TAM A320 crash at Congonhas, Brazil
Join Date: Dec 2001
Location: Finland
Age: 44
Posts: 121
Likes: 0
Received 0 Likes
on
0 Posts
marciovp,
There is nothing peculiar about the acceleration of the other engine. It is only the autothrust trying to maintain the selected speed. As the other engine was put to IDLE, the remaining engine fights the sudden loss of energy, as the autothrottle still remained connected and tried to achieve the selected approach speed.
Tero
There is nothing peculiar about the acceleration of the other engine. It is only the autothrust trying to maintain the selected speed. As the other engine was put to IDLE, the remaining engine fights the sudden loss of energy, as the autothrottle still remained connected and tried to achieve the selected approach speed.
Tero
Join Date: Sep 2000
Location: Bielefeld, Germany
Posts: 955
Likes: 0
Received 0 Likes
on
0 Posts
Originally Posted by donstim
Originally Posted by PBL
Southwest went off the end of a snowy runway at Chicago Midway and killed a car passenger because deployment of reversers was significantly delayed over the assumed deployment time in their landing-distance calculation.
It is ironic that the NTSB itself is the most influential remaining proponent of the "single cause" explanation.
PBL
Join Date: Sep 2000
Location: Bielefeld, Germany
Posts: 955
Likes: 0
Received 0 Likes
on
0 Posts
bsieker has performed a Why-Because Analysis (WBA) of this accident from the information available so far. He has kindly made his interim WBA , including most importantly the Why-Because Graph (WBG) derived from the facts known so far, publically available. One may find it on the Compendium on Computer-Related Incidents with Commercial Aircraft, on the RVS WWW site at the University of Bielefeld:
http://www.rvs.uni-bielefeld.de/publ...s/TAM3054.html
This page also includes links to the CVR and FDR data, as well as this PPRuNe thread, since I judge the quality of technical discussion on the thread as sometimes very high, and some of it has obviously contributed to the analysis.
PBL
http://www.rvs.uni-bielefeld.de/publ...s/TAM3054.html
This page also includes links to the CVR and FDR data, as well as this PPRuNe thread, since I judge the quality of technical discussion on the thread as sometimes very high, and some of it has obviously contributed to the analysis.
PBL
Join Date: Jul 2007
Location: Germany
Posts: 556
Likes: 0
Received 0 Likes
on
0 Posts
PBL,
thanks for announcing my Why-Because-Analysis here and making it publicly available.
I'd like to thank my colleagues at Causalis Limited, Jan Sanders, Jörn Stuphorn and PBL, and the many contributors on this PPRuNe thread for their valuable input!
PBL's comments on the case on the web site are worth reading, as well as an introduction to WBA, which can also be found there.
I'm looking forward to hearing some comments on this analysis.
Bernd
thanks for announcing my Why-Because-Analysis here and making it publicly available.
I'd like to thank my colleagues at Causalis Limited, Jan Sanders, Jörn Stuphorn and PBL, and the many contributors on this PPRuNe thread for their valuable input!
PBL's comments on the case on the web site are worth reading, as well as an introduction to WBA, which can also be found there.
I'm looking forward to hearing some comments on this analysis.
Bernd
Last edited by bsieker; 7th Sep 2007 at 12:44. Reason: Added URL to WBA.
I support PPRuNe
Join Date: Jul 2007
Location: Belo Horizonte, Brazil
Posts: 162
Likes: 0
Received 0 Likes
on
0 Posts
News today
TAM mechanics made a deposition at a CPI of the House, and said that they did work on the maintenance of the A320 that crashed fours days before. That they did have the part that was needed for the right reverser and that it would take three hours to replace it. But it had to be done at night and it would be noisy because the engine would have to work. So he consulted his supervisor who advised him to just lock the right reverser because according to Airbus the plane could fly ten days without it working.
Also the technical guy from National Agency for Civil Aviation (ANAC) said that it was decided internally that planes to land in Congonhas had to have both reversers and that this direction was published in their site. He believed that this was to be followed but for some reason another director of ANAC said that this was not so and TAM said that they followed the Airbus directives, not ANAC. What a mess...
Before the three wheels touched the ground?
Thanks.
