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

AF 447 Thread No. 5

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

AF 447 Thread No. 5

Old 4th Aug 2011, 14:09
  #1521 (permalink)  
 
Join Date: Jun 2009
Location: Paris
Posts: 691
Hi HN39,
Originally Posted by HazelNuts39
With reference to this question and your reply, how do you explain (on page 112 of the french report, also on page 113) the cycling of 'Mach selecté' between 02:00:00 and 02:09:55? Is this Mach_1, Mach_2 or Mach_3?
None as it is unrelated to ADR Machs.
Mach "selecté" or "selected Mach" is entered manually by the pilot in "selected" mode or managed by the FMA in "managed mode". Of course, one needs an autopilot working in order to use the flight guidance.
Consequently, this graph value represents the "Mach" targeted currently selected/used before autopilot kicked off.
Last change was at 0209:58 when "managed" mode was switched to "selected" mode, and pilots set it at Mach 0.80 (they were reducing targeted Mach). So before this point, Mach variations should have been automatically managed.

Looking at graphs pages 112-113, Mach_selecté is the blue straight line at M.80 (barely visible on the second one). The green value should be ADR_Mach_1 (as there is no ISIS_Mach). The blue blocks should be periods of flight in managed_Mach.
Mach_selecté do not move at all during the event like altitude_selectée ; the last one was rather the active mode (vertical)... hence, maybe this "je te mets sur ATT" would rather be "je te mets sur ALT".
But I'm not sure, I need to verify this last point.
takata is offline  
Old 4th Aug 2011, 14:13
  #1522 (permalink)  
 
Join Date: Jun 2009
Location: VA, USA
Age: 54
Posts: 561
3holelover:
I can only imagine RudderRudderRat... but can it's stall really feel like an overspeed?
How about:

Can it's stall really SOUND like an overspeed?
GarageYears is offline  
Old 4th Aug 2011, 14:37
  #1523 (permalink)  
 
Join Date: Feb 2001
Location: right here inside my head
Age: 61
Posts: 178
"Can it's stall really SOUND like an overspeed? ... If it did, that could certainly explain the confusion on the part of both the PNF and the Captain, but the PF?.... I have difficulty understanding how he could have interpreted motions and responses to his inputs as an overspeed vs a stall. ...even if it did sound that way.
3holelover is offline  
Old 4th Aug 2011, 14:59
  #1524 (permalink)  
 
Join Date: Jun 2009
Location: VA, USA
Age: 54
Posts: 561
Page 108 English version:

Why is the Capt sidestick shown as INOP from 2:14:20?

Did the FO side priority switch get pressed?

Or am I misunderstanding?
GarageYears is offline  
Old 4th Aug 2011, 15:13
  #1525 (permalink)  
 
Join Date: Jun 2009
Location: Paris
Posts: 691
Originally Posted by GarageYears
Page 108 English version:
Why is the Capt sidestick shown as INOP from 2:14:20?
Did the FO side priority switch get pressed?
Or am I misunderstanding?
You are not mistaken at all. PF disabled PNF stick, using priority, during the last 10 seconds or so. Note that the last imputs from the PNF reverted also to nose down. The vast majority of his few imputs were ND anyway.
takata is offline  
Old 4th Aug 2011, 15:23
  #1526 (permalink)  
 
Join Date: Jul 2008
Location: Milwaukee WI
Age: 68
Posts: 35
"Can it's stall really SOUND like an overspeed? ... If it did, that could certainly explain the confusion on the part of both the PNF and the Captain, but the PF?.... I have difficulty understanding how he could have interpreted motions and responses to his inputs as an overspeed vs a stall. ...even if it did sound that way.
In an aircraft with unpowered controls, the pilot gets important tactile feedback re airspeed from how much force is required to move the controls. I doubt that an experienced Cessna pilot, for example, would mistake a stall for exceeding Vne in a light aircraft, even in the absence of external references or air data.

