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

AF447

Old 8th Jul 2009, 22:06
  #3341 (permalink)  
 
Join Date: Jul 2009
Location: France - mostly
Age: 84
Posts: 1,682
Likes: 0
Received 0 Likes on 0 Posts
Translation issues

Re: #3353 (Squawk ident)

Quote:
Or should we have to start to endlessly discussion on how to translate "level attitude" in French?
/Unquote

That would be pointless, because the french version IS probably a translation from the english version which, as far as I can see, is (almost) a carbon copy of the US airworthiness regulation FAR Part 25.

Regards,
HN39
HazelNuts39 is offline  
Old 8th Jul 2009, 22:14
  #3342 (permalink)  
 
Join Date: Jun 2009
Location: Colorado
Posts: 41
Likes: 0
Received 0 Likes on 0 Posts
einhrvr,

After the calcs guys did earlier on the tail's horizontal velocity based on a 4sec/rotation ballpark, I too would wonder about having/not having evidence of sideways movement on the fin at impact. That's why I've been backing away from a standard spin or a flat-spin with high rotation rate.

One thing the 'failed slightly forward' represents to me was either slightly nose-down and/or with a slight forward horizontal travel at impact.
These seem to conflict with a flat-spin for an A330s as far as I know(I HAVE seen A330s fly over in the middle of the night going between Cancun and Vancouver BC).

Edit: One thing the 'slightly twisted to the left' represents to me was some form of rotation, I'll admit a spin isn't the only way for that to happen.

If it were a low-altitude stall, it would really need to kill forward travel. Can be done. Not my favorite scenario.

Last edited by ttcse; 9th Jul 2009 at 00:23.
ttcse is offline  
Old 8th Jul 2009, 22:34
  #3343 (permalink)  
 
Join Date: Jun 2009
Location: Colorado
Posts: 41
Likes: 0
Received 0 Likes on 0 Posts
JD, not unless the flight continued from 02:10 and beyond 02:14 for some more minutes. The zone of east drift challenges and has the potential to replace the backwards drift projections.

Edit:If east drift replaced backwards drift projections, it might make more (or perfect) sense if af447 came down further west. Then the currents initially took debris eastward and then shifted to north-northeastward.

Otherwise, you would need to address and pull in all the posibilities for explaining why acars stopped at 02:14 but the plane continued on for some time.

Last edited by ttcse; 9th Jul 2009 at 00:28.
ttcse is offline  
Old 8th Jul 2009, 22:45
  #3344 (permalink)  
 
Join Date: Jun 2002
Location: Orlando, FL, USA
Posts: 108
Likes: 0
Received 0 Likes on 0 Posts
Einhverfir,

Well, there are a couple problems with that, still. Note, I am not discounting entirely the possibility of a spin. I am just saying it is not consistent with the findings of the report.
Agreed. It seems also that your arguments against a flat spin [not a normal spin] are based entirely on the credence you give to the report. You have accepted mentally that the fin separated from the rest of the structure in the manner that they surmised -forward on impact. Even they only surmised - you are taking it as fact.

Where would the Board get the time to make a thorough scientific analysis of the recovered VS and where would they physically do that? On an island off the coast of Brasil? On the deck of a ship? Did they fly it back to France before the report?

If you had a spin from altitude, I would expect the plane to hit at a near vertical vector with very little forward/backward acceleration. Also unless the rotation was very, very slow, the sideways inertial forces would still be a major (rather than minor) component to the tail failure. However the report found that the rudder failed forward with a slight left-ward twisting element (it is possible that this element could have been sufficient to throw it clear of the fuselage).

At 4 secs per rotation in a flat spin from altitude, I would expect sideways and forces to be a substantial fraction of the vertical forces and the forward deceleration to be negligible. Hence the VS would have failed sideways (perpendicular from the fuselage) with perhaps some other slight vector added representing what little was left of the original fight vector. Remember that the VS is close to the extremity of the spin, so the centrifugal forces are very high there.
All of that is quite logical but again, you are basing it on the VS being attached to the rest of the airplane during the flat spin. Is there some reason why you have decided that it could not have separated before that flat spin began and thus caused the flat spin? Is it impossible that the forces on it after the upset could have caused its rearmost attach point to fail first? What if the horizontal stabilizer failed downwards and tore loose from it then causing it to come off? There are so many possibilities.

