Go Back  PPRuNe Forums > Flight Deck Forums > Tech Log
Reload this Page >

AF 447 Thread no. 4

Tech Log The very best in practical technical discussion on the web

AF 447 Thread no. 4

Old 6th Jul 2011, 12:39
  #861 (permalink)  
 
Join Date: Aug 2009
Location: Texas
Age: 64
Posts: 7,190
Received 384 Likes on 238 Posts
BOAC:

To answer your pilotage question, Mr Occam suggests to me a scan breakdown (and PNF heads down in a display/ECAMS series of tasks) in the style of

artificial horizon/attitude indicator
+
altimeter/altitude strip

dropped out of the PF's scan in favor of something else (something to do with airspeed or secondary effects of that system having a hiccup).

Attention thus diverted, a small input in the stick over time became substantial THS deflection ... hence the 30 seconds of inexplicable nose up input, and beginnings of roll series.

This guess derived from a few years of testing two seat crews with a variety of malfunctions and seeing some focus (as a team) on the malfunciton, and others dividing tasks, one flying, one troubleshooting.

The CVR info might be helpful in understanding how close a guess this is to actual events.

@bearfoil: it does not appear that a/p is required to penetrate turbulence, though I think every AB pilot would prefer it. From what the AB pilots have shared in the various threads, one could (gasp) hand fly by using conventional scan pattern and gently setting and keeping an attitude (nose) and wings as level as possible as one rides out the bumps.

Last edited by Lonewolf_50; 6th Jul 2011 at 12:57.
Lonewolf_50 is offline  
Old 6th Jul 2011, 12:51
  #862 (permalink)  
 
Join Date: Jun 2009
Location: France
Age: 76
Posts: 196
Likes: 0
Received 1 Like on 1 Post
Originally Posted by CONF iture
Starting with the autotrim under manual flying ... who needs anything like it ?
To have to trim under manual flying has always been part of the pleasure and understanding of what it's all about piloting.
Raise your hand if you thought that an autotrim under manual flying was a necessity (Remember your time on the other types ...)
Auto-trim is necessary in order to implement a load factor demand pitch law. The aim is to get a constant path free stick behavior, i.e. a path independent of speed, thrust or configuration. I didn’t try it myself but it could well be that with a sidestick controller such a behavior is better and preferred compared to a more conventional “trim to speed” system. However, you need to add protections at both ends of the longitudinal flight envelope if you want the system certified for commercial operations, hence the high speed and high angle of attack protections.

Problems appear at less than Normal Law due to protections removed but then basic certification requirements don’t apply. You are left with stall warning only in ALT 2 law.

AF447 has been autotrimmed to the max up position under STALL WRN ... something is awfully wrong here !
At least, leave to the pilots that responsabillity if such was really their wish ...
100% agreed.

There is something paradoxical with this new “space aged” flight control system. Certainly the airplane is easy to hand-fly in normal mode, perhaps a bit more “touchy” in alternate mode where default gains are used and again more sensitive in direct mode. Easiness has been used as a commercial argument to customers, suggesting savings in training. We, old pilots, having used “dinosaur” flight control systems regularly disconnected automation, even at altitude, when we felt we could do better manually, which was often the case. At the time, we were never really in need to “update” our skills in this respect and today the background experience is still there.
Now think about the current situation of young pilots. They are just given a taste of direct law in the sim, not necessarily at altitude, a few minutes once every 4 or 5 training session and the rest of their hand-flying life is Easy Law, follow-the-FD-it-is!!! Yet, they are supposed to have a decent level of flying skills when they are abruptly passed a less than optimal system. The paradox is that making the task easier erodes the necessary skills i.e. does not systematically result in enhanced safety.
DJ77 is offline  
Old 6th Jul 2011, 12:53
  #863 (permalink)  
 
Join Date: Jun 2009
Location: Bedford, UK
Age: 70
Posts: 1,319
Received 24 Likes on 13 Posts
Roll control was maintained which indicates attitude data OK.
Mr Optimistic is offline  
Old 6th Jul 2011, 13:02
  #864 (permalink)  
 
Join Date: Aug 2009
Location: Texas
Age: 64
Posts: 7,190
Received 384 Likes on 238 Posts
Mr O.

Ideally, roll control being maintained would mean to me the PF remained wings level when that was the called for flight regime: maintain course and speed to get to next navigation point.