Also the technical guy from National Agency for Civil Aviation (ANAC) said that it was decided internally that planes to land in Congonhas had to have both reversers and that this direction was published in their site. He believed that this was to be followed but for some reason another director of ANAC said that this was not so and TAM said that they followed the Airbus directives, not ANAC. What a mess...
There is nothing peculiar about the acceleration of the other engine. It is only the autothrust trying to maintain the selected speed. As the other engine was put to IDLE, the remaining engine fights the sudden loss of energy, as the autothrottle still remained connected and tried to achieve the selected approach speed.
Thanks.
Guest
Posts: n/a
Before the three wheels touched the ground?
Thanks
Thanks
At the flare at least one of the TLs (the left one) was brought to idle.
At the same time the other one either was brought to idle but its TLA reading remained in CLB or remained also phisically at CLB.
In such a situation be the wheels on the ground or not the AutoThrust will rely in just one engine to keep assigned speed. Then the right engine will show a thrust increase.
I don't know whether this thrust increase should be limited by the right TLA reading or not in this case.
Join Date: Jul 2007
Location: Germany
Posts: 556
Likes: 0
Received 0 Likes
on
0 Posts
Originally Posted by marciovp
TAM mechanics made a deposition at a CPI of the House, and said that they did work on the maintenance of the A320 that crashed fours days before. That they did have the part that was needed for the right reverser and that it would take three hours to replace it. But it had to be done at night and it would be noisy because the engine would have to work. So he consulted his supervisor who advised him to just lock the right reverser because according to Airbus the plane could fly ten days without it working.
Also the technical guy from National Agency for Civil Aviation (ANAC) said that it was decided internally that planes to land in Congonhas had to have both reversers and that this direction was published in their site. He believed that this was to be followed but for some reason another director of ANAC said that this was not so and TAM said that they followed the Airbus directives, not ANAC. What a mess...
Originally Posted by teropa
There is nothing peculiar about the acceleration of the other engine. It is only the autothrust trying to maintain the selected speed. As the other engine was put to IDLE, the remaining engine fights the sudden loss of energy, as the autothrottle still remained connected and tried to achieve the selected approach speed.
Originally Posted by marciovp
Before the three wheels touched the ground?
Bernd
Guest
Posts: n/a
In the case of the TAM 3054, it seems that A/THR thrust assignment to the right engine wasn't limited by the right TLA reading.
Marcio,
Does it have an identifier (portaria, resolucao, etc) and number ?
Also the technical guy from National Agency for Civil Aviation (ANAC) said that it was decided internally that planes to land in Congonhas had to have both reversers and that this direction was published in their site. He believed that this was to be followed but for some reason another director of ANAC said that this was not so and TAM said that they followed the Airbus directives, not ANAC. What a mess...
Does it have an identifier (portaria, resolucao, etc) and number ?
Join Date: Jul 2007
Location: Germany
Posts: 556
Likes: 0
Received 0 Likes
on
0 Posts
In the case of the TAM 3054, it seems that A/THR thrust assignment to the right engine wasn't limited by the right TLA reading.
Some paragraphs in the FCOM, where the thrust-limiting function of the thrust lever position with autothrust active is described in the singular form, confirm that these limits are independent.
Bernd
Last edited by bsieker; 7th Sep 2007 at 16:07.
Guest
Posts: n/a
Well, the right engine thrust would be limited to the right engine TLA reading (thus, to CLIMB power), I suppose you meant limited by the left TLA.
But before A/THR desengagement the right engine thrust was (apparently) above climb power when the left TL was brought to idle.
Join Date: Aug 2002
Location: Switzerland, Singapore
Posts: 1,309
Received 0 Likes
on
0 Posts
Please, don't start it all over again!
It is absolutely per design of an A320 (as of any aircraft with autothrottle) that the aircraft wants to maintain it's ordered speed, even when there is one engine not providing any forward trust anymore.
Imagine the same aircraft in another scenario: aircraft shortly before landing, pilots see that they cannot stop it, so they order a Go Around and shuffle in some power. Unfortunately, one engine quits, or even worse, it goes to revers (a very feared case which has happened a few times, although never on a A320). So they are hanging barely in the air, 20 ft above the runway, the end rapidly approaching. You can be happy that the other engine is providing trust, i.e. the ordered speed, and both pilots prey that it will be enough to get away from the ground.
