Yeti Airlines Crash

[The black sheep]

Hi all,
horrid thing!
I am not a pilot, but get some things about planes. It appeared to me that the plane was at a very slow speed, and that the pilot initiated a left turn, which instantaneously caused the left wing to stall. It is always the wing on the inside of a turn which stalls first. Once that wing stalled, at that altitude, there was no other outcome.
Without any insider knowledge, unless there was engine failure, overloading has to be the primary suspect.

 

[nancypants]

Husband and son flew on this flight a few years ago. He has a rather terrifying video of that flight but It seems that is par for the course on this journey.

I dunno, I did this exact flight a couple of times in 2019 (and a few others but they were 1900D/Buddha air rather than yeti air/atr) and all seemed as toward and professional as say a European or Australian operation

 

[jb747]

I am not a pilot, but get some things about planes.

...

Without any insider knowledge, unless there was engine failure, overloading has to be the primary suspect.

I look forward to hearing your logic for this conclusion.

 

[Pushka]

I dunno, I did this exact flight a couple of times in 2019 (and a few others but they were 1900D/Buddha air rather than yeti air/atr) and all seemed as toward and professional as say a European or Australian operation

It wasn't a comment about the pilots skill but the actual conditions of the airport runway and general environment. Given those and the weather they probably have excellent skills and experience. This was in 2013.

 

[albatross710]

Looking at the wing, there is flap, but not a great deal. The obvious suspicion would have to be that they've missed a stage of flap. Insufficient flap will most certainly lead to a stall. The wing drop is consistent with that.

@jb747 thinking more about the events that lead up to stall what do you think about mis-perceptions as to the slope of the runway and surrounding terrain? possibly making them think they had a lower AoA than actual and putting them above/below glideslope

I found a reference saying a Pokhara runway had a 1.67% slope which doesn't sound enough to mislead them. I didn't find which runway or new/old airport though. Maybe the surrounding sloping terrain might also confuse a visual approach.

Thoughts?

Uphill or downhill terrain before the runway threshold: − An uphill slope in the approach zone or a drop-off of terrain at the approach end of the runway creates an illusion of being too high (i.e., impression of a steep glide path, as shown on Figure 1), thus:
/snip
A downhill slope in the approach zone creates an illusion of being too low (i.e., impression of a shallow glide path, as shown on Figure 2), thus: • • Possibly inducing a correction placing the aircraft above the intended glide path ; or, Preventing the flight crew from detecting a too steep flight path.

from https://www.skybrary.aero/sites/default/files/bookshelf/177.pdf

 

[jb747]

@jb747 thinking more about the events that lead up to stall what do you think about mis-perceptions as to the slope of the runway and surrounding terrain? possibly making them think they had a lower AoA than actual and putting them above/below glideslope

I found a reference saying a Pokhara runway had a 1.67% slope which doesn't sound enough to mislead them. I didn't find which runway or new/old airport though. Maybe the surrounding sloping terrain might also confuse a visual approach.

Looking at the terrain charts, and I'd place the new airport as about 100-150' lower than the old one. I don't think that any slope to the terrain is significant. The crew would have been pretty familiar with the effects anyway.

FR24 have put up a chart showing altitude data, but sadly its resolution makes it somewhat hard to decipher. It looks like there was a short period there the aircraft maintained roughly 4000', and that might fit in for part of a downwind leg. It then starts a reasonably consistent descent. But, towards the end, and this is where the resolution fails us, there seems to be a very short period of level, and it would seem the stall took place at the end of that. Such a pity that the normal FR24 data isn't available.

 

[jb747]

I don't know the validity of this, but the latest I've read on this accident claims that both propellers were feathered.

 

[flyer89]

I don't know the validity of this, but the latest I've read on this accident claims that both propellers were feathered.

I hope this isn’t a case of the old “god let’s you feather the wrong engine just once”.

 

[jb747]

I hope this isn’t a case of the old “god let’s you feather the wrong engine just once”.

You feather a propellor, not an engine. That’s relevant because it’s quite possible that a propellor could be feathered, and with a free turbine (which the ATR has) the engine could still be humming away. In that case, unfeathering should be just a push of the condition lever away. It’s certainly an interesting finding, but now the question is why.

 

[flyer89]

The Aviation Herald

Aviation Herald - News, Incidents and Accidents in Aviation

avherald.com

Preliminary report has been released. Unpleasant reading.

“At 10:56:27, the PF disengaged the Autopilot System (AP) at an altitude of 721 feet Above Ground Level (AGL). The PF then called for “FLAPS 30” at 10:56:32, and the PM replied, “Flaps 30 and descending”. The flight data recorder (FDR) data did not record any flap surface movement at that time. Instead, the propeller rotation speed (Np) of both engines decreased simultaneously to less than 25%1 and the torque (Tq) started decreasing to 0%, which is consistent with both propellers going into the feathered condition.”

