JAL incident at Haneda Airport

[andrasz]

Originally Posted by framer

...if the JAL crew had blown 8 doors within 30 seconds of coming to a halt, would everyone have survived? Pretty hard to Monday morning quarterback these events.

Agreed, but there is a difference between blowing all doors without thinking and waiting for 8 minutes for permission with flames visible outside. The fact that the two front doors were open but only one in the rear suggests to me that it was the cockpit crew who verbally instructed those in the front to evacuate.

[FlyingRoland]

Originally Posted by MichaelOLearyGenius

Similar incident happen in LAX in the 80s. A USAir 737 landed on top of a metro liner. Different circumstances though, ATC cleared the metro liner to line up and wait then forgot about it, they then cleared the 737 to land. All on board metro liner were killed but survivors on the 737.

i believe this was US1493 in 1991.

[skytrax]

Originally Posted by monkey.tennis

strong disagree. 8 minutes to initiate and 18 minutes to complete an evacuation of a clearly burning aircraft is a catastrophic failure.

There are many aspects to be clarified during the investigation. We shall see if there was a delay and what led to that.
As of now, facts are that cabin crew managed to get everyone out using 3 slides out of 8. LR1 were at a angle that slew down the sliding and L4 was barely touching the ground due to nose being collapsed.
This is a successful evacuation in very difficult circumstances, far from a catastrophic failure.

EDIT: It is easy to talk now after the events....that crew had a very different perspective when the event took place. It was an unplanned emergency, at night, engine still on after stopping....many factors to be considered.
Let's also remember SQ368 in 2016. Plane was on fire during landing and no evacuation took place. They waited several minutes inside while emergency services took care of the situation.

[dr dre]

Originally Posted by Claybird

About the HUD... I am reminded of my electrician: He said, plug the power cord directly onto the socket. No extensions etc. the more you put between the machine and the power line the more could go wrong. You might say WTF does this have to do with the accident?

HUD my ass. Eyes on the runway and the instruments that you're leveled off. Runway - Instruments. The HUD is a nonsensical distraction

You’re seriously going to blame the HUD for this accident, or suggest it degrades safety?

It actually does “put the cord into the socket”. It reduces times spent looking up and down and concentrates eyes on only critical information right in front of you.

While don’t we go back to analog gauges, flight engineers and non precision NDB approaches while we’re at it.

[MartinM]

Originally Posted by skytrax

There are many aspects to be clarified during the investigation. We shall see if there was a delay and what led to that.
As of now, facts are that cabin crew managed to get everyone out using 3 slides out of 8. LR1 were at a angle that slew down the sliding and L4 was barely touching the ground due to nose being collapsed.
This is a successful evacuation in very difficult circumstances, far from a catastrophic failure.

Another contributing factor was likely the running engine number 2 at idle thrust. This could have easy sucked in a pax while evacuating. I am not sure how long they used the forward right side for evac under that condition.

[andrasz]

Originally Posted by skytrax

Let's also remember SQ368 in 2016. Plane was on fire during landing and no evacuation took place.

In my book that incident amounts to dereliction of duty. With a fuel-fed aircraft fire, if there are any usable exits, you evacuate. No ifs, no buts. There is no guarantee that CFR can contain the fire, or that it does not spread. Even with hindsight, that aircraft should have been evacuated immediately via the left side exits. They were simply lucky that things did not turn south.

[VHOED191006]

Originally Posted by framer

The thing that stands out to me is that there was no loss of life on the two aircraft mentioned above, could that be because in neither case did crew go off half-cocked blowing doors into dangerous environments? For example, if the JAL crew had blown 8 doors within 30 seconds of coming to a halt, would everyone have survived? Pretty hard to Monday morning quarterback these events.

Yes, you may be onto something there. Whilst of course we weren't in that cockpit and weren't the ones in command (therefore indeed we can't really Monday morning quarterback these events),here are my thoughts. Something tells me that when you hear from both ATC and the cabin crew that there is fire/a bright orange glow coming from the belly of the aircraft and there's another one behind you on the runway, along with smoke in the cabin, plus the fact that you're no longer on the runway and that nose gear has collapsed (I would think that the crew would have been told all of these things fairly quickly), I don't think making such a decision would take that long. Just my 2 cents on that.

