SOP calls for condition levers full forward for t/o and landing. Even using reduced np, levers still go high.

A few words on autofeather systems, again from a DH8 perspective (that comes with the same PW120 series engines the ATR uses as well).

Generally, they are only armed for takeoff, as this is when an unfeathered propeller will have the most performance impact (and yes, an engine failure during go around will consequently by default require manual feathering). Arming conditions are (roughly) the following:

- Selection by flight crew
- Both power levers above a certain angle
- Both torques above a certain threshold.

That fulfilled, the system will feather a propeller in case of one engines torque dropping below a defined threshold for a defined time; on the Dash 8-400 it was 25% for 3 seconds IIRC (it has been a while). This will trigger:

- a feather command to the affected propeller control unit, also involving the alternate feather pump,
- most importantly, a cutout signal to the other, good engine to inhibit autofeathering on that side.

So even if autofeather should fire by error on one engine, it would take additional system malfunctions for it to affect both engines at once. Even in the previously quoted case of the Taiwanese ATR, I understand that the malfunction affected only one of the engines and the systemic safety barriers worked, keeping the fault manageable in principle.

If indeed both propellers ended up feathered on the accident flight, this would most likely not be due to autofeather imho. Rather, some flight crew action would have to have been involved to achieve this. The other ways of feathering a propeller would be pushing the associated alternate feather push button (activating an oil pump drawing engine oil from a dedicated reservoir to force the prop into feather) or pulling the condition lever back to low RPM, lifting it out of the reached detent and further pulling it to start/feather position. But none of this seems appropriate in final approach; as has been mentioned, the condition levers are either (type and situation dependent) left in low RPM or pushed forward into max RPM position.

That's on the Classics. On the -500 and -600, the condition levers are generally left in the AUTO position for take-off and landing.

It's far too early in the investigation to positively confirm on this. A double spurious auto-feather is not that likely, though the Taiwan crash did IIRC also have "only one" engine going into auto-feather. Enough to confuse the flight crew and subsequently crash the Taiwan ATR.

The Nepal crash also happened in a low-level/low-speed regime, so very little room to make mistakes, once "something auto" goes haywire. When a full power on the remaining engine is not immediately applied, the 20-30% reserve above the stall speed is energy wise lost pretty fast (a couple of seconds). Sinking starts and the natural tendency to pull-up completes the crash preparation into unavoidable.

Did this happen: I don't know, though it is certainly a plausible mechanism, this kind of accidents develop very fast from "normal" to "irrecoverably", especially, when "the other engine" is also feathered due to confusion.

Yes. https://assets.publishing.service.go...995_HA-LAJ.pdf

The reason that the DHC 8 autofeather was switched off after climb power was set, was due to malfunctions in the dual channel system that produced intermittent uptrim as the system got transient engine fail indications. My understanding is that it was intended to be on during flight but SOP changed after the fault was found, which I think went unchanged?

What is SOP on the ATR?

Double Auto Feather
 

If both engines fail, both props feather - If no fuel, both engines fail...

I think we'd have heard very quickly, and certainly by now, if fuel starvation had been an issue.

Yes plus the rather long post accident fire kind of contradict this. Plus in fuel stavation accidents ,including the previously mentioned Tunisair ATR, engines normally do not stop at the same time .

Excuse my ignorance, never flown a turboprop. On flame out is the ATR designed to autofeather automatically without crew input? What are the implications when it comes to restart? If they've gone into full feather you'd need a starter assisted restart?

I am trying to relate to the recent Dash 8 in Norway who had several flame outs on approach as ice shed into the air intakes. The engines were out for about 25-35 seconds each but the auto-ignition turned the fire back on. How would that have worked if they'd gone into feather upon loss of torque?

For the Autofeather to happen certain conditions have to be met. Those conditions are there during the T/O, and during the APP, after an action that happens after putting the Gear Down. Other than that, the Autofeather will not happen automatically.

Correct !
It is worth noting that the feathering of the propeller does not shut down the engine. The engine can still be running with the propeller feathered.

About the Dash 8, don't know anything about it, but my understanding is that it has more power than the ATR.

It has been established that both engine propellers went into feather. We will have to wait to learn if that was caused by system malfunction, or if it was crew error when intending to reach for the Power Levers, or even for the Flap Lever, they instead reached for the Condition Levers.

Thanks zerograv!

172, props that are feathered do not need a starter to un-feather,depending where the `condition lever/button is,.The act of unfeathering should start the engine as well..

@ Zerograv :
Not familiar with the ATR quadrant, but looking at a photo, do you mean that : 1) retarding the conditions levers to Fuel shut off would automatically feather the props and 2) there is no push or pull button or notch to pass to get to the fuel shut off position ?

That may be true with single-shaft engines like the Dart or similar, where a spinning propeller will turn the compressor as well. On the PW120 series, the propeller is not connected to the turbomachinery proper except via the power turbine. Here, unfeathering the engine will only result in drag and not in a restarted engine.

Again, only speaking for the DH8, but every engine restart in flight required the use of the electric starter both on the -300 and -400 series. A windmill relight is no option. The propeller is only unfeathered after the engine restart has been successful and the turbomachinery is running normal again. The associated procedure is rather long (2 full pages in the QRH) and takes much longer to complete than available when already in a circling approach at rather low altitude. I´d guess that facing such a predicament, using the available energy to glide to some sort of landable field would be the better option.

Which is of course rather academic and leaves aside the startle factor and other things on the crews mind at such a moment.

1) Under normal circunstances for flight the Condition Levers would be 'AUTO' position. Retarding it from the 'AUTO' position, the Condition Lever will come to the 'FTR' (feather) position.
2) As you correctly mention there is a Botton that has to be used to pass it then from the 'FTR' position to the 'FUEL SO' ('Fuel Shut Off', and therefore Engine Shutdown). This is to prevent inadvertently shuting down the engine.

The Condition Lever of the ATR has 4 positions. '100 OVRD', 'AUTO' (characterised by a notch), 'FTR', and 'FUEL SO'.
In the older models (300) there is no '100 OVRD', or 'AUTO' positions, but the operation principle is the same.

Yes, although not recommended (:O), you can in fact hold the prop of a PW120 (or PT6) while the engine is starting.

With some helicopters you can start the engine with the rotor held stationary, once at idle release the rotor brake, provided rapid rotor acceleration which was useful to avoid rotor blade flapping in high winds.

Am I the only one who raised his eyebrows upon learning this?

My wrong ...
Mixing the older versions Condition Levers (no Notch) of the 300, with the Condition Levers of the newer versions of the 500 and 600 series. It's been a while.
In reality in the newer Condition Levers, with a Notch at the 'AUTO' position, it is also necessary to use the trigger (a pushbotton) to remove it from the 'AUTO' rearwards towards the 'FTR' position.
Far less likely that the Condition Levers would have been moved from the 'AUTO' by mistaking the the Levers.
Sorry !!! :O