase of flight :

1. PF Maintain 2.5 degrees of pitch
2. PNF Set the thrust parameters as they are usually in CRZ
3. Wait for improvement

It is a suggestion that would clarify the situation ...

PJ2

"Do nothing" has been my position since July 30th 2009 in a response to Stepwilk, (Post #3990), and I have consistently maintained this position in the face of considerable opposition since.

HN39's work shows the effects of a slight increase in pitch, (2.5deg increase) and while slightly destabilizing, are benign in the short term. The point has always been, when in already-stable flight, why bother? It adds work, complexity as you say, and clearly most crews ignored the drill.

In my view, the opposition is primarily because the UAS drill and checklist is indeed confusing even to trained crews and was to this point that I have continually addressed myself, even re-designing the drill. Reverting to the simplest solution is also the safest one - do nothing - maintain pitch and power, remain calm and get out the books; even the video to which you provided a link now says this, yet the BEA maintains a position now at odds with this thinking from an ordinary A330 line pilot. What is to be done?

We're arguing the same point.

NeoFit

Undoubtedly, your position is right when UAS event occured in a "shinning blue sky tempest".

But would you please consider that at 02:10 UTC, the vessel was flying in ITCZ (also called F.I.T - Front Inter Tropical in another language).

AlphaZuluRomeo

+1 NeoFit

Also I remember one fellow poster (can't remember who precisely, sorry...), who said as a captain he sometimes asked his F/Os: "look at me, not the instruments, and tell me without cheating what is our pitch right now" (aircraft in cruise). IIRC many couldn't say precisely.

Perhaps that's one reason? (is it good or bad, I'm not sure)
5° & CLB, as a memory item. That is what you must know, for the time needed by the PNF to get the QRH at the right page, and crawl through the table for your aircraft current weight/alt.

I wonder if it would not be better (and simpler?) if any crew member could tell what are the current stabilized pitch & power settings, in order to maintain those should the A/P and/or A/THR drop off, for whatever reason. But as the shared experience of the above quoted captain teached me, this knowledge seems not (enough) widespread...

[edit] would like to add that, even if this question is interesting, I see it as unrelated to AF447 where the crew neither maintained standard cruise pitch&power, neither set 5° & CLB as per the published procedure if they had condidered the safe conduct of the flight was affected, which I don't think was the cas, at this point.
(and BTW yes, I'm aware of the "unsafy condithingy" brandished by totally un-agended people, based on a memo from an avionic manufacturer ; this memo was twofold IMO: to alert about potential danger and (in the mean time) to cover said manufacturer's assets)

bubbers44

"Are there pilots that follow the flight director blindly? Didn't happen when I was flying"

Sorry to follow my own quote but when did pilots not become pilots but follow the magenta line?

I always looked at the flight director but sometimes it gets programmed wrong as in one of my previous posts when my check airman got it so screwed up I just ignored it and flew my flight plan. He eventually caught up. No you do not follow the flight director blindly. You verify everything it is doing. To just follow it with no verification makes you not a real pilot, soory.

PJ2

Neofit;
Many of us here have flown in that kind of weather, (light to moderate chop or turbulence) and we know that the pitch will vary slightly as will the thrust...in most circumstances not much but it will vary. The requirement to "set" five degrees in the checklist may not be exactly the result in the aircraft but it will be close, while the other crew member gets out the QRH and the PF. Five degrees is about 2 to 2.5 degrees above normal cruise pitch and perhaps prevents the airplane from descending if the drill required, say, "3deg" and not 5, thus preventing a possible overspeed situation from developing - it guarantees that the aircraft will at least climb and not descend. So there are some reasonable reasons behind 5deg providing it isn't held too long.

My "do nothing" view means roughly the following: Once aware of the problem, (and it may take a few moments), call the abnormal so the other crew member is aware you're aware and is alert for next actions, wait for a few moments, observe the altitude, get the FMC onto the GPS page and see if the airplane is climbing or descending and adjust pitch accordingly. In the Airbus at least, "thrust lock", which occurs when the autothrust is disconnected involuntarily, will prevent any change in the power setting. If the thrust setting was lower than cruise power when the autothrust disconnected, that will be corrected with the QRH pitch-and-power tables. In the meantime, control is maintained even if the airplane wanders a bit. It certainly isn't going to lose speed fast if the thrust is a bit low. But like the training notes say, get on with the QRH checklist smartly. The checklist provides guidance on how to ensure that the correct pitch and power are set and it takes a while as the settings slowly affect airplane performance and it stabilizes with the changes. "Do nothing" doesn't mean literally just sit there!

