TAM A320 crash at Congonhas, Brazil

[barit1]

Sometimes exhibits itself as an "orbiting" engine - at certain speeds the engine and FCU get into a sympathetic vibration or oscillation, usually not too severe.

Generally remedied by changing the FCU, even though the offending unit passes a bench check OK.

[marciovp]

Airbus

I have a 13 pages PDF file with all the parameters at the black box. I donīt know how to enter it here but I can send it by e-mail for the asking.

FROM AIRBUS INDUSTRIES:
ACCIDENT INFORMATION TELEX - ACCIDENT INFORMATION TELEX
The data which follow have been approved for release by the Brazilian investigation authorities.
It is confirmed that the aircraft was dispatched with the Engine 2 thrust reverser inoperative as authorized by the MEL.
It is confirmed that the associated operational procedure of TAM MEL was updated according to current MMEL page 02-78 p1 SEQ 001 REV 29 which reminds the crew to select both thrust levers to idle before touchdown and requires to select both reversers at touchdown.
The following is the sequence of events according to the recorders:
Final Approach phase
· The aircraft was approaching runway 35L.
· The last wind information given to the crew by the ATC was 330°/8kts.
· The runway condition given to the crew by the ATC was wet and slippery.
· Landing configuration was established with Slats/Flaps fully extended, gear down, ground spoilers armed, autobrake selected to MED.
· Approach speed was 145 kts
· The final approach was performed with Autopilot OFF - disconnected at about 370 feet (radio-altitude), Flight Directors ON, Auto-Thrust (ATHR) ON.
· The CM1 was the Pilot Flying.
· The crew approach briefing included a reminder that only the left engine thrust reverser was available.

Flare and touch-down
· During the flare, the "RETARD" call-out has been normally triggered
· The "RETARD" call-out has been triggered 3 times, ending at the selection of the engine 1 reverser.
· Before touchdown, the engine 1 throttle was retarded to idle.
· The engine 2 throttle is recorded in the Climb position and remained in this position to the end of recording.
· Preliminary trajectory computation indicates that the aircraft landed in the touch-down zone.
Landing roll
· Just after touch-down, idle reverse was selected on engine 1, followed within 2 seconds by the selection of max reverse which was kept to the end of recording.
· Following reverser 1 selection, the ATHR disconnected as per design and remained disconnected to the end of recording.
· With the engine 2 throttle being in the Climb position: 1/ the engine 2 EPR remained at a value of approximately 1.2 corresponding to the EPR at the time of ATHR disconnection; and 2/ the ground spoilers did not deploy and the autobrake was not activated.
· Maximum manual braking actions began 11 seconds after touch-down.
· Rudder inputs and differential braking have been applied during the landing roll.
· The aircraft overran the runway at approximately 100 kts.
DFDR and CVR data show no evidence of aircraft malfunction.
At this stage of the investigation, and as already indicated in the previous AIT n°3,
Airbus remind all operators to strictly comply with the following procedures:
A- During the flare at thrust reduction select ALL thrust levers to IDLE.
B- For the use of the thrust reversers when landing with one Engine Reverser inhibited
refer to :
· For A318/A319/A320/A321 MMEL 02-78 Page 1 Rev 29
· For A310 MMEL 02-78 Page 1 Rev 17
· For A300-600 MMEL 02-78 Page 1 Rev 15
· For A330 MMEL 02-78 Page 1 Rev 17
· For A340 200/300/500/600 MMEL 02-78 Page 1 Rev 19

[GlueBall]

marciovp: . . . obrigado

[armchairpilot94116]

Taipei vs Congonhas overrun comparisons

Taipei: Touch Down at 146kts at 1750 foot mark into 8550 foot runway (considered normal TD zone ) with 300 foot RESA

Congonhas: Touch Down at 145kts at X feet (but considered normal TD zone) onto 17R/ 35L runway of 6365 feet with 200 foot RESA.

TAipei: #1 T/R rev max, no spoilers, #2 at 1.08 EPR until shutdown.

