Go Back  PPRuNe Forums > Flight Deck Forums > Rumours & News
Reload this Page >

TAM A320 crash at Congonhas, Brazil

Rumours & News Reporting Points that may affect our jobs or lives as professional pilots. Also, items that may be of interest to professional pilots.

TAM A320 crash at Congonhas, Brazil

Old 14th Aug 2007, 09:45
  #1621 (permalink)  
 
Join Date: Jul 2007
Location: BRU
Posts: 82
Rob wrote:

Sorry if I got carried away. I am learning a lot from many of you and would like to continue enjoying your company.
The same goes for me. What astonishes me more and more is that the AB logic tries to prevent a lot of human errors (eg extending GS when airborne-is this then not a matter of good airmanship?) but didn't do anything to prevent the possible HF we are confronted with in the case of TAM 3054. (except for the adapted MMEL, which apparently didn't get through to every pilot) The logic doesn't avoid this omission and, what is worse, it is even misleading in that it stops the RETARD call after TD. It puts pilots in the wrong mode, because they know that the ac tries to prevent HF, so when suddenly GS don't extend, they start looking for a fault in the ac, not in the human factor.
It would be impossible to predict every possible HF, but by eliminating some and not adressing others, a logic like this has to lead to major problems during operations. I hope AB, and certification bodies, will eventually be modest enough to aknowledge this.
In the meantime the bodies of the two pilots have been identified. May they rest in peace.
borghha is offline  
Old 14th Aug 2007, 09:53
  #1622 (permalink)  
 
Join Date: Dec 2001
Location: Finland
Age: 40
Posts: 121
The more I think about this accident, and the likeliness of a human mistake, the more clear it becomes.

The psych expert described one possible chain of events from the POV of human mind tendencies of reaction during the landing.

If you think about it. The previous landing was OK. The TLs were not an issue. The FO managed to idle both TLs and everything was just fine. Then comes the second flight, with the capt as PF. They are engaged in thinking about a slippery and short runway, with TR2 inop. Very little energy (I think) is contributed in both of the pilots' minds to what to do when TR2 is inop. But probably more is directed towards: slippery rwy, short rwy, flying the app 1 dot low, getting the plane to STOP.

Now imagine this. With all the precautions and preparations done to enable a safe landing in these less than ideal conditions, and then you hear "No spoilers" after touchdown. Right at this moment you are facing a brutal slap in the face against your plans. As the seconds fly by, you are faced with the inability to decelerate. Nothing happens even though your operable TR is reversing, and you know it since you pulled it back yourself. Speed is not decelerating, and just a couple of seconds later you know that this time tea and bisquits at the Chief's office is not going to save the plane. Something far worse is going to happen, but you are not thinking about crashing into a building at 100kts just yet. What do you do? There's only one answer to this. You will try to manipulate ALL the controls that instinctively would allow you to STOP the plane, make the speed come down and prevent a complete disaster. You will stand on the brakes. Allow me to speculate that IF the plane had had some sort of facility to A) WARN the pilots of the situation where one ENG brakes, and the other pushes forward and B) ALLOW the Ground Spoilers to be extended and possibly C) WARN the pilots more clearly that there are NO Ground Spoilers and NO Autobraking, then this crash would have been prevented.

To put a long story short: In a catastrophic situation, where you have to make decisions in a SPLIT second, the human mind can and will lock up just too often to just ignore this in the operation of the aircraft. There is no way in hell that the crew could have detected the misplaced TL during the situation where all they were concerned over was the fact that the plane is NOT decelerating.

What to do?

A) is already covered (partly) by AB by issuing the new ECAM msg and the CRC for this situation.

B) is absolutely shameful, that the pilots have no manual control over the GS whatsoever. Just amazing. On a Boeing plane all you had to do was pull the SB all the way back.

