TACA aircraft crashed in Honduras

[Roy Bouchier]

Experience

As an old timer, It's refreshing to read Bubbers44 essentially practical comments on this topic rather than the airy fairy nonsense we hear so often.The remark that I feel so telling was that 'in the circling, perhaps they forgot.'
Now there speaks the voice of experience!

[Dream Land]

'in the circling, perhaps they forgot.'

Experience not on the Airbus I'm afraid, forgetting final flap is possible but not likely due to the display.

[Lemurian]

Clandestino

My perf manual shows ALD for 60t A320, on wet rwy, conf full, 10kt tailwind, at 3000 ft amsl to be 1600m. However I operate under JARs and RLD for me in this case would be 1.92xALDdry and that's 2120m.

Agreed !

Also 60t is an estimate. At the time being, we don't know if it was the real weight.

Of course not, but that is about the minimum weight I could imagine, not willing to go into intricate calculations based on pax weights, estimated cargo or so. I didn't even consider the weight of the deadheading crew.
All the performance derive from that - estimated - baseline value.
The perf margins I took went further as I took the touch-down Vref = Vls and the Vref speed will only vary by 1 kt / ton.
I have to say here that there is absolutely no value judgement on the crew from me. I am just saying that the margins for mishandling were narrow. Very narrow.

As for the -very popular on Prune - theory of a desperate attempt at a late go-around, the maths don't bear it out :
Estimated landing speed : 148 kts
Estimated speed abeam the camera : 100 kt
Estimated speed at the crash site : less than 40 kt
All point to a continuous, albeit insufficient deceleration, and use of brakes (smoking hot as per one of the initial videos, did they start a fire ?).

Now the theory that they had not full flaps. It is very possible, in a HF sense as the procedures on a circling approach call for config 2 / Gear down at the straight-in approach break and final selection of full flaps on base / final turn. Some airlines, though, call for full flaps throughout the circling. Need to knowe TACA's SOPs on this.
Bubbers theory could well be possible here (I don't believe I would ever see the day I'd agree with him ).

[DingerX]

[blockquote]Estimated speed abeam the camera : 100 kt[/blockquote]
To those who would suggest that the movement of the camera would exaggerate the speed, or that the perspective of the runway (going from abeam to moving away) would make it appear to be slowing down, here's how that figure is arrived at: in the 4 seconds it's on camera, the aircraft travels 5 plane lengths. Length of an A320 is 37 meters and change, so back-of-the-envelope gives 170 kph, or approximately 100 kt.

[Austrian Simon]

Originally Posted by DingerX

[blockquote]Estimated speed abeam the camera : 100 kt[/blockquote]
To those who would suggest that the movement of the camera would exaggerate the speed, or that the perspective of the runway (going from abeam to moving away) would make it appear to be slowing down, here's how that figure is arrived at: in the 4 seconds it's on camera, the aircraft travels 5 plane lengths. Length of an A320 is 37 meters and change, so back-of-the-envelope gives 170 kph, or approximately 100 kt.

We strongly disagree with the assessment by Flight Global, that the airplane was travelling at 100 knots at that point. If that was true, the airplane would definitely have stopped before the runway end, given that the airplane had already slowed down from their touch down speed of around about 150 knots (tailwind 10 knots plus reduced IAS due to density altitude taken into account). If those 100 knots were true, the airplane would have slowed down at a rate of 2.4 meters/second/second at minimum (touch down at the threshold) up to 4.5 meters/second/second in maximum (touch down at the aiming markers, 4.5 m/s/s or 9 knots/s being quite some more than maximum autobrakes) since touch down, suggesting that the airplane would have stopped between 450 to 700 meters before the runway end with that (continued and constant) deceleration.

See the computations and considerations here:

http://avherald.com/h?article=4077cedf/0022

If speed was indeed a factor, the airplane must have been quite a lot faster than 100 knots at that point 650 meters past the threshold and 1000 meters before the runway end.

I would rather wait for the data off the FDR ...

Servus, Simon

[RWA]

Originally Posted by j32wreck

The spoiler panel deployed appears to be panel 5. Panels 2-5 are used in flight to varying degrees.

Thanks. Can someone help me with this quotation for Eric Parkes' 'A320 Notes'?

