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

AF447 Thread No. 3

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

AF447 Thread No. 3

Old 30th May 2011, 15:27
  #721 (permalink)  
 
Join Date: Oct 2006
Location: East Sussex
Age: 86
Posts: 277
Likes: 0
Received 0 Likes on 0 Posts
Machinbird. You must have flown the Gnat!
pontifex is offline  
Old 30th May 2011, 15:43
  #722 (permalink)  
 
Join Date: Jul 2009
Location: Australia
Posts: 21
Likes: 0
Received 0 Likes on 0 Posts
skirtz - first post was a stellar post, I couldn't have said it better.

It does appear to me that these pilots may've been deceived naturally believing that their flight path was not that of a stall, not as serious as it was, until the RA warnings kicked in. Severe weather, at night, at altitude, and having instrumentation (their only interface to maintaining safe flight) now questioning the validity of their concious interpretation of the situation.

If the turbulance was severe; after loss of reliable air data, possibility the initial stall accelerations could have been masked by turbulance?

No AoA, no redundancy-by-design for air data - clearly not a software fault. The first fault was invalid continuous air data from the sensors due to significant meteorological conditions (icing). You could also argue flying into such weather was a contributing factor and I would agree. Treat the disease, or treat it's symptoms.

P.S. Trolls who have an axe to grind regarding software and/or programmers, cough cough they know who they are, I suggest they walk the walk and start removing all software from their daily life, start now with their own computer... If they like I can send a program that can assist.
davionics is offline  
Old 30th May 2011, 16:04
  #723 (permalink)  
 
Join Date: Jul 2009
Location: IAH
Posts: 58
Likes: 0
Received 0 Likes on 0 Posts
JD-EE
promani, (and others) this foaming at the mouth to find somebody to blame, particularly a deceased somebody, is unseemly

I have just read your post. Sorry for the late reply. I challenge you to find anywhere where I have insinuated that the cause was 'pilot error'. All I have done is to extract published information and queried it. So do not accuse me of finger pointing at the flight crew. Maybe you should ask yourself if the BEA is finger pointing, as they have said that they found no fault with the aircraft, and then they published the Note.
Sometimes I get the impression that some of you may think that you are in a private members club, outsiders forbidden. But you are not.
Who was it who said to search near the LKP? Moi and someone else. Who was it who asked if the 3rd FO presence was significant? Moi. (Notice I am NOT finger pointing).
I hope this clarifies everything. Thank you.
promani is offline  
Old 30th May 2011, 16:12
  #724 (permalink)  
 
Join Date: May 2010
Location: Boston
Age: 72
Posts: 443
Likes: 0
Received 0 Likes on 0 Posts
Sensor_validation:
BEAs statement is a careful selection of established facts which must have been checked out carefully by legal opinion - so I think I think it is worth looking carefully at the words used - albeit in the English translation - does the French read the same? Maybe reading in too much?


Quote:
the Captain woke the second co-pilot
Suggesting he was asleep, and could have been for a couple of hours - but within 5 minutes he was in front of the controls as PNF.
I totally agree that the BEA "note" has more information than may appear if simply read as a document.

The BEA probably has a very good idea of the facts -and- causes of the acciedent chain at this point. Given the legally influenced approach (good post many pages back I can't find at the moment) i would propose this:

A: Every fact in the document has a stong consensus to be true.

B: Every given fact will match up with the final report as one of the "holes in the cheese"

C: Ommited data such as what happened after the dual control input are still in analysis as to either certainty or relevance.

Given the above here are 2 observations:

1: I suspect the narrative was ended at the point (~10K feet) where it is not yet known if the aircraft was recoverable.

2: On the cpt waking the relief pilot, not only was he asleep he was deeply asleep. This can be inferred from "waking" the pilot as opposed to calling him. (I welcome native French speakers to commnet, this could be a translation artifact.)

On waking from a deep sleep the recovery time to full attention is much longer than if one wakes naturally at a normal point in the sleep cycle.

<< releavant personal story, skip if skimming posts>>

The only car accident I have had in 40+ years of driving occured while I was driving my brother in law's hot air ballon cross country.

