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

AF 447 Thread No. 11

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

AF 447 Thread No. 11

Old 23rd Nov 2013, 09:51
  #901 (permalink)  
 
Join Date: Jan 2005
Location: W of 30W
Posts: 1,939
Likes: 0
Received 0 Likes on 0 Posts
Originally Posted by rudderrudderrat
If you are feeding "bad data" (intermittent / silent stall warning) onto the flight deck - then you'll probably get bad diagnosis out.
It surely brought CONFUSION in the CPT's mind.
All AoA vanes were following a 100 knots airflow and they were all measuring an AoA greater than alpha stall. The STALL WARNING had no reason to quit.
The STALL WARNING logic is one of the contributory factors in this accident.
CONF iture is offline  
Old 23rd Nov 2013, 11:10
  #902 (permalink)  
 
Join Date: Jan 2008
Location: Blighty (Nth. Downs)
Age: 76
Posts: 2,108
Received 4 Likes on 4 Posts
Instrument-Flying Skills

Quote from Clandestino:
"I'll try different approach this time"

If this is a spot-the-difference test, Clandestino, I admit defeat.

As usual, you have quoted snippets of text - sometimes a phrase, rarely more than one sentence - and then dismissed them, in your instinctively adversarial style, without acknowledging their context in the writer's broader argument.

Before I respond to your latest rejection of such a policy, let me reproduce the key points of substance in our discussion about the possible merits of PFs routinely flying jet airliners, in what you neatly describe as friendly skies, with the AP disengaged, and also - where practicable - the A/THR and/or FD.

Clandestino:
"Again: issue with AF447 is not the pilot unable to handfly the aeroplane; it was pilot unable to understand the situation, implication of his actions and pretty precisely handflying just the wrong way, while his assisting pilot was, unfortunately, as lost as him. Nothing of it can be prevented by practising eye-to-hand coordination in friendly skies."

Me:
"I could not disagree more. All that needed to be done in AF447 was to keep the wings level, and MAINTAIN a suitable pitch attitude and thrust for high-altitude cruise. If you are accustomed to monitoring and understanding your a/c in all flight phases, you may learn ball-park figures even without ever disengaging the AP. But there are at least two snags to that as a policy.
"Firstly, you will not learn the very gentleness of any corrections that need to be made on the side-stick if you unexpectedly find yourself without the AP at high altitude. (Rather like driving a car at over 200 kph.) Secondly, human nature means that hands-on practice concentrates the mind in a way that mere observation does not."

The most recent reply from Clandestino:
(1) "Yes but this is all very well when you either know what happened or maintain presence of mind not to do harm when you are not sure what exactly you should be doing."
(2) "You suggest that A330 is sensitive beast that has to be treated gently. I say that it is exaggeration and you can throw her around hamfistedly at cruise altitude, in alternate law for 4 minutes with normal acceleration varying between +1.96 and -0.26G with end result being no worse than just utter mess in cabin and galleys."
(3) "Why are we not discussing some other LOCs like Birgenair? 20 000 hrs could not prepare the skipper for simple recognition that rejecting the takeoff when airspeed is not working is a must or just to crosscheck the three ASIs. Or close call at EHAM where the guy who taught others how to perform V1 cut just couldn't make it properly when it happened for real? That's what AF447 is about: the crew that just couldn't perform when it needed to."

Re (1), you have identified the very point I've been making: the PF needs to know how to avoid doing serious harm to the flight profile, and unnecessarily exploring the limits of the flight envelope.

Re (2), I am saying that, to maintain a safe and sensible flight profile at cruise altitude, you have to handle any large jet with kid gloves. What you may have to do to recover from an unusual attitude is another matter. The fact that airliners are stressed in extremis to about +3.75G and -1.5G is nice to know, but irrelevant to normal operations. The AF447 L.O.C. started with an a/c in normal, stable cruise-flight; albeit in routine, moderate turbulence. Notwithstanding Machinbird's and my own previously-stated reservations about the Airbus FBW downgrade combination of load-factor control in pitch with stick-to-aileron for roll, the only way to develop a scan and stick-dexterity that includes more than one parameter is to practise it regularly.

Regular hand-flying practice reinforces the understanding that in level cruise-flight, regardless of turbulence or windshear, any pitch attitude more than 2 or 3 degrees above or below the norm is unsustainable unless you want to climb or descend. Recovery from any startle factor would be quicker if the PF is confidently familiar with the sort of picture that needs to be maintained or restored on his attitude indicator, and how to achieve it. The same applies to thrust.

