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

AF 447 Thread No. 9

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

AF 447 Thread No. 9

Old 6th Aug 2012, 17:21
  #1061 (permalink)  
 
Join Date: May 2011
Location: BOQ
Age: 75
Posts: 484
It is not being addressed in training as their are no algorithm's in the sims for high altitude handling. ( most only up to 14000ft) which risks negative training.
IcePack:

What's your source for that statement?
OK465 is offline  
Old 6th Aug 2012, 17:39
  #1062 (permalink)  
 
Join Date: May 2011
Location: BOQ
Age: 75
Posts: 484
CONF:

It may depend on how the instructor inputs the airspeed variation.

If the indicated aircraft speed is slewed directly, you'll get an effective windshear.

Check and see if you've got a 'Variable Airspeed Drift' malfunction available in the ADIRS section.

Using this malfunction can give a nice approximation of the Air Caraibes situation with appropriate ADR disagree messages and no windshear warning.
OK465 is offline  
Old 6th Aug 2012, 18:08
  #1063 (permalink)  
 
Join Date: Sep 1999
Location: United Kingdom
Posts: 907
Sims are set up to meet the equivalent test schedule for the type. Most sims handle the same at 35000ft as they do at 14000ft which is definitely not the case for the aircraft. Sorry would have no idea of the source document for that.
Trouble with negative training is that it may be of benefit to get pilots to hold attitude & power but could lead them to believe that the handling is easy which in an A330 at height with an aft C of G due trim tank it is not. Hence in a surprise situation (in real a/c) they may over control or disbelieve the indications.
Whatever, the simulator industry needs to address this short coming IMHO.
IcePack is offline  
Old 6th Aug 2012, 18:14
  #1064 (permalink)  
 
Join Date: May 2011
Location: BOQ
Age: 75
Posts: 484
Sorry would have no idea of the source document for that.
Then how did you find out about this?
OK465 is offline  
Old 6th Aug 2012, 19:06
  #1065 (permalink)  
 
Join Date: Aug 2011
Location: Grassy Valley
Posts: 2,123
Question. The aircraft (F-GZCP) was in an airstream that was actively changing directions, from the VSI, her accelerations were brisk in the vertical. This was happening at 2:10:05, both sides. ( before and after). So since I've not seen an answer, let me ask again. The ACARS show a WindShear event of some description. Is there a possibility Bonin was reacting to this? Was there no W/S alert on the ECAM? Seems to me WindShear can cause discrepancies in Pitot reporting, sufficient to not only cause autopilot to drop out, but to degrade the Control Law as well....
Lyman is offline  
Old 6th Aug 2012, 20:29
  #1066 (permalink)  
 
Join Date: Aug 2007
Location: The Netherlands
Age: 63
Posts: 287
Answers

Lyman,

Answers to your 4 questions:

NO
NO
NO
NO (actually the answer may be wrong, 'cause it might just be that to you it really seems so)

By the way, windshear cautions or warnings show up where they belong, on the pilots' Primary Flight Displays (same in Airbus as in Boeing).
EMIT is offline  
Old 6th Aug 2012, 20:39
  #1067 (permalink)  
 
Join Date: Jan 2005
Location: W of 30W
Posts: 1,939
IcePack,
I cannot tell about the 14000ft stuff, you may be right, but I can tell it is still pretty easy to over control at 35000ft in a simulator. IMO the handling would still require a LOT of attention.
For the record, despite the fuel in the tail, the CG for AF447 was still very much fwd compared to the ideal target for fuel saving.