Obviously that's missing from this aircraft, although that begs the question of how control "feel" could be correctly added in the absence of reliable airspeed. Presumably Boeing has such an algorithm; whether or not it would have worked in this situation is beyond my ken.

But it does go to the problem of how far the pilot is separated from flying the airplane (as opposed to flying the computers) in 4th generation air transport aircraft.
bratschewurst is offline  
Old 4th Aug 2011, 15:36
  #1527 (permalink)  
 
Join Date: Jan 2008
Location: PARIS FRANCE
Age: 73
Posts: 53
Thank you Takata for your expanation about the flight directors.
I have been as all of us, appaled by the entire sequence of events.
I firmly beleive (and it has been said elsewhere) that the trim wheel and the THS staying at 13° until the crash is the key explanation to their not regaining control of the aircraft, but I keep wondering about the FD's. As you said, when reengaging, they take the altitude of the time as their new reference, and the horizontal bar will (with the plane diving down ) give an UP order to the pilot, who is trained to follow it.
They(the FDs) should have been switched off on both sides from the very moment when the loss of valid airspeed indications became clear to the pilots : emergency procedure on page 81.
Strangely (to me…they must have a good reason !) Airbus does not ask for the Fds to be switched off in their procedure « Stall Alarm » (page 64) when asking to set ,with TOGA, an attitude of 12.5 degrees (taking off) or 5 at altitude. In those cases, a different order by the horizontal bar could be misleading and unhelpful or confusing.
In fact they (the Fds) were not cut off, so the bars on the displays kept coming on, then off, and one wonders if the horizontal bar, which one is trained to following, was not one of the reasons the flying pilot kept pulling upwards ?
NARVAL is offline  
Old 4th Aug 2011, 15:45
  #1528 (permalink)  
 
Join Date: Jun 2009
Location: California
Age: 51
Posts: 203
Originally Posted by JD-EE
xcitation, might a 10 degree nose down attitude give them a rather distinct impression they are in a dive, even after asking for full nose up? He asked for nose up. Nothing happened. He asked for more nose up. Nothing happened. He kept this up to hauling back on the stick all it would go. THEN the plane goes nose DOWN 10 degrees over a few seconds?

I suppose if he'd ever been properly trained about stalls in A330s he might have recognized "the real thing" and not a warning. Perhaps if he'd had a real AoA report to look at he might have recognized "the real deal" and not a mere warning.

The plane quit doing what he asked for. It went nose down, and stay there depressingly long, when he had full up on the elevator.
I think at that point they were in a total state of confusion which would challenge even the best Ace. I have slightly modified my view on the overspeed.
Consider the start when AP disconnect the PF made aggressive nose up. After only 4 secs stall warning comes on. He reduces his nose up a bit however continues his ascent to FL375. By that point he is in a nightmare having traded his velocity energy for altitude potential energy.

Why did PF nose up initially? It was accidental. No verbal announcement, no increase in thrust etc. At that it appears there is no reason for a climb. Only AP disconnect and speeds becoming invalid. No indication of overspeed or stall until 4 seconds into his aggressive zoom climb manouver.
If you look at PF verbal transcript he sounds IMHO like he is in shock. It takes him a long time to even acknowledge the climb despite PNF nagging him 3 times to get back down. IMHO he is not aware of his own stick input from the onset. IMHO he is in the medical terminology "altered" i.e. mentally impared perhaps due to shock. Epinephrine surge can do strange things to the body, cold sweat, tense muscle, max heart rate and max respiration, blood draws to core and you can look pale. If you have been in this state you know how it feels and it's not good. By contrast the PNF appears to be much more collected and even frustrated by PF lack of response.
xcitation is offline  
Old 4th Aug 2011, 15:55
  #1529 (permalink)  
 
Join Date: Nov 2008
Location: Crew lounge
Posts: 87
At that it appears there is no reason for a climb
What about loosing instantly 300 feet when IAS quit (speed related altimeter correction) ?
GerardC is offline  
Old 4th Aug 2011, 16:09
  #1530 (permalink)  
bearfoil
Guest
 
Posts: n/a
With respect, PJ2 is precisely on point. However, the focus of the investigation seems to be closing on Human Factors, and rightly so, if the Data released is accurate.