Re (3), this discussion clearly reads across to other types, many of which might be more of a handful at the beginning of a similar UAS event than the short-body A330 in Alternate 2B, with its advantage of retaining load-factor control in pitch. The two events you mention were on T/O, presumably involving the requirement for instant decision and rapid action? That did not apply in AF447 - quite the contrary. Excessive reliance on automation may well be eroding the basic instrument-flying skills that my generation, and our forebears, took for granted.

Last edited by Chris Scott; 23rd Nov 2013 at 15:10. Reason: Re-insertion of italics and underscoring. Last para expanded. Title added.
Chris Scott is offline  
Old 23rd Nov 2013, 11:38
  #903 (permalink)  
 
Join Date: Jul 2009
Location: France - mostly
Age: 83
Posts: 1,688
Likes: 0
Received 0 Likes on 0 Posts
It is much better to use tactile (vibration / stick shaker).
How about the airplane shaking so badly that the PNF couldn't hold a handset?

From the judicial expert's final report:

7.4.3.1.2 BUFFET ONSET ET DETERRENT BUFFET
Le Buffet Onset est un état transitoire qui n'a pu être identifié lors de la décélération a lktlsec, il a
été perçu lors d'une prise d'assiette franche vers 12°. Il consiste en une légère vibration
aérodynamique de basse fréquence (environ 4 à 5 Hz).
Le Buffet Deterreni est un phénomène totalement inconnu pour un pilote de ligne, seuls les pilotes
d'essai l'ont rencontré.
La dimension "Deterrent" c'est a dire Effrayant, Dissuadant est liée au Mach, cette dimension
diminue fortement dès que le Mach diminue.
Le Buffet Deterrent que nous avons subi est un phénomène impressionnant qui ne s'apparente pas
totalement à de la turbulence, il s'agit plutôt d'une vibration turbulente (passage à vitesse élevée sur
les bandes de ralentissement à l'approche d'un péage routier par exemple)
Le phénomène a été mesuré au centre de gravité et au siège pilote, il a été également évalué à la
partie arrière de la cabine.
L'effet 'Effrayant, Dissuadant" est très net au poste de pilotage, nettement moins au centre de la
cabine et important à l'arrière de la cabine.
L'intensité du phénomène est liée au mach d'une part et au taux de décélération (ici lktlsec). Les
valeurs étant différentes pour le vol AF447, il est probable qu'il y ait eu une diminution sensible et
rapide des effets, ce qui explique que l'équipage technique n'en ait pas fait état et que l'équipage
commercial n'ait pas appelé le cockpit à ce sujet.

GOOGLE translation (corrected to the best of my ability):

7.4.3.1.2 BUFFET ONSET AND DETERRENT BUFFET
Buffet Onset is a transient state that could not be identified during the deceleration at 1 kt/sec, it was perceived when taking the pitch attitude to 12 °. It consists of a slight aerodynamic vibration at low frequency (about 4 to 5 Hz).
The Deterrent Buffet is a totally unknown phenomenon for an airline pilot, only the testpilots have encountered it.
The dimension "Deterrent" that is to say Scary, Dissuading is related to Mach, this dimension strongly decreases when the Mach decreases.
The Deterrent Buffet we have suffered is an impressive phenomenon that is totally unlike turbulence, it is rather a turbulent vibration (moving at high speed over the 'deceleration strips' when approaching a toll road, for example).
The phenomenon has been measured at the center of gravity and the pilot seat, it was also assessed in the rear part of the cabin.
The effect "Scary, Dissuading" is very clear in the cockpit, much less in the center of the cabin and important in the rear part of the cabin.
The intensity of the phenomenon is related to Mach on one hand and the deceleration rate (here 1 kt/sec). Since the values ​​are different for flight AF447, it is likely that there was a significant and rapid decrease of the effects, which explains that the technical crew did not mention it and the cabin crew did not call the cockpit about it.

Last edited by HazelNuts39; 23rd Nov 2013 at 13:53.
HazelNuts39 is offline  
Old 23rd Nov 2013, 15:39
  #904 (permalink)  
 
Join Date: Jan 2008
Location: Blighty (Nth. Downs)
Age: 76
Posts: 2,108
Received 4 Likes on 4 Posts
Hi HN39,

Is there any chance you could improve the definition of that chart, please?

In the translation of the judicial report, I suggest that "important" may mean "considerable", or "significant", but one can see how easily any tranlation of empirical expressions can be misleading...

"l'approche d'un péage routier" means "the approach to a toll plaza" (i.e., at the end of a toll road).