You have to learn that a machine doesn't know the intention of a pilot nor does it know the outcome of the case. It's just there, designed to work under every thinkable circumstance.
Likewise it has to be repeated that all data given in the published data recorder are completly consistent and logical with each other. There is hardly any other convincing theory other than that the system itself worked completly as it was designed. There most probably (I know that some are attacking me for early conclusions) was never any technical problem on this aircraft (except the unservicable reverser 2), and that all went down the drain with one single action, the TL 2 being in CLB detent.
I understand that our Brazilian friends try to find also other channels of information. Lots of them are in a hectic manner to produce "evidence" and strong action so it can never happen again. This is understandable. Still I believe rather a set of data of a FDR than an commision of a parliament or an "enhanced audio tape". Bear in mind that the FDR data has been "issued" by the Brazilian Air Authority. There might be the case of falsifying it, but as previous said, the data is perfectly consistent and logical.
Dani
It is absolutely per design of an A320 (as of any aircraft with autothrottle) that the aircraft wants to maintain it's ordered speed, even when there is one engine not providing any forward trust anymore.
Imagine the same aircraft in another scenario: aircraft shortly before landing, pilots see that they cannot stop it, so they order a Go Around and shuffle in some power. Unfortunately, one engine quits, or even worse, it goes to revers (a very feared case which has happened a few times, although never on a A320). So they are hanging barely in the air, 20 ft above the runway, the end rapidly approaching. You can be happy that the other engine is providing trust, i.e. the ordered speed, and both pilots prey that it will be enough to get away from the ground.
You have to learn that a machine doesn't know the intention of a pilot nor does it know the outcome of the case. It's just there, designed to work under every thinkable circumstance.
Likewise it has to be repeated that all data given in the published data recorder are completly consistent and logical with each other. There is hardly any other convincing theory other than that the system itself worked completly as it was designed. There most probably (I know that some are attacking me for early conclusions) was never any technical problem on this aircraft (except the unservicable reverser 2), and that all went down the drain with one single action, the TL 2 being in CLB detent.
I understand that our Brazilian friends try to find also other channels of information. Lots of them are in a hectic manner to produce "evidence" and strong action so it can never happen again. This is understandable. Still I believe rather a set of data of a FDR than an commision of a parliament or an "enhanced audio tape". Bear in mind that the FDR data has been "issued" by the Brazilian Air Authority. There might be the case of falsifying it, but as previous said, the data is perfectly consistent and logical.
Dani
Join Date: Jul 2002
Location: UK
Posts: 3,093
Likes: 0
Received 0 Likes
on
0 Posts
Not quite, flyingnewbie10,
The TL was in the CLIMB detent, which is at the top end of the A/THR active zone of travel. SPEED mode kept that engine's thrust at approach thrust, not climb. If the engines were at climb thrust the aircraft would likely have sailed straight over the runway.
If I understand this right:
While A/THR was still connected, the thrust of the right engine crept up to balance the reduction in thrust from the left, which is normal for SPEED mode. When the A/THR disconnected, the actual engine thrust remained where it was at disconnect and did not change until impact.
The TL was in the CLIMB detent, which is at the top end of the A/THR active zone of travel. SPEED mode kept that engine's thrust at approach thrust, not climb. If the engines were at climb thrust the aircraft would likely have sailed straight over the runway.
If I understand this right:
While A/THR was still connected, the thrust of the right engine crept up to balance the reduction in thrust from the left, which is normal for SPEED mode. When the A/THR disconnected, the actual engine thrust remained where it was at disconnect and did not change until impact.
Guest
Posts: n/a
The TL was in the CLIMB detent, which is at the top end of the A/THR active zone of travel. SPEED mode kept that engine's thrust at approach thrust, not climb. If the engines were at climb thrust the aircraft would likely have sailed straight over the runway.
First someone said here that the TLA reading limits A/THR maximum thrust for each engine even in speed mode
After the left Tl was brought to idle and then to reverse, you can see in the graphs a considerable EPR increase in the right engine. It leaves approach thrust and surpasses climb EPR (apparently).