If this is too quick, then my apologies. The CVR shows at its very basic a lack of underdstanding of the Flight Path, and its stimuli, both human, and mechanical, (Electrical?).

The Human dynamic will eventually be broken down by Psychologists, Behavioural, Physiological, and probably Psychiatric. This seemingly dysfunctional mess after loss of Autoflight will be dissected, and rightly so.

First, and not least, is the SS. The SS is only lacking in 'performance' relative to its Human counterpart. There is no need for each Pilot to see his counterpart's stick, period. Not if CRM, and a strictly no nonsense environment is followed. That is how I see it. Does PJ2 agree?
 
Old 4th Aug 2011, 16:23
  #1531 (permalink)  
 
Join Date: Oct 2009
Location: UK
Posts: 1,271
Hi PJ2,
The pilot is no more separated from his aircraft because it is wire-control than he is under hydraulic control.
I disagree.
AB FBW Normal Law feels completely different to any conventional aircraft I've flow. (VC9, 707, 737, TriStar, 747-400).
Occasionally, I am still tempted to "pick up the wing drop" on the approach in gusty conditions. Unfortunately the FBW decides to pick up the wing itself and the roll rate is now faster than I wanted. Hence PIO.
The PF was handling Roll Direct in ALT LAW which he obviously found very difficult (from the recorded inputs). Meanwhile Pitch ALT Law is unknowingly trimming out the elevator load with the PF's errors in back stick commands. (They have lost the natural speed stability of the aircraft).

No conventional aircraft feels anything like AB FBW in ALT LAW.

Hi Bear
There is no need for each Pilot to see his counterpart's stick
True if the trajectory is what you want. If it's not - then you have no idea if it's PF's inputs or CB activity throwing you around.
rudderrudderrat is offline  
Old 4th Aug 2011, 16:34
  #1532 (permalink)  
 
Join Date: Jun 2009
Location: florida
Age: 77
Posts: 1,217
control "feel" and mach versus stall

For the subsonic wing shape you can "feel" buffet and "stiffening" of the stick at the critical mach, whether via artificial control stick mechanisms or just "in the seat of your pants". You can also have control reversal for the ailerons, as the shock waves are messing around with the pressures that normally result in roll right for right stick and vice versa. Some planes have a pronounced nose-down "tuck", and the only way out is to reduce power and deploy spoilers or speed brakes.

The average "heavy pilot" without military experience rarely sees this. Additionally, the military planes usually have a greater structural margin per the design specification. So we could "push the envelope" in a way you would never try in a heavy. I only exceeded the Vne once, but the jet was robust and my biggest clue was that with a great amount of back stick the thing was not moving as much as I expected. Glanced at the airspeed and sure enough, I was 40 or 50 knots above the placard ) gotta admit that I was pulling off a heavily defended target and was getting hosed down big time by bad guys). You can't do that in a modern airliner, that is, gradually recover from the overspeed without ripping the wings off.

As Doze and others have repeatedly stated, most military and commercial heavies have not had the traditional control "feel" since the 50's. Move the hydraulic valve with the stick/yoke and the pressures went to the control actuators. Clever engineers used feedback devices to "stiffen" the controls at high "q" and to limit the control movement, while providing a sense of the Cessna/Champ/Cub "feel". Big deal.

Comparing control "feel" with the military FBW systems is not fair. We flew at AoA's and speeds and gees not applicable to the heavies. We flew by our own "feel", and the FBW confusers allowed us to ask the jet to do as much as it could without being too worried about stalling or exceeding the structural limits.