AlphaZuluRomeo;
It's a good exercise. But not being able to tell what the pitch and power was does not mean that all is lost and "5deg" is the only answer.

To me, it means leave the airplane alone. By all means do what is obviously necessary to maintain stable flight, but having to change something in these circumstances would be extremely rare. The energy state of the airplane was fine using the pitch-power settings just seconds before the event. Unless sudden and unexpected entry occurs into turbulence that is severe to extreme, the pitch and power settings which existed just before the event are going to be very close to, if not exactly what is needed in terms of temporary pitch and power settings while the other pilot gets out the QRH. If the airplane starts to wander downhill, gently squeeze the stick back a tiny bit and wait for the input to take effect. Tiny movements on the stick is key, regardless of the method used.

I agree that five degrees is safe, providing the crew response is timely. But I think destabilizing the airplane when it was fine before, especially when almost no one is practised at high-altitude manual flight, requires some careful thought.

bubbers44

Why not just hold the pitch attitude of 2.5 and cruise power until airspeed returned. The altimiters were working fine. They would have been fine if they hadn't pulled up into the obvious stall. Experienced pilots would not have had a problem. Cheap new hire pilots will.

mm43

Probably worthwhile reminding the current punters that A33Zab provided a comprehensive spreadsheet layout of essential data back in AF447 Thread No.4 post#691. Those with a screen resolution width of 1024 pixels or less should view the image directly here.

We don't know how long the UAS condition would have remained if the aircraft had maintained FL350, but at least they would have eventually "flown" out of the ITCZ.

PJ2

mm43;

Thanks for the link. IIRC, the time for the airspeed to return to normal was painfully short...less than a minute? but by that time the airplane was established in the stall. In the original sim exercise, (I did two), it took about 4 minutes to slow down from 270kts to about 186kts, in level flight. At the time I never contemplated the possibility of a pitch-up and it is remarkable how quickly the energy is lost. That said, Owain Glyndwr has posted data which posit a successful recovery as low as 6000ft. He readily acknowledges that few if any transport line pilots would point the airplane down the necessary -10deg pitch to do so but the higher-level scenarios, which still require just a -10deg pitch attitude, recover successfully.

bubbers44;

Indeed, why not just hold 2.5 and make that the target in the drill?

For almost all conditions that would be fine and that is how I have argued - perhaps not "2.5", but the last-known. Clearly it is what all other crews did, to a greater or lesser degree.

What I can see out of this extended conversation is a subtle need to at least protect the airplane from going downhill. What perhaps wasn't anticipated was someone pitching up to 15deg in cruise and that has motivated my argument all along.

To me, and perhaps to most transport pilots, if it started going downhill a bit we'd just bring it back up, hopefully gently, (but that assumes that the altimeters are working!...but there's still the GPS.

Five degrees caters to all weight/altitude conditions, but I know from having looked for years and years at the ACMS data in cruise flight (A330-300) that the pitch doesn't vary much from 2.0 to 3.0 degrees even in the kind of turbulence AF447 was in. So keeping it at 2.5 and then working from there, fine-tuning it just in case that was a transient pitch attitude or power setting, is very doable.

I'm pretty sure we have to look beyond the "cheap new hire pilot" notion here. A great deal of commentary has already been offered by a number of contributors on the topic.

AlphaZuluRomeo

Certainly not, indeed. But 5°/CLB has apparently been deemed stable enough. Why not 2.5? Perhaps because you don't have an associated stable-in-all-conditions thrust setting?

I'm not advocating 5°/CLB should be applied at all time when UAS. Neither am I sure the authors of the procedure meant it that way.

If the safe conduct of the flight is affected => if above FL100 => 5°/CLB => QRH
If the safe conduct of the flight is not affected => maintain current => QRH

What means "safe conduct of the flight is affected" is, IMO, where the "problem" lies...

Once again, this is not related directly to AF447 as we have no clear indication they tried to follow any (part of) procedure.
That doesn't mean the procedure should not be improved. It should.