TAM A320 crash at Congonhas, Brazil

[BOAC]

OK - let me re-phrase #1493: what EPR would you expect to see on a full reverse?

[Zeke]

Originally Posted by PBL

I have been informed that it has been released to the public by the Brazilian aviation authorities.

Link to the FDR http://download.uol.com.br/fernandor...r_graficos.pdf (requires PDF reader)

[discountinvestigator]

With regards to deceleration, I think you would expect 1.7 m/s/s (or 0.17g in real units) and 3.0 m/s/s (or 0.3g) for Lo and Med on an Airbus, I seem to remember but I do not have the manuals in front of me. With Medium brakes selected you should still get 0.3g on a dry runway even with one engine EPR around 1.20. Don't go near the tyres afterwards and look at the colour of the carbon brakes! Same effect with full manual braking on a dry runway, both make the assumption that anti-skid is working.

Please be aware of the time frame and data recording buffer differences in this data. Please be aware of draft comparisons of FDR and CVR time.

PBL, I support your reasoning for the comments relating to making the investigation more public. There are others too, but I am not willing to discuss them here.

Secondly, there have been a few studies which show that pilot reaction time to engine failure on take-off for application of full brakes is around four seconds. I can see that figure and logical brain processing time being similar here for spoiler non-activation to brake application. As an academic, you might be able to find an old NASA contractor report, or similar, done in the late seventies or early eighties on this. They were investigating why so many aircraft fell off the end of runways with V1 RTOs. The answer was because the certification decision time was one second and the pilots took four seconds to react. The good news is that they increased certification reaction time for the pilots to two seconds.... now does this explain why we need runway strips and runway end safety areas?

As a side issue, they also found that pilots did not press the brake pedals fully, but that issue was included in training afterwards. Boeing lets you apply more brakes and only disconnects when you apply more than the autobrake setting. I think that Airbus disconnects on sensing foot pedal pressure, so actually applying manual braking can release the brake pressure and then apply more.

Finally, if you are going to do a WBA on this, can I join in please?

Discount

[marciovp]

Confusion and question

Let me first thank you all, especially the pros, for tolerating my presence here, a not too young pilot of Tripacers and Cherokees in the weekends.
But in my favor let me say that aviation remains my first love...

The CPI of the House heard the representative of Airbus that stated in no uncertain terms that there was nothing wrong with the A-320...and then said that CENIPA (The Air Force Investigative and Preventive Agency) was in agreement with the statement that Airbus released about the accident. The Representatives in the Committee were surprised by this support from CENIPA that kepp telling them that the causes of the accident have not been found with certainty... They are now asking CENIPA to come back for them to ask questions.

I found it strange that nobody asked the Airbus guy about the improvements in the warning system after the Taipei accident and why that was not made mandatory.

Finally, going from my amateur experience with small planes it is difficult for me to believe that when close to landing a pilot would move only one T/L to neutral. It seems to me so odd. And a question: what would happen if he had moved both T/ls to REV?

Many thanks to you all.

[Lemurian]

PBL, ELAC, bsieker

Please compare the data on
1/- the mag heading
2/- the localiser
3/- the lateral acceleration
4/- the rudder - and brake - pedal inputs
5/- the ground speed, i.e. the deceleration.

It looks to me that the airplane started skidding as early as :30, some 20 degrees to the left.
That runway was indeed very slippery !.

Another remark is the control inputs . They are coming, later on the roll from the " wrong " side, i.e the F/O's. The left side stick was left pretty much neutral after touch-down.

Final remark : the touch-down occured -even if we took the #2 glide antenna -just or before passing abeam the glide path transmitter, confirming that the HP had managed to effectively land short (the one dot low call-out ).

Just my two € cents.

[PAXboy]

marciovp

I found it strange that nobody asked the Airbus guy about the improvements in the warning system after the Taipei accident and why that was not made mandatory.

At a guess, by an outsider: because the House of Representatives would ask questions about their own liability and to make it look as if they are doing something.

It will be for the formal inquiry to ask the detailed technical questions. It is usual for the final report on a major event, like this, will take two years to research and produce, but it can be longer.