You are probably right, but the BEA reports roll excursions. One need not fault the attitude indicating system for that ... other explanations have been offered that do not point to a failure, for which various data to date have lacked one labeled "inertial reference/AI" failure.
Lonewolf_50 is offline  
Old 6th Jul 2011, 14:02
  #865 (permalink)  
bearfoil
Guest
 
Posts: n/a
A/P is "required" by SOP via Airbus, not in some mechanical requirement. Leaving the controls alone is not unwise in turbulence, where "maintaining" trumps "maneuvering" yes?

There is one other interface in the mix that gets little attention, that being auto v. mechanical?

The Sidestick would seem to be an attempt at "feel" retention. As such, perhaps jcjeant's touchscreen would be a more honest approach? The THS is powered by a jackscrew via hydraulic pressure, not the speediest method of response. Trim implies "relief", not control.

TRIM accomplishes a stability in pressure, independent of deflection. This is a valuable resource when "maneuvering", for it communicates the a/c demands and wishes to the flyer. Does the FCS have need of this "feel"?

Such a valuable resource to eliminate from the endeavor! As to certification, why would a THS be required to plant at max deflection and stay there whilst a recovery is attempted? And if not required for certification, the a/c must be recoverable with elevators only, I would think. So why is the HS even allowed such an excursion?

This thread is generally peopled by those who would grant the system a pass on criticism in the LOC/event arena. Yet the pilots are referred to as "Why in the World...?"

Either the PF knew not the position of the THS, or he disregarded it.

Again, the answer, for me anyway is to be found in the seconds leading to a/p disconnect, and those few that elapsed at the initial climb.

The PF completely ignored the TRIM WHEEL? He commanded a ridiculous climb at low energy? He held backstick seven miles down into the SEA?

Absolutely not. From the outset, BEA has been manipulating public opinion re: this accident in the worst of ways, impugning the pilots, to include those who were open to continued UAS events. Does anyone doubt the motive behind the "Note"? Airbus got, and then sent, the message, n'est-ce pas?

bear, It is impossible to slander without abusing the TRUTH.
 
Old 6th Jul 2011, 14:41
  #866 (permalink)  
 
Join Date: Aug 2009
Location: Texas
Age: 64
Posts: 7,190
Received 384 Likes on 238 Posts
bear, in re words and terms, and a little too much "all or nothing" in your reply:

A/P is "required" by SOP via Airbus, not in some mechanical requirement.
Leaving the controls alone is not unwise in turbulence, where "maintaining" trumps "maneuvering" yes?
No. You make an assumption there, that not having A/P available means that controls are left alone. I respectfully disagree, having flown in turbulence more than once, with and without an autopilot function (though not on an Airbus, of course).

There is one other interface in the mix that gets little attention, that being auto v. mechanical? The Sidestick would seem to be an attempt at "feel" retention.
Not as described in a myriad of posts by pilots who fly Airbus airliners, no. You and I may have gleaned different messages from same posts. Perhaps I am the one misreading.
As such, perhaps jcjeant's touchscreen would be a more honest approach?
No idea. Haven't tried to fly a plane by touch screen.
The THS is powered by a jackscrew via hydraulic pressure, not the speediest method of response.
It seems to work pretty well, doesn't it? I don't know if speed of response is the critical metric, given that rate damping and control damping are features common to various FBW aircraft, as well as some AFCS and EFCS suites in non-FBW aircraft.

I also recall that the elevator is generally involved in THS lift adjustment under all laws but direct. The way I read the block diagram, when you use trim wheels, the elevators don't do much, so you are moving the THS, but I may be misreading that block diagram. That would explain why those who have used it have to be very careful with trim wheel control of pitch, as the fine pitch trimming, or fine pitch control, via elevator adjustment isn't to hand in that case.

If I have misread the block diagram, my apologies.

Beyond that, any semantic argument about what trim should mean isn't as important as what you learn in type training, and follow on training, about how your flight control system and auto pilot functions work, in full up modes and in degraded modes.

The first fleet helicopter I flew had a different trim system, and different AFCS, than the second fleet helicopter I flew some years later. Just because the trim systems didn't work identically doesn't mean that I could not trim the aircraft, didn't understand how to, or was somehow at a loss.

You learn to fly your aircraft.
It's what good pilots do.
It's what professional pilots do.
If for one reason or another there is a systemic obstacle to such systems familiarity, that must be addressed.

EDIT: I may be dating myself. AFCS "automatic flight control system" and EFCS "electronic flight control system" (now DFCS??) which abbreviations may no longer be in general usage.