The trailer had no (effective anyway) brakes, I had been sleeping in the back of the small pickup truck and we had just swapped drivers.

Time was early morning, light rain had just started. I was going down the ramp from the rest area and concentrating on picking a gap to merge with the moderate traffic. The rig did not have much acceleration so this was fairly important.

The big buick in front of me stopped at the bottom of the ramp.

The brakes were clearly not going to stop the rig in time so I aimed squarely at the back. A swerve could well have rolled the rig even if there was room.

Fortunatly no injuries and once we pried the fan out of the radiator the pickup was driveable.

The point of this is that while there were a large number of factors, brakes, weather and the idiot 16 year old driver who stopped suddenly it was still very much my fault and I strongly suspect the narrow concentration (picking my merge) was a result of not being fully awake.

BTW: I think the poor kid got the worst of it, the entire time we were waiting for police to finish reports etc his father was yelling at him that he had warned him not to do that in the past...

Last edited by MurphyWasRight; 30th May 2011 at 16:23. Reason: Typos, clarity.
MurphyWasRight is offline  
Old 30th May 2011, 16:27
  #725 (permalink)  
 
Join Date: Jun 2009
Location: Bedford, UK
Age: 70
Posts: 1,329
Received 15 Likes on 9 Posts
Was a successful recovery possible with the a/c “en ligne de vol” at 35000 ft after the climb and descent back to cruise level, within structural load limits, given an 'optimal' response ? If it was, can anyone say at this stage how much time margin they had for delay? Presume some 40 odd degrees of nose down pitch would be necessary at some stage to unstall the wings. Hell of a scarey thing to do when you are already descending at those rates.
Mr Optimistic is offline  
Old 30th May 2011, 16:33
  #726 (permalink)  
 
Join Date: Nov 2001
Location: Los Angeles
Posts: 26
Likes: 0
Received 0 Likes on 0 Posts
With regard to the "deep stall" idea here, if the PF applied ANU command all the way to the surface, why is it necessary to invoke a hypothetical (and unlikely) unrecoverable deep stall? Between the stabilizer trim setting and the up elevator, isn't that enough? In fact, when the PF did momentarily start to put the nose down, the airplane responded, but then the misbegotten stall warning went off and discouraged him from continuing. Has there ever been a proven instance of an unrecoverable deep stall in a conventionally-configured (ie low tail) transport?
Garrison is offline  
Old 30th May 2011, 16:35
  #727 (permalink)  
 
Join Date: May 2011
Location: Vancouver Canada
Posts: 2
Likes: 0
Received 0 Likes on 0 Posts

Look at the time line and the position of the a/c. There should be no doubt as to what the crew were facing in the midst of a 55,000+ cell.

The point in question is how did they find themselves there in the first place. What sort of professional judgement created this kind of situational awareness? This has all the early appearances of 2 pilots in way over their heads from the time the Captain left the cockpit.

AVIATE, NAVIGATE, COMMUNICATE....there appears to have been problems in all 3 areas...... with devastating consequences
PDBK is offline  
Old 30th May 2011, 16:52
  #728 (permalink)  
 
Join Date: Jun 2009
Location: Paris
Posts: 691
Likes: 0
Received 0 Likes on 0 Posts
Hi Machinbird,
Originally Posted by Machinbird
Originally Posted by Takata
Nonetheless, despite having a THS trim at +13 deg and being already in a full stall situation, it seems that those imputs are effective by reducing the AoA... hence, elevators were still effectives, don't you think?
Absolutely yes, but they ran out of authority against that great big THS.
I respectfully disagree with the second part of the sentence. So far, nothing in the BEA note is telling us that "they ran out of authority against" the THS at this point (0212:17). Such fact could only be established once we'll know how much nose down imput was applied. From the note, it looks like it was effective by reducing their AoA but it may also be the result of N1 reduction to 55%. In fact, you seem to agree with me later:
Originally Posted by Machinbird
We don't yet know how long the nose down input was held, do we? Only that it brought on the return of the stall warning. The stall warning may have stopped the crew's nose down input, or maybe it didn't.
Considering the low energy state they were in (airspeed was down to 185 kias at 0210:51), I'm prone to believe that those elevators imputs were still effective despite this THS trim... up to the point that their AoA would become an issue for recovering any control at all.

