TransAsia ATR crash in Taiwan.

[JuniorD]

Cobalt Blue said:

That won't stop all and sundry trying to deflect blame on the pilots. The simple fact is that if three very experienced pilots can make such a fundamental error, and with such tragic consequences, then something is badly wrong with the way information was presented to them by their instruments. For such a critical event, the state of the engines has to be shown in a way that is impossible to ignore. I read that the similar Fokker twin-turboprop has lights that flash red on the power lever of a failed engine - hardly rocket science and probably/possibly simple to install as an upgrade.

Supplementary warning lights on the throttles are all well and good, but what if they fail to work? You'll have to rely on the other instruments, which is what you should be doing anyway. Even when a master warning goes off, you still have to cross check the engine instruments to ensure they corroborate a problem.

As for simplicity of install as an upgrade; nothing is that simple. Improvements to aircraft systems take years to design, validate and install across a fleet. In this case, all for what? For the once in a 25 year event that someone manages to be inclined to shut down the wrong engine during a critical phase of a flight? (I think its less than 20 recorded instances since 1946 of wrong engine shut down in airline and military ops)

You can implement every possible way to prevent human error, and someone will still manage to err.

The voice recording on this flight will make for very interesting listening. There must be some reason why they elected to do what they. None of us can ever appreciate the full circumstances of the flight that day.

Edited by JuniorD on Sunday 8th February 22:33

[croyde]

Did Ginetta Girl describe earlier possibly why the pilots would have thought that the wrong engine was at fault.

Maybe that's why they acted instinctively rather than looking at instruments.

I'm no pilot but I know take off is the most dangerous phase of flight where a split second decision, when you feel that power is failing, either saves the day or puts you into the ground.

Edited by croyde on Monday 9th February 12:40

[Eric Mc]

Were the crew also regular ATR-42 pilots?

[DonnyMac]

I think I read 72 hours Eric.

[Hooli]

Ginetta G15 Girl said:

HoHoHo said:

To those of us who may not understand the wiggly lines etc., can you please explain WTF was going on?

I'll try now that I have had time to sit down and write something.

Eric Mc said:

Ssshhh - it's a secret within the brotherhood and sisterhood. They don't REALLY want to explain anything - just demonstrate their detailed knowledge of the subject.

Those of us who don't understand the code are outsiders and unworthy of inclusion.

Oh do grow up!




Looking at the graphical data there appear to be 4 distinct timing points (I have interpolated the timings as best I am able from the limitations of the display).

Note No1 engine is the port, No2 the Starboard.


At 02:52:35 the following happens:

Master Warning Caption No2 Engine Flameout.
No2 Prop starts to go into the Beta range (indicating that it is correctly autofeathering.)
No2 Turbine Inlet Temperature (ITT) falls rapidly (indicating engine shut down).
No2 Overall RPM begins to rapidly drop (NP).
No2 Fuel flow shuts off.
No2 Prop torque begins to fluctuate indicating that the blade angle is rapidly changing (ie autofeather is occurring).
No2 LP Compressor RPM begins to fall (NL).
No2 HP Compressor RPM begins to fall (NH).

At 02:53:05 the following happens:

No1 Power Lever is retarded slightly (PLA - Power Lever Angle reduces), No2 is left at full power.
No1 Turbine Inlet Temperature (ITT) reduces .
No1 Overall RPM reduces slightly (NP).
No1 Fuel flow reduces.
No1 LP Compressor RPM reduces slightly (NL).
No1 HP Compressor RPM reduces slightly (NH).


At 02:53:25 the following happens:

No1 Condition Lever is moved to Fuel SHUT OFF (CLA - Condition Lever Angle).
No1 Prop starts to go into the Beta range (indicating that it is feathering).
No1 Power lever is fully retarded.
No1 Turbine Inlet Temperature (ITT) drops to minimum (idicating engine shut down).
No1 Overall RPM begins to rapidly drop (NP).
No1 Fuel flow shuts off.
No1 Prop torque begins to fluctuate indicating that the blade angle is rapidly changing (ie feather is occurring).
No1 LP Compressor RPM begins to fall (NL).
No1 HP Compressor RPM begins to fall (NH).


At 02:54:19 the following happens:

No1 Condition Lever is moved to the AIRSTART position (fuel shut off valve opens).
No1 Turbine Inlet Temperature (ITT) starts to rise.
No1 LP Compressor RPM begins to rise (NL).
No1 HP Compressor RPM begins to rise(NH).


So, No2 engine fails at around 02:52:35, about 30 seconds later power on No1 engine is retarded. About 30 seconds after that, No1 engine is shut down. About a minute later a relight attempt is started on the No1 engine.



If we couple this with the report of a left wing drop at around the time the No2 failed, as well as with the speed read out from Flight Radar 24 it is apparent that:

The crew mis-identified the failed engine, retarded the live, then shut down the live engine. They then tried to relight the engine they had shut down.

Up to the failure of No2, the rate of climb is around 1500 fpm dropping by about half fom the point of failure until the live engine is shut down. More significantly, the airspeed is decaying to 100 kts when the No2 fails, it then rapidly decays further before stabilising(ish) - looks like the nose was lowered.



This is mere supposition but I would suggest that it is possible that they mis-handled the climbout and possibly stalled the left wing at around the time the engine Master Caption illuminated (accelerated stall). The wing drop led them to believe it was a No1 failure so they shut it down. From that point the inevitible happened.

Edited by Ginetta G15 Girl on Friday 6th February 14:31

Thanks. I couldn't make sense of the data & that explains it brilliantly.

[saaby93]

Confirmed turned off the wrong engine again

http://www.bbc.co.uk/news/world-asia-33358707

[Brother D]

http://www.asc.gov.tw/main_en/docDetail.aspx?uid=3...

[dvs_dave]

Why during such a critical stage of flight would a manual engine shutdown procedure even be considered? Surely you would want to stabilise and verify the situation before thoughts of manually cutting power are even an option? After all, the faulty engine will safely shut itself down, cut the fuel supply and feather the prop all automatically as it did in this case. Why would you then need to pull the throttle back to shutdown position? That's just asking for trouble and risking, as in this case, shutting the wrong one down?

Sounds to me much more like a procedural problem rather than an instrumentation one.

[KTF]

When the FDR data was posted, Pprune called them pulling the lever on the wrong donkey way in advance of the report.

The sad thing is, you can see they almost managed to relight it again but were too close to the ground.