On the other hand I am suggesting that there would not be a flat spin if the VS remained attached. What would cause the flat spin would be the failure of the entire tail cone [aft of the pressure bulkhead], or of the VS alone. In turn, either one of those would likely cause the engines to leave their mounts. What was then remaining would be the fuselage and the wings. They would then fall near vertically, somewhat like a leaf, with very little rotation, until impact.

On the other hand, in a low altitude stall, we would expect strong forward and vertical forces, along with some minor twisting as the airframe hits the water not quite level (due to wind). This seems most compatible with the findings to date. Not that the findings are infallible at this juncture.
How do you get the airplane from where it was in cruise, high altitude, to the point at which this low altitude stall occurs? Does it just glide or fly on down smoothly? If so, why; and wouldn't that take about 90 miles? Are the engines running in your scenario or are they flamed out? What exactly do you mean by low altitude - how low?.

If it did not stall all the way down from FL350, why would it suddenly stall at low altitude? If it made a last minute pull up into an accelerated stall - how did the pilots manage to see the ocean and judge when the should pull up? Was it diving at very high speed before this happened?

There are just too many IF's for me.
surplus1 is offline  
Old 8th Jul 2009, 23:45
  #3345 (permalink)  
 
Join Date: Mar 2006
Location: England
Posts: 997
Likes: 0
Received 6 Likes on 3 Posts
surplus1, {post 3358} you appear to have misunderstood the difference between the flight control system (FBW) and the autoflight system. The airbus follows a similar concept as conventional aircraft, except as stated that the technology used in flight control system is FBW.
See A330 Flight Controls.

With the autopilot engaged, if the aircraft’s reaction to a gust in turbulence is beyond certain boundaries (pitch +25, -13, roll 45, speed, AOA, well before the constraining ‘g’ limits of the FBW system), the autopilot disengages (with alerts) and hands a fully functioning aircraft (controls) to the crew, there is no change in control law. Your assumption of a change of law is incorrect.
In manual flight the control system limits the manoeuvre within the constraining boundaries.
See A330 Auto Flight.

Changes in flight control law are generally the result of computational system failure and/or input failure, this includes electrical failures.
In this accident the comparator functions of the ADC inputs appear to trigger the conditions which would cause a change in control law.
See A330 Flight Laws.

The limits of the normal law (protections) are no more likely to be exceeded when flying in any other law, than they might be exceeded in any conventional aircraft, excepting gross differences in the underlying aerodynamic designs. ‘IF’ turbulence caused an upset then it could have happen to any aircraft, and even exceeding the stated limits, it does not imply that control will be lost.

Your subsequent speculation about airframe failure is just too extreme, both from the scarce evidence available, and extensive research and previous turbulence encounters, which form the basis of certification standards.
This does not mean that ‘it’ cannot happen; just that it is most unlikely. In order to strengthen your speculation, then the reported weather would have to be correlated with other aircraft upset and failure events, and the assumptions within the certification requirements.
Although the reported storm cells appear to have been very large, they were not as large as some that have been recorded. There is no evidence that the aircraft entered a Cb.

Re post 3368, you might find your tenuous argument easier if you described the aircraft motion after a structural failure as ‘spin like’; who knows what the actual flight path modes would be without a fin or other large pieces of structure. Please do not cite the A300 accident; there are not enough similarities to be valid.
PEI_3721 is offline  
Old 8th Jul 2009, 23:48
  #3346 (permalink)  
 
Join Date: Jun 2009
Location: Paname
Age: 55
Posts: 21
Likes: 0
Received 0 Likes on 0 Posts
The man, the machine and the training