OK365,
You are most probably correct.
I'm not too aware of the different possibilities for the sim as I have never occupied that back seat. What I can tell is that the instructor seemed surprised too.
CONF iture is offline  
Old 6th Aug 2012, 20:45
  #1068 (permalink)  
 
Join Date: Jul 2009
Location: France - mostly
Age: 80
Posts: 1,689
Originally Posted by Lyman
The ACARS show a WindShear event of some description.
The ACARS message you're referring to has the Meaning (ref. Interim#1, p.49):
"This message indicates unavailability of the reaction to windshear detection function"
HazelNuts39 is offline  
Old 6th Aug 2012, 21:37
  #1069 (permalink)  
 
Join Date: Jun 2009
Location: Germany
Age: 66
Posts: 782
IcePack
Sims are set up to meet the equivalent test schedule for the type. Most sims handle the same at 35000ft as they do at 14000ft which is definitely not the case for the aircraft. Sorry would have no idea of the source document for that.
Trouble with negative training is that it may be of benefit to get pilots to hold attitude & power but could lead them to believe that the handling is easy which in an A330 at height with an aft C of G due trim tank it is not. Hence in a surprise situation (in real a/c) they may over control or disbelieve the indications.
Whatever, the simulator industry needs to address this short coming IMHO.
Read it somewhere, however couldn´t find it. Stumbled on another pdf though.

Validation of stall recovery technique and training from Airbus March 2011

edit: another look from april 2012
aviation week

Last edited by RetiredF4; 6th Aug 2012 at 21:56.
RetiredF4 is offline  
Old 6th Aug 2012, 21:42
  #1070 (permalink)  
 
Join Date: Jan 2005
Location: W of 30W
Posts: 1,939
Originally Posted by clandestino
I am afraid that Australian ATSB has shown utter disregard for your ideas that fix is inadequate and in its final report of upset to A330 near Learmonth has found actions by Airbus quite acceptable.
Where did I write inadequate ?
It is only a temporary fix where the pilot needs to be quicker than the malfunction to badly bite one more time.
But as usual you redirect the subject as you cannot assume your writing :
Many a claim was made on this rumour network that oh-so-complicated-Airbus-took-the-controls-away-from-me-when-it-shouldn't-have. None of them were substantiated except St.Johns and Bilbao - that's fixed now.
CONF iture is offline  
Old 6th Aug 2012, 22:23
  #1071 (permalink)  
 
Join Date: Aug 2009
Location: Texas
Age: 60
Posts: 5,361
CONF iture
Same reply that 1) The conservative approach would be to make sure that automation stays out of the way when data are identified as unreliable.
I can see the appeal of that (as I am very much a fan of stick and rudder flying). I think henra's response answers some of the "why" the system is as it is.

TTex
Make sense to who, or better said, to WHAT?
In this accident we have an aircraft designed to be flown by computers that suffers a GIGO computer issue that convinces the computers to give up on their job.
Which leaves the pilots with a system designed to be flown by computer and they now have no computer or at least not all of the computer.
BUT, it still wants to be flown like it is a computer because only PART of the computer system gave up, the rest is still working and it only knows computer sense.

I ask again, make sense to who or what?

Forgive my english/grammar, Texan is my native language and i'm in a hurry.
I understood you perfectly.
It makes sense to whomever bought the plane with the objective being a smooth ride in a modern airliner. I will suggest to you that, for the most part, as evidenced by most of the 228 casualties being strapped in, that even as it fell from the sky it was a smoother ride than one might expect for out of control flight. (Then again, many of those strapped in might have strapped in, and then dozed off, a buffet here or there not noticed, be it turb or stall ... )

The Airbus flies already very well under Direct Law, and manual trim is part of that direct, just like the Cessna, or the Airbus 310
The only question I'd have on that is: how often do you get to practice that?

Lyman
Easier to descend than climb.
Gravity assisted flying! What will they think of next! :-)

Sorry for that, I could not resist.

Last edited by Lonewolf_50; 6th Aug 2012 at 22:24.
Lonewolf_50 is offline  
Old 6th Aug 2012, 22:33
  #1072 (permalink)  
 
Join Date: Feb 2005
Location: Correr es mi destino por no llevar papel
Posts: 1,387
Originally Posted by rudderrudderrat
If you are going to be picky, during roll out, you are actually in a sub mode of Direct Law called Ground Mode.
Thank you for being nitpicky for me.

Originally Posted by RR NDB
I don´t like the approach to diagnose UAS by System output.
Neither do I. BEA is also unhappy but it issued no recommendation on fixing it. What you fail to understand is...

