Cobalt – you are correct, losing an engine at night in IMC is going to be a challenge as well. A lot of twins cannot be flown on the A/P on one engine so you may well be back to hand flying and managing the actual engine failure could be very distracting. Timothy mentioned earlier that for this sort of flight there is a reason for two engines, dual crew and commercial training. When things go wrong he is absolutely right. Managing an engine failure in these conditions is a completely different situtation in a twin with two suitably trained pilots, whereas with one, who is not all that current it is a whole different ball game. I also think there are a lot of pilots who talk about managing an engine failure in a single over an undercast with sea or mountains below estolling the likely good outcome, are deceiving themselves. It may work out, but its going to be a tough chellenge.
A lot of twins cannot be flown on the A/P on one engine
That is astonishing. If I had one of those it would be on Ebay same day. Who could possibly certify such junk?
Having said that, I don’t think all SEPs will run the autopilot when an engine fails (to manage a glide while you are busy sorting other stuff). If the engine is windmilling you should still have enough vacuum to drive the KI256 and the autopilot needs only that, and electrical power. But if the engine actually stops, the AP will disconnect when the AI topples.
Who could possibly certify such junk?
From memory something like a KFC200 and 150 needs to be disengaged, pilot action taken and then re-engaged before continuing. Basically you still need a pilot who is proficient and knows what they are doing. But at the same time a simple wing leveller driven off a rate gyro can fly a mulit-engine aircraft, if the pilot configures the aircraft correctly after the failure, and he has regained/maintained control.
I think that it’s a bit more nuanced than whether an autopilot can cope with an engine failure in a twin. Firstly, it may be physically capable of it without being certified to do so and secondly, circumstances vary enormously. For example there is a big difference between an engine running down in the cruise and one going bang in the go around.
My (reasonable wide) experience is that a/ps will work quite satisfactorily in an engine failure in a twin, and you can dial in some rudder trim to bring the ball in the middle.
Just for fun, I have pulled an engine in a DA42 at altitude without making any other input and it has done remarkably well holding a very weird attitude (essentially a massive slip) but still flying.
I have also wanted a stable platform for photographing a solar eclipse from my Aztec and found that it flew very well at high power on a large amount of rudder trim.
So I don’t think that they should generally be consigned to eBay. At the same time, I don’t think that I would do a single engined go-around at minima on a/p.
Timothy – my experience is similiar, but I think the POH specifically says no A/P? Did you reset the A/P – I assume so after the aircraft was retrimmed? I think most pilots would probably try and use the A/P regardless in these conditions once trimmed. Of course it all adds to the immedaite work load and I am guessing you would be flying for some short while manually while sorting out the trim and dealing with all the usual engine failure drills, including seeing whether a restart was feasible and sensible. I think at that level at night in IMC this is a reasonable workload not to be undertaken lightly without currency, which I think was your earlier point.
I can imagine the 42 in the cruise would behave itself ok, but in the climb it barely goes up at MTOW on one unless it is perfectly trimmed and there doesnt seem to be exactly a surfleit of horses, unlike the Aztec which performances much better. The Aztec on the other hand depending on the version and the engine is also going to loose its hydraulics so while not immediately adding to the workload in due course someone is going to be pumping away which is fine with another pilot but might well be a struggle with a passenger, especially if he is sitting in the second row.
Which on the point of the thread means that a twin suffering an engine failure in these conditions (night and IMC) is still going to be a high workload for single pilot ops, albeit not beyond for anyone current. On the other hand on something a little more sophisticated with auto feathering and trim and A/P approved for single engine ops (oh and dual pumps) it should be less stressful. I guess that is the issue with any aircraft not exactly designed for public transport work – there are compromises! I guess the whole point is that in IC if everything were thrown at you could you cope is the vital question – engine failure, loss of AI, loss of A/P, electrical failure. I think of the things that are manageable, these are the most difficult.
I think you can turn peters point around and ask what GA autopilots will continue to work on a single after engine failure, without at least some mode change, or without some hopefully obvious restriction ? AFAIK there are no mutli-engine autopilots for GA as such.
It all depends on the phase of flight, and how sudden the engine failure is.
With a sudden failure in the climb, the autopilot has to come off immediately, since the pilot needs to adjust the pitch quickly to a sensible attitude, while identifying and feathering the dead engine, and re-trim the aircraft. If you leave it on, you will fight the servos and the autopilot will trim the wrong way, too!
Once stable and trimmed in the single engine climb, I see no reason why the heading mode of the autopilot should not be used to fly straight, assuming the aircraft climbs fine with wings level. Any vertical mode (other than airspeed mode, which is rare) requires a lot more caution; vertical speed mode especially is problematic because any reduction of speed below blue line will lead to a pitch up runaway and stall. A nav mode (or using heading mode for turns) might also be fine, but again requires caution because it might use bank angles that are not appropriate for a single engine climb.
In cruise or descent, sudden action is not required; personally I would still turn the AP off, then re-engage it once the aircraft is set up and trimmed for single engine flight, and then re-engage.
I think at that level at night in IMC this is a reasonable workload not to be undertaken lightly without currency, which I think was your earlier point.
There is a big difference between “game over” versus having something to hang on (e.g. height in a SEP, OEI in a MEP)
I never flew piston twins, but I think any MEP on one engine can be degenerated into a SEP with no engine very easily, if the latter is easier to fly in IMC?
FYI, same debate between 0 engine and 1 engine for gliders, I know which one to pick depending on currency/mission
For direct Nantes-Cardiff, having that 2nd engine option is a plus even when it is not at full power (as long as it can stretch my glide beyond what I get in SEP for the same height it has some value ) but again this was not the ultimate cause of the accident !
I must be old fashioned but if the AFM has an AP restriction, and most puddle jumper twins have a restriction on using AP OEI, why disobey it?
Timothy’s example of the DA42 rolling with aileron to maintain flight in a side slip is producing a high beta on the rudder with potential fin stall. The DA42 prohibits use of the AP OEI.
I think any MEP on one engine can be degenerated into a SEP with no engine very easily, if the latter is easier to fly in IMC
Once the failure is dealt with and the aircraft is in trim, there are three things that make the twin “harder” to fly
– less excess power to climb —> small pitch deviations can lead to loss of climb and even loss of control if mishandled
– asymmetry, funny trim and less excess power need more caution when turning, especially when slow
– power changes require rudder action
So there is no reason to reduce power on the good engine, with one exception: A sudden failure below or close to the red line. But you have to be VERY quick to prevent the aircraft from rolling over – although that I only know from theory, because for good reasons simulated engine failures near or below red line are NOT practices, even at altitude.