[monkey.tennis]

Originally Posted by skytrax

There are many aspects to be clarified during the investigation. We shall see if there was a delay and what led to that.
As of now, facts are that cabin crew managed to get everyone out using 3 slides out of 8. LR1 were at a angle that slew down the sliding and L4 was barely touching the ground due to nose being collapsed.
This is a successful evacuation in very difficult circumstances, far from a catastrophic failure.

EDIT: It is easy to talk now after the events....that crew had a very different perspective when the event took place. It was an unplanned emergency, at night....many factors to be considered.
Let's also remember SQ368 in 2016. Plane was on fire during landing and no evacuation took place. They waited several minutes inside while emergency services took care of the situation.

Another question that needs to be asked is why this fire wasn’t extinguished before taking hold? From what I’ve seen it looked like a very slow response from the fire service who perhaps were struggling to manage two sites at once.

You’re right that with a successful outcome it is hard to be too critical but I’m sure this will be a major topic in the investigation. I can tell you one thing for sure though, there’s no way the CC I work with would wait for 8 minutes to initiate an evacuation in this scenario.

Also, the aircraft is certified to be evacuated using 50% of the slides in 90 seconds. They had 38% of the slides open and it took 10 minutes. Nothing about this evacuation suggests it was well executed to me.

[FlyingRoland]

Questions...

Questions to dig deeper into:

1. What part of the A350 hit the DHC-8 first?

2. Why did the nose gear collapse?

3. What did cause the radome damage?



To answer that I checked the landing geometry of the A350 verses the static DHC-8 Q315 and tried to visualize 3 different scenarios:



A: the A350's RADOME hit the top part of DHC-8's TAIL first.

B: the A350's NOSE GEAR hit the top part of the DHC-8's TAIL first.

C: the A350’s AFT FUSELAGE hit the top part of the DHC-8's TAIL first.



Assumption made is that the A350's attitude was around 3 degrees nose up, in line with a just initiated flare to land.



The red line is the runway surface. The green lines are 3 degrees descent angles.



My analysis is as follows:



A: the A350's RADOME hit the top part of DHC-8's TAIL first.

- if this would be true the A350 would have flown through the Dash-8

- the A350 would have been severely damaged on impact

- the A350 would have dragged the Dash 8 beyond the point where it was lined up.



B: the A350's NOSE GEAR hit the top part of the DHC-8's TAIL first.

- if this would be true the A350's main gear would have caught the wings of the Dash-8

- the A350's main gear would be in a much worse shape than it appeared on available images

- the A350 would have dragged the Dash 8 beyond the point where it was lined up.



C: the A350's AFT FUSELAGE hit the top part of the DHC-8's TAIL first.

- if this is true the nose gear would have cleared the Dash 8's nose gear by a hair.

- At a ground speed of around 120 knots it would take another 0,4 seconds for the A350’s MAIN GEAR to impact the top of the Dash 8’s WINGS, most likely rupturing the fuel tanks, and setting both the Dash 8 and A350 on fire.

(The drawing showing scenario C shows the moment where the A350’s MAIN GEAR impacts the top of the Dash 8’s WINGS. You can see the tail in RED which was hit 0,4 seconds before…)

- the Dash 8 would remain at the same position as it was lined up and no debris would be found between the Dash 8's final position and the A350's final position

- the A350 would have gone into a nosedive immediately after impact and hit the runway nose gear first causing instant collapse of the nose gear.

- the radome then was most like damaged after the A350 left the runway and hit the PAPI.





Looking at all available evidence my speculation is that C is the most likely scenario.



Please shoot...

[andrasz]

Originally Posted by dr dre

You’re seriously going to blame the HUD for this accident, or suggest it degrades safety?

HUD is a great tool to aid low visibility operations, and does all the things you mention. However in this instance (if it does turn to be the case), the SOP of always using it, even in full visual conditions, may have prevented the JL crew from spotting the DH3 on the runway. I'm not saying that without the HUD they would have certainly noticed (probably not), but it could have contributed. As always, some safety improvements may have freak unintended consequences. E.g. the reinforced cockpit door has already killed several hundred people, and very doubtful if it saved any.

[jaytee54]

I have been retired a long time, but I used to operate into Narita where the ATC was good.

From the transcript I can imagine it happening. JAL516 comes on frequency, hears DAL 276 taxiing full length for T/O, and then he’s Cleared to Land. In my day that meant, great, runway is all mine (except in USA where it meant, ‘I expect the runway to be clear when you get there’).