In the very last sentence, my interpretation in English is slightly different from yours, as follows:
"The values being different from flight AF447, it is probable that there would have been a noticeable and rapid reduction of the effects..."

Now, can you confirm that the chart is from a post-accident flight test? In which case, was it considered that the "deterrent buffet" experienced on the test flight was LESS severe than on AF447, or MORE severe?
Chris Scott is offline  
Old 23rd Nov 2013, 16:44
  #905 (permalink)  
 
Join Date: Jul 2009
Location: France - mostly
Age: 83
Posts: 1,688
Likes: 0
Received 0 Likes on 0 Posts
Hi Chris,

Thanks for your refinements of my translation. On many french 'peages' you pass a toll station at entry where you take a ticket, and again at the end where you pay the toll. The 'strips' are also present on lesser roads to force drivers to slow down before dangerous sections.

I reduced the size of the graphic only slightly, the resolution of the original is hardly better. The grid is 1 sec by 0.2 g, the scale markings on the left are from 1.0 to 1.4 g. The upper trace is for the pilot seat, the middle one near the center of gravity and the lower one in the rear of the cabin.

Yes, the judiciary experts participated in a test flight of an A340, an A330 rated pilot in the left seat and the Airbus pilot in the right seat. The testflight was preceded by a simulator session and the objectives were to evaluate:
-- experience "Deterrent Buffet"
-- for the three 'expert' pilots to appreciate manual piloting in normal, alternate and direct law at low and high altitude
-- apply the procedures for "IAS Douteuse" and "Stall Warning" at high altitude.

As to the test flight buffet being LESS or MORE severe as AF447 I leave the interpretation of the French text to you. In the "Commentaire d'experts" at the end they write:
"Le ressenti du deterrent buffet varie en cabine selon le positionnement des passagers. On ne peut pas affirmer qu'a l'approche du decrochage du vol AF 447, les conditions de vitesse et de taux de prise d'incidence aient genere de facon similaire les sensations du deterrent buffet. Celles-ci n'apparaissent pas de facon evidentes sur le FDR".
HazelNuts39 is offline  
Old 23rd Nov 2013, 16:47
  #906 (permalink)  
 
Join Date: Jun 2011
Location: West of Offa's dyke
Age: 87
Posts: 476
Likes: 0
Received 0 Likes on 0 Posts
Hi Chris

I see I crossed with HN39 in cyberspace, but still :-

Now, can you confirm that the chart is from a post-accident flight test? In which case, was it considered that the "deterrent buffet" experienced on the test flight was LESS severe than on AF447, or MORE severe?
I don't think you are going to do any better than section 1.16.4.2 of the final report. That says AI did special post-accident tests at as near as possible to the accident conditions and the test pilots established the AoA at which buffet reached deterrent levels.

If that chart is the result of those tests, as it presumably is, then it shows values appropriate to AF447, neither much more nor much less.
Owain Glyndwr is offline  
Old 23rd Nov 2013, 16:51
  #907 (permalink)  
 
Join Date: May 2011
Location: BOQ
Age: 78
Posts: 542
Likes: 0
Received 1 Like on 1 Post
...the judiciary experts participated in a test flight of an A340...
Why a 340? Or maybe better put why not a 330?
OK465 is offline  
Old 23rd Nov 2013, 17:19
  #908 (permalink)  
 
Join Date: Jul 2009
Location: France - mostly
Age: 83
Posts: 1,688
Likes: 0
Received 0 Likes on 0 Posts
Why a 340? Presumably because Airbus had one available and it is aerodynamically and systems-wise similar to an A330. One doesn't want to fool around in deterrent buffet in an airplane that is being readied for delivery to a customer.
HazelNuts39 is offline  
Old 23rd Nov 2013, 18:05
  #909 (permalink)  
 
Join Date: Dec 2007
Location: Pasadena
Posts: 633
Likes: 0
Received 0 Likes on 0 Posts
Deterrent buffet, test in A340 vs A330

While I'd agree that the A340 and A330 are - intentionally - very similar, this buffet appears to shake the cockpit much more than the cabin, yet is presumably excited by the air over the wings as the flow starts to change, rather than by the air over the forward fuselage.

As a result, the number and shape of the engines, affecting the airflow around the wing near the engines and the elastic and inertial properties of the wings, might be a relevant thing to look at. The weight distribution in the forward fuselage might matter too for how the cockpit wags around.