Originally Posted by RR NDB
Block garbage to enter System would be better?
...the pressure that hit the total pressure sensor was not the real one due to clogging of the pitot, but such reduced pressure resulted in calculating airspeed of about 52 kt at the minimum and this is valid speed during t/o run or landing roll and cannot be rejected by simple algorithm! This is not readily identifiable to computer as a garbage! It can be identified as such only by someone who has a whole picture of the aeroplane state in his CPU and can recognize that sudden drop from 270 kt to 52 without aeroplane starting to fall or being torn apart in severe turbulence actually means speed display is totally wrong and we call that entity "a pilot". You just can't teach the computer to compare all the parameters and come to intelligent solution, if you try, you can only end up with set of instructions so convoluted its execution will be totally unreliable. Farthest Airbus came is displaying "CHECK GW" message when computer detects discrepancy between input weight, airspeed and AoA - it can not determine which is which so output is made to warn about the most common failure, to input wrong weight into FMGS. A lot of guesswork but it's best that can be done for the time being.

Originally Posted by RR NDB
Sadly, the (non adequately trained) crew of AF447 were:
Question of adequate training is another can of worms. How come other AF crews dealt without much fuss with UAS if AF training is inadequate? It is not just about training, it is also about selection. Starting from the first flight in glider, ending when the logbook is closed for good.

Originally Posted by bubbers44
I don't think the test pilots that approved the 5 degrees nose up and climb power tried it with blocked pitot tubes.
Do you have a slightest idea who writes and approves the procedures?

Originally Posted by bubbers44
The trapped pitot pressure and the decreasing static pressure would cause an indicated airspeed increase and at some point an overspeed warning
So it would! So freaking what? Pilot who would seriously consider acting IAW overspeed warning in UAS situation must be totally "destructured", to use BEA term.

Originally Posted by Lyman
HN.... Why would the boffins select a climb, at all, in the first place?
Panic, caused by sudden realization that something is wrong but one has no clue what it is or what should be done. So doing first thing that crosses his mind and disregarding absolutely every clue it wasn't good idea in the first place.

Originally Posted by RR NDB
R&D must be done to improve current situation.
What situation?

Originally Posted by RR NDB
Why not protect the plane to REC MAX apogee. Hard limit (Airbus SAS approach)
I have read many a misconception about Airbus protection, this one goes into top tier. No altitude protection is offered ever on Airbus as AoA makes it irrelevant. No hard air data based protection is available in alternate laws as there's no sufficient redundancy to guard against the unwarranted activation of the protection.

Originally Posted by TTex600
Poor training. As I've stated more than once on this topic, my training was no better.
How come other 36 crews went through UAS ordeal unscathed? Better training than yours?

Originally Posted by syseng68k
If the systems fail and give no information, then fine
So it is. No electronic part of the system failed in AF447 case. It just wasn't fed the right pressure.

Originally Posted by syseng68k
but to fail and give incorrect information is extremely dangerous
Problem is that many a PPRuNer has noticed it is dangerous only when it resulted in 228 deaths and all of a sudden there are righteous demands, very similar to ones made by certain young F/O

Originally Posted by Ernest K Gann
Captain Hughen, this is all an anachronism. We belong in the modern world and should therefore be secure. Please arrange a remedy for this grievous mistake at once.
At least he was so forthright to humour himself once the fright loosened his grip.

Originally Posted by RR NDB
Simply because Pitot´s may fail SIMULTANEOUSLY
So what? I have flown aeroplane with single pitot and then it got blocked. There's always way out.

Originally Posted by rudderrudderrat
I would agree if the aircraft has natural positive dynamic stability, then it is a non event. If left alone after a disturbance, it will naturally seek its trimmed angle of attack.
I have to warn everyone that trying to explain some aspect of Airbus FBW while not being versed in basic aerodynamics tends to produce entertaining, if not true or otherwise useful , results. What you are referring to is static stability and that airbus lacks it in normal and alternate law is total non-event.