Last edited by Lonewolf_50; 6th Jul 2011 at 14:57.
Lonewolf_50 is offline  
Old 6th Jul 2011, 15:19
  #867 (permalink)  
bearfoil
Guest
 
Posts: n/a
Lonewolf 50

This thread is impossibly iterative, and requires, imo, a lot of leeway in particular opinions. For example, SmilinEd said ages ago that a/p in turbulence may need a look see. It may not keep up with its own mission, relative to trend, and rapidity of surface deflections required, which may not be appropriate. Airbus strongly recommends using the a/p at all times, and each operator generally demands the same of each flight crew.

Is there an "EASY!" switch on otto? A mode that will tolerate serious bumps and puckering ride in patience, while anticipating some relief soon?

AF447 was in the bump city, by BEA portrayal, via Captain duBois. Yet the crew remained in a/p, as expected, and when it quit, there was the mundane handover, "I have the controls". No big.

The THS' jackscrew works fine, it is appropriate for a large airliner that needs gentle and deliberate deflections. That it took sixty seconds to travel from three degrees to thirteen seems appropriate for its size, that of a large dancefloor used to manipulate 150 tons of mass.

Without assigning any rating to each post, relative to experience and knowledge, my goal is to be critical of the failures here in man and machine, and learning as much as possible in the process. Please continue to correct where necessary,
 
Old 6th Jul 2011, 15:24
  #868 (permalink)  
 
Join Date: Jan 2005
Location: W of 30W
Posts: 1,916
Likes: 0
Received 0 Likes on 0 Posts
Originally Posted by DJ77
Auto-trim is necessary in order to implement a load factor demand pitch law.
But, IMO, direct law would have done just fine whenever AP is OFF.
Anyone here complaining about flying the 320 in direct law ?


Originally Posted by DozyWannabe
But you can't infer from that information that the PNF was unaware of what was going on because he wasn't getting any feel through his sidestick. Perhaps he felt that the other pilot was doing what was necessary, perhaps he didn't feel confident enough to intervene (see Birgenair again), perhaps he got absorbed in the "PNF" role to the extent that he was more focused on the ECAM messages than his PFD or - and I know this will be controversial - perhaps he'd fallen asleep and was woken by the alarms. All of which are possible, but we don't know - we'll find out with the arrival of the report.
From my personal experience, I just don't know what my partner is doing with his sidestick. At best, and it's not necessarily better, I'm guessing ...

When the captain came back, he probably took the middle jumpseat, from where, you hardly see any of the sidesticks ... How can he evaluate the flight control inputs ?

And forget about falling asleep at that time. Did the captain fall asleep as well after entering the cockpit ? Is it why the BEA could not publish any comment from him ...
CONF iture is offline  
Old 6th Jul 2011, 15:41
  #869 (permalink)  
 
Join Date: Jan 2008
Location: Blighty (Nth. Downs)
Age: 77
Posts: 2,107
Received 4 Likes on 4 Posts
Initial pitch-up

bearfoil, BOAC, and others,
At risk of unnecessary repetition, the BEA Interim Report #2 (Page 48) shows that, with a sudden drop in the perceived dynamic pressure (blocked pitots) the ADR will infer a drop of altitude of the order of 300ft. So, while maintaining FL350, the altimeters might have dropped suddenly to indicate 34700ft erroneously. It's conceivable that this might have been the reason for the PF's initial pitch-up command.

In their most recent Update, the BEA says that ASI1 dropped from 275kt to 60kt, but I have to admit that at the moment I don't know which altimeter reading that would have affected. Perhaps one of our ADR students will shed light?
Chris Scott is offline  
Old 6th Jul 2011, 15:44
  #870 (permalink)  
 
Join Date: Jan 2008
Location: Blighty (Nth. Downs)
Age: 77
Posts: 2,107
Received 4 Likes on 4 Posts
ASI strip on PFDs

PJ2,
Thanks for the PFD and centre-pedestal pictures yesterday. At what stage would you say the ASI strips on each PFD might have ceased to indicate the current IAS/CAS calculated by the ADRs, and what would cause that? Must admit to being confused about this.

Are you hinting that its disappearance on PFD2 may have provoked the PF's remark at 02:12:02:
"I don't have any more indications" ?
Chris Scott is offline  
Old 6th Jul 2011, 15:51
  #871 (permalink)  
 
Join Date: Jun 2011
Location: Devonshire
Age: 96
Posts: 297
Likes: 0
Received 0 Likes on 0 Posts
We have seen a number of threads speculating on "how to stall" an AB 330.
Does anyone have access to a suitable simulator to see if or how the Captain might have recovered the aircraft, when he arrived on the flight deck ?
(AB are said not to have any data about a full stall - or it is not available.) A few attempts with our own or somebodies "Made-up" data might show - something instructive, even if it needed several attempts.
If there is some success whilst Pitot and Static are used, then try again, this time without. What is there to lose - for us ?
A couple of 747s have glided successfully ( if not intentionally).
Wasn't Cully's an AB...
Linktrained is offline  
Old 6th Jul 2011, 15:52
  #872 (permalink)  
bearfoil
Guest
 
Posts: n/a
Chris Scott, would AS1 be LHS? Pitot #1 is directly out the Captain's glass, and above Pitot #3? Straightforward, but I could not say for sure.
 