[bsieker]

Originally Posted by Lemurian

Please compare the data on
1/- the mag heading
2/- the localiser
3/- the lateral acceleration
4/- the rudder - and brake - pedal inputs
5/- the ground speed, i.e. the deceleration.

It looks to me that the airplane started skidding as early as :30, some 20 degrees to the left.
That runway was indeed very slippery !.

Interesting observations. Here's my take on it.

1/ the magnetic heading:

It only deviates significantly in one direction (as opposed to just fluctuating in both directions in the time before) at :43, which is most likely where the aircraft started veering off the runway.

2/ the localizer:

Where exactly is the transmitter located? Any deviation from the localizer path would be bigger the closer you get towards it. And a slight deviation from the centreline would be shown exaggerated. I stand corrected on this one, if it turns out to be otherwise.

3/, 4/ rudder/brake pedals/accel-lat:

I think these are normal deviations we'd expect to see given a wet (but not slippery) runway, with asymmetric thrust, and the PF trying to keep the aicraft centered with the rudder, which interferes with the braking action.

5/ deceleration:

the deceleration is best seen directly at the accel-long graph. It is within the range I'd expect on wet surface with less-than-optimal downforce (high-lift devices extended, but no spoilers). Remember that deceleration force ("friction") is proportional to the downforce on the wheels. There is also no indication that anti-skid was inoperative. The brake pressure graph might shed light on that, but I think it is not among the released data. (And it would be difficult to interpret, since it is only sampled every 4s.)

Another remark is the control inputs . They are coming, later on the roll from the " wrong " side, i.e the F/O's. The left side stick was left pretty much neutral after touch-down.

Yes, I was wondering about that, too, but I have no idea what it might mean.

Bernd

[stickyb]

Spoilers

Could somebody that knows comment on the traces for Spoiler position and Ground spoiler armed status on page 4 ?

[bsieker]

Could somebody that knows comment on the traces for Spoiler position and Ground spoiler armed status on page 4 ?

Spoiler surface position is the roll-function that the spoilers have (supporting the ailerons.) It coincides with the lateral stick inputs. On the ground, attitude control is in direct law.

No idea why they would disarm the ground spoilers.

[bsieker]

Originally Posted by marciovp

I found it strange that nobody asked the Airbus guy about the improvements in the warning system after the Taipei accident and why that was not made mandatory.

This accident lets it appear in another light, since it was a catastrophe, but the previous accidents with "forgotten" thrust levers were less serious.

One (of two) wasn't even a hull loss, and in the other one, no-one on board died. Three people on the ground died, which is a tragedy, and three too many dead people, but still maybe the problem wasn't considered serious enough to mandate an airworthiness directive, specially since it involved violation of SOPs. This may change now.

[PBL]

Originally Posted by Sdruvss

"...Assim, há a hipótese de que o manete teria sido movido corretamente, mas por algum motivo isso não foi detectado. A falha pode ser eletrônica, dos sensores ou até mecânica."

The chances that both those thrust lever angle sensors were working perfectly throughout the aircraft's history, up to that point of touchdown, and then both simultaneously failed with exactly the same incorrect reading, without giving any sign of failure, are so small that I don't imagine people could give it much credence, unless something else turns up that indicates that this was the case. The chances that people leave the thrust lever up by mistake are much higher. After all, we already have two previous instances of that.

PBL

[leilas23]

Airbus deposition

Yannick Malinge was questioned in Congress about the new alarm system that would be developed after Taipei. Authorities there considered it a risk (but not a cause) that the retard signal ceased before TL´s were properly moved to idle. Airbus promissed the new alarm (a new warning light and aural signal) and Malinge said that certification authorities did not agree with this because, in the space of a few seconds during flare and touchdown, the alarm was not believed to contribute to an understandable quick reaction. There were no further explanations. Also, he made it clear that moving both TL´s to rev max (including inop side) was included in operational procedures in april 2006 -- before that, one would leave TL for inop reverser at idle. I found this to be quite an interesting bit of information.

