FZ981 -- 61 die; plane explodes after crashing at high speed on landing

[HJ1an]

Monitoring is not done by subjective listening to engines or so. It is an instrument scan of the most important items on the pfd: speed, horizon, altitude and vertical speed. If any of these change in an unexpected way then something is wrong, and it should alert the pilot to take action. Other items are looked at too, but they are not so important as these 4 parameters.

Let's assume a pilot does scan his instruments - say at about 10+- miles* from the runway and about 2500ft* and something happened to the engines, without any visual or noise indicating something was wrong, the pilot was looking at - say - his or her checklist. Before looking at that check list the speed was about 150knots*. Then he or she looked at the checklists. Then looks back at the speed. By which time, it's showing about 110knots* and the plane had started dropping at an increasing rate, wasn't felt at first, but feels it now. "Uh oh" time. Quickly reacted, full thorttle, but too late, stalled or loss of altitude at that point and the curve of the descent was beginning to go the point of no return.. would this scenario be possible?


*just an estimate-Insert the appropriate numbers corresponding to section as needed

[KL-666]

Such thing developes in many seconds (a minute or more), like in the case of Turkish. Not in the one second you looked away. Then you would only have lost maximal 10 kts and reacted immediately.

To be precise, Turkish went all the way down to 85 kts in 100 seconds from about that 150 kts. 100 seconds of not scanning is an insane amount of time, considering the ease and speed of such scan. There is no excuse thinkable for that. I bet they even did not plan on scanning on second 100. They just felt the plane falling from under them.

My mistake, they did not feel anything, the stick shaker activated.

Kind regards, Vincent

[HJ1an]

Such thing developes in many seconds (a minute or more), like in the case of Turkish. Not in the one second you looked away. Then you would only have lost maximal 10 kts and reacted immediately.

To be precise, Turkish went all the way down to 85 kts in 100 seconds from about that 150 kts. 100 seconds of not scanning is an insane amount of time, considering the ease and speed of such scan. There is no excuse thinkable for that. I bet they even did not plan on scanning on second 100. They just felt the plane falling from under them.

My mistake, they did not feel anything, the stick shaker activated.

Kind regards, Vincent

That makes sense. But I guess there is definitely a lapse in concentration in that particular case for both of them.

Personal experience - I was once so tired while driving, that I simply lapsed for far longer than I would have probably liked when stopped at an intersection when I got a several distractions (phone ringing, and another car was cutting in front of me, some suicidal pedestrians running across the roads and probably a few other things happening at the same time.) My brain simply stood still and blanked out for a considerable of time. Fortunately, the default reaction is to step on the brake and I was just simply stopped for longer, and nothing happened during that time.

There's also something called tunnel syndrome or tunnel vision(?) that the concentration on one particular area others are completely neglected even though alarm bells are ringing or the house is burning down etc.

[KL-666]

What you say exists, but in the report on this case there is no evidence of fatigue whatsoever. Believe me, it is airline culture and training.

Kind regards, Vincent

[jwocky]

Did you find anything about how far the flaps were out? I see nothing.

Lets just speculate a moment ... if she is light, on full throttle with flaps out, she has an enormous lift as soon as she gets some speed. Only the extreme climb would make sure, she gets no speed. And the A/P would try to hold the nose where ...? At an angle to maintain the climb and use for that h-stab trim. So, if the pilot goes manually, he has to take the trim back and do what? Because the lift is still there, the nose goes even more up, which leads in a stall. Which afterwards could explain the steep descend.
So ... I am sure, this is not procedure, but if I want to bring her under control, in FG, I go back to a moderate flap setting, climb a bit slower, then correct the h-stab or, in most situation, if the A/P didn't drop, the A/P does that for me. So the crucial point appeas to be flaps. Because the CoG shouldn't be too far from maybe what 40-50 inches MAC? I remember something from the Boeing documents? Unless ..
This steep descend makes me think, she was maybe not entirely out of trim but out of balance? Some weight shift? Luggage? Cargo? Anything, that may can slide?

[jwocky]

Did you find anything about how far the flaps were out? I see nothing.

Lets just speculate a moment ... if she is light, on full throttle with flaps out, she has an enormous lift as soon as she gets some speed. Only the extreme climb would make sure, she gets no speed. And the A/P would try to hold the nose where ...? At an angle to maintain the climb and use for that h-stab trim. So, if the pilot goes manually, he has to take the trim back and do what? Because the lift is still there, the nose goes even more up, which leads in a stall. Which afterwards could explain the steep descend.
So ... I am sure, this is not procedure, but if I want to bring her under control, in FG, I go back to a moderate flap setting, climb a bit slower, then correct the h-stab or, in most situation, if the A/P didn't drop, the A/P does that for me. So the crucial point appeas to be flaps. Because the CoG shouldn't be too far from maybe what 40-50 inches MAC? I remember something from the Boeing documents? Unless ..
This steep descend makes me think, she was maybe not entirely out of trim but out of balance? Some weight shift? Luggage? Cargo? Anything, that may can slide?

[HJ1an]

Did you find anything about how far the flaps were out? I see nothing.

Good point, anyone knows if the flaps automatically retract at TOGA?

As for balance, it's also possible the cargo isn't tied down correctly or the passengers are somehow all sat at the back..

[legoboyvdlp]

No, but usually they are at 30 for landing; set to 25 for go around, and set to either 5 or 15 for takeoff.

[KL-666]

Slowly more official information is trickling through.

On Apr 5th 2016 Russia's Ministry of Transport reported that on first approach to runway 22 the aircraft went around, the crew reported a wind shear on final approach, climbed to FL050 then FL080 and entered a hold to wait for improvement of weather. The crew subsequently requested and was cleared to FL150 in the hold. When air traffic control provided information that the visibility was 5000 meters, cloud ceiling at 630 meters/2000 feet, winds from 230 degrees at 13m/s gusting 18m/s (25 knots gusting 35 knots), no wind shear, the crew requested another approach clearance. On final approach at about 220 meters/720 feet the crew went around again and climbed, at 900 meters/3000 feet the stabilizer moved nose down causing the aircraft to stop climbing at about 1000 meters/3330 feet and entering a descent. The aircraft impacted ground about 120 meters from the runway threshold. Preliminary examination results of flight data and cockpit voice recorder do not reveal any evidence of a technical malfunction of engines or aircraft systems or any evidence of an explosion. The investigation is focussing on how the pitch control system works and crew actions during the go around. The captain (ATPL, 5,965 hours total, 2,597 hours on type) was certified for CATIIIa approaches.

This clearly disqualifies any stall theories and any resemblance with Kazan and other pitch-up upsets. The stabilizer trimmed down between 900 and 1000 meters in order to level off. A very normal thing to happen when autopilot levels off. After that it will go to trim normal for level flight. But it did not go to trim normal. To me this can mean only one thing: the autopilot was disengaged while still leveling off. Either by the pilots or by the autopilot itself because of an unusual situation. After that the plane continued "leveling off" with the trim still down, so continuing to move the nose down, and the pilots never caught it.

Kind regards, Vincent

[jwocky]

I still wonder about the flaps. To force the A/P to trim the elevator full down just to level out means there must be a lot of lift. So this would indicate a lot of flaps, right? Which would also explain why the steep climb didn't cause a stall. They were on the second approach already down to 220 meters, so at this point, the flaps had to be out and maybe they didn't pull them in during the climb or very late and disengaged the A/P at the same time? Like qa communication error in the cockpit?