TACA aircraft crashed in Honduras

[bubbers44]

Our procedures on the B757 say to use max braking any time we feel it will make a safer landing. I never did because I didn't want to be the first pilot to try it on that airplane. Going to manual braking does the same thing after touchdown by using maximum braking. We were required to use position 4 or max, if required. Another pilot used max sometimes but I didn't want to be the test pilot since max is almost never used. Something must have happened after the 70 knot call when they were reducing reverse thrust to cause this accident.

[RWA]

Originally Posted by bubbers 44

if they were reducing max reverse thrust at 70 knots they were at that time satisfied that they didn't need that much reverse thrust any more.

I think the operative word in that is 'if,' bubbers44. Thanks to the 'wall of silence' that the investigators have erected (particularly as to what the DFDR shows) all we have to work on is a (plainly-incomplete) CVR transcript.

15:45:21.8 [CAM] (sound similar to engine spool up)
15:45:26.4 [CAM] (soun similar to engine spool down)
15:45:27.2 [SIC] SEVENTY Knots
15:45:28.3 [PIC] quíteme quíteme el….!
15:45:28.8 [CAM] (sound similar to throttle movement)

We don't know for certain that the 'sound similar to throttle movement' was in fact the pilot reducing or cancelling reverse thrust. And the sound came after he called for the F/O to 'cancel' something....

Originally Posted by bubbers44

Sometimes it is difficult to know how well the brakes are working until reversers are stowed.

Thanks for clarifying that.

I think that the most likely explanation, on the limited information we have, is the one you suggest; namely that they only realised after they reduced the reverse thrust that the brakes weren't working properly. In which case the captain might have been asking for the 'Anti/Skid' to be turned off?

Another (less likely) possibility is that he was trying (too late) to go around.

It's even possible that he did indeed cancel the reverse thrust, but (under pressure) neglected to double-check that both throttle levers were firmly in idle? Thus getting some highly-unwelcome forward thrust?

Further to that - just 'thinking aloud' - as I understand it, if both levers are not firmly in 'idle' the spoilers don't deploy on touchdown. Does the same thing apply the other way round; that is, if the levers aren't firmly in idle after cancelling the reverse thrust, would the spoilers retract?

[bsieker]

safetypee, bubbers44,

Autobrake MAX setting is different between Airbus and Boeing.

For Boeing it is the highest setting used for landing, and the setting used for takeoff is called "RTO", and is in a different position on the rotary switch (on the other side of "OFF"). That's how it is for the 737 classic and NG, and I assume it is similar on the 757.

On Airbus there are only 3 settings total, LO, MED and MAX. Approach SOPs say:

Originally Posted by A320 FCOM 3.03.18, SEQ 001, REV 36

[...]
AUTO BRK ..... AS RQRD

Use of autobrake is recommended.
Use of MAX mode is not recommended for landing.
On short or contaminated runways use MED mode
On long and dry runways, LO mode is recommended.

safetypee,

So I don't know about "poorly qualified", but I'd like to forward that compliment to the Airbus FCOM authors.

As bubbers44 has also pointed out, on a slippery runway, MAX autobrake will not produce harsh deceleration, the difference to MED perhaps (I could not find any information about this) being that MAX will commence braking at MLG touchdown, whereas MED has a 2 second delay. So, it may upset the cabin by quick derotation and possible hard NLG touchdown, even on contaminated runways. Manual braking at MLG touchdown, however, is permitted if deemed necessary.

On dry runways, at typical landing weights, although I agree that MAX autobrake would probably not damage the brakes and/or tyres, it would upset the cabin.

It still holds that I would not consider a landing "normal", if it is anticipated that dry patches (between wet/contaminated areas) capable of achieving in excess of 3m/s/s (~ 0.3 G) deceleration are required. For any surface not capable of achieving more then .3 G of deceleration, MAX will make no difference.

Again, in extreme cirumstances, say, landing on a very short runway in strong crosswinds (making immediate full manual braking more difficult), autobrake MAX can be used.


bubbers44,

do you know if MAX autobrake on Boeing does in fact maximum possible braking, and that perhaps the difference between RTO and MAX autobrake settings is only a delay after MLG touchdown (or other conditions, such as Reverse thrust / Idle thrust plus tyre spin)?

I apologize for my ignorance on the matter of autobrake systems on Boeing aircraft.


Bernd

[bsieker]

Originally Posted by RWA

It's even possible that he did indeed cancel the reverse thrust, but (under pressure) neglected to double-check that both throttle levers were firmly in idle? Thus getting some highly-unwelcome forward thrust?