Spoilers – Five spoilers are installed on each wing. From the wing root to wing tip they are numbered 1 through 5. All are used as ground spoilers. Numbers 2 through 5 (the 4 outboard spoilers) provide roll control. The middle three (2 – 4) provide in-flight speed brakes. If a SEC fails the spoiler(s) it controls is automatically retracted (if extended) and that spoiler(s) deactivated. There is no reversion to other computers.

Spoiler priorities:

Spoilers 1 & 2 - SEC 3, Yellow and Green
Spoilers 3 & 4 - SEC 1, Yellow and Blue
Spoiler 5 - SEC 2, Green

http://www.chipsplace.com/helpful/Ai...FlightControls

My main problem is that, as an old-time, strictly amateur aviator, I haven't the faintest idea what an 'SEC' is.

But the implication of the passage is that there is some sort of possible systems failure that can result in only Spoiler 5 being extended? As appears to have happened in this case?

[bubbers44]

Our procedure was to be fully configured on speed downwind for rwy 02. We rarely did the VorDme 02 approach because we arrived from the north. Occasionally we had to do as they did and go around and fly the 02 approach to descend to 5700 ft instead of 6000 to get below the cloud bases. I don't have any B757 charts since retirement but we landed there at max landing weight a lot when runways were dry because of the high fuel charge there. The wet landing weights were the same on 02 with a 5 knot headwind as landing on 20 with a 5 knot tailwind. I always landed on 20 with any tailwind on 02 because the wind changed a lot at TGU. We routinely taxied to 02 for take off with a 15 knot tailwind and waited for it to drop to what we needed that day depending on weight and temperature. It always worked out fine and 5 or 10 minutes was all it took. I appreciate the transcript of ATC communications. It clears up a lot of unknowns.

We used position 4 autobrakes on every landing there. It works the same as Airbus autobrakes and sets a desceleration rate. However that desceleration rate only applies if the runway conditions allow it to without activating the antiskid system. Reports of braking problems there since resurfacing the runway may play a part in the investigation. Thanks Lemurian. I always appreciated your input even when we didn't agree.

[J.O.]

It's highly unlikely that only one spoiler was deployed during the landing, unless they were facing a significant defect which should have precluded a landing on such a short runway, particularly when it was wet. It's more likely that the one deployed spoiler as seen in the pictures has just not drifted out of the extended position as hydraulic pressure bled off. It's also possible that the accident caused damage which blocked the hydraulic lines to that spoiler, thus negating the possibility for the fluid pressure to bleed off.

As for the speculation that they did not have full flaps for landing, as was mentioned above, "Flap 3" and "Flap Full" are the normal landing configurations for the A320. Prior to doing a Flap 3 landing, we are required to select LDG FLAP 3 on the EGPWS panel. If one doesn't make that selection, the EGPWS will be screaming "TOO LOW FLAPS" rather loudly on final approach. The crew would have to be in a coma to miss this call.

[Strongresolve]

Again I must say, that the exact translation of the word that the controler used to alert of runway condition is Damp, he said "pista humeda."
He didnt said wet. "Mojada." o "Contaminada."
I think that this is important because makes the diference between Required Landing Distance in Dry and Wet coditions, and it makes the diference between normal operation and pilot error in calculations of Required Landing Distance.

I dont know if he landed long or short, but the fact is that he pased the camera with three landing gears on the ground, and the reverses didnt were deployed. I´m anoyed about that.

I suppose that everybody affected by TGU closure and Airbus are going to blame the pilot for the accident based in his evaluation of landing runway conditions. Always is easier to blame the usual suspect to calm down the economical repercussions of an airplane crash.
Firsts declarations of the honduran Civil Aviation Authoroties said that the problable cause was a failure of the plane, when the notice about Toncontín closure was know they started to blame the pilot for landing in a wet runway with a lot of tail wind.

In this airport pilots usually prefer to land on the runway 02. For the runway 20 you need a very high rate of descent, and with a MDA of 2700 feet above the runway you dont have a good chance to go in if the day is cloudy. In Circling things are not much better, the turn is to the captains bad side, and you fly closer to the mountains. Another fact is that recomended rate of descents are not published in the VOR/DME App chart for runway 20. The exceed the margin for a stabilished approach.
Runway 02 is safer always, instead of tail wind.