Originally Posted by rudderrudderrat
Any deviation must be recognised and corrected early. Once AF447 had deviated by that much Altitude, it didn't matter how hard they pulled trying to regain FL 350 - it would't work.
What?!? AF447 did regain FL350 by pulling, if only fleetingly as it passed through it on her final descent! Now, if only they pushed...

Originally Posted by TTex600
Bubbers, stop being dismayed about the pilots and start being outraged at the system that trained them and placed them there.
If we assume that Airbus is dangerous by design, shouldn't we be praising the system for always putting the super-pilots in the cockpit that made it home and failing just once? Of course not.

Originally Posted by CONF iture
I don't think they realized they had to deal with an UAS scenario
You are right. How could they realize when they were suddenly faced with 200+ kt drop in IAS?



Originally Posted by CONF iture
But do we know what IAS2 was displaying ... ?
BEA does. 88, 93, 94.

Originally Posted by CONF iture
It is only a temporary fix where the pilot needs to be quicker than the malfunction to badly bite one more time.
Dear CONF iture, by know you should have really known better than assuming I would link to the report without reading it. So: 1. QF72 was single occurrence in 28 million hours of flight 2. temporary fix of switching off the faulty ADIRU was later supplemented by rewriting the FCPC algorithms. I will suggest you should stop burying yourself, just because I know you will not listen to me.

Last edited by Clandestino; 6th Aug 2012 at 22:35.
Clandestino is offline  
Old 6th Aug 2012, 22:35
  #1073 (permalink)  
 
Join Date: Jun 2009
Location: Germany
Age: 66
Posts: 782
Another read i could find concerning Loss of Control in Flight
Training Foundations and Solutions
.


Some bits out of it:

What Changed: Uncovering the Problem
Part of the predicament has nothing to do with LOC-I at all; in fact it is the reduction of other accident causes that left LOC-I as one of the last remaining causal factors to be “tamed”. To be clear, the overall number of accidents and fatalities due to LOC-I have not been increasing. It is the improvement in other accident categories that has resulted in the emergence of LOC-I as the leading cause of fatal accidents in air transportation worldwide. It’s as if when the “swamp” of the aviation accident pool was drained of other causal factors, the “snakes” of LOC-I were exposed. It is only in the past five years that LOC-I has overtaken Controlled Flight Into Terrain (CFIT) as the leading fatal accident category among airline
aircraft worldwide. While other accident areas have improved, LOC-I has stagnated at an unacceptable rate.


The lack of improvement in LOC-I is evidenced by it’s relative contribution to the overall fatality rate. From the period of 2001 to 2008, LOC-I experienced a 29% increase in the contribution of this category
towards the overall fatal accident rate. For the ten year period ending in 2008 the next closest accident category was CFIT, at 20% (roughly one half of LOC-I fatalities).


Loss of Control in Flight as a percentage
of Overall Fatalities:
Year percentage
2001 30.4%
2006 30.6%
2008 39.5%
The Relative Threat: Higher risk, less effective training
To get a perspective on whether or not we should be concerned about a reduction in all-attitude experience in the cockpits of the world’s airliners, we will examine the relative statistical threat of a fatal accident from several categories of accidents.

The worldwide chances of an LOC-I death by comparison to other accident categories

Runway Excursion (Combined Take Off & Landing): x 2.3
Non-engine systems failure: x 5.6
Runway Incursion (By a Vehicle, Aircraft, or Person) : x 9.1
Engine Failure: x 879.5
RetiredF4 is offline  
Old 6th Aug 2012, 22:53
  #1074 (permalink)  
 
Join Date: Jun 2009
Location: Germany
Age: 66
Posts: 782
Clandestino: What you are referring to is static stability and that airbus lacks it in normal and alternate law is total non-event.
You should tell BEA. They obviously dont know about it.