One way to "help" the pilots with the FBW system is to employ some of the PC game technology and have forces applied to the controls according to the flight conditions. In other words, use actual pressures detected upon the control surfaces themselves to be "felt" by the pilots. Can also use basic air data like dynamic pressure as a baseline - go fast, stiff stick, slow down and get sloppy and need more control displacement.

enuf war stories, and I continue to read the report again and again, as well as the neat posts here.
gums is offline  
Old 4th Aug 2011, 16:36
  #1533 (permalink)  
 
Join Date: Jun 2009
Location: VA, USA
Age: 54
Posts: 561
First, and not least, is the SS. The SS is only lacking in 'performance' relative to its Human counterpart. There is no need for each Pilot to see his counterpart's stick, period. Not if CRM, and a strictly no nonsense environment is followed. That is how I see it. Does PJ2 agree?
Bear
GarageYears is offline  
Old 4th Aug 2011, 16:40
  #1534 (permalink)  
bearfoil
Guest
 
Posts: n/a
rudderrat

How do. It is especially important in emergency situations that the "other stick" be unavailable, visually. It is a distraction, a potentially lethal one.

If things are bad enough in the cockpit to doubt your counterpart's actions, it is CRM that must rule, not a "second guessing" that is a detriment to one's own responsibilities. Everything up to the simultaneous (and Dangerous) actions was defensible, from a command standpoint. Imo.

I might be wrong, but at this point I believe it might have been Airbus' purpose to isolate the two sticks, and if so, I can see why? Only one can LEAD, and only one can CONTROL, at any given time. It is at the point where these separate functions start to "Blend", that Satan starts to grin?
 
Old 4th Aug 2011, 16:44
  #1535 (permalink)  
 
Join Date: Sep 2001
Location: Toronto
Posts: 2,252
There is no need for each Pilot to see his counterpart's stick, period. Not if CRM, and a strictly no nonsense environment is followed.
in an ideal piloting environment I could perhaps be persuaded.

But when humans are involved, errors will happen.

PNF may have clued in had he seen his ss sitting on the back stop.

My underlying concern is that when stuff goes wrong, good information is needed for the crew to put things right.

Among other factors, ss position and some red on the altitude/VSI display might have alerted PNF even though it seems he was too nice a guy.

PNF does not seem to have had any positive ideas about what needed to be done -- nor did CDB.
RatherBeFlying is offline  
Old 4th Aug 2011, 16:50
  #1536 (permalink)  
 
Join Date: Feb 2001
Location: right here inside my head
Age: 61
Posts: 178
Q regarding language:

Forgive my ignorance of French please, but is there different language used in French for "Pitching" up/down vs "climbing/descending"? The use of phrases such as "you're going up" and "you're going down", sounds (in the transcript) to me to have added to confusion.

Pj2;
There is a difference in FBW in the busses... The "feel" in your hand is simply spring pressure which won't change at all with speed, whereas in older, hydraulic and cable beasts, the control column and cables would actually change the force required to move them with speed. (through the "trim, feel and centering mechanisms") ...and I think (but can't seem to locate my books at the moment) still in the 777, while being the same FBW style electron flow output, the force req'd at the column is adjusted with speed... somehow (damn, where are those books!) force xducers(?) -or is it the backdrive actuators? (I can't remember, dangit. ...maybe I'm wrong about that )

In any case.... none of that changes how a human pilot would "feel" the changes in g acting on his body from his actions during a stall vs during an overspeed.
3holelover is offline  
Old 4th Aug 2011, 16:55
  #1537 (permalink)  
bearfoil
Guest
 
Posts: n/a
RatherBeFlying...

Do you see how "Two Pilots-One Stick" is a recipe for disaster? Take your own, only. It is not an instruction flight. "What's he Doing?" Is as Lethal as "What's it doing?"

This is not......"Francois, follow me through...."

3hl....