Then EPR in the right engine starts to reduce again until it gets fixed at (I suppose) CLB power.
It looks like that in a first moment the A/THR is somehow trying to compensate reverse thrust in the left engine, surpassing CLB thrust in the right one.
Maybe this "anomalous" thrust in the right engine is just derived from a momentaneous difference between ACTUAL EPR and TARGET EPR.
Guest
Posts: n/a
Still I believe rather a set of data of a FDR than an commision of a parliament or an "enhanced audio tape"
The pedestals have stops that probably make some noise when the TLs pass by them. This noise could be isolated in the analysis.
Join Date: Jul 2002
Location: UK
Posts: 3,093
Likes: 0
Received 0 Likes
on
0 Posts
And the Eng2 increase in thrust is notable, but it doesn't reach anything like climb thrust, If I'm reading that chart right.
I don't think such evidence would be admissible in court though.
Is it such a terrible thing to admit that the likeliest explanation is that the pilots made a mistake?
J.
My apologies but an Audio Spectrum Analysis of the CVR could show whether both TLs were brought to idle or not.
Is it such a terrible thing to admit that the likeliest explanation is that the pilots made a mistake?
J.
Join Date: Jul 2007
Location: Germany
Posts: 556
Likes: 0
Received 0 Likes
on
0 Posts
After the left Tl was brought to idle and then to reverse, you can see in the graphs a considerable EPR increase in the right engine. It leaves approach thrust and surpasses climb EPR (apparently).
It never exceeds climb power. It cannot with autothrust. (except Alpha-Floor, and one-engine-out, in which case the A/THR limit is MCT) .
Then EPR in the right engine starts to reduce again until it gets fixed at (I suppose) CLB power.
It is frozen at the power last commanded by autothrust. Ca. EPR 1.18
It looks like that in a first moment the A/THR is somehow trying to compensate reverse thrust in the left engine, surpassing CLB thrust in the right one.
Please compare the EPR values with the ones from the one Take-off that is also provided in the EPR graphs. (p11).
FLEX power is set for take-off, in this case delivering around EPR 1.35, at thrust reduction levers are brought to CL, activating autothrust, and setting (still in climb) CLIMB power, which in this case is around EPR 1.28.
During this landing, EPR never exceeded EPR 1.26, well below CLIMB power.
This has been discussed before, but glad to be of help, again
Bernd
Last edited by bsieker; 7th Sep 2007 at 16:16. Reason: typos
Guest
Posts: n/a
No. It is frozen at the power last commanded by autothrust. Ca. EPR 1.18
From what you said above I can suppose that the right engine thrust remained locked.
Is that it ?
Join Date: Jul 2007
Location: Germany
Posts: 556
Likes: 0
Received 0 Likes
on
0 Posts
From what you said above I can suppose that the right engine thrust remained locked.
Is that it ?
Is that it ?
In case of an "involuntary disconnect" of autothrust, i. e. any mode of disconnect other than bringing both levers to idle or pressing the pushbutton on the lever side, thrust remains locked. Until the levers are moved.
Bernd
Last edited by bsieker; 7th Sep 2007 at 21:03. Reason: Typo
Join Date: Oct 2005
Location: USA
Posts: 21
Likes: 0
Received 0 Likes
on
0 Posts
Quote:
Originally Posted by donstim
Quote:
Originally Posted by PBL
Southwest went off the end of a snowy runway at Chicago Midway and killed a car passenger because deployment of reversers was significantly delayed over the assumed deployment time in their landing-distance calculation.
There are many other factors that also contributed to that accident. Let's be careful in stating a single "cause," especially when the final accident report has not yet been published.
Let's also be careful in reading things into statements that aren't there. The word "because" does not imply "only".
Originally Posted by donstim
Quote:
Originally Posted by PBL
Southwest went off the end of a snowy runway at Chicago Midway and killed a car passenger because deployment of reversers was significantly delayed over the assumed deployment time in their landing-distance calculation.
There are many other factors that also contributed to that accident. Let's be careful in stating a single "cause," especially when the final accident report has not yet been published.
Let's also be careful in reading things into statements that aren't there. The word "because" does not imply "only".