I'd like to go back a few days ago, to a previous exchange between singpilot and PJ2:
Sorry, I think Occams' Razor applies here.
Yep, and it doesn't paint a pretty picture, frankly.
So there's everything to learn, and nothing to learn...depends upon who's doing the learning. There is the pitot issue and the response, and the broader issues of cascading failures possibly overwhelming crews.
I probably have too much imagination, because when I red this, I felt like I was just a kid, watching two grown-ups discussing serious matters in a hushed voice...So I wondered what they were talking about, and I came up with a theory.
Maybe the computers have become so good at flying the planes that manual flight is now discouraged by SOPs and that training time is now reduced...
Maybe pilots nowadays don't accumulate as much manual flight time and hands-on practice as they used to...
And maybe sometimes it might make a difference between life and death...

The way I see it, if there is a lesson to draw from AF447, it could be that computers will never replace pilots, and that software test procedures will never replace manual flight training hours...
Anyway, since all this is derived from a couple of posts exchanged between grown-ups, it's likely that I just have too much imagination...

Last edited by JuggleDan; 9th Jul 2009 at 00:01. Reason: Fixed typo
JuggleDan is offline  
Old 8th Jul 2009, 23:53
  #3347 (permalink)  
 
Join Date: Mar 2006
Location: England
Posts: 997
Likes: 0
Received 6 Likes on 3 Posts
JuggleDan, see http://www.pprune.org/safety-crm-qa-...ml#post5041334 and other posts in the thread.
PEI_3721 is offline  
Old 8th Jul 2009, 23:55
  #3348 (permalink)  
 
Join Date: Jun 2009
Location: Chelan, WA
Age: 48
Posts: 43
Likes: 0
Received 0 Likes on 0 Posts
Surplus1:

Sorry if I wasn't clear on this case.

My view right now is that I lean somewhat against spin hypotheses and somewhat towards the idea that the VS separated during impact (there are a number of reasons it might not have impacted the fuselage in this case). I reserve the right to change my mind as more information comes in. In general I assume that BEA has somewhat more info than we have from just the photos. I assume the visual inspections in the reports are better than the photos we have been seeing. It seems that in the case of the VS, they had to modify the examination based on photos, so this area is more weak than I would like.

One issue here also is that the argument that the plane hit in a flat spin is, in my estimation, based in some cases on the BEA report too. So I figure if we can clear up points about what is consistent with the report, we can move on to further questions, such as whether the report is accurate.

Some elements of the BEA report are very important in this regard and some of these are important and well put together. For example the discussion of uniform compression damage across various sections of the plane struck me as quite odd. Other elements are far less certain at this point.

If we give the report credence, as to the direction of failure on the VS, then I think we would also have to assert that it separated on impact, because it would seem highly unlikely that the VS failed forward unless this was the result of inertial forces (again inertial forces would have been down + forward in a low altitude stall, and down + sideways in a flat spin).

Certainly some theories can be solidly dismissed at this time. For example, the idea that the plane was in a flat spin at 0210Z as this would not be consistent with impact even 4 minutes later.

Other theories, like entering a flat spin at 0214Z are harder to dismiss. At this point, they seem inconsistent with the BEA report (while a stall at low altitude with significant forward momentum is not). But as you say, the BEA report could be wrong about key things like the tail.
einhverfr is offline  
Old 9th Jul 2009, 00:00
  #3349 (permalink)  
 
Join Date: Jun 2009
Location: Paname
Age: 55
Posts: 21
Likes: 0
Received 0 Likes on 0 Posts
Problem with autopsy results?

The reason why the autopsy results are still missing may be explained by the French newspaper Le Monde (To Be Confirmed, obviously):
Selon un haut magistrat français qui n'a pas souhaité être identifié, les autorités brésiliennes refusent de transmettre leurs conclusions portant sur 51 corps retrouvés dans l'océan Atlantique car elles n'ont pas encore ouvert d'enquête judiciaire mais seulement une procédure administrative.
In other words (thx Google Translate): Brazilian authorities refuse to forward to the French the autopsy results because they have opened no penal inquiry yet, the ongoing procedure only being an administrative/civil inquiry.