Congonhas: #1 T/R rev max, no spoilers, #2 at 1.20 EPR until contact with building.

Taipei: Manual braking commenced after 15 seconds at 4800 feet into 8550 foot runway. A/C departed runway end at estimated 67kts.

Congonhas: Manual braking commenced after 11 seconds at X feet into 6365 foot runway. Departed runway at estimated 100kts.

Taipei: speed reduction of 79kts achieved before departure from end of runway at an average rate of about 47.5 feet used per knot reduction in ground speed, after application of manual braking. This is assuming fairly ineffectual speed reduction just with #1 reverse at max, until application of manual braking. Another 3183 feet of runway or so needed in theory to come to complete stop if using this rate of speed reduction.

Congonhas: similar aircraft, similar wet runway (except Taipei was not newly paved ), similar predicament. TAM A320 probably wouldve needed around another 4700 feet of runway to come to complete stop (give or take a wide margin due to different runway and not totally exact situ).

Taipei: RESA of about 300 feet (and not even EMAS) was enough to stop the A/C (with the aid of a ditch the NLG fell into). 67 knots to zero within 300 feet of RESA. AT the average rate of about 4.4 foot/knot ground travel to ground speed.

Congonhas: At this same rate of speed reduction would mean only about another 440 feet of RESA would've likely been enough to bring the 320 to a halt. Not exact of course, an estimation.

Thus the IFALPA recommendation of 240 meters x double runway width would have been enough to stop the TAM jet in its tracks. And especially if the RESA was EMAS.

17R/35L should be extended to 10,000 foot and with EMAS type RESA of 240 meters x double runway width to allow Congonhas to adequately contain a similar over run and avoid another similar disaster.



TAipei figures from final report (except my own foot used per knot calculations). Congonhas figures from whats been posted on this thread.

Please feel free to comment (even to tell me how wrong I am)

Looks like PF somehow left throttle 2 at the wrong spot, and the jet performed as DESIGNED accordingly.

Prevention would entail adequate training just for this exact scenario plus some revisions to the warning system perhaps. And certainly Congonhas could be a lot safer then it is, nobody can doubt that. thanks your attention

Point being? Runway excursions are going to happen again to AB and Boeing jets for various reasons. But runways need to be as long as possible and with EMAS type RESA for a chance at containment.

p.s. AB calculations for length of runway needed to stop in Taipei is substantially less then my guesstimate and my calculations are very simplistic. But what if AB calculations were actually exceeded a bit in the real world? My point was that it was going to take a lot more runway then was available in TAipei to stop, and dont even mention Congonhas. There was no way that jet was going to stop in time with what it was facing.

[armchairpilot94116]

Question for those in the know. I am really going to show my ignorance so forgive me. But I was wondering since the grass slows the aircraft better then the tarmac does apparently. And I am certain it is fraught with danger and is a total unknown. Aircraft running in the grass are totally out of control and can lose engines and wings and rolll upside down and catch fire , among other things. But if you were piloting that plane and found yourself in that very unfortunate and desparate situation of speeding down the tarmac and knowing that the jet wont stop and cant be stopped before you run out of runway. And knowing that the end is not going to be a good one. Would you steer your plane off the tarmac and into the grass alongside at say the last third of the runway and hope that this last desparate measure would work?

Or is that :

A. beyond stupidity to even consider
B. wouldnt possibly be a consideration because of A and because no training or such a thought would even cross your mind.

And finally would you think it would be preferable then to run straight off the end of the runway at Congonhas and into either the freeway below or buildings or both.

not a happy choice anyone should have to make, but would it work to save some lives?

[HotDog]

Yes, sometimes it works. Luft Hansa Cargo 742F did a groundloop at Kai Tak in 1983 after an aborted take off on RW13. It stopped him from going into the harbour.

[NigelOnDraft]

Armchair

17R/35L should be extended to 10,000 foot and with EMAS type RESA of 240 meters x double runway width to allow Congonhas to adequately contain a similar over run and avoid another similar disaster.