The J722A comes onto Tower and cleared to C5, you’re number 1 (for departure, not for the runway, unclear). That JAL179 comes on, and is no 3, (so that DAL ahead must be no 2, and I'm no 1).

He arrives at C5, the red bar has been turned off already (didn't read the notam), he’s number one, so what does he do?

I know I’d have stopped and asked for verbal confirmation, and I’d have read it back loud and clear.

But he has heard nobody on approach, there are lights at 5 or 6 km, for 34L at a guess, he’s number 1, and 2 and 3 are taxying to C1 so the Coastguard crosses the line and lines up, waiting. JAL166 is told to slow down, for a departure (that must be me! gives me time) and still he waits, 166 is told ‘min app speed, (gives me more time) and then the world caves in.

JAL516 was convinced cleared meant runway is all mine and this happens. This flight has not been mentioned since Coastguard came on frequency.

I think in that situation (when stop-bars are u/s) they should erect a row of red lights across the holding point and all aircraft must use C1.

[andrasz]

Originally Posted by FlyingRoland

To answer that I checked the landing geometry of the A350 verses the static DHC-8 Q315 and tried to visualize 3 different scenarios:

Great visualization, thanks for the effort.

Looking at the videos to me it appears that the 350 touched down a second or two before the collision, which makes A the likely scenario. Also the damage to the 350 engines (and the intense fuel-fed fire in both) is consistent with hitting the outer wings of the DH3, matching scenario A. In both B & C the engines would have avoided hitting the DH3. Also on the videos you can see parts of the DH3 (wings & engines) dragged under the 350 for some time, which is not consistent with C.

[MartinM]

Originally Posted by dr dre

You’re seriously going to blame the HUD for this accident, or suggest it degrades safety?

It actually does “put the cord into the socket”. It reduces times spent looking up and down and concentrates eyes on only critical information right in front of you.

While don’t we go back to analog gauges, flight engineers and non precision NDB approaches while we’re at it.

Well, yes. The eyes will be focused more on the HUD Infos than on the surroundings. Specially under heavy workload you get sort of a tunnel view, I would call it that way, based on experience in military jet HUD during dogfight. You are going toi be extremely focused on this and to nail the target. Any other instruments are seldom checked.

However, this must be said, upon landing, even in a F16 or F18, you see if something is moving on the runway, while landing given that the lightning is on. For me the DH3 from behind was not illuminated properly until the very last moment where the landing light of the A350 would light the obstacle up. But thats probably at a speed of 160 knots about 1 or 2 secs away from collision. In short, it leaves the A350 crew exactly to say "s...t" and thats it

[FlyingRoland]

Originally Posted by andrasz

Great visualization, thanks for the effort.

Looking at the videos to me it appears that the 350 touched down a second or two before the collision, which makes A the likely scenario. Also the damage to the 350 engines (and the intense fuel-fed fire in both) is consistent with hitting the outer wings of the DH3, matching scenario A. In both B & C the engines would have avoided hitting the DH3.

Thanks for info. If A would be true then that does not explain why the radome just has a big dent.... the damage to the front of the A350 would have been much more severe.

[MartinM]

Originally Posted by jaytee54

I think in that situation (when stop-bars are u/s) they should erect a row of red lights across the holding point and all aircraft must use C1.

C1 thru C14 were u/s, bare that in mind

[xcris]

Originally Posted by KKoran

If you haven't heard 'cleared for takeoff', 'line up and wait', or 'cross', you should not cross the holding position marking. Exception for specific instructions to taxi on the runway, but, if you hear anything else, you should hold short and ask for clarification. Period, full stop.

yeah, "full stop". Like it has not happened before, right?
The theory is one, the facts (!) are: instruction to taxi and hold, readback correct, and the a/c finished sitting on a runway, with the captain convinced that he has the right (aka instructions) to do so.

[andrasz]

Originally Posted by FlyingRoland

Thanks for info. If A would be true then that does not explain why the radome just has a big dent.... the damage to the front of the A350 would have been much more severe.

The two aircraft were not necessarily perfectly aligned. Looking at the damage, the radome probably hit the tip of the left HS, or you could be right, that damage may have happened later during the slide when taking the PAPIs out. With a tad steeper flare the HS of the DH3 would have slid under the nose of the 350.

