Air France crash at YYZ (Merged)

[NigelOnDraft]

Never trained go around due long land ! And agree that unprepared most would, given the situation over threshold, end doing the same as happened here.

I don't agree that most people would try to land out of a long land situation rather than go around

FC - your situation was more typical of being trained for - a poor outcome to the landing, a bounce, drifting to edge of runway....

We did actually do some training in this area a few sims back. Came after the QF BKK accident... and an incident a little closer to home. The key was a comment made in both cases along the lines of "get it down", and trying to impress that maybe the correct response to this was "GA....". As I said, I have subsequently had the (mis)fortune to put this into practice, only at the cost of my Pax's nerves, and the subsequent PA "Welcome to ABZ for the second time this evening..."

NB this crew may have had the following factors to add to this:
1. High App Speed anyway due full aircraft
2. May have added a few knots for the weather - in fact, almost certainly did...
3. Tailwind may have spoilt their day in the very late stages, and Thrust response added a few more knots...
4. All the above leads to a very high groundspeed...
5. The (localised) poor vis will have removed many of the visual cues of a long flare / float...
6. And that's before we get into fatigue / LH lack of handling etc.

The lack of visual cues, and high GS, could mean that the float did not necessarily seem excessive...

All I am saying is that there is no definitive measure we have, or are trained for, as to what constitutes an excessively long float that demands a GA. As with most accidents, suspect we'll find a lot of "a few % here" and "a few % there" led to a situation that got out of hand...

There are various technological solutions thay could help. One of the more complex (!) would be those boards they have at Mil Airfields... you know the ones that have a single digit every 1000' e.g. '5' means 5000' LDA left. Not difficult to brief "if we have not touched down by 6000' to go, we must GA..."

NoD

PS I have always maintained it is easier to stop, then land, than v-v

[BOAC]

And that's before we get into fatigue

- NOD - Paris -Toronto - Air France - are you serious

[411A]

<<Point of this is to show that a GA will almost always save you - although like you, we never practised it so late in the sequence>>

A very similar situation to yours, Few Cloudy, only in my case, the First Officer never allowed the aircraft to actually touch...just kept on trying to land, half way down the 11,000 foot runway.

Accelerating (never mind not decelerating) in ground effect ain't good for ones health.

This was a very unfortunate accident, and a lot I suspect will truly be learned from the final report.

And, certainly the cabin crew did an outstanding job.

PS: the possible 'fatigue' issue is a red herring, at best.

[Doors to Automatic]

FC - just out of interest what did you say to the passengers after this incident?

I would have been sh*tting myself down the back even as a seasoned traveller.

NigelonDraft - those distance-to-go boards that you mention exist at all US civil airports - I'm not sure about Canada/YYZ though.

[exeng]

facelac

Never trained go around due long land ! And agree that unprepared most would, given the situation over threshold, end doing the same as happened here.

I've folowed this thread with great interest from a safety perspective.

Your comment concerns me. Are there other Pilots out there who would continue with a 'long land' because they have never been 'trained go around due long land'.

I disagree with facelac's statement that unprepared most would etc, etc.

I'm sure training departments do teach the importance of landing within the touchdown zone. I'm also sure that most training departments will also train a 'go-around' if a touchdown has not been achieved in the zone.

A go-around from the touch down is something I have experienced in the sim and real life.

Please note that my comments here are in respect to 'facelac' comments and are not in relation to the incdident where we await the outcome.


Kind regards
Exeng

[604guy]

those distance-to-go boards that you mention exist at all US civil airports



Are you sure about that? I frequent many US civil airports and I guess they must be really small because they sure aren't noticible. Yes, I have seen them at the odd couple of places but they certainly aren't everywhere as you suggest.

The only places in Canada that I have see them in recent years are at military fields.

[av8boy]

those distance-to-go boards that you mention exist at all US civil airports

Are you sure about that? I frequent many US civil airports and I guess they must be really small because they sure aren't noticible.

They're optional (and therefore not nearly at "ALL" of them), but not uncommon...


KFJK 31L (on the left side, starting with 11 at the threshold)

KLAX

KORD (Hey! They've got radar! Can anybody telll me, is that an ASR9 on a new tower sitting next to the old ASR7 tower in the background? I have not been keeping up...)

KBOS 27 (I prefer to see the numbers start larger than six though...)

KSFO

KDFW

KDCA

KSMO 21 (Santa Monica, California)

Dave

[nyt]

New unverified facts published by Le Figaro (in french at http://www.lefigaro.fr/france/20050915.FIG0005.html):
-reversers have been deployed 12s (700m) after touchdown, possibly because of a disagreement on the flight deck (CVR transcripts to be published on oct 4th)
-ILS on the longer runway (24R) just had been fried, leaving the crew with no choice as per SOP (favour ILS).
-windshear detector and speed measures also fried, the FDR reported 23K at touchdown instead of 10K indicated by ATC

[ATC Watcher]

Watch out for that "Figaro report". This newspaper is well known for " his master's voice " links.