C) The "No spoilers" on ECAM I think is not enough to communicate such a critical shortness in the braking ability of the aircraft. Especially so that this shortness also meant that the autobrakes did nothing. In the heat of the action the pilots didn't realize that there was no deceleration. The LEAST the thing could do would be to say "No spoilers" together with "No autobrakes" with proper warning chimes so that the PNF could call these items immediately upon touchdown and the PF could then slam the brakes.

What I see here is an airplane built so tightly around desired logical chain of events, where there is not enough communication towards pilots to let them know that THIS time we have departed from the desired chain of events, and we are NOT stopping at all. So do something. It's simply not enough to expect a human mind to "COMPLETE" an unfinished input map to enable the desired output in a situation where each second matters. In that kind of case the pilot would need to be told CLEARLY and concisely what is NOT happening that he expects to happen, AND allow him all the manual control needed to get the most out from the airplane systems (manually) to try to make it right. And forget about the logic maps at that split second.

If the investigation reveals the mistake of the PF of leaving the TL in the CLB detent, then so be it. Humans err. But the airplane in question did almost NOTHING to help the human in the situation. Especially AFTER the mistake had already been done. And I will also add that people should forget about the mumbojumbo of what is or isn't expected of a pilot at each phase of the flight. The simple fact is that IF a mistake can be done, someone will make it someday. And the machine better help the person then to overcome the mistake with the least negative results. Not like this. As was said before in this thread, IF the mistake happens... the system should "fail gracefully". The Airbus did not do this, and 200 people lost their lives as a result. Think about that.

Tero
teropa is offline  
Old 14th Aug 2007, 10:23
  #1623 (permalink)  
PBL
 
Join Date: Sep 2000
Location: Bielefeld, Germany
Posts: 955
Teropa,

if you look at the 2nd previous landing at Congonhas (p 13 of the FDR) you will see that the crew armed Autobrake MED but then used manual braking as soon as WoW on 3. That seems to me to be a good idea.

I understand from a note on this thread that the same crew accomplished that previous landing at Congonhas (denoted "2nd previous" on the FDR). So it looks as if their procedure was to arm autobrake MED but intending to use manual braking.

If that is so, one wonders why they did not then do this.
If you intend to stop at all costs (as you are suggesting the main motivation would have been), and you have in any case planned to use manual brakes at WoW, then why not just brake as you intended? Spoilers or no.

I am wary of speculation into human factors such as you proposed. There are so many things that could have happened with human minds in this unfolding situation. And everything one proposes is based on assumptions which we also have no further way of checking. This is not stuff we can ever get to be sure about in this case. Nevertheless, I think it a very good idea for professionals to think through what they might do in such a situation. (I was very glad way back when, when I spun a C152 inadvertently on solo with about 20 hours TT during stall practice, that I had seen one demonstrated a few flight hours before. )

PBL
PBL is offline  
Old 14th Aug 2007, 10:41
  #1624 (permalink)  
 
Join Date: Dec 2001
Location: Finland
Age: 40
Posts: 121
PBL,

I understand what you mean by speculation of HF being something rather "uncertain" by nature. However, I would employ the KISS method here, ie. Keep It Simple, Stupid. Basic assumptions like:

- We are on approach, and expecting to land. Thus we are required to STOP on the runway.
- We are taking off, expecting to climb away from the Airport. Thus we are required to not stop on the runway.

Are just examples of two rather basic situations, where the individual is oriented towards some expected action. I will not start speculation as to how and why for example in the takeoff situation, the mind must not think "Go" until v1 with the obvious detailed conditions of course, and how a pilot has to act based on facts at hand rather than previous situation.

But I can tell that if an airplane is not assisting a person in the most rudimentary, dare I say instinctive actions, then it's not doing its job right. We can naturally speculate as to what is instinctive to whom etc. etc., but even psychology has many sets of assumptions as to what happens inside the mind in certain situations, so I am speculating that it's safe to say that IF a pilot is landing a plane, and is unable to decelerate, then his/her mind is thinking: "I want to stop, not crash and die".