Obviously, this news is dedicated to Will Fraser
JuggleDan is offline  
Old 9th Jul 2009, 00:11
  #3350 (permalink)  
 
Join Date: Jun 2009
Location: various areas
Posts: 225
Likes: 0
Received 0 Likes on 0 Posts
PJ2 Thank you once again.
Under normal (design) circumstances, all AC and DC busses remain powered; there is some load shedding, (galleys) in loss of generation.

As you say, the ECAM would display the failures and the required actions. FLR/WRN category messages would be generated and recorded by the FIDS, (Fault Isolation and Detection System), collated by the CMS (Central Maintenance System) and sent to the ACARS for maintenance action. In short, they would be a part of any transmitted messages.
I take from that, it is almost impossible that a power plant failure occurred at or before 02h14z, therefore AC1 and ACARS were most likely operating and able to report any anomalies in engine operation (whilst the ACARS aerial was pointed skyward at least). What of ACARS (and AC1) after 02h14 if the 'controlled flight' theory were at all possible?

Which brings me neatly to another issue (amongst the many possible variables). The NW FO comment as provided by Greybeard
Here's what the FO wrote in an email the day after:

... Entering the cloud tops we experienced just light to moderate turbulence. (The winds were around 30kts at altitude.) After about 15 sec. we encountered moderate rain. We thought it odd to have rain streaming up the windshield at this altitude and the sound of the plane getting pelted like an aluminum garage door. It got very warm and humid in the cockpit all of a sudden. Five seconds later the Captains, First Officers, and standby airspeed indicators rolled back to 60kts. The auto pilot and auto throttles disengaged. The Master Warning and Master Caution flashed, and the sounds of chirps and clicks letting us know these things were happening.
MY bolding

Assuming no 'recollection' error (event sequence and/or timing), the AP and AT disengaged AFTER the speed indications had rolled back to 60kts, which based on exhaustive information provided here (including the ADR thinking time before DISAGREE), means the cascade of faults (in the AF447 case) might suggest the last recorded rudder travel limit prior to ALTN would have been a very low speed index (high travel available)? I followed that discussion very carefully, and it is still not clear in my mind if the above were possible/probable!? Which brings me back to your comment
In cruise, an engine failure would require an immediate but small application of rudder, (gently!). As with all high-altitude flight one must be careful with the use of controls because thin air provides so little damping effect. Providing rudder is used judiciously, roll-due-to-yaw is not a problem.
You see my line of query i.e. due speed related upset prior to 02h14, or assem thrust related rudder application after 0214z following upset!


Along the same lines:-

surplus1
All of that is quite logical but again, you are basing it on the VS being attached to the rest of the airplane during the flat spin. Is there some reason why you have decided that it could not have separated before that flat spin began and thus caused the flat spin? Is it impossible that the forces on it after the upset could have caused its rearmost attach point to fail first? What if the horizontal stabilizer failed downwards and tore loose from it then causing it to come off? There are so many possibilities.
On the other hand I am suggesting that there would not be a flat spin if the VS remained attached. What would cause the flat spin would be the failure of the entire tail cone [aft of the pressure bulkhead], or of the VS alone. In turn, either one of those would likely cause the engines to leave their mounts. What was then remaining would be the fuselage and the wings. They would then fall near vertically, somewhat like a leaf, with very little rotation, until impact.
ARFOR is offline  
Old 9th Jul 2009, 00:23
  #3351 (permalink)  
 
Join Date: Jun 2009
Location: California
Age: 55
Posts: 203
Likes: 0
Received 0 Likes on 0 Posts
xcitation, WHEN would the passengers or crew have put on life jackets?
Not sure you are asking the right person the right question. Please re-read my posts. My apologies if it was that badly written that a lawyer couldn't discern my meaning.
If I interpret your question to be when would the AF447 cabin crew have instructed the PAX to don jackets. I think the SOP for this equipment would be for instructions to happen after ditching in the sea.

If the plane was in extreme turbulence please describe the sequence of events needed to don life vests and then tell me if that is feasible in a violently buffeting airplane.