Take one look at theis airfield and say where on earth you are going to fit in all the extra runway Even if you give up extending the runway, but still want the EMAS, are the operators going to be happy with the loss in runway length

The "solution" to accidents is not to try and stop the particular occurance occurring again at e.g. CGH... it must be procedural / process changes etc. worldwide. So are you really advoctating closing all runways <10,000' long

[BOAC]

I would restate my call a few years ago for CVRs to be capable of holding up to 1 hour rather than the 30 minutes as now. With digital technology this should be easily achievable. That way we might have an insight into the briefing for the landing.

[NigelOnDraft]

http://www.tc.gc.ca/tcss/tsb/air/199...WISSAIR_E2.htm

Current technology easily accommodates increased CVR recording capacity. In fact, the majority of newly manufactured solid-state memory CVRs have a two-hour recording capacity, and there is a worldwide industry move towards two-hour CVRs. The European Joint Airworthiness Requirements specify that aircraft first certified after 01 April 1998 be fitted with two-hour CVRs. There is also a proposal to include such a requirement in the Standards and Recommended Practices of the International Civil Aviation Organization (ICAO). The ICAO Flight Recorder Panel, consisting of experts from a number of States, met on 12-20 November 1998, and recommended to ICAO's Air Navigation Commission that aircraft manufactured after 01 January 2003 be fitted with two-hour CVRs.

Sounds like they listened to you

[ELAC]

Congonhas: similar aircraft, similar wet runway (except Taipei was not newly paved ), similar predicament. TAM A320 probably wouldve needed around another 4700 feet of runway to come to complete stop (give or take a wide margin due to different runway and not totally exact situ).

Armchair,

I'm certainly all for arguments supporting EMAS, but the stopping distance numbers you've calculated are not correct. You are relying on a fixed number of knots per second deceleration rate to come to your conclusions but aircraft don't stop that way. First you need to bear in mind that this is an exercise in absorbing/dissipating the energy of the aircraft. Going back to basic physics Energy is given by E=mvē, where m = mass and v=velocity. So, an airplane at 140 kts. has just about twice the energy of one at 100 kts. (140ē/100ē = 19600/10000 = 1.96). Having slowed from 140 kts to 100 kts you haven't reduced the energy by 29% (40/140) but rather 49% (9600/19600).

The second point is that in a normal stop the effect of the various retarding forces varies depending on speed (reversers and spoilers have more effect at higher speed, brakes increase in effect as the aircraft slows) and the total available decelerating forces is consequently variable as the aircraft slows, not linear.

Having said that, in this instance the only stopping forces were the #1 reverser and the brakes. The effect of the #1 reverser would be effectively nil below 100 kts. so the remaining distance from there to fully stopped would rely totally on the brakes. Assuming maximum brake application their effect would normally be a function of the surface area of the tire in contact with the runway and the coefficient of friction depending on the runway contaminants. Lower coefficients could mean more incipient skidding, more anti-skid releases and so less brake pressure applied.

Compared to the approximate average of 1.5 m/sē deceleration (as per bsieker's calculation) that the longitudinal g readout shows on average from 140 kts to 100 kts the remaining deceleration would likely have been at a greater rate as reducing speed decreased lift produced, increasing the weight on the wheels and the tire surface area in contact with the runway. However, with the additional thrust from #2 it's also possible that the stopping energy required might have exceeded the brakes capacity. If that were to occur the deceleration rate would drop off at that point.

Also, be careful what conclusions you draw from the Airbus comparisons in the Taipei report. From page 144 of the report you have extracted the figure of 67 kts. as the over-run speed, but you aren't looking at the other columns of data, specifically that for scenario #2 ground stopping distance required is computed as 2380m versus 2072m distance available (it actually says 2154m, but the data for scenarios 2 & 3 were accidentally transposed). 2380m-2072m = 208m = 676' extra runway required, nowhere near the 3183' you had calculated based on a fixed 47.5 feet/kt rate. As you noted the aircraft actually came to a stop within the RESA after a 90š turn and a nosewheel in the ditch. That suggests that the Airbus numbers are probably pretty good if you ask me.