An AF Captain that spoke to both crew on the Toronto A340 strongly refutes what the newspaper reported; According to him , there was no disagreement at all between the 2, and if the 12 seconds are factual, the suggestion given by the newspaper that it is crew indecision is also not correct according that same source.

Let's wait for the CVR transcript ( due Oct 4th if one beleives the BEA) and most interestingly the FDR for the correct timing and position / activation of the various items..

[PanPanYourself]

They're saying its pilot error.

http://edition.cnn.com/2005/WORLD/eu....ap/index.html

This doesn't really constitute pilot error as the cause of the accident does it? How long is an acceptable time to wait before deploying thrust reversers? It usually seems like it takes about 5 or 6 seconds after touchdown. But hey, what the hell do I know? I'm just a lowly, miserable, economy class passenger, so somebody please shut me up and put me in my place. But let me ask one more stupid question... If the pilot had deployed thrust reversers immediately after landing would that have prevented the accident? Thrust reversers only help with the deceleration process right? The brakes do most of the work? No?

[RoyHudd]

AF A/THR SOP?

Just a rumour, peut-etre, but do AF have an SOP which recmmends the use of manual thrust on landing an Airbus?

Frankly difficult to imagine, but this rumour is hard.

[Idle Thrust]

posted 15th September 2005 18:30
------------------------------------------------------------------------------

"Let's wait for the CVR transcript ( due Oct 4th if one beleives the BEA)"

ATC Watcher,

Are you sure about that? The Canadian TSB has a record of not releasing even a transcript of the CVR, citing privacy laws. The SR111 CVR is an example. Only the TSB knows what really went on in that cockpit.

[Doors to Automatic]

PanPan

On the Airbus an autobrake setting gives a set decceleration rate so reverse only helps in taking work off the brakes. On a wet runway this would probably help in bringing the plane to a stop quicker so yes, applying reverse quicker would have brough the plane to a halt quicker.

Also didn't the pilot override the autobrakes? If so, again the reverse would have added to the decceleration progress.

[ATC Watcher]

Idle,

Only the TSB knows what really went on in that cockpit.

The 2 recorders were received by the BEA in le Bourget on August 2nd , decripted and the results sent to TSB on August 8.
So at least 2 agencies know the contents ( and I suspect also Airbus Industrie ) but the BEA will not publish nor comment on those.

As to the date theCVR contents will be revealed ( Oct 4th) that info comes from Canada , and is confirmed by the BEA.

Let's wait. You might be right and it might not be a word for word transcipt but an indication of what and when based on the CVR.

[Flight Safety]

At the touchdown speed of the aircraft, it woud have used about 2900 feet of the remaining runway before thrust reverser deployment 12 second later. However the aircraft would have slowed to some degree within 12 seconds (by how much will have to come from the FDR), so it actually would have used somewhat less runway than this before reverser deployment.

I'm sorry if I missed this (if it was posted earlier), but has anyone posted the braking performance of this aircraft, at the landing weight used, in those conditions? Since there doesn't seem to have been any anomilies with the aircraft systems or the landing configuration, was it actually possible to stop the aircraft after it touched down 4000 feet from the threshold?

[atakacs]

was it actually possible to stop the aircraft after it touched down 4000 feet from the threshold?

Given the friction available the answer is NO.

I would try to avoid any "arm chair investigation" but it seems quite clear that the pilots landed WAY too long and somewhat over speed. The only mystery, if any, is why did they fail to realize it in due time…

[Zeke]

CYYZ Runway 24L
Surface : ASPHALT
True Heading : 227.0
Latitude 43.675292 : 43° 40' 31.05" N
Longitude -79.597236 : 079° 35' 50.05" W
Elevation 547.0 feet : 167 meters
Slope -0.2°
Landing Distance 9000 feet : 2743 meters
Takeoff Distance 9000 feet : 2743 meters
Lighting System CL : HIRL : A3 : PAPI


A340-300 Landing performance ...

From the A340 FCOM 2 :

LANDING RECOMMENDATIONS
For most cases of abnormal landing configuration the increased Actual Landing Distance does not exceed the Required Landing Distance in normal configuration. However, the addition of several of these factors can very quickly lead to an overrun. Special notice should be taken of the runway condition. A slippery runway is the most common reason for overrun at landing. The combination of a slippery runway with any other factor such as tailwind, or increase in approach speed, without reason, should be avoided. As far as possible, the combination of any failure affecting the braking capability of the aircraft (spoilers, reversers) with landing on a contaminated runway should either be avoided or carefully prepared, with a check of the available runway length against the forecasted landing distance. During a visual approach, all means of monitoring the flight path should be used : first of all use the ILS together with available visual aids such as VASI or PAPI. Approach speed must also be carefully monitored, along with the wind and ground speed in particular during final approach.


VLS: is 1.23 VS1g in config 3 or full. It is the same as VREF for config full in the QRH.