 

[Mattg]

The Aviation Herald

Aviation Herald - News, Incidents and Accidents in Aviation

avherald.com

Preliminary report has been released. Unpleasant reading.

“At 10:56:27, the PF disengaged the Autopilot System (AP) at an altitude of 721 feet Above Ground Level (AGL). The PF then called for “FLAPS 30” at 10:56:32, and the PM replied, “Flaps 30 and descending”. The flight data recorder (FDR) data did not record any flap surface movement at that time. Instead, the propeller rotation speed (Np) of both engines decreased simultaneously to less than 25%1 and the torque (Tq) started decreasing to 0%, which is consistent with both propellers going into the feathered condition.”

In other words, the pilot inadvertently moved the levers for the propellers when they intended to instead lower the flaps?

 

[jb747]

Preliminary report has been released. Unpleasant reading.

At 10:56:27, the PF disengaged the Autopilot System (AP) at an altitude of 721feet Above Ground Level (AGL). The PF then called for “FLAPS 30” at 10:56:32, and the PM replied, “Flaps 30 and descending”. The flight data recorder (FDR) data did not record any flap surface movement at that time.

So the PM was asked for more flap, acknowledged it, but did not select it...

Instead, the propeller rotation speed (Np) of both engines decreased simultaneously to less than 25%1 and the torque (Tq) started decreasing to 0%, which is consistent with both propellers going into the feathered condition.

Didn't move the flap lever, but instead moved the condition levers?

At 10:56:54, another click was heard, followed by the flaps surface movement to the 30 degrees position.

Almost 30 seconds after being asked for the flap, moves them without comment, or apparently wondering what he'd moved the first time.

 

[AviatorInsight]

In other words, the pilot inadvertently moved the levers for the propellers when they intended to instead lower the flaps?

Looks that way. There’s obviously 2 condition levers and only one flap lever. How could they have gotten that so wrong to pull both?

 

[NM]

Apart from an engine failure condition, are there any other situations where propellers would normally be feathered during flight? For example, does re-engaging the propeller via the condition lever adjustment allow faster re-application of thrust than spooling up from engine idle?

 

[Quickstatus]

How could they have gotten that so wrong to pull both

Maybe the Human Factors issue?

Oh no...
The propellor condition control lever is between the engine power control levers and the flap levers on the right
The feather position is one detent lower than the auto position (the position in the picture) - therefore it only takes moving one lever one detent to completely remove engine thrust. and a second detent to completely cut off fuel....

Is the tactile nature of the flaps lever compared to the propellor condition lever possibly not all that different?

Why is the design of adjacent levers so similar?
It is not the first time that similar design of controls in an aircraft have lead to crashes until the controls were redesigned to look and feel different

 

[jb747]

The thrust levers just move forward and aft.

Condition levers apparently have a lock release button on the side that has to be pressed to allow movement.

Flap levers have to be lifted before they can be moved fore/aft.

Possibly a better way to look around the coughpit:

About The ATR Aircraft Family | ATR Aircraft

The ATR-600 series has the best environmental credentials compared to other turboprops and regional jets in the sector.

www.atr-aircraft.com

 

[jb747]

Apart from an engine failure condition, are there any other situations where propellers would normally be feathered during flight? For example, does re-engaging the propeller via the condition lever adjustment allow faster re-application of thrust than spooling up from engine idle?

In flight…no, it exists to reduce drag. A rotating, unpowered, prop is basically a huge speed brake. Feathering turns the blades into the direction of motion, and reduces the drag to a minimum. On the ground, if you feather a prop, it helps to stop it rotating at shutdown.

These engines are free power turbines. The ‘engine‘ part is not physically connected to the propellor. They already have virtually instant power response compared to a geared system like the RR Dart.

 

[flyer89]

Almost 30 seconds after being asked for the flap, moves them without comment, or apparently wondering what he'd moved the first time.

Maybe a chance to save it at this point. Not sure why pushing everything forward wasn’t a muscle memory for either of them.

Additionally, there’s actually some technical talk going on at pprune (for once) and someone who flies the ATR is not able to explain how this was even done accidentally.

 

[jb747]

Maybe a chance to save it at this point. Not sure why pushing everything forward wasn’t a muscle memory for either of them.

Additionally, there’s actually some technical talk going on at pprune (for once) and someone who flies the ATR is not able to explain how this was even done accidentally.

Pushing everything forward might work for a piston engined aircraft but it's not really the action needed here. The PF did push the thrust levers up. The condition levers wouldn't really be something that is touched often enough to become part of that muscle memory, nor would you normally have hands on them at all. That's another reason that it's strange that the PM moved them at all. They should have felt wrong! I guess another question would be just how often the PM had actually flown in the right hand seat.

 

[Quickstatus]

One of the consequences with feathering is the engine torque will also reduce to prevent the engine over torque due to reduced propellor load from the feathering. How long would it have taken for engine torque to increase if feathering was corrected? I can't imagine it would be instant