For this accident bsieker has computed a guess http://www.pprune.org/forums/showthr...69#post3466669 at stopping distance based on the average long g as being 1728m or perhaps a bit less. My own guess http://www.pprune.org/forums/showthr...03#post3465103 which was based on the published performance data and a complete guess about the effect of the #2 engine came to 1520m ground stopping distance required. Both are guesses, but since they derive from completely different data sets (observed long g vs. published performance + WAG) and yet are fairly consistent it's likely that the actual value that is eventually computed will be somewhere in the same range.

The amount of additional distance you've suggested of 4700' (~100 kts @ 47.5'/kt) is not logical. Assuming roughly 800m (2600') was used to stop prior to the runway end your total ground stopping distance would be 4700' + 2600' = 7300' (2246m). Though they are all guesses, that value would not conform to either the observed values or the published performance we've got at hand.

I'd suggest that you may want to rework your figures. The case for EMAS is a strong one, but you wouldn't want invalid calculations undermining your argument.

ELAC

[BOAC]

Originally Posted by ELAC

The effect of the #1 reverser would be effectively nil below 100 kts.

- not sure where you get that from? Certainly not true in my experience. Brakes would certainly be MORE effective below 100kts but reversers still provide significant reverse thrust down to a standstill.

NOD - Hmm! Loads of 'recommendations' but will it happen? According to my info the 737-600 and 800 operated in JAA should have it. What was the date of that rec?

[NigelOnDraft]

All BA's Airbuses are 2 hour, except the "Classics"

[ELAC]

Originally Posted by ELAC
The effect of the #1 reverser would be effectively nil below 100 kts.

- not sure where you get that from? Certainly not true in my experience. Brakes would certainly be MORE effective below 100kts but reversers still provide significant reverse thrust down to a standstill.

BOAC,

My observation is taken from the Airbus performance data shown in FCOM 2.03 along with a bit of practical judgement as to how it might apply to this unusual configuration. To give an example using the A330 data I have at hand:

Effect of 2 Reversers Operative on Min ALDdry (i.e max effort stop): -2% (RR Engines)
Effect of 2 Reversers Operative on Min ALDdry: -1% (GE Engines)
Effect of 2 Reversers Operative on Min ALDwet: -6% (RR)
Effect of 2 Reversers Operative on Min ALDwet: -4% (GE)
Effect of 2 Reversers Operative on A/B Med ALDdry: 0% (RR)
Effect of 2 Reversers Operative on A/B Med ALDdry: 0% (GE)
Effect of 2 Reversers Operative on A/B Med ALDwet: -1% (RR)
Effect of 2 Reversers Operative on A/B Med ALDwet: -2% (GE)

So for the above cases the maximum effect of both reversers operating was 6% for a max braking effort on a wet runway. There is obviously variation between different engines as the comparison between the RR & GE engines shows, and I don't have the exact values for the V2500 engined A320 at hand. However, my practical experience with A320s (CFM engined mind you) was that the reversers mostly made noise and didn't seem to contribute as much to a stop as the engines on the A330. That's a personal observation, I can't prove it at the moment.

However, using the above numbers as a rough guidline the most you could expect out of 2 reversers would be a 6% reduction in ALD, and hence the maximum contribution for 1 reverser would be 3%. In fact in the configuration as at CGH it's actual effect would likely be less as control issues at lower speeds (wet with thrust assymetric) would probably cause you to cancel it or reduce braking on that side.