JD-EE
Was that a rhetorical question (equally )
If not then for the first part I defer to cabin crew qualified on similar equipment; for the second part I would say it would be feesible but very unlikely especially if you are working with the flat spin hypothesis.
xcitation is offline  
Old 9th Jul 2009, 00:31
  #3352 (permalink)  
 
Join Date: Jun 2009
Location: Colorado
Posts: 41
Likes: 0
Received 0 Likes on 0 Posts
einhrvfr
Other theories, like entering a flat spin at 0214Z are harder to dismiss. At this point, they seem inconsistent with the BEA report
Better still, upset and perhaps flat spin entry at 02:13:2x when the acars was lost for about 15-20 secs, as pointed out in the report.
ttcse is offline  
Old 9th Jul 2009, 00:41
  #3353 (permalink)  
 
Join Date: May 2002
Location: Canada
Posts: 155
Likes: 0
Received 0 Likes on 0 Posts
The ability to take the Airbus into Alternate or Direct Law resides within the design of the autoflight system but is a profoundly non-standard approach with this design and is entirely within test-pilot territory.
Considering the scenario a Qantas crew recently found themselves in, I suspect there are some new Murphy's Law test-pilot members who have filed away the method in their mental toolbox. I filed it from day one on the Airbus but fortunately never had to use it.
Tree is offline  
Old 9th Jul 2009, 02:05
  #3354 (permalink)  
 
Join Date: Jun 2009
Location: NNW of Antipodes
Age: 81
Posts: 1,330
Received 0 Likes on 0 Posts
PEI_3721;

With the autopilot engaged, if the aircraft’s reaction to a gust in turbulence is beyond certain boundaries (pitch +25, -13, roll 45, speed, AOA, well before the constraining ‘g’ limits of the FBW system), the autopilot disengages (with alerts) and hands a fully functioning aircraft (controls) to the crew, there is no change in control law.
Thank you for that concise description.

It is worthy of note that the GS between 0200z and 0210z AOC positions decreased from 464KT to 463KT, possibly indicating that the chop conditions remained similar. However, GS may have significantly decreased just prior to 0210z as the a/c penetrated the Cb up-draught. A significant pitch up and latent heat transfer through condensation forming ice could probably be directly responsible for the AP "giving up". This in just a few seconds. G forces and sudden hell breaking out in this manner, is not the best way to start a recovery, especially as the pilots minds were probably engaged on why they couldn't log on to DAKAR using ADS-CPDLC, a procedural requirement 20 minutes before TASIL. Never mind not noticing the bad stuff on the WX radar.

mm43
mm43 is offline  
Old 9th Jul 2009, 02:12
  #3355 (permalink)  
 
Join Date: Jun 2009
Location: I am where I am and that's all where I am.
Posts: 660
Likes: 0
Received 0 Likes on 0 Posts
xcitation, I probably should have said "could" instead of "would", my bad. You had stated that SOP was not to put them on. I was amplifying your answer with the question when could the passengers have put them on given the suspected extreme turbulence would severely bounce any passengers who unbuckled seat belts to start putting one on.

You stated it was SOP not to. I intended to append that it very probably was not physically possible. My wording got sloppy. Please excuse me.

JD-EE
JD-EE is offline  
Old 9th Jul 2009, 02:21
  #3356 (permalink)  
 
Join Date: Jun 2009
Location: Brazil
Age: 76
Posts: 35
Likes: 0
Received 0 Likes on 0 Posts
BEA

I can imgine the BEA committee sitting arguing in the same way we are , with nobody agreeing on anything, except that here we probably have more qualified people
Gringobr is offline  
Old 9th Jul 2009, 03:00
  #3357 (permalink)  
PJ2
 
Join Date: Mar 2003
Location: BC
Age: 76
Posts: 2,485
Received 1 Like on 1 Post
JuggleDan;
Your comments sat pretty maturely for me - I think you get it. Not everyone does, usually because they fly a desk or a laptop. Many writers and safety experts don't fly but they sure get it as evidenced in their excellent contributions to flight safety. People who think that SMS is the answer to keeping accident rates down don't get it. 'nuff said.
it's likely that I just have too much imagination...
For a beancounting manager, yes. For a pilot, no. For present times, also a very mild, 'yes' but like I said, it depends upon who's doing the learning.