Regarding the 2nd previous landing. I was under the impression that it was a different crew at the controls then?

Tero
teropa is offline  
Old 14th Aug 2007, 10:43
  #1625 (permalink)  
 
Join Date: Jan 2001
Location: UK
Posts: 2,045
B) is absolutely shameful, that the pilots have no manual control over the GS whatsoever. Just amazing. On a Boeing plane all you had to do was pull the SB all the way back
Works both ways.... if the B757 had auto retraction on the spoilers a la Airbus, then Cali might not have happened

Edit: "shameful" is a strong phrase The designers and certification authorities have had years to devise the design, refine it, refuse it etc. Please try and justify why a designer should allow GS to be obtained without TLs at idle (or failed etc.) ?? It is not, IMHO "predictable" that pilots will not select idle at touchdown... we have all learnt of a systemmatic HF issue. It has been addressed, maybe enough (?) but seemingly too late / inadequately publicised for CGH....
NigelOnDraft is offline  
Old 14th Aug 2007, 10:45
  #1626 (permalink)  
 
Join Date: Dec 2001
Location: Finland
Age: 40
Posts: 121
NoD,

That is true.

Tero
teropa is offline  
Old 14th Aug 2007, 10:47
  #1627 (permalink)  
PBL
 
Join Date: Sep 2000
Location: Bielefeld, Germany
Posts: 955
Originally Posted by teropa
Regarding the 2nd previous landing. I was under the impression that it was a different crew at the controls then?
If that was so, then my observation does not apply, of course.

PBL
PBL is offline  
Old 14th Aug 2007, 11:12
  #1628 (permalink)  

Mach 3
 
Join Date: Aug 1998
Location: Stratosphere
Posts: 622
B) is absolutely shameful, that the pilots have no manual control over the GS whatsoever. Just amazing. On a Boeing plane all you had to do was pull the SB all the way back.
I just don't understand what some pilots/contributors/critics want?

They want more control devolved to the pilot (presumably this means less "automation"?) yet when the system state is ambiguous, they want it to be able to disambiguate between the possibilities (presumably this means more "automation"?) ....

Which do you want? More or less?

Do you want the FBW FCS to be more capable or less capable?

Personally, I think it is a very clever system already, which, in the presence of this particular ambiguous state, did leave ultimate control to the pilot.

(Kind of like when you press the brake and accelerator together in your car...for those of you who are inclined to do such things.)

I don't understand why anyone would think that, in the circumstances a pilot would have been any more inclined to pull a speedbrake lever, than retard a thrust lever?

Cognizant of the fact they were flying an Airbus, with one action, system faults notwithstanding, the machine would have given the pilots all the stopping assistance they needed.

To coin a phrase, "all they had to do" was retard the additional thrust lever.

I don't know why they didn't.

But I do know that I have a job because altimeters don't set themselves, pressurisation panels don't reconfigure themselves, landing gear don't deploy themselves...amongst other things.

Does the fact that some are compaining about the system logic post accident bother anyone else?
SR71 is offline  
Old 14th Aug 2007, 11:21
  #1629 (permalink)  
 
Join Date: Dec 2001
Location: Finland
Age: 40
Posts: 121
SR71

Which seems more "pro-cognitive" / logical to you in a situation, where you just don't have the time to "analyze".

1. Retard a TL from an "A/T engaged" detent, but maybe be a little hesitant not to retard it below IDLE to account for the TR inop dispatch. Or could I retard it all the way to MAX REV...

or

2. Use one or all of the 3 stopping facilities of the aircraft: Brakes, Ground Spoilers and Thrust Reversers?

AB is not a bad airplane. It's not perfect either. And if there's room for improvement somewhere, one place has clearly been identified during the discussions (...and repercussions) of the TAM accident.