Anyway, accepting that 3% was the likely maximum reduction in landing distance for the single reverser, and noting that the effect of reverse is most significant at higher speeds, it's probably reasonable to guess that 2/3's of the effect is gained in the high speed part of the landing above 100 kts. So, that would leave us with 1/3 or 1% of the landing distance as a probable maximum remaining contribution of the #1 reverser below 100 kts. By my calculation the total ALD for the situation was about 1800m. If effect of reverse below 100 kts. is 1% of the ALD then that comes to 18m. And that value is only what it is because of the increased ALD due to the #2 engine. In a normal wet runway case the effect would come to 11m and change.

Given that we are looking at the contribution to landing distance of the #2 engine at 1.2 EPR being somewhere in the range of 500m-700m, 18m one way or the other is not significant to the result as it is far below the reasonable margin of error in the baseline assumptions. Perhaps "effectively nil" was a poor choice of words on my part. There is likely (barring cancelling due to control issues) some contribution that would be made by the reverser below 100 kts, it just isn't going to change the calculation to any significant degree.

That said, I'd agree that if those 18m were all that was seperating you from the drop at the end of the runway you'd consider the effect of the reverser very significant indeed.

ELAC

PS - What appears to be significant in a normal landing with modest use of brakes and what is actually significant when all braking is applied to the maximum are two different things. We pilots often make inaccurate estimations of performance by what we observe in normal operation. Next time you take off somewhere high, hot and with obstacles (say LAS) ask your colleague if he thinks the A/C would meet the SID climb gradients on 1 engine. Odds are he'll say yes without any supporting evidence since it isn't easily accessible information. In fact, a lot of the time the answer will be no, but we are so used to the airplane's normal behavior that we don't estimate the limiting situations very well.

[NigelOnDraft]

Next time you take off somewhere high, hot and with obstacles (say LAS) ask your colleague if he thinks the A/C would meet the SID climb gradients on 1 engine.

Sometimes couldn't make them in an A340 on 4 Engines at a low cool airport

[atakacs]

@ELAC

enjoying your most educated and well redacted posts

noting that the effect of reverse is most significant at higher speeds

Not sure to fully understand why that would be the case ?

[fly nice]

1.20 Epr

Quote:
With the engine 2 throttle being in the Climb position: 1/ the engine 2 EPR remained at a value of approximately 1.2 corresponding to the EPR at the time of ATHR disconnection; and 2/ the ground spoilers did not deploy and the autobrake was not activated. (thread # 1529)

I assume this EPR value was retained till the end of the recording?

The only EPR aircraft that I am familiar with are either low bypass (JT8D) in which case 1.20 EPR would not be significant, or high bypass (RR Trent 556) where 1.20 EPR is quite a handfull.

Could someone give me an indication what kind of N1 / thrust this gives you on the A320?

[Gigajoules]

Post 1537:

The physics shown are not quite correct. Accelleration is not equal to velocity squared. Accelleration is change in velocity per unit time, thus, units for velocity is distance per unit time (i.e m/s) and thus, accelleration will have units (distance per unit time) per unit time (i.e. m/s^2). Energy is given by E=mv^2, where m = mass and v=velocity. So, the calculation gave the right answer but using F=ma will not give that answer.

[ELAC]

noting that the effect of reverse is most significant at higher speeds

Not sure to fully understand why that would be the case ?

atakacs,

Not an expert on this, but I think there are 2 reasons.

First, for a fixed deceleration rate you will spend more time in the high speed regime as a function of total stopping time versus at lower speeds.

Second, at higher speed more air volume passes the fan for a given rotating speed during a given time and hence with reversers redirecting the flow more reverse airflow is generated.

ELAC

[ELAC]

Energy is given by E=mv^2, where m = mass and v=velocity. So, the calculation gave the right answer but using F=ma will not give that answer.

Gigajoules,

Thanks for the correction. Post edited to reflect the correct formula.

ELAC

[fly nice]

Ok E=MC2

But how much forward thrust does 1.20 EPR produce. You are all doing arbitrary math, but at a given drag coefficient, 1.20 EPR will keep that speed indefinitely; OR NOT?

Could someone enlighten us as to how much thrust 1.20 EPR produces on the A320? N1 or otherwise. . .