Anyway, my main point was to first focus on what information they lacked/missed in order to understand their situation until this point of the upset as I'm pretty convinced that this sequence as described at 0212:17 doesn't prove any understanding of their real situation.

As you said, this interuption of the "stall warning" until the very moment where the PF "started" to do something effective against the stall could have added to the overall confusion. But, IMO, it doesn't explain what signal didn't reach the PF brain (or induced him into doing the wrong thing) during the previous two minutes. This is what I'm focusing on, now, as this THS training seems a secondary issue when, at first, the pilot seem to be unable to aknowledge correctly what was primarily going on: a stall.

Last edited by takata; 30th May 2011 at 17:03.
takata is offline  
Old 30th May 2011, 17:11
  #729 (permalink)  
 
Join Date: Jul 2009
Location: UK
Posts: 134
Likes: 0
Received 0 Likes on 0 Posts
Originally Posted by paull
Rather than worrying about the state of the PNF, if we are going to read things literally then 'THE PILOT woke the Co-Pilot', not "had him woken" so it is a good job this did not all happen at 1:55am with just one pilot, one asleep and the captain out of the cockpit.

It seem illogical to criticise how awake someone is, when there are probably numerous occasions when you are happy to accept just one in the cockpit.
Interesting this is not yet clearly stated - I assumed the 2nd co-pilot was sleeping in the cockpit in the "Third occupant seat", where he would have been at take-off - or just an intercom call away. There was a post, probably by PJ2, in an older thread detailing the likely seat moves, and who was controlling from where - I don't think there's anything new here - although the tabloids seem to have it wrong.

I still feel that the BEA have started to paint a picture hinting at the crew interactions by their publication of selective facts.

Sullenberger always praises his co-pilot when talking about the Hudson ditching, the Singapore A380 engine incident was worked through by all (5?) in the cockpit. The first officer in the recent fatal air India crash missed his opportunity to abort his sleepy but senior captain's landing.
sensor_validation is offline  
Old 30th May 2011, 17:11
  #730 (permalink)  
bearfoil
Guest
 
Posts: n/a
takata

Greetings. Trying to avoid the shotgun approach evident here, I'd like to continue focusing on the autopilot, a/thr drop.

"... (like the RTLU was at 0210:05, remember? we told you that it was its last valid airspeed before ADRs faults)..."


This time point is gotten loaded up with data that cannot have happened all at once. The loss of auto pilot and autothrust is claimed at this time stamp, and also the PF's control inputs to correct for low right wing, and what appears a reaction to ND or excess speed, or other.

The pilots noted reaction to loss of speeds and "alternate law" is indexed at 11 seconds later?

I am unable, but will some kind soul post an accurate log?

In this sequence is all we need to know re: upset and unrecovered descent.

From some seconds prior to loss of autopilot, to TOC, the data waits to explain the crash to us all. It most certainly carries the preponderance of understanding, imho.

Until this is divulged, there will be an incessant and uninformed tug of war between the pilot's side, and the others. This is not tolerable, and BEA is struggling with a model of (for) mission that is beyond antiquated. For what it's worth, I don't find fault without mercy with their approach, they appear to be clueless, though trying their best to do some things they are ill equipped to do.

A traveller who does not check weather, and have a passing acquaintance with his/her carrier's business model is a trusting soul. It would upset the mercantile no end, but an informed public ultimately would be good for business.

When Mr. Joyce rode that 380 back from LA, I'd like to have known his heart rate, and how many times he glanced at the TRENTs. Flying is an exquisitely wonderful way to get around. A touch of discretion is not ill advised.


ventus 45

"In both cases, the HS is immersed in / operating in turbulent flow which radically reduces it's control authority."[/B] (and may exaggerate it)

What you describe is wake turbulence, certainly. This is (can be) lethal, and not just to the extent of Roll/Pitch. (use one's imagination?)

@Flight Safety

"...Right Way Up, about as likely as 3 engines failing at the exact same time in the exact same way, not very likely..."