Why does the discussion bother you post accident? Do you mean that afterwisdom should not exist? I fail to see your point. Transport aviation, with humans in the back, in the scale it exists today.... With the equipment it exists today. You bet there is the need for analysis post accident, and proper actions.

Tero
teropa is offline  
Old 14th Aug 2007, 11:38
  #1630 (permalink)  
 
Join Date: Jul 2002
Location: UK
Posts: 3,182
1. Retard a TL from an "A/T engaged" detent, but maybe be a little hesitant not to retard it below IDLE to account for the TR inop dispatch. Or could I retard it all the way to MAX REV...

or

2. Use one or all of the 3 stopping facilities of the aircraft: Brakes, Ground Spoilers and Thrust Reversers?
But 'all the pilots had to do' to achieve 2 was retard the misconfigured TL to at or near the idle position. If A/THR is going to be cancelled on landing anyway I don't see what they had to gain by leaving it in a forward position.
DozyWannabe is offline  
Old 14th Aug 2007, 11:52
  #1631 (permalink)  

Sun worshipper
 
Join Date: Dec 2001
Location: Paris
Posts: 494
A/THRinvoluntary disconnect

bsieker,
You seem to have a better understanding of the thrust control system's terms, I'd greatly appreciate it if you could read p148 of the incident report of the Taipei-Sungshan-Overrun (http://www.asc.gov.tw/acd_files/189-c1contupload.pdf), and tell me what I understood wrong.
Goodness ! After reading the gooble dee gock of that report, I can now ubderstand your puzzlement as to the working of the thrust system. I also understand that there has been some cross-references between your posts and mine ( I couldn't understand some of the vaklues you quoted ).
I need to go back to each and every sentence of that text for it to make some sense.
I'll come back to you this evening.
Regards.
Lemurian is offline  
Old 14th Aug 2007, 12:46
  #1632 (permalink)  
Paxing All Over The World
 
Join Date: May 2001
Location: Hertfordshire, UK.
Age: 63
Posts: 9,104
bubbers44
I have never flown an aircraft with this type of logic so do not understand the loyalty so many posters here have to giving this kind of control to a computer and accepting what ever it gives you as better than what a pilot could do if allowed to override the logic.
Non-pilot speaking: All the AB pilots have said that the a/c will give you what you ask of it and that, as far as can be understood at this stage: The a/c WAS giving the pilots EXACTLY what they asked for. Unfortunately, it appears that they asked for the wrong things.

Once again, I repeat the suggestion I made many pages ago in this thread: People will want to reconsider having the two small video cameras on the f/d being part of the recording process. Then we would see what they did and did not do. We can never know what went through their minds.

It was well put by SR71
I don't understand why anyone would think that, in the circumstances a pilot would have been any more inclined to pull a speedbrake lever, than retard a thrust lever?

Cognizant of the fact they were flying an Airbus, with one action, system faults notwithstanding, the machine would have given the pilots all the stopping assistance they needed.

To coin a phrase, "all they had to do" was retard the additional thrust lever.
We do not know why and, in all likelihood, will never know why. However, we do know that some flight crew have left TLs forward instead of retarding them when all the instructions on ALL aircraft tell you to bring them back to idle.
PAXboy is online now  
Old 14th Aug 2007, 12:50
  #1633 (permalink)  
 
Join Date: Jan 2004
Location: LPPT
Age: 54
Posts: 431
Please allow me to go back at my previous suggestions that could help untie critical situations as this.

a) Engine Control: IF TL #1 is retarded to Reverse AND the TL #2 is above Idle and bellow Flex/MCT/TOGA just set Idle thrust on Engine #2 (clear last info on TL on A/THR disconnect) and return to TL indication if moved again.

b) Spoilers: their automation logic is very cool indeed but let them out IF the pilot moves the handle all the way down (mechanical override).

I know, its not been tested, approved, too much money to investigate, certify, install, etc... but wouldn’t you like that on your plane?