This is an extremely complex accident with a depth of possible analysis far greater than any recent ones I can recall - there are a huge number of issues "at rest", latent, in this accident and only a few of them are technical.

Last edited by PJ2; 9th Jul 2009 at 16:02. Reason: Edited with apologies to Sidney Dekker who is checked out on the MD80, (thank you S). My original point was, it doesn't always take an aviator to recognize these safety issues.
PJ2 is offline  
Old 9th Jul 2009, 06:51
  #3358 (permalink)  
 
Join Date: Jun 2009
Location: Paname
Age: 55
Posts: 21
Likes: 0
Received 0 Likes on 0 Posts
PJ2,

Thank you very much for your kind answer. Quite frankly, a lot of my thinking was actually derived from your post #2628. While I fully understand your position, your post gave me a feeling that having the SOPs discourage hand-flight looks like a pretty dangerous double-edge sword...

Back to the point,
For a beancounting manager, yes. For a pilot, no. For present times, also a very mild, 'yes' but like I said, it depends upon who's doing the learning.
Actually, even the bean counters may actually learn a few good things: due to the crashes in June, insurers are likely to lose a lot of money. See for example Les assureurs aériens accusent leurs plus grosses pertes depuis 2001 - Transport - E24.fr ("Airflight insurers acknowledge their biggest losses since 9/11").
Thus, they may put the airflight companies under pressure for more safety and risk reduction. I admit this is partly wishful thinking, but who knows...

BTW, I was positively surprised by some passages in yesterday's interview of AF's CEO:
Il n'y a pas de contradiction entre la sécurité et l'économie. Quand on améliore la sécurité, on améliore l'image de la compagnie et on améliore logiquement ses performances économiques.
[SNIP]
Nous allons regarder à nouveau le contenu de la formation des équipages, la qualité des informations météo disponibles et la mise à jour des informations par des échanges avec le sol. A l'occasion d'un incident et à fortiori d'un accident, le retour d'expérience fait apparaître des possibilités de défaillance. Qu'elles aient ou non un lien avec la cause de l'événement, notre devoir est de les corriger.
In other words (thx Google):
There is no conflict between safety and economic performance. When safety improves, it improves the company public image, and thus mechanically improves its economic performance.
[SNIP]
We are going to review the contents of crew training, the quality of the available weather information and the ways information is updated through radio contacts with ground. Whenever an incident occurs, and all the more when it's an accident, analysis shows latent failure pathways. Whether or not they are connected with the cause of the event, it is our duty to correct them.
I daresay there is hope yet!
JuggleDan is offline  
Old 9th Jul 2009, 07:27
  #3359 (permalink)  
 
Join Date: Jun 2009
Location: SPAIN
Age: 65
Posts: 50
Likes: 0
Received 0 Likes on 0 Posts
Hi,

We are going to review (again) the contents of crew training, the quality of the available weather information and the ways information is updated through radio contacts with ground.
This must had been a basis day work .. to perform those duties.
Seem's Air France was sleeping on their laurels and this last crash awaked them..
Yet many alarm bells were heard for years in AF (from unions .. from pilots)

They were dreaming and sleeping well .. and now they have short nights and make nightmares .....
Some AF heads are accountable for the bad outcome of this airline.

Bye.
LeandroSecundo is offline  
Old 9th Jul 2009, 07:39
  #3360 (permalink)  
 
Join Date: Jun 2009
Location: SPAIN
Age: 65
Posts: 50
Likes: 0
Received 0 Likes on 0 Posts
Hi,

BEA to bring Airbus down?
BEA is not there for put Airbus or AF down or support AF or Airbus.
BEA is there for make a IMPARTIAL investigation with all their knowledge and publish the results and the eventual corrections or practices to be amended.
I hope this will be this way !

Bye.
LeandroSecundo 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.