[Dysag]

RWA

Since no-one replied to your question, here goes:

SEC is the Spoiler Elevator Computer. Since these moving surfaces are not connected physically to the sidestick, it's the SEC which controls which surfaces will deflect, and by how much, depending on the conditions. So it's best to have lots of active SECs!

You probably worked out that yellow, green and blue are the three hydraulic sytems.

[Austrian Simon]

Originally Posted by Strongresolve

Again I must say, that the exact translation of the word that the controler used to alert of runway condition is Damp, he said "pista humeda."

Thanks, I have put your remark in a comments in the translation. Was this already your second remark - didn't see your first!

Originally Posted by DingerX

in the 4 seconds it's on camera, the aircraft travels 5 plane lengths. Length of an A320 is 37 meters and change, so back-of-the-envelope gives 170 kph, or approximately 100 kt.

That approach is seriously, seriously flawed!

Let's make this clear in an assumed picture: the camera looks in a right angle to the runway, is fixed and the lens system makes sure, that all angles in the pictures are depicted correct from corner to corner. The camera looks from 30 degrees right to 30 degrees left, the camera is say 150 meters off the runway.

As the airplane comes into the view at the right hand corner of the picture, it is therefore at a distance to the camera of 150/cos(30 degrees)=173 meters, and it has to travel a distance of 150*sin(30 degrees)=75 meters to reach the center of the picture. Because of the larger distance the airplane appears smaller than it would be in the center due to the perspective, and because of the angle the distance of 37 meters in reality appears on the picture as 37*sin(30 degrees)=18.5 meters. The same applies for the left hand corner of the picture. The total distance on the runway centerline from right to left corner would be 150 meters or 8.1 times the airplane length measured at the corner or roughly 4 times the airplane length measured in the center of the picture.

Now, let's put the camera at an angle to the runway, say 30 degrees to the left. The center line of the runway at the middle of the picture would now be 173 meters away (the camera still 150 meters off the runway in the same position, just rotated). The airplane appears in the picture at its closest position to the camera, 150 meters distance and departs the picture at an angle of 60 degrees (as we have another right angle in the triangle, this time not in the center but at the right hand corner of the picture, we can use Pythagoras without any doubt), the airplane travels therefore 86.5 (173*sin(30)) meters until it hits the center of the picture. The distance of the runway center line at the left hand corner of the picture would appear at 300 meters (150/cos(60)), the total distance travelled from right to left hand corner of the picture would therefore be 260 meters (compare 86.5 meters from right corner to center point!). You can not apply any linear measurement in this view at all.

One can not simply say therefore, the airplane could be put 5 times into the picture from appearance to disappearance, 4 seconds, the airplane is 37 meters long and thus we see at a speed of 5*37/4 meters/second or 166 kph. The perspective of the picture and the angles of even a still camera (not to mention a panning one!) introduce gross errors to that approach.

Now, we aren't even able to determine the angles between the camera and the runway centerline with any degree of certainty in this scenario, but we would need to do so for every single frame of the video to be able to use those frames for the measurements. So the panning introduces a lot of additional difficulty into that determination.

All of this with an optimal camera, which a surveillance camera usually isn't (that aims to view a broader angle than would appear on the pic). You'll note, that the exact distance to the airplane is needed and the exact angles to be able to compute the actual distances covered. And we need to know the exact time stamps of the pictures to be able to really compute the time - that video doesn't provide times at all. We don't know, whether frames have been inserted or removed for a "smooth video", what time intervals were between the various pictures of the camera (maybe it was indeed 1 picture per second as it may look like).

There is a reason, why photogrammetry is a science!

Servus, Simon

[bubbers44]

To land straight in on 20 it was necessary to descend as soon and as rapidly as conditions allowed. At 15 DME descend to 7,000 ft and call the airport in sight or before 15 DME if able to get visual contact. Then get established so at 3miles, about 8 DME you are at 4300 ft. Sometimes S turning to a right base entry helped if you were a bit high. Otherwise you had to overfly the airport and make right traffic to 20 which was difficult from the left seat if you weren't familiar with the valley. 20 is an upslope runway so it is easy to get low if you just judge your descent by runway appearance alone. On really hot days you were weight restricted landing on 20 because of go around climb gradient restrictions because of terrain. Landing on 20 is like landing at any other airport if you start your visual approach early. Check airmen didn't count landing on 20 as meeting the requirements of a TGU check out so either they had to land on 02 with a legal tailwind or do another qualification flight.