All three pitots are placed carefully to be in similar air. All three are the same marque, and "the icing" may have been benign (conformal) in its presentation.

BA038. Both TRENT7's rolled off at the 'same time', for the exact same reasons.....caution advised.

Last edited by bearfoil; 30th May 2011 at 17:43.
 
Old 30th May 2011, 17:29
  #731 (permalink)  
 
Join Date: Jun 2009
Location: Paris
Posts: 691
Likes: 0
Received 0 Likes on 0 Posts
Originally Posted by BEA note
The airplane approached the ORARO point. It was flying at flight level 350 and at Mach 0.82 and the pitch attitude was about 2.5 degrees. The weight and balance of the airplane were around 205 tonnes and 29% respectively. Autopilot 2 and auto-thrust were engaged.
I'm not sure if the MAC noted 29% in the note is not a typo for "around 39%". When looking at the "target CG" as it should be computed by the fuel-computer (FCMC) for a 205 tonnes ground weight aircraft, it will give a MAC around 37.75% (target). Hence, it should vary in this part of the flight between 37.25% and 37.75% (while maximum certified AFT should be above 39.75% with a 2% safe margin).


takata is offline  
Old 30th May 2011, 17:34
  #732 (permalink)  
 
Join Date: Oct 2009
Location: berlin
Posts: 152
Likes: 0
Received 0 Likes on 0 Posts
ventus45 This research seems to be suggesting that a "conventional empenage mounted HS" at grossly excessive angles of attack (say greater than 30 degrees) can become locked into a "super stall" by being immersed in a fuselage generated wake consisting of two unsteady counter rotating vorticies.

the HS is immersed in / operating in turbulent flow which radically reduces it's control authority.
the A330 is a wide-body, that will generate even bigger vortices



is stall protektion the reason why an other widebody has so much fins between the main-wing-fin and the side-fin?
grity is offline  
Old 30th May 2011, 17:45
  #733 (permalink)  
bearfoil
Guest
 
Posts: n/a
The Shark was designed by Burt Rutan?
 
Old 30th May 2011, 17:52
  #734 (permalink)  
 
Join Date: Jul 2009
Location: UK
Posts: 134
Likes: 0
Received 0 Likes on 0 Posts
Don't shark's tend to dive when they fall out of the sky?



Headington, Oxford
sensor_validation is offline  
Old 30th May 2011, 17:53
  #735 (permalink)  
 
Join Date: Nov 2001
Location: Los Angeles
Posts: 26
Likes: 0
Received 0 Likes on 0 Posts
Why is there all this talk about being "locked in" a "super stall"? If the stick had been held forward and the trim wheel rolled AND and the plane had failed to recover, this would make sense; but none of that happened.
Garrison is offline  
Old 30th May 2011, 17:54
  #736 (permalink)  
 
Join Date: Dec 2002
Location: Where the Quaboag River flows, USA
Age: 71
Posts: 3,387
Received 3 Likes on 3 Posts
Garrison

I cannot think of one, but transport category stall accidents are extremely rare and the most recent one (B737 THY at AMS) was so low that recovery was not possible, once recognized.

That said, what I find impressive is the A330 here maintained a completely stalled condition, with some rolling and slow yaw component, for more than 3 minutes. I suspect the high thrust on the underslung engines combined with the stab position made the elevators only marginally effective, even if used more persistently than it appears happened.

While we know the attitude presentations were accurate, does anyone have an speculation on moonlighting and/or visual conditions?
galaxy flyer is offline  
Old 30th May 2011, 17:58
  #737 (permalink)  
bearfoil
Guest
 
Posts: n/a
Garrison

One thinks you are pedantic here? What you quote with exception to nomenclature happened, and for want of proper verbiage, (will you supply some?) it is merely an error of device, not reality. Further, your critique seems based on assumptions the a/c could somehow have recovered? It may be so, but your stuck with 'assumption'?
 