This is a possibility, and I think it has happened before. The ergonomics of the unique A320 thrust levers are such that, coming out of reverse, it often happens that the thrust levers advance well into the forward thrust range. Most pilots therefore have the habit to pull the levers firmly back once out of reverse to avoid that.

However, there has been some confusion about the SOP action at the 70kts call. It is not to cancel reverse thrust, but rather to reduce reverse
thrust from maximum reverse to idle reverse, to keep open the option of engaging full reverse again, if necessary. Reversers are routinely stowed at taxi speed.

Further to that - just 'thinking aloud' - as I understand it, if both levers are not firmly in 'idle' the spoilers don't deploy on touchdown. Does the same thing apply the other way round; that is, if the levers aren't firmly in idle after cancelling the reverse thrust, would the spoilers retract?

This is not strictly true.

Short version:

Thrust levers below 15: -> Spoiler Deployment
Thrust levers above 20: -> Spoiler Retraction


Long version:

Thrust levers have to be "at or near idle or in reverse" for the spoilers to deploy.

"Near Idle" in this case means below 15 degree thrust lever angle.

To put that in perspective: Forward idle is "0" degrees, Maximum climb power ("CLIMB detent" in airbus-speak, also the fixed setting for normal autothrust operation) is at 22.5 degrees.

Effectively the spoilers deploy as soon as both thrust levers are well out of the CLIMB position.

Once extended, Spoilers only retract again if thrust levers are subsequently advanced above 20 degrees, i. e. almost into the CLIMB detent.


Bernd

[FlexibleResponse]

Again, in extreme circumstances, say, landing on a very short runway in strong crosswinds (making immediate full manual braking more difficult), auto brake MAX can be used.

Please be careful in what you say...this is in direct conflict with the Airbus FCOM recommendation not to use Auto brake Max for landing. I will remind all of an accident in DragonAir where an A320 with Auto brake max selected for landing at Kai Tak, departed runway 31, skidded across the grass and ended up on the parallel taxiway...heading for the "Nullah".

As an aside, having being the victim of the of the PNF selecting Auto brake Max switch instead of his display switch to Terrain on approach (once as operating crew and once as a pax) I can tell all that the autobrake only operates after about 3-5 seconds after touchdown and the deceleration is SEVERE!

[VONKLUFFEN]

???mmmhhhh!!!

Air France A340
Iberia A340
Brazil TAM A320
Tegucigalpa TACA A320
Chicago Mexicana A320

Five accidents involving Rwy runoffs in the last 2 years. Are pilots stupid or does Airbus Landing logic is not pilot friendly when work load is high? Nothing against the Bus but is fair to question why so many and in similar conditions.

[bsieker]

FlexibleResponse,

you're absolutely right.

What I meant with "Exteme circumstances" is an emergency which requires landing on the first possible hard surface available, such as a fire in the cabin or something like that.

I note with some amusement that I was being criticised for quoting Airbus procedures that say not to use MAX autobrake for landing, and now I'm being reminded to choose my words carefully (which I do) for saying that there may be circumstances in which MAX for landing can be justified.


Bernd

[FlexibleResponse]

Bernd,

And you are absolutely correct as well!

However, in extreme circumstances, my experience would guide me to pre-select AutoBrake Med, but immediately after touchdown, follow up with max manual braking and rudder pedal steering as the actual situation dictated.

Max Autobrake on landing with asymmetric residual lift on the wings seems to have a lateral directional control mind of it's own when it encounters differing runway friction coefficients between the left and right main landing gears.

(This doesn't seem to be a problem on RTO).

Regards!

[bsieker]

Originally Posted by VONKLUFFEN

Air France A340
Iberia A340
Brazil TAM A320
Tegucigalpa TACA A320
Chicago Mexicana A320

I think you are suffering from selective perception. Or have you forgotten about these runway excursions with Boeing aircraft, all within the last year and a half. Take away the first in the list, and it's 6 in under one year. Compared to your 5 Airbus incidents in 2 years. Note that I'm not at all implying that one type is safer than the other.


- March 2007: B737, Garuda Indonesia Flight 200, Yogyakarta
- November 2007: B737, Mandala, Malang
- January 2008: B737, Atlas Blue Deauville
- February 2008: B737, WestJet, Ottawa
- March 2008: B737, Adam Air, Batam
- March 2008: B737, RyanAir, Limoges
- May 2008: B747, Kalitta Air, Brussels (Not a landing, but still an overrun)


That's after just a brief search. I'm sure there were more, with other types, too.


Bernd

[Dream Land]

Vonkluffen

And at least the first two have nothing in common with the others.