GD&L

P.S.- I’m with SR71 when he says that this is one of the best discussions ever seen on PPRuNe. My applause.
GearDown&Locked is offline  
Old 14th Aug 2007, 12:52
  #1634 (permalink)  
 
Join Date: Aug 2007
Location: Brazil
Age: 67
Posts: 131
A bit more about computer logic, if I may.
If we allow computers to make such an important decision (I WON'T GIVE YOU GS NOR A/B IF YOU DON'T MOVE BOTH TLs TO IDLE), we need to teach computers also how to put "hands on" and really correct our (pilots) mistakes.
The computer on-board TAM's A-320 went in to a "stubborn mode" (DON'T MATTER WHAT YOU DO, I WON'T GIVE YOU GS).

IMO, on-board computers are crossing a very dangerous line. They "evolved" from a "friend" that warns pilots, to a proudly "I got it" mode.

Yes, I know. Computers tried for decades to find efficient ways to warn us (pilots) that we forgot to put down the landing gear, for example . Beeps, horns, nice female voices, and we still landed with the gear up. One computer I knew even tried to cover the airspeed indicator with a red flag, stating CHECK GEAR!!. And we landed gear up complaining we couldn't see our airspeed...
Yes, I know it's not easy to fly with us. We do a lot of mistakes, but with proper training we are able to learn how to avoid most of them.
But, IF computers need to take over the controls in order to "prevent" us from doing mistakes, then computers need "proper training" also.

I remember of some incidents (and one fatal accident) where a crew conflict happened, one wanted to stop, the other wanted to go around.

Resuming, I believe that is quite obvious a conflict happened inside the cockpit. Not between the two pilots, but between the PF and the CF (computer flying)...
How to prevent this "conflict" from happening again?
Simply by only giving the controls to computers when we want to do so. And if we don't like what they (computers) are doing, allow us to get ALL controls back...

best rgds,
Rob
Rob21 is offline  
Old 14th Aug 2007, 13:11
  #1635 (permalink)  
 
Join Date: Jul 2002
Location: UK
Posts: 3,182
Rob21:
The computer on-board TAM's A-320 went in to a "stubborn mode" (DON'T MATTER WHAT YOU DO, I WON'T GIVE YOU GS).
Where do you get this from? The computer would cheerfully have given them ground spoilers if the thrust lever was in the correct position. It would also have stowed the TR, dropped the spoilers and given them TOGA power if they had shoved the levers forwards into the TOGA position. Whether this would have given them enough thrust soon enough to get them back into the air we'll never know.

There's a fundamental minsunderstanding of what the A320 computer does and does not do going on here. It does not 'prevent' the pilots from doing anything and never has. The alpha floor protection was designed to prevent the aircraft from inadvertantly falling out of stable flight when above 100ft, but that's about it for overriding manual input.
DozyWannabe is offline  
Old 14th Aug 2007, 13:53
  #1636 (permalink)  
 
Join Date: Aug 2007
Location: Brazil
Age: 67
Posts: 131
Dozy,
Yes, but do we really know that engine #2 TL was not at iddle? That was the computer reading, and he act accordingly. There is no doubt about that, and what we see on FDR's graphics is what the computer "saw"...
And at this point of the investigations, a "wrong" computer reading of TLA is not out of the question, I would imagine. I also believe it is one of the "probable causes" to be investigated.
I know, it's not very often we see on-board computers fail. We (humans) need to do something to them. Simple things, like a coffee spill can make a computer go "blind" or crazy.
Thanks for your inputs,
Rob
Rob21 is offline  
Old 14th Aug 2007, 13:56
  #1637 (permalink)  
 
Join Date: Jul 2002
Location: UK
Posts: 3,182
And at this point of the investigations, a "wrong" computer reading of TLA is not out of the question, I would imagine.
I'd say it's a fairly remote possibility at this stage, if the printed and leaked information proves to be correct.
DozyWannabe is offline  
Old 14th Aug 2007, 14:04
  #1638 (permalink)  
 