Old 6th Jul 2011, 16:01
  #873 (permalink)  
 
Join Date: Jan 2008
Location: Blighty (Nth. Downs)
Age: 77
Posts: 2,107
Received 4 Likes on 4 Posts
bearfoil,
Yes, that's my understanding. But I'm not sure that ADR1 displays both CAS and altitude on PFD1 (the L/H PFD). Now, where was that schematic that someone kindly posted an eon back?
Chris Scott is offline  
Old 6th Jul 2011, 16:02
  #874 (permalink)  
 
Join Date: Jun 2009
Location: Paris
Posts: 691
Likes: 0
Received 0 Likes on 0 Posts
Hi bear,
Originally Posted by bearfoil
why would a THS be required to plant at max deflection and stay there whilst a recovery is attempted?
The more important questions we are left with are:
- Was a stall recovery really attempted? If, yes, when did it take place?
So far, I really can't say anything about it as we'll need the full cockpit conversations and many other data in order to settle this matter.

Same remark about the PNF "late" acknowledgement of "speed lost, Alternate law". Was it ever acknowledged by the PF?
What the PNF was trying to say to the PF then? Maybe his message was: "hey! no speed, alternate, no protections... what are you doing? we are climbing!"

Here, the question of each F/O aircraft experience seems critical to me. If the PF, like I think was the most senior with five times A330 more hours than the junior one, the PNF could have no choice left but to call for the Captain in support rather than taking-over. This may explain his early emergency calls.

All we can suspect, so far, is that the PF was attempting to do something as we have those stick imputs recorded. But what exactly was he really attempting to do? For me, it looks like he was trying to fly alpha-max as, at each STALL alarm, he reacted the same way by applying back stick imputs. It is called tunnel vision, sticking with the wrong choice by thinking that it is the good one.
We'll see.
takata is offline  
Old 6th Jul 2011, 16:12
  #875 (permalink)  
 
Join Date: Aug 2009
Location: Texas
Age: 64
Posts: 7,190
Received 384 Likes on 238 Posts
All we can suspect, so far, is that the PF was attempting to do something as we have those stick imputs recorded. But what exactly was he really attempting to do? For me, it looks like he was trying to fly alpha-max as, at each STALL alarm, he reacted the same way by applying back stick imputs. It is called tunnel vision, sticking with the wrong choice by thinking that it is the good one.
I wonder why you call it tunnel vision instead of a conditioned response to procedural training, pattern mapping, or learned behavior.

PJ2 has time and again pointed out that a malfunction need not be treated as an emergency, nor turned into one, and I agree with that with the following caveat: once stall warning horn goes off, perhaps the crew went into emergency mode rather than malfunction mode. What was the last training session (that this crew participated in) involving stall warning horns dedicated to?

Recency and proficiency are training factors that must be considered.

(Once again, the pitch and power chorus will ask
'where was that conditioned response when warranted?'
in a lyrical, multipart harmony.)

takata:

When's the last time you taught someone how to do something?

How did you do it?
Lonewolf_50 is offline  
Old 6th Jul 2011, 16:14
  #876 (permalink)  
 
Join Date: Jan 2005
Location: France
Posts: 2,315
Likes: 0
Received 0 Likes on 0 Posts
Originally Posted by Linktrained
We have seen a number of threads speculating on "how to stall" an AB 330.
Does anyone have access to a suitable simulator to see if or how the Captain might have recovered the aircraft, when he arrived on the flight deck ?
With over 800 posts even here you're excused....
The answers, already discussed earlier, are
a) no, there is no "suitable" simulator, since there are not enough aerodynamic data in the "full stall" domain to program a simulator accurately. Simulators are programmed for this domain by extrapolation, and "best guesses". And yes, it IS being looked at by work groups, etc. ;
b) we have nowhere near enough info about what happened in the AF447 cockpit to be able to write a "scenario" to try out what could have been done, even if we had that "suitable" simulator.
ChristiaanJ is offline  
Old 6th Jul 2011, 16:22
  #877 (permalink)  
 