[Leodis737]

VONKLUFFEN - pilot confusion / Airbus logic

I think there's enough knowledge about the Air France A340 to put our minds at rest on that one. The aircraft responded fine to the pilot's commands - the problems were: landing into a thunderstorm; rapidly changing winds; landing very long.

Tegucigalpa - despite the best efforts of those of us on this thread (!), I would say we genuinely don't know what the problem was yet, so this one should be left open for now.

Mexicana - again, I don't think there is enough information out - what's known so far is that there were CB clouds over the airport shortly before the incident, conditions were described as raining and gusting winds, and the plane finished with its nose gear in the EMAS and its main gear still on the runway - I have also seen one report that the ILS was out of service at the time on that runway (4R), although without further information it is hard to know what to do with that.

That leaves two:
- Iberia A340 - landed hard and broke the plane, brings into question whether the systems on that craft could usefully be made more robust or tolerant of such damage;
- TAM A320 - it seems the pilot left one thrust lever at climb power - changes to the Manuals have been made, will surely feature extensively in future training - raises the question of whether some further aural/visual warnings would be valuable.

Going back to your point about 'pilot confusion', it appears the pilot was confused in the TAM case, but I am not sure in any of the others. I have not seen a transcript for Quito (Iberia A340), which might reveal how much the pilots were aware of what was going on after their hard landing (but the video of the final stages of their rollout leaves me in awe - as a non-pilot).

[RWA]

Originally Posted by bsieker

"Near Idle" in this case means below 15 degree thrust lever angle.

To put that in perspective: Forward idle is "0" degrees, Maximum climb power ("CLIMB detent" in airbus-speak, also the fixed setting for normal autothrust operation) is at 22.5 degrees.

Hi Bernd, good to see you.

Must confess to being mystified by all these references (presumably emanating from Airbus) about 'degrees.' Might be easier to visualise things if they said how many centimetres (or millimetres) say 15 degrees of lever movement represents. At Congonhas, if I recall correctly, the DFDR apparently 'said' that one lever was around 20 degrees forward; at Phoenix it said that the captain had actually shoved one lever into TOGA (he denied it but the investigators chose to believe the DFDR).

It happens so often that one has to wonder how accurate the DFDR sensors are. And ALSO how on earth pilots manage to taxi A320s if you have to shove the throttle levers almost as far as the 'Climb Detent' before you get any power.

In any case, as far as wrong throttle lever positions are concerned, I believe that the problem is that if one is left out of idle, the autothrust remains engaged and reverts to the last commanded thrust setting; which would presumably be the power required to maintain maybe 140 knots on final approach with gear and lots of flaps down (again presumably, about 55% N1).

The point is well covered in Eric Parks' (to my mind) excellent no-nonsense guide to flying A320s - "Do not carry thrust to the flare as the autothrust will begin to command climb thrust as speed deteriorates if you do not bring back idle. This will cause a “thrust bump” that will have you floating down the runway with excess energy."

Thanks for the info about the spoilers. Though the fact remains that, when the aeroplane finished up, only one out of ten spoiler panels was deployed. I find it difficult to believe that no less than nine of them were somehow jolted into retracting by the impact. Given that they are built to withstand a 140-knot-plus slipstream, they are presumably pretty 'positively' locked in place - it's hard to believe that they can so easily be knocked flat in a crash.

Parks also has a word to say about the need to check that the throttle levers are solidly in 'Idle' after cancelling reverse thrust, too. He recommends stowing at 60 knots, but adds a caution about double-checking - "As the aircraft slows through 80 knots slowly push the thrust levers back toward idle reverse so as to be at or near idle reverse at 60 knots. Be sure you push the thrust lever all the way back through the detent into forward idle. Then retard the lever again against the stop to ensure minimum forward thrust in idle."

Airbus A319/320 Notes

Tony

[bsieker]

Hi Tony,

Originally Posted by RWA

Must confess to being mystified by all these references (presumably emanating from Airbus) about 'degrees.' Might be easier to visualise things if they said how many centimetres (or millimetres) say 15 degrees of lever movement represents. At Congonhas, if I recall correctly, the DFDR apparently 'said' that one lever was around 20 degrees forward; at Phoenix it said that the captain had actually shoved one lever into TOGA (he denied it but the investigators chose to believe the DFDR).

It happens so often that one has to wonder how accurate the DFDR sensors are.

The thrust levers rotate about an axis, so degrees seems like a useful way to measure their travel. As it happens, the travel of the grips is perhaps very roughly about half a centimetre per degree. The DFDR readout from Congonhas showed both levers at 22.5 degrees during all those flight phases where they would be expected to be at the CLIMB detent (i. e. from thrust reduction after takeoff, to the flare). So there is no doubt about the accuracy. It also showed one to remain unchanged at that position during the entire landing; the other being retarded and pulled into full reverse, as expected.