Presumably the test pilots were confident that the A340 results would apply to the A330, and convinced the investigators and the magistrates. Perhaps at 1-10Hz the differences are small - that certainly seems to be more of a shake than a bend.

If it's their own test A340, then it might also be safer to do this extreme test on board a very familiar aircraft.
awblain is offline  
Old 23rd Nov 2013, 19:00
  #910 (permalink)  
 
Join Date: Oct 2013
Location: Wengen
Age: 52
Posts: 380
Likes: 0
Received 0 Likes on 0 Posts
A340 V A330 : Buffeting

Hazelnuts39 : votre thèse sur le buffeting me laisse perplexe.
So are you stating that the buffeting standard applied to all Airbus models is based on the A340?
Winnerhofer is offline  
Old 23rd Nov 2013, 19:56
  #911 (permalink)  
 
Join Date: Dec 2012
Location: Ramona, CA
Age: 65
Posts: 37
Likes: 0
Received 0 Likes on 0 Posts
>The experts are surprised that the auto trim doesn't stop trimming the plane up once alpha prot is reached, as it is the case in ALT1.<
That functionality only exists in Normal law.
In Alternate law, (if airspeed is available) the protections will provide Low Speed Stability, which is essentially direct law pitch (pitch trim stops, the airplane's natural tendencey to pitch down if faster than the trimmed speed).
If there is no valid airspeed, then that can't work, and you're left with g-load demand with no bottom end.
Bpalmer is offline  
Old 23rd Nov 2013, 20:56
  #912 (permalink)  
 
Join Date: May 2011
Location: BOQ
Age: 78
Posts: 542
Likes: 0
Received 1 Like on 1 Post
The Airbus widebody family of flight test aircraft, the A380, A330 and the all-new A350 XWB, took off from Toulouse today flying together for the first time before continuing on separate flight test missions.
This is dated 20 Sept, 2013.

A350 XWB joins the A330 and A380 for an Airbus ?Xtra-widebody? family flight*| Airbus News & Events
OK465 is offline  
Old 23rd Nov 2013, 21:18
  #913 (permalink)  
 
Join Date: Jul 2002
Location: UK
Posts: 3,182
Likes: 0
Received 0 Likes on 0 Posts
Originally Posted by rudderrudderrat View Post
If someone could please explain how an A330 could ever have a CAS<60kts (at 1 g) whilst airborne - then I'll give up on this point.
If you are feeding "bad data" (intermittent / silent stall warning) onto the flight deck - then you'll probably get bad diagnosis out.
I'm pretty confident that the NCD value was to keep the FCS computations predictable and consistent - the FWC probably didn't come into it. I don't think the impact on Stall Warning was considered - and I'd be surprised if it was considered on other types either.

Audio clues tend to get filtered out by the brain during high work load. It is much better to use tactile (vibration / stick shaker).
You'd think so, but the record doesn't bear this out - shakers and pushers have been ignored or dumped pretty regularly. Even if we confine the data set to accidents after the launch of the A320, there have been more ignored shaker warnings than there have been ignored aural "cricket" warnings.

Originally Posted by CONF iture View Post
The STALL WARNING logic is one of the contributory factors in this accident.
No argument there, but let's not beat about the bush - it's a comparatively small factor when looking at the incident as a whole. Also, as I said before - we don't know how stall warnings on other types would respond when so far outside the accepted flight envelope.

Last edited by DozyWannabe; 23rd Nov 2013 at 21:38.
DozyWannabe is offline  
Old 23rd Nov 2013, 22:06
  #914 (permalink)  
 
Join Date: Oct 2009
Location: UK
Posts: 1,270
Likes: 0
Received 0 Likes on 0 Posts
Hi DozyWannabe,
Even if we confine the data set to accidents after the launch of the A320, there have been more ignored shaker warnings than there have been ignored aural "cricket" warnings.
If you have a look at the CVR transcript from page 23 onwards:
http://www.bea.aero/docspa/2009/f-cp...nexe.01.en.pdf you will see that the "C-chord" (Altitude Alert) was on continuously until impact. It was only interrupted by "cricket" "Stall Stall" and GPWS warnings (higher priority).