Old 30th May 2011, 18:06
  #738 (permalink)  
 
Join Date: Jul 2009
Location: Not far from a big Lake
Age: 81
Posts: 1,461
Likes: 0
Received 0 Likes on 0 Posts
Takata
I agree I have overstated the BEA proven case so far with regard to control authority, however, if you had ever flown an aircraft like the A-4 Skyhawk (which was a wonderful and fun flying machine by the way) you would understand the limitations of nose down elevator against a large THS in (nearly) full nose up trim. As I stated earlier, the elevator is more like a trim tab in comparison of actual relative control authority.

The reason the crew did not understand what they were looking at was this:

This illustration come from Davies excellent book: "Handling the Big Jets."

The crew is trying to figure out if they are seeing a pinball show from hell caused by crazy computers or the real thing.

The attitude is probably fairly stable in pitch by the time the Captain re-enters the cockpit and the likely lateral imbalance problem probably seems not much worse than it was initially.

The only big clue should have been the rapidly unwinding altimeter. BTW, I don't remember at the moment but:

The standby altimeter. Is it the old fashioned kind with the pointer that goes around for every 1000 feet lost, or is it just another digital display with rapidly changing digits? Under stress, the old fashioned kind is far better for drawing attention to itself. Sometimes digital is not better for absorbing information.

Are you able to answer the question about having to fly the aircraft laterally while staying off the pitch axis that I asked earlier?

I get the feeling that this aircraft needed almost continuous lateral input to fly wings level, and that the forces required on the stick were higher than just fingertip level. With the RH seat flying and a heavy right wing, the tendency might be to cup the stick with the palm, and in this mode, it would be very easy to make aft stick inputs as well.

Why right wing heavy? Who knows (Maybe a change in rudder trim as they switched rudder control modes), but if you will note, all turns made after the initial loss of control were to the right.
Machinbird is online now  
Old 30th May 2011, 18:08
  #739 (permalink)  
 
Join Date: Jun 2009
Location: florida
Age: 81
Posts: 1,608
Received 52 Likes on 15 Posts
c.g./THS effectiveness/elevator effectiveness/ etc

I can't conclude that the 'bus THS and the elevator were "blanked" or rendered ineffective to to the downwash or vortices from the main wing without seeing more data. Flight test results would be supreme, but who wants to take a big jet beyond the design limits to prove the point?

Further, and sadly, Airbus has assured us it is almost impossible to "lose control" of that jet. That assertion seems to permeate the crew training and design corrective actions by Airbus following certain incidents.

I'll let that aspect of my observations recede and prolly not raise it again for a long time.

to Garrison, Tk, et al:

I initially thought a "deep stall" such as we had in the Viper could explain the flight path of the jet ( not crew actions). But looking at many charts and manuals causes me to reconsider that theory.

Tk now raises the issue of the actual c.g. of the jet at onset of the episode. I had already looked at the graph Tk has just posted ( many moons ago), and was initially surprised that the aft c.g. values were as large as they were. And we now question the actual c.g. values used by the cosmic flight control computers?

All that being said, I will throw my vote in with 'bird in that the elevator control authority is drastically reduced when the THS is near max limits. It is even less effective at very high AoA and aft c.g.

I also go with Garrison on the point that recovery would/might have been possible using available system inputs and control surface authority. Might have taken a long and constant stick input and maybe touching the forbidden trim wheel, but possible. With that being said, it unfair to require the crew to be NASA "golden arms" or Chuck Yeager clones.
gums is offline  
Old 30th May 2011, 18:13
  #740 (permalink)  
 
Join Date: Jan 2005
Location: France
Posts: 2,315
Likes: 0
Received 0 Likes on 0 Posts
Originally Posted by NigelOnDraft
You might just see a fair performance in the sim (and is the sim model correct? unlikely) in a "set" training package - to see that replicated 2-3 years later, at night, circadian low, with no warning I think unlikely to see a good % of well handled recoveries
You've got a point. I like it....
Call Captain X at three a.m. for a sim session, "now". Then throw a few nasties at him.

I think the unions would go berserk.... even if it might produce some "interesting" results.

Maybe an idea for research?

Last edited by ChristiaanJ; 30th May 2011 at 22:23. Reason: bolded "circadian low" in the quote, to better explain my answer.
ChristiaanJ is offline  

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.