BEA Final report
However, positive longitudinal static stability on an aeroplane can be useful since it allows the pilot to have a sensory return (via the position of the stick) on the situation of his aeroplane in terms of speed in relation to its point of equilibrium (trim) at constant thrust. Specifically, the approach to stall on a classic aeroplane is always associated with a more or less pronounced nose-up input. This is not the case on the A330 in alternate law. The specific consequence is that in this control law the aeroplane, placed in a configuration where the thrust is not sufficient to maintain speed on the flight path, would end up by stalling without any inputs on the sidestick. It appears that this absence of positive static stability could have contributed to the PF not identifying the approach to stall.
Clandestino: You are right. How could they realize when they were suddenly faced with 200+ kt drop in IAS?
You should have helped BEA with their report. They got it wrong again.

BEA Final report 2.1.2.4 Identification of the situation
Three seconds after the autopilot disconnection, the ECAM displays no information that is likely to point to a speed indication problem:
The ECAM mentions a maximum speed that should not be exceeded but does not mention a minimum speed. This could lead crews to suppose that the main risk is overspeed. In the absence of any reliable speed indication, this might lead to a protective nose-up input that is more or less instinctive.

..........
Thus, having identified the loss of airspeed information, the PNF turned his attention to the ECAM, undoubtedly in an attempt to refine his diagnosis and to monitor any actions displayed. He started to read the messages, and consequently called out the loss of autothrust and the reconfiguration to alternate law. The successive display of different messages probably added to the confusion experienced by the crew in its analysis and management.
In the absence of a specific message expressing detection of unreliable speed by the systems, the crew was unable to identify any logical link between the symptoms perceived and these ECAM messages. The impression of an accumulation of failures created as a result probably did not incite the crew to link the anomaly with a particular procedure, in this case the “Vol avec IAS douteuse” procedure.

Last edited by RetiredF4; 6th Aug 2012 at 22:54.
RetiredF4 is offline  
Old 6th Aug 2012, 22:58
  #1075 (permalink)  
 
Join Date: Jun 2009
Location: Oxford, England
Posts: 297
The last time we did practice a scenario of UAS at low level,
the initial clue that something was wrong came under the
warning : WINDSHEAR WINDSHEAR WINDSHEAR

My airspeed was getting in the red, which made sense to me
in case of windshear.

As per SOP my call was "WINDSHEAR TOGA" which implies to
follow the FD, not to turn them OFF. That's only when I heard
the instructor word : "eh guys ..." that I took a quick look
at the other PFD and could remember that one of the exercises
for the day was UAS not WINDSHEAR ... The appropriate call was
then only made : "UNRELIABLE SPEED"
So, in the case of UAS, you could get a windshear warning and / or
a stall warning. Neither of which are correct. Wonder what else ?.

In the case of uas, one would think that the warning system would
at least consult the IRU and other available sources to correlate
the data before presenting it to the crew.

I guess that's all ok though, yes ?...
syseng68k is offline  
Old 7th Aug 2012, 00:20
  #1076 (permalink)  
 
Join Date: Aug 2005
Location: fl
Posts: 2,561
I flew with pilots that would not have done what they did. I wouldn't. If we lost airspeed we would maintain pitch and cruise power as they should have. Get out the UAS checklist and press on. No stall, no panic, but they didn't. Experience helps when you didn't spend your whole short career monitoring an autopilot.
bubbers44 is offline  
Old 7th Aug 2012, 00:30
  #1077 (permalink)  
 
Join Date: Feb 2011
Location: Nearby SBBR and SDAM
Posts: 873
UAS diagnosis

Hi,

Clandestino:

Originally Posted by RR NDB
I don´t like the approach to diagnose UAS by System output.


I put:

I don´t like the approach to diagnose UAS by System output. As showed in an Airbus SAS paper mentioned in an earlier post. This can be improved and in AF447 case seems very important.
In an earlier post i observed:

IMHO this is dangerous. Digital signal processing of air speed analog data can inform UAS onset to the crew before System start to process garbage.

Neither do I. BEA is also unhappy but it issued no recommendation on fixing it. What you fail to understand is...
The recommendations will generate a better design. I hope an UAS DSP subsystem will be eventually used.