Join Date: Jun 2009
Location: Paris
Posts: 691
Likes: 0
Received 0 Likes on 0 Posts
Hi Lonewolf,
Originally Posted by Lonewolf_50
I wonder why you call it tunnel vision instead of a conditioned response to procedural training, pattern mapping, or learned behavior.
I respectfully disagree with such statement.
So far, there is no such "conditioned response" measured behavior, including several other AF crews, as those many previous cases won't show any pattern of similar behaviors.
The more I can say, as there is not such a pattern during similar events, is that a specific training was lacking in most recorded case.
takata is offline  
Old 6th Jul 2011, 16:37
  #878 (permalink)  
PJ2
 
Join Date: Mar 2003
Location: BC
Age: 76
Posts: 2,484
Received 0 Likes on 0 Posts
Chris;
But I'm not sure that ADR1 displays both CAS and altitude on PFD1 (the L/H PFD). Now, where was that schematic that someone kindly posted an eon back?
This the one?




The photographs of the PFD without speed and altitude information have caused confusion for some - please accept my apologies.

In the post I did indicate that the altitude information would have remained, that the THR LK wouldn't have been displayed and that the FPV would not have been available, that there would have been a red "FD" flag and it was just the speed information that was missing, (removed if there is there is no data to display). I wanted to convey a sense of the display rather than portray it exactly as it was which of course isn't possible until we have more information.

The image below of the PFD is perhaps a more realistic representation of what the PFD may have looked like at the start of the climb. I caution everyone that this image is a composite, manipulated, Photoshopped picture and is just a guess and NOT related to anything other than my imagination. It is intended only to represent what the PFD may have looked like given what we know from the BEA Update.





Lonewolf_50 - yes, agree...My view has always been that the UAS event was not an emergency but it slowly turned into one the moment the pitch-up began and of course a stall is a full-blown emergency requiring an immediate response.

Notwithstanding notions that the aircraft was pitched down and the "correction" was to pull back on the stick, (doubtful I think), pitching-up and leaving "stable" flight meant that the last known-and-established variables, "pitch-and-power", were themselves intentionally varied, leaving no way back to stable, level flight, not, at least, without immediately setting pitch and power to the QRH settings and waiting (a very long time) for the aircraft to re-stabilize.

Others have said, and I agree, that even at the apogee, a full nose-down-stick-held-to-the-stops-power-off response, perhaps even without rolling the THS forward to something around zero degrees, would likely have led to recovery if not outright avoidance of the stall, (someone else said the aircraft was still at about 215kts and not stalled at the apogee).

Last edited by PJ2; 6th Jul 2011 at 17:21.
PJ2 is offline  
Old 6th Jul 2011, 16:42
  #879 (permalink)  
 
Join Date: Jan 2005
Location: France
Posts: 2,315
Likes: 0
Received 0 Likes on 0 Posts
Total red herring question probably, from a Concorde AFCS ancient.

When the air data part of an ADIRU goes 'belly-up' (in this case because of UAS) does that mean that the possibly still valid data such as altitude and vertical speed (static pressure data only) are also thrown out with the bathwater, or are there separate 'flags' (F/Ws) for separate data such as IAS, Mach, TAT, Alt and VS ?

IIRC, on Concorde all we (i.e., the AFCS) got from the ADCs was an ADC (1 or 2) failure warning. Period.
The AP and/or AT would drop out from internal command/monitor discrepancies between the ADC data, and in most cases also on an ADC F/W.
Since the INS was a totally separate system, it didn't really enter into the air data "discussion".

For clarity I should maybe add, that, for various (mostly irrelevant) reasons, between Concorde (the first FBW airliner....) and A320, the separate ADS (air data system) and inertial reference system (IRS) ended up inside the same system/box, now better known as the ADIRU.

Last edited by ChristiaanJ; 6th Jul 2011 at 17:03.
ChristiaanJ is offline  
Old 6th Jul 2011, 17:00
  #880 (permalink)  
bearfoil
Guest
 
Posts: n/a
takata

howdy sir. I think the CVR reports thus: (2:10:16?)

PF: "So, the speeds are lost."

PNF: "Alternate Law"

So, No mystery twixt the two at this point? If "simultaneous" mutual acknowledgment of the problem, a strong case exists for this being the first 'arrival' of Alternate Law (subject to a 1-2 second P(N)F recognition?)

If this voice is one only pilot, still, the recognition is established as well after autopilot drop.
 

Thread Tools
Search this Thread

Contact Us - Archive - Advertising - Cookie Policy - Privacy Statement - Terms of Service

Copyright © 2024 MH Sub I, LLC dba Internet Brands. All rights reserved. Use of this site indicates your consent to the Terms of Use.