The crew didn't acknowledge the continuous "C chord" (which could have been easily cancelled).
Does that, plus all the ignored "Stall Stall" warnings not suggest to you that they never "heard" any of the audio warnings?
rudderrudderrat is offline  
Old 23rd Nov 2013, 22:18
  #915 (permalink)  
 
Join Date: Jul 2002
Location: UK
Posts: 3,182
Likes: 0
Received 0 Likes on 0 Posts
Originally Posted by rudderrudderrat View Post
Does that, plus all the ignored "Stall Stall" warnings not suggest to you that they never "heard" any of the audio warnings?
Of course it raises the possibility, and I'm not debating that aural overloading is a potential hazard. However, from BEA548 through NWA6231, Birgenair 301 and West Caribbean 708 - to name just a few, there are a slew of fatal accidents where a tactile shaker (or pusher) was either ignored or dismissed by the crew. I'm with you - theoretically, buzzing the stick in your hand should get through where aural warnings can't, but experience shows that if the situation is confused enough, it doesn't make as much difference as one might think.
DozyWannabe is offline  
Old 24th Nov 2013, 01:12
  #916 (permalink)  
 
Join Date: Jan 2005
Location: W of 30W
Posts: 1,939
Likes: 0
Received 0 Likes on 0 Posts
Originally Posted by Dozy
it's a comparatively small factor when looking at the incident as a whole
Small or big, it is one more hole to line up for the accident to take place.

we don't know how stall warnings on other types would respond when so far outside the accepted flight envelope
And ... ?

Originally Posted by Bpalmer
If there is no valid airspeed, then that can't work, and you're left with g-load demand with no bottom end.
Which is totally aberrant.
The auto trim goes almost to the max value when it should have quit at the same time the AP and A/THR have.
CONF iture is offline  
Old 24th Nov 2013, 19:55
  #917 (permalink)  
 
Join Date: Dec 2012
Location: Ramona, CA
Age: 65
Posts: 37
Likes: 0
Received 0 Likes on 0 Posts
If someone could please explain how an A330 could ever have a CAS<60kts (at 1 g) whilst airborne - then I'll give up on this point.
If you are feeding "bad data" (intermittent / silent stall warning) onto the flight deck - then you'll probably get bad diagnosis out.
At extremely high angles of attack (e.g., 45°) the air striking the pitot tubes and the static ports (located well below the midline of the fuselage) approach the same angle. Indeed, at one point (at least), the error reported on the ISIS showed that the static pressure exceeded pitot pressure!
Since CAS is a measure of the difference between these two pressures, the airspeed indication will be erroneously low.
This happened as the stall deepened and the angle of attach exceeded 45°. You can easily correlate the periods of reduction in angle of attack during the descent with the stall warning sounding periodically.
Bpalmer is offline  
Old 24th Nov 2013, 20:08
  #918 (permalink)  
 
Join Date: Dec 2012
Location: Ramona, CA
Age: 65
Posts: 37
Likes: 0
Received 0 Likes on 0 Posts
The auto trim goes almost to the max value when it should have quit at the same time the AP and A/THR have.
It should have quit based on what? Your opinion?
In the absence of airspeed to switch the pitch mode to the Low Speed stability(i.e., Direct law -trim becomes manual only), the pitch mode remained in g-load demand. As the g-load reduced due to the increasing descent rate when the stall happened, the flight controls atttempted to bring the g-load back to the commanded value in the only means it had: more elevator. When the elevator is held at other than neutral, the stab trims, and so it went.
Bpalmer is offline  
Old 24th Nov 2013, 20:41
  #919 (permalink)  
 
Join Date: Oct 2009
Location: UK
Posts: 1,270
Likes: 0
Received 0 Likes on 0 Posts
Hi Bpalmer,
Nice description of IAS.
CAS is IAS corrected for instrument and position /installation errors.

Last edited by rudderrudderrat; 24th Nov 2013 at 21:10.
rudderrudderrat is offline  
Old 24th Nov 2013, 22:35
  #920 (permalink)  
 
Join Date: Nov 2013
Location: Beverly Hills
Posts: 2
Likes: 0
Received 0 Likes on 0 Posts
CVR specifics

Hi, Everyone. Been reading these 11 threads from the beginning of this tragedy. Finally registered in order to ask what (hopefully) is not taken as a stupid question: is the CVR audio purely mic driven or are the C-chord, stall cricket and other audio alerts wired directly into the box? I doubt they are wired in, but if it were the case and the electronics were going haywire, it could explain why the crew never acknowledged the alarms (requiring speaker failure) while still existing on the CVR... Speaking of that, I do still wonder why the BEA neglected to include the "one hour sleep" reference and if there is more to the "privacy" excuses re CVR, legit or not.
Armchair is offline  

Thread Tools
Search this Thread

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

Copyright © 2023 MH Sub I, LLC dba Internet Brands. All rights reserved. Use of this site indicates your consent to the Terms of Use.