Join Date: Mar 2007
Location: In my head
Posts: 687
While we wait for Lemurian to get back, the penultimate sentence of that Taipei-Sungshan-Overrun report on page 148 just in language terms alone does 'look' rather leading by use of unusual punctuation, and use of that vaguely vocabulated statement 'allowed the thrust to be frozen' which leaves further questions, does it not?
That involuntary ATHR disconnection allowed the thrust to be frozen on engine2 whose lever was at CLB notch !
Those are Toulouse's own supposedly carefully chosen words, right ?

And in the Taipei-Sungshan-Overrun the "frozen thrust on engine2" was never resolved by the pilots? Once it was in that particular "frozen" state (whatever that really means), was there truly any possibility that it could have been unfrozen in anyway at all? I can hear Airbus drivers now shouting "Yes, just move TL #2 to IDLE!!!", but can we actually assume that because there is no record in the report of the pilots disputing it, that 22.5 degrees was indeed its physical position throughout? Has that been established by means other than FDR? (by asking them). Surely the answer is yes, and if they had moved TL #2 they would remember?

TLA and TRA has been discussed but how is the figure that appears in the FDR actually derived?

Forgive me especially if I have failed (again) to notice something that has been asked before and is now reconciled, but what I am asking is, leaving aside what the FCOM says would be the effect, is there any way that TLA on TL #2 could fail to be recorded in the FDR due to the data derivation being obtained not from the lever itself but instead backwards from 'the' circuit which commands actual (frozen) Thrust or from the 'one' that resolves the angles or even from the 'one' that handles ATHR perhaps? Please forgive and correct any over-simplification ... but we do have "resolvers" for at least one operational purpose and learned that inconsistent 'resolved' values (more than a quarter degree) are rejected ... how many discretely-derived instances are 'stored' in the 'buffer(s)' ? Is there scope for a flawed derivation of TLA being plotted on the FDR?

Last edited by slip and turn; 14th Aug 2007 at 14:52. Reason: Dozy and Rob21 thinking same lines obviously - beat me to it ! Now edited again to improve grammar a bit !
slip and turn is offline  
Old 14th Aug 2007, 14:27
  #1639 (permalink)  
 
Join Date: Jul 2007
Location: BRU
Posts: 82
Doz wrote
The computer would cheerfully have given them ground spoilers if the thrust lever was in the correct position
Doesn t this contradict your own statement? The computer (decides to) 'gives' GS IF the pilots do/had done ... So it is the computer that makes the decision here, ignoring completely that one of the human inputs was incorrect... And allowing them just a few seconds to figure everything out, to correct their error (obey my logic!) because overriding the logic isn't possible!
I re read the Taipeh report, and it is cristal clear in its conclusions and recommendations. (although I m amazed about the poor English) It was AB's decision not to make the adaptation mandatory, and the aviation authorities apparently didn't see it fit to override this decision. Had they overriden it, perhaps this accident would never have happened.




Last edited by borghha; 14th Aug 2007 at 14:32. Reason: Typo/complement
borghha is offline  
Old 14th Aug 2007, 14:52
  #1640 (permalink)  
 
Join Date: Jul 2005
Location: btw SAMAR and TOSPA
Posts: 565
Little have we read here about any specific training to land with 1 t/r inop at the airline concerned. Is there any? Did it apply to the crew concerned?

Nor have we read anything about the pre-landing briefing that may have included the post-touchdown procedure, because it seems the cvr script has a gap of 20 minutes?

Or wasn't there a briefing of the style, "I tell you what I am gonna do, when we are down" and the one pilot landed entirely on his own in Palegre, without sharing his procedure, and the other pilot landed on his own at CGH, without sharing his procedure?
threemiles is offline  

Thread Tools
Search this Thread

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

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