[borghha]

sdruvs wrote

“[...]Mas a leitura indica que o manete ficou em quase 25º --isso equivale a uma forte aceleração ("max climb"). Enquanto isso, o manete esquerdo foi trazido para reverso máximo, posição a -20º, como ocorrera em Porto Alegre. "Esquecer" o manete nessa posição durante um pouso é considerado por pilotos um erro primário e improvável. Assim, há a hipótese de que o manete teria sido movido corretamente, mas por algum motivo isso não foi detectado. A falha pode ser eletrônica, dos sensores ou até mecânica.”

Here you are:

But the reading indicates that the lever remained at almost 25° - this amounts to a strong acceleration (Max climb).- Meanwhile, the left lever was pulled to maximum reverse, position -20°, as occurred in Porto Alegre. To 'forget' the lever in this position during landing is deemed by pilots to be a serious and unlikely error. Therefore there is an hypothesis that the lever was moved in the correct way, but for some reason this was not detected. This fault could be electronic, in the sensors or even mechanical.

[lomapaseo]

I found it strange that nobody asked the Airbus guy about the improvements in the warning system after the Taipei accident and why that was not made mandatory.

This is not an Airbus decision, this is a regulator decision. Airbus is like the farmer who can only lead the horse to the water but is unable to make him drink

[agusaleale]

Everything resumes in a pilot error or plane that fails, don´t you agree?

I think that we should pose two questions then:

1) If pilot error, what happened to them, what factors contributed to the fatal error (given the good previous landings with the same plane)

2) If not, then, what happened with the plane, what was the problem?

Almost everything I read, conducts to the hypothesis of a pilot error, as it has been stated (too premature according to me) that the Airbus had no failure.

[barit1]

This is not an Airbus decision, this is a regulator decision. Airbus is like the farmer who can only lead the horse to the water but is unable to make him drink

I've processed my share of service bulletins, both FAA-mandated and OEM-recommended. While the OEM may not be able to bring the same forces to bear on an operator, he can make bulletin compliance a condition for continued guarantee or warranty coverage. This is a pretty powerful lever to use on an operator when the manufacturer feels strongly about the change.

If Airbus didn't bring this force to bear on operator, they evidently didn't view it a significant safety issue.

[bubbers44]

The previous DFDR post is interesting that in three consecutive landings, three different procedures were used regarding the TR inop #2 engine. First one #2 idle, second one #2 fullreverse TLA as per SOP, the last showing not bringing #2 to idle at all. Either they forgot or they did and the sensors thought they were not at idle.

In training I have never flew an approach with a reverser locked out so doubt if any other airline emphasizes it either. All my reverser locked out approaches were flying the line, usually on normal runways.

Since reversers don't normally count for landing distance requirements they don't get much attention. If they work fine, if they don't work fine.

When is the last time any pilot here had a sim session landing with one reverser locked out?..................I rest my case.

[SyEng]

Ref p9 of the FDR plots. There seems to be a disparity in recorded TLA when you would expect them to be on the idle stop (0° on 1 vs. ~-3° on 2) during flare and at the end of the landing roll. This is the previous landing at Porto Alegre where the PF apparently used the latest SOP with Rev Max on both i.e. no stagger expected at any time.

[Mad (Flt) Scientist]

Small offsets on instrumentation are to be expected, especially on something that isn't recalibrated frequently. A couple of degrees is also not significant relative to the "idle band" used by the GS deployment logic, nor significant relevant to the discrepancy between measured TLA position and where it "should have been".

While I can't speak for AB, small discrepancies like that are pretty common - and are one reason why plots like those released/leaked, which invariably don't contain the data one would require to properly calibrate any of the parameters, must be read cautiously. When reviewing a real FDR properly, the first task is to ensure that the data has not offsets or calibration problems, and to account for them if it does. This 'advance' data may be 'raw' and not adjusted, so only gross information should be read from it until its properly validated.

[Lemurian]

flyingnewbie 10

a pilot flying the same a/c reported that when he started descending from FL 260 he detected a extreme vibration of the engine (which engine ?) at idle, without any indication that the A/THR was inactive, and that this vibration stoped when Thrust was increased.

And your conclusion is ...?