A sudden failure of the sensors/recording mechanism is not strictly impossible, but was deemed extremely remote, especially in the light of two known previous incidents/accidents, where the same thing happened, and the crew survived.

And ALSO how on earth pilots manage to taxi A320s if you have to shove the throttle levers almost as far as the 'Climb Detent' before you get any power. :-)

I don't know where that comes from. It's not that you don't get any power, it's only that the Spoilers retract only at higher power settings, which are normally only applied during a landing if you're going to perform a touch-and-go. Normally idle or slightly above idle power settings are sufficient for taxiing.

In any case, as far as wrong throttle lever positions are concerned, I believe that the problem is that if one is left out of idle, the autothrust remains engaged and reverts to the last commanded thrust setting; which would presumably be the power required to maintain maybe 140 knots on final approach with gear and lots of flaps down (again presumably, about 55% N1).

Not quite. What happens is this:

One thrust lever remains above idle, thus autothrust remains engaged, and spoilers do not deploy. Autothrust tries to maintain approach speed, and spools up the remaining engine that is not limited to idle (thrust in autothrust operation is limited to the thrust lever position). If the other engine is put into reverse, an error condition triggers autothrust disconnection (as can be read in the appendix to the Taipei runway overrun report), freezing the engine's thrust at the current thrust.

The point is well covered in Eric Parks' (to my mind) excellent no-nonsense guide to flying A320s - "Do not carry thrust to the flare as the autothrust will begin to command climb thrust as speed deteriorates if you do not bring back idle. This will cause a “thrust bump” that will have you floating down the runway with excess energy."

Autothrust will not command climb thrust, it will only add enough thrust to keep approach speed.

Thanks for the info about the spoilers. Though the fact remains that, when the aeroplane finished up, only one out of ten spoiler panels was deployed. I find it difficult to believe that no less than nine of them were somehow jolted into retracting by the impact. Given that they are built to withstand a 140-knot-plus slipstream, they are presumably pretty 'positively' locked in place - it's hard to believe that they can so easily be knocked flat in a crash.

Spoiler surfaces are not locked.

They retract if electric control is lost to a surface. If hydraulic power is lost, the surfaces stay at their current deflection unless pushed down by aerodynamic forces. So the most likely explanation is that the electric control was lost, and the spoilers retracted, except for the one, that might have been jammed.


Bernd

[alf5071h]

NVpilot, thanks (#381).
My approach in this thread has been to ‘devil’ the issues (mischievous), in order to understand some of the thoughts and logic being discussed – e.g. asking ‘why’ over five descending levels of explanation often uncovers interesting rationale.

You (and others) promote the use auto-brake, but not to be applied at every airport. Thus the circumstances / situation need to be established before the decision can be considered, and then, a decision might (should) consider several other variables, e.g. wind, rain, slipperiness and runway overrun. This involves an assessment process requiring an understanding of the nature of risks, e.g. what risks are there in the use auto brake (see previous discussions).
So exactly how do pilots judge when and where to use auto brake?
If pilots have to be proficient in manual braking in the more difficult conditions then why not ‘practice’ in normal conditions – at least a point for debate and perhaps speculation in the context of this accident.

Similar points can be made about the procedures when using thrust reverse (TR). The Airbus checklist does not require delay, select MAX at touchdown.
Why would a pilot feel committed with TR deployed on a short runway?
How many of the everyday normal operations on non limiting runways might not stop if the spoilers failed – how do you know; particularly if the runway is wet (or slippery)?
If the reluctance to select REV immediately is the fear of not being able to go around from the runway (spool up / TR not canceling), then consider probability leading up to that event – the combination of no spoiler and TR failure to cancel; I suggest that this is extremely improbable in comparison with landing 10kts fast and accepting a 10kt tailwind where the resultant landing distance on a wet runway might be a similar to that of not getting spoilers.
Those operators who use factored distances might have some additional margin over those who do not (crew additive), but any margin can be quickly eroded if braking or reverse is delayed.
If the concern about difficult runways remains – poor or no overrun area, then why not apply your own safety distance. Reduce the landing distance available in your decision making. This ‘addition’ is no more than a pilot is expected to decide on when adding a suitable (safe) margin, a margin which might cope with an improbable failure.

kwick, a new reference with a view on the human contribution in fatal accidents
Global Fatal Accident Review 1997-2006. (2.2mb)

[bubbers44]

Landing at TGU requires strict attention to runway conditions and tailwinds. All of these were ignored landing downwind on a 5400 ft runway with a cliff on the end wet with a 10 knot tailwind. It probably doesn't matter much if coming out of reverse at 70 knots finished the accident because it probably was going to happen anyway.