The mis-alignment is also supported by the different degree of damage to the two engines of the 350. The right engine had the entire fan missing (with the core continuing to run), possibly due to hitting the right engine of the -8, while the left engine remained essentially undamaged with only some dents in the cowl.

[jaytee54]

Originally Posted by andrasz

Great visualization, thanks for the effort.

Looking at the videos to me it appears that the 350 touched down a second or two before the collision, which makes A the likely scenario. Also the damage to the 350 engines (and the intense fuel-fed fire in both) is consistent with hitting the outer wings of the DH3, matching scenario A. In both B & C the engines would have avoided hitting the DH3. Also on the videos you can see parts of the DH3 (wings & engines) dragged under the 350 for some time, which is not consistent with C.

I agree with Andrasz, scenario A perfectly describes the damage to the nose of the Airbus (and the nacelles).

[MartinM]

Originally Posted by FlyingRoland

Questions to dig deeper into:

1. What part of the A350 hit the DHC-8 first?

2. Why did the nose gear collapse?

3. What did cause the radome damage?



To answer that I checked the landing geometry of the A350 verses the static DHC-8 Q315 and tried to visualize 3 different scenarios:



A: the A350's RADOME hit the top part of DHC-8's TAIL first.

B: the A350's NOSE GEAR hit the top part of the DHC-8's TAIL first.

C: the A350’s AFT FUSELAGE hit the top part of the DHC-8's TAIL first.



Assumption made is that the A350's attitude was around 3 degrees nose up, in line with a just initiated flare to land.



The red line is the runway surface. The green lines are 3 degrees descent angles.



My analysis is as follows:



A: the A350's RADOME hit the top part of DHC-8's TAIL first.

- if this would be true the A350 would have flown through the Dash-8

- the A350 would have been severely damaged on impact

- the A350 would have dragged the Dash 8 beyond the point where it was lined up.



B: the A350's NOSE GEAR hit the top part of the DHC-8's TAIL first.

- if this would be true the A350's main gear would have caught the wings of the Dash-8

- the A350's main gear would be in a much worse shape than it appeared on available images

- the A350 would have dragged the Dash 8 beyond the point where it was lined up.



C: the A350's AFT FUSELAGE hit the top part of the DHC-8's TAIL first.

- if this is true the nose gear would have cleared the Dash 8's nose gear by a hair.

- At a ground speed of around 120 knots it would take another 0,4 seconds for the A350’s MAIN GEAR to impact the top of the Dash 8’s WINGS, most likely rupturing the fuel tanks, and setting both the Dash 8 and A350 on fire.

(The drawing showing scenario C shows the moment where the A350’s MAIN GEAR impacts the top of the Dash 8’s WINGS. You can see the tail in RED which was hit 0,4 seconds before…)

- the Dash 8 would remain at the same position as it was lined up and no debris would be found between the Dash 8's final position and the A350's final position

- the A350 would have gone into a nosedive immediately after impact and hit the runway nose gear first causing instant collapse of the nose gear.

- the radome then was most like damaged after the A350 left the runway and hit the PAPI.





Looking at all available evidence my speculation is that C is the most likely scenario.



Please shoot...

Awesome work! Thank you so much.

Scenario A is for me the correct one. There was still a little nose up. That saved the pilots of the A350. The radome hits the tail and crushing it, the nose gear of the A350 collapses while rolling over the DH3 at full force 160knots, the main gear rolls over the DH3 debris at will weight literally flattens the DH3 remainders, JAL crew is unable to deploy the thrust reverser, neither engine 1 nor 2 on any photo have the reverser deployed and given that engine 2 was running, indicated that engine 1 sucked in debris from the DH3 resulting in e flame out while engine 2 did not have a flame out. That leads to the conclusion that there was potentially a lateral offset to the right upon collision, meaning that either the DH3 was not on the centerline but more to the left or the A350 was off the centerline to the right. Another observation about a slight nose up position - spoilers were not deployed yet. Unless something has changed from my A320/A330 times, spoilers deploy automatically on Airbus and would have been in armed position.

[51bravo]

Yes, from (one of) the footages it is very clearly scenario A. Thanks for the visualisation of geometries.

On the use of "number one, two, etc." I always asked myself: does it help at all? Or just adds a risk. Maybe this will be removed from phraseology in the aftermath.