VAPP : When the tower wind has been inserted in the Landing Performance Page of the FMGS VAPP becomes VLS plus one third of the headwind component within the limits of VLS + 5 to +15 KTS.
When strong downbursts or windshear is expected up to VLS + 15 KTS.

Dry (Landing Mass x1000 kg - ACTUAL LANDING DISTANCE (METERS)) (autobrake MED)
170-1450
190-1580
210-1700
230-1830

Wet (Landing Mass x1000 kg - ACTUAL LANDING DISTANCE (METERS)) (autobrake MED)
170-1510
190-1690
210-1880
230-2050

6.3 MM (1/4 INCH) WATER (Landing Mass x1000 kg - ACTUAL LANDING DISTANCE (METERS)) (autobrake MED)
170-1820
190-2070
210-2330
230-2570

12.7 MM (1/2 INCH) (Landing Mass x1000 kg - ACTUAL LANDING DISTANCE (METERS)) (autobrake MED)
170-1710
190-1920
210-2150
230-2370

Elevation correction : per 1000 ft above SL (autobrake MED) (autobrake MED)

Dry : +3%
Wet : +4%
6.3 MM (1/4 INCH) WATER : +4%
12.7 MM (1/2 INCH) WATER :+4%

Headwind correction : per 10 kt headwind (autobrake MED)

All conditions : No correction for headwind due to wind correction on approach speed

Tailwind correction : per 10 kt tailwind (autobrake MED)

Dry : +17%
Wet : + 21%
6.3 MM (1/4 INCH) WATER : + 26%
12.7 MM (1/2 INCH) WATER : +24%

Reversers correction - 4 operative reversers (autobrake MED)

Dry : 0 %
Wet : 0 %
6.3 MM (1/4 INCH) WATER : − 13%
12.7 MM (1/2 INCH) WATER :− 10%


Speed correction - Per 5 kt speed increment (and no failure)

All conditions : add 8 % (all runways)

FLUID CONTAMINATED RUNWAY LANDING PROCEDURES . . . . . . . . . . . . . . . . . . CONSIDER

— Avoid landing on contaminated runways if the antiskid is not functioning. Use manual braking and maximum reverse. However, although manual braking is recommended,there is no constraint preventing the selection of autobrake LOW or MED on approach to ensure early brake initiation.
— Approach at the normal speed.
— Make a positive touchdown after a brief flare.
— As soon as the aircraft has touched down, lower the nose wheel onto the runway and select maximum reverse thrust.
Do not hold the nose wheel off the ground.
— If necessary, the maximum reverse thrust can be used until the aircraft is fully stopped.
— If the runway length is limiting, apply the brakes before lowering the nose gear onto the runway, but be prepared to apply back stick to counter the nose down pitch produced by the brakes application. (The strength of this pitching moment will depend on the brake torque attainable on the slippery runway).
— Maintain directional control with the rudder as long as possible, use nose wheel steering with care.

[Flight Safety]

Thanks Zeke, that was the information I was looking for. It's interesting that the braking performance is better with 1/2 inch of water on the runway compared to 1/4 inch of water. I guess the extra water helps to slow the aircraft down.

So we know the following facts about this accident, as they relate to braking performance:

1. All systems were working normally, including reversers and anti-skid.
2. The landing configuration was set to FULL, the autobrake was set to MED.
3. The touchdown airspeed was 143K, the ground speed at touchdown was 148K, indicating a 5K tailwind component at touchdown.
4. The investigators have said the touchdown speed was close to normal (I can't find what the exact VRef should have been).
5. The touchdown point was approximately 4000ft (1219m) down the runway from the threshhold, leaving 5000ft (1524m) to stop (oops, from Zeke's data, we're probably already going off the end of the runway aren't we).
6. The spoilers deployed normally when the MLG touched down.
7. The pilots applied full brake pressure almost immediately after touchdown, thus overiding the autobrake. However the anti-skide worked normally.
8. The pilot didn't deploy the thrust reversers until 12 seconds after MLG touchdown.

For braking performance, the main piece of information that's still missing, is the landing weight of the aircraft. I scanned all 538 posts from this thread, and no one has posted this data. This information will tell us what braking performance should have been expected, how tight the stopping conditions actually were on this runway under these conditions, and perhaps if the pilots should have even attempted the landing.

[Zeke]

Flight Safety,

The actual landing distance is from 50 ft, not the touchdown point.

Based upon the speeds, I would have an educated guess the weight was around the 200t mark, Vref at 200t is 141 kts.

Without autobrake, the actual landing distance is
190t - 1160 dry / 1510 wet
210t - 1280 dry / 1710 wet

I would expect the corrected landing distance from 50 ft onto a the wet runway at CYYZ to be
180t - 1515 m
190t - 1610 m
200t - 1718 m
210t - 1825 m

[Flight Safety]

The actual landing distance is from 50 ft, not the touchdown point.

That's a bit grey isn't it, given that you could float before touchdown, etc. How do know if your touchdown point will be too long?