Please could you explain: "What you fail to understand is...". What you mean?


Question of adequate training is another can of worms. How come other AF crews dealt without much fuss with UAS if AF training is inadequate? It is not just about training, it is also about selection. Starting from the first flight in glider, ending when the logbook is closed for good.


Even morons could deal with UAS if are trained to use a DSP UAS detector.

(Air speed is important. Pitot´s seems obsolete. Thales version used in F-GZCP was obsolete)

Air speed anomalies are frequent and potentially affect the stability of the System. The current situation is not good.

The 3 redundant (Pitot´s and subsystems, CPT, FO and Stand By) System elements failed SIMULTANEOUSLY. No redundancy, therefore. Better to use a single element.

On the REC MAX protection please don´t consider it so seriously. Actually good pilots can unload the wing above REC MAX and save the day.
RR_NDB is offline  
Old 7th Aug 2012, 00:37
  #1078 (permalink)  
 
Join Date: Feb 2011
Location: Nearby SBBR and SDAM
Posts: 873
Hi,

bubbers44:

Experience helps when you didn't spend your whole short career monitoring an autopilot.


RR_NDB is offline  
Old 7th Aug 2012, 01:12
  #1079 (permalink)  
 
Join Date: Aug 2011
Location: Grassy Valley
Posts: 2,123
Thumbs up

BEA...

"Thus, having identified the loss of airspeed information, the PNF turned his attention to the ECAM, undoubtedly in an attempt to refine his diagnosis and to monitor any actions displayed. He started to read the messages, and consequently called out the loss of autothrust and the reconfiguration to alternate law...."


In the prior paragraph, they reference a three second interval post a/p loss, and note the ECAM has not displayed the Law reconfig. It was sixteen seconds later the PNF called "Alternate Law".

Three seconds after the autopilot disconnection, the ECAM displays no information that is likely to point to a speed indication problem:

What is the point of This timeline, and the two time intervals? Thay mention the possibility of nose up, but make no inference other than, "possible confusion..."


The ECAM mentions a maximum speed that should not be exceeded but does not mention a minimum speed. This could lead crews to suppose that the main risk is overspeed. In the absence of any reliable speed indication, this might lead to a protective nose-up input that is more or less instinctive.


Sensory overload? The pilots? Or BEA? What a jumble...

Last edited by Lyman; 7th Aug 2012 at 01:51.
Lyman is offline  
Old 7th Aug 2012, 01:41
  #1080 (permalink)  
 
Join Date: Feb 2011
Location: Nearby SBBR and SDAM
Posts: 873
Fault tolerance and graceful degradation

BEA Final report 2.1.2.4 Identification of the situation
Three seconds after the autopilot disconnection, the ECAM displays no information that is likely to point to a speed indication problem:
The ECAM mentions a maximum speed that should not be exceeded but does not mention a minimum speed. This could lead crews to suppose that the main risk is overspeed. In the absence of any reliable speed indication, this might lead to a protective nose-up input that is more or less instinctive.

Thus, having identified the loss of airspeed information, the PNF turned his attention to the ECAM, undoubtedly in an attempt to refine his diagnosis and to monitor any actions displayed. He started to read the messages, and consequently called out the loss of autothrust and the reconfiguration to alternate law. The successive display of different messages probably added to the confusion experienced by the crew in its analysis and management.
In the absence of a specific message expressing detection of unreliable speed by the systems, the crew was unable to identify any logical link between the symptoms perceived and these ECAM messages. The impression of an accumulation of failures created as a result probably did not incite the crew to link the anomaly with a particular procedure, in this case the “Vol avec IAS douteuse” procedure.


K.I.S.S. has it counterpart: Keep It Complex Stupid

Question: How they could diagnose the UAS? (The so called, non event)



We still today have two problems:

1) Important (redundant) elements prone to fail simultaneously
2) Diagnosability of UAS
RR_NDB is offline  

Thread Tools
Search this Thread

Contact Us Archive Advertising Cookie Policy Privacy Statement Terms of Service

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