[Rananim]

I think you are suffering from selective perception. Or have you forgotten about these runway excursions with Boeing aircraft, all within the last year and a half. Take away the first in the list, and it's 6 in under one year. Compared to your 5 Airbus incidents in 2 years. Note that I'm not at all implying that one type is safer than the other.

Runway excursions on a conventional aircraft are down to pilot error or met conditions or combination of both.This is simply not the case with the AB.Too many cases where design has been cited in the probable cause and software updates been implemented.Warsaw(imagine a skipper landing his aircraft and he cant get wheelbrakes..crazy),Bilbao(a skipper commands a go-around and the computer says NO..more craziness),Adelaide(skipper gets no TR/spoiler because the computer relying on radalt readings decides they're in the wrong mode..)..just a few examples,there are more out there,all you have to do is google them.I dont accept Lemurian's explanation of the Cardiff incident..and neither does the AAIB who investigated it.Likewise at Ibiza,the exceptional actions of the skipper were what saved them from a watery end.AB's answer is always to blame the crew for misunderstanding the design.ie Bilbao,inputs to sidestick are additive..but the crew pulled back instinctively when sink rate increased,what would the designers have them do.Its an instinctive reaction.
And the latest two crashes in Congonhas and Toncontin..yes,pilot error will figure highly I suspect but what kind of error?The land-at-all-costs get- home-itis that we see in conventional aircraft accidents as in Burbank or Little Rock or something just a little more intriguing...what part did static thrust levers and no tactile feedback play in Congonhas?what part did brake(auto or man)anomalies play in Toncontin..why did he cancel rev at 70 if everything was going to hell...autobrakes,never use them,bar an exceptional cross-wind,and I always end up over-riding them.The pilot flies the plane,not the computer.

[DL-EDI]

In any case, as far as wrong throttle lever positions are concerned, I believe that the problem is that if one is left out of idle, the autothrust remains engaged

You seem to be imagining the crew having to place the thrust levers accurately at 14.9 degrees to allow the spoilers to deploy, thus making it sound hard to judge. As far as I'm aware, they can be pushed back "against the stops", pretty much every time (if not every time). In the cases discussed above, I'm pretty sure the errant levers were not left "slightly" out of idle but a long way out of idle.

[kwick]

alf5071, thanks for the link, now in my human performance information files.

By now, just read that work has started on the runway at Toncontin, some machinery is getting dirt out of surrounding hills in the south sector (rwy 02 approach) to have an additional 300 meters of usable asphalt.

[VONKLUFFEN]

Great...

feed back after my last post. Thanks for reminding me about Boeing incidents and accidents. Nice and "clear" explanations about AB logic. However, us sitting here with a cup of coffee in our hand is hard to believe and understand (literally)so many open windows and the ones that might get open if software takes over our humble brain. Let me explain.
If it is dark in my room I turn on the light. If it's hot I use air con. If I need to eat I cook on the stove or use the microwave. I don't need to think what if. I just do what it is easy to understand and then I can read, refresh my body and eat a nice meal. If I had to think how electricity works and where it comes from and how it gets to my house and if I don't want to read anymore what do I need to do turn it off correctly and not create a short circuit I would stumble with the table on a dark room because I would rather not turn it on in the first place. Same with the air con and cooking appliances, I would get a heat stroke and would be burned with the fire or cooked with the microwave just trying to get my food.
Again , it is NOT my intention to fight AB but still there must be something out there that keeps the pilot out of the loop completely in easy and complex situations. EX 15 deg. thrust setting! Who will check that ever or have the accuracy to do it when landing in the middle or a hurricane, storm , typhoon, bounce landing, you name it.
Many incidents and accidents happen because of unstable approaches. It would be interesting to consider only the ones where all was "normal" , try to compare them and see if we all get a clue where the problem might be either with Boeing or Ab.
Happy landings!

[Dream Land]

why did he cancel rev at 70 if everything was going to hell...

My own theory is that too many new pilots are programed like little robots to do things at certain phases of flight without thinking, I see it all the time.



Von, it's not as complicated as you make it out to be, until the TAM accident (and the one in the Philippines) I would have thought that closing the thrust levers when landing is a no brainer, I can assure you that no one is bringing up the 15 degree TLA in training and there is no real need to know this, also it's not a factor in any of the landing incidents.

I agree with Bubbers that this accident occured when a decision was made by the captain, nothing to do with events taking place on the runway.

Thanks Zeke for the diagram.