[PJ2]

Post #144, Tree, Bubbers 44;

Quote:
We Boeing pilots don't have to delay reversing for a ground spoilers call because we have a handle that the PNF verifies or pulls so ground spoilers don't go through computer logic. We can reverse at touchdown. It is a nice safety feature.
I agree bubbers 44. Not being able to manually select ground spoilers with Config full is a deal breaker. Using Config 3 on short and slippery runways was to me an unsafe "workaround" due to increased speed and increased tendency to float.
My record: 10,000 + on short and slippery/icy runways in 732, about 1,000 in the 320. Caveat emptor.

It is a nice safety feature but there are also accidents which have resulted from manual selections of both systems on the 732 (reverse) and the DC8, (spoilers - more than one accident). It can be assumed that Airbus' intentions were to deal with the possibility of inadvertent deployment while ensuring their availability once the a/c was solidly on the ground and they used the technology around which the design concept was created to do so rather than defaulting to cable & pulley systems with associated mechanical interlocks.

I'm not trying to compare the two as such a discussion is profoundly pointless and a waste of time given the success of both models/types but merely trying to clarify why the design might be as it is on the Airbus.

I discussed the work-around here during the TAM accident discussion last year. The technique was discussed in terms of bringing the flap lever up one step after touchdown, (ie, Conf Full to Conf 3) to enable spoilers/brakes then reverse. Those that disagreed I don't think were aware or had not flown the 727 in which a one-step-up flap re-configuration was done routinely after touchdown. I agree it ought not to be a work-around but a manufacturer-sanctioned and available technique for crews faced with short, contaminated runways. I agree that there may be more politics and operational or technical risk involved and that the design was a point raised in the Lufthansa Warsaw overrun accident, (Conf 3 vs Conf Full on contaminated runways). I believe that Lufthansa (and others) went to Conf 3 as a standard flap configuration for landing shortly after. Config 3 is the "recommended" flap setting but we rarely see it in the data.

[RWA]

Originally Posted by Dysag

Since no-one replied to your question, here goes.

Many thanks, Dysag.

You probably worked out that yellow, green and blue are the three hydraulic sytems.

No chance, mate. Thanks to getting married early in life, I could only afford gliding for most of my flying 'career.'

So most often I didn't even have a bloody ENGINE to work with, leave alone hydraulics! Spoilers, yes - vital, landing a glider..........

[DingerX]

Dear Simon,

I was merely giving the methodology for "how" those numbers were arrived at, not making any statement about their veracity; indeed I did note the problems with moving cameras and angular speed.
For the record, the "plane-lenth" measurement involved counting those lengths traveled with reference to the ground, in each frame, not taking a general measurement.
I also agree that as the aircraft moves away from the camera, it will look slower than what it suggests.
I do not find your initial arguments conclusive, namely that if it were going at 100 kts at that point, it would not have overrun. That's putting the horse in front of the carriage, or using analysis to drive the facts.
You did, however, inspire me to go back and take a close look at the data.

First, you are factually wrong. The tape is time-stamped to the thousandth of a second. Second, the camera's perspective is good enough that, even with the panning, you can overlay two images without noticeable problems from distortion, which I've done. Third, you are intuitively correct: at the point it comes on the screen, it is going notably faster than 100 kts.

I took the first and third frames where the aircraft appears. The time-stamp of the first frame ends with 18.928; the third frame is stamped 20.421, so the interval is .593 seconds. Matching the vertical section of the first image that contains the aircraft with the scene in the second image, it is clear that the aircraft has traveled on the runway over a plane-length.
Assuming conservatively that's forty meters, that would come out 243 kph or 131 kts. I'm sure greater precision (and perhaps an even higher ground speed) can be had.
If you'd like to play with the image, and make the necessary angular motion calculations, or just see for yourself or post for others, drop me a line and I'll mail it to you.

So: 1) camera motion can be compensated for, 2) The image gives us precise time measurements, and 3) you'd still need to calculate angular effects, but back-of-the-envelope shows it's at least 130 kts.