Again, in defense of IMC, feel is not necessarily desirable. It would seem, in a very real way, that PF may have allowed his own kinesthetics to drive his decisions? And to the contribution of LOC? Can we entertain a new descriptive? "Loss and MAINTENANCE of LOSS of control" ? Feedback is in some way an enemy of Instrument flight? If there is a total breakdown of systerms, well, ok, but that isn't happeneing here? If, in an emergency, a pilot has ONLY feel, well, ok, but otherwise there is a definite lack of engineered blend of seat of pants/FCS "Co-Operation"? More's the Pity? Will there need be hardwired ButtCheeks?

Last edited by bearfoil; 4th Aug 2011 at 17:24.
 
Old 4th Aug 2011, 17:01
  #1538 (permalink)  
 
Join Date: Jun 2009
Location: VA, USA
Age: 54
Posts: 561
I think (but can't seem to locate my books at the moment) still in the 777, while being the same FBW style electron flow output, the force req'd at the column is adjusted with speed.
Well, not sure that would help a dime once suffering UAS.... or stalled and doing 170knots at FL380... next we'll be suggesting alternate gains that come into play. It gets complicated doesn't it?
GarageYears is offline  
Old 4th Aug 2011, 17:13
  #1539 (permalink)  
 
Join Date: Mar 2006
Location: Choroni, sometimes
Posts: 1,975
There is a difference in FBW in the busses... The "feel" in your hand is simply spring pressure which won't change at all with speed, whereas in older, hydraulic and cable beasts, the control column and cables would actually change the force required to move them with speed.
I have mentioned that before.

This accident wouldn't have happened in e.g. 727 just of the feeling on the elevator.

AND

it would not have happened with a reliable stall warning, stickshaker/- pusher.


AIRBUS, where is your extraordinary sience?

A bungee for monkeys?
hetfield is offline  
Old 4th Aug 2011, 18:00
  #1540 (permalink)  
 
Join Date: Jul 2011
Location: Northern Hemisphere
Posts: 195
Altitude drop, roll, turbulence right before AP disconnect.

Originally Posted by xcitation View Post
...

Why did PF nose up initially? It was accidental. No verbal announcement, no increase in thrust etc. At that it appears there is no reason for a climb.
xcitation,

I've asked the question myself, and have contemplated several speculative explanatory scenarios, due to the limited official information at the time. I've posted one of the scenarios a week, or two, before the BEA report was made available. The posted scenario was that at AP disconnect, the AP was in the process of responding to a drop in altutude, and roll due to heavier turbulence, which was interrupted by the AP disconnect. So the PF reaction was to continue the AP's action with NU, and roll.

The BEA Report made available recently brings more information.
The graphs at the end of the document seem to provide several concurring elements to support the scenario I've considered:

1. there is a slight altitude drop on the Altitude graph, right before the AP disconnect - see page 108 of the English version of the report. The drop is very small, and it's hard to tell how much it is - maybe around 100 ft or so - as the graphs scale corresponds to 35000 ft or so, but Zooming IN, helps seeing it a bit better, within the limits of the graph resolution, and scale.
The altitude drop coincides with a reduction in thrust - see the N1 Actual Eng1% and Eng2 % graphs - and increased roll - see Roll graph.

2. the Normal Acceleration graph, which if I understand correctlty is a reflection of the Turbulence, shows increased up/down air turbulence activity right at that time.

So, I think, turbulence, drop in altitude (with reduced thrust) and roll, were the factors that the PF was responding too with Roll and NU.

I have a vague recollection of seeing posts tangentially, or directly mentioning the same thing, but can't recall exactly which ones.

Last edited by airtren; 4th Aug 2011 at 19:42.
airtren is offline  

Thread Tools
Search this Thread

Contact Us - Archive - Advertising - Cookie Policy - Privacy Statement - Terms of Service - Do Not Sell My Personal Information

Copyright © 2018 MH Sub I, LLC dba Internet Brands. All rights reserved. Use of this site indicates your consent to the Terms of Use.