[j32wreck]

As for the -very popular on Prune - theory of a desperate attempt at a late go-around, the maths don't bear it out :
Estimated landing speed : 148 kts
Estimated speed abeam the camera : 100 kt
Estimated speed at the crash site : less than 40 kt
All point to a continuous, albeit insufficient deceleration, and use of brakes (smoking hot as per one of the initial videos, did they start a fire ?).

Only one board deployed and engine 1 nowhere near full reverse.

[Austrian Simon]

Originally Posted by DingerX

I was merely giving the methodology for "how" those numbers were arrived at, not making any statement about their veracity; indeed I did note the problems with moving cameras and angular speed.

Alright, taken note of, and we are in agreement here.

I do not find your initial arguments conclusive, namely that if it were going at 100 kts at that point, it would not have overrun.

Not quite ;-)

If the airplane had slowed down to 100 knots already, from an approximate 148 knots or so, it would have exercised quite some braking already. If that braking action continued for some time, it would not have overrun. In fact, even on an ungrooved wet runway a deceleration of 1.5 meters/second/second (3 knots/s) is well normal, and I used that deceleration rate.

First, you are factually wrong. The tape is time-stamped to the thousandth of a second. Second, the camera's perspective is good enough that, even with the panning, you can overlay two images without noticeable problems from distortion, which I've done. Third, you are intuitively correct: at the point it comes on the screen, it is going notably faster than 100 kts.

Yep, absolutely - we are in agreement of the speed being higher than 100 knots.

And you are right, that there are timestamps overlaid by the copyright notice of the TV station that obviously broadcast it. I hadn't seen those until your remarks.

Can you recheck your time stamps, especially as you write 18.928 and then compute 0.593 seconds to 20.421, please?

The timestamps are hard to decipher, but the airplane is not visible at ...19.421, is fully visible at ...20.028, is last visible at ...22.031 and is entirely out of the picture at ... 22.431, so that makes it maximum 3 seconds visible.

Now add the angle between runway and camera to that scenario ...

If I however use 130 knots as a speed at that point 650 meters past the runway threshold with 1000 meters to go, a deceleration rate of 1.5 m/s/s would be insufficient to stop, stopping distance would be 1500 meters (departing the runway end at a speed of 38 meters/second=74 knots). A deceleration rate of 2.2 meters/second/second (~4.4 knots/s) would be necessary to stop at the runway end (which may not be available on a wet runway).

Assuming conservatively that's forty meters, that would come out 243 kph or 131 kts. I'm sure greater precision (and perhaps an even higher ground speed) can be had.

Photogrammetry can certainly clear that up. That was my main point.

So: 1) camera motion can be compensated for

No doubt about that.

2) The image gives us precise time measurements, and 3) you'd still need to calculate angular effects, but back-of-the-envelope shows it's at least 130 kts.

Agreed.

Servus, Simon

[Doors to Automatic]

What deceleration rates do Autobrake "low" and "medium" give on an A320 just out of interest (assuming dry runway; typical landing weight) and what sort of ground run distances do these translate to?

[DingerX]

In calculating the speed, I initially read the stamp as 19.928. In working it up, I looked at it closer, and I saw 19.828. Sorry, that's the source of the problem. I think we can put it at "at least" 130 kts.

[Austrian Simon]

Originally Posted by GobonaStick

Austrian Simon, why is it that the Spanish ATC clearly talks about runway "cero dos" but your translation says "two zero"? Some Spanish language quirk?

Thanks for pointing that out, I'll need to check with the Spanish source, as it is clear from the later discussion, that they were going for "dos cero", e.g.

15:26:02Z......
[TA390] TONCONTIN TORRE. TACA TRES NUEVE CERO. SOBREVOLANDO EL CAMPO.
PARA CINCO MIL QUINIENTOS.
[TGU TOWER] TACA TRES NUEVE CERO. CONTINUE APROXIMACION PISTA DOS CERO.
VIENTO DOS CERO CERO GRADOS. SEIS NUDOS. NOTIFIQUE EN FINAL.
[TA390] NOTIFICAREMOS EN FINAL PARA LA DOS CERO. TACA TRES NUEVE CERO.
GRACIAS.

and only switched to "cero dos" for their second approach after the missed approach procedure.

Servus, Simon