Inspired by the Barry Schiff "the very best of Proficient Flying" DVD and the NASA research in the 70s on this topic, I did the Impossible Turn in the Cirrus SR22 full motion simulator in Poznan, Poland this week. OK. I did not wait x seconds after the engine failure at 1000 feet AGL to count for the confusion and misbelieve when you have a real engine failure. I turned into the wind at 45 degrees bank angle and 85 knots IAS back to the runway. Once I 'secured' the runway I have set full flaps.
And now at 300-500 ft...
Commander: that won't work. The Cirrus needs some good speed and sinks quite fast.
What is impossible about a turn at this altitude? 1000ft AGL is pattern altitude and the runway was large. Hardly anything worth mentioning I'd say. You can easily make it to the airfield and if required do some steep s-turns to lose altitude or do a slip with full flaps (forbidden but who cares in case of an engine failure).
When you go back to the simulator, configure it for an engine failure at 400ft AGL and OVC002 as well as a strong wind which gives you a tailwind.
But weren't you the one saying one should always pull the chute in case of an engine failure and not try a forced landing. That video would have been more interesting...
What is impossible about a turn at this altitude?
A "possible turn" so to say... But as AeroPlus pointed out: In real life you will have lost several seconds and 20KT of airspeed until you realise that something is wrong (at least my students in the procedures trainer always do - if they notice at all before the stall warner wakes them up), another several seconds to put the nose down and regain those 20KT, all the while traveling forward and downwards. So by the time you start your turn you may well be at 500ft and a mile away and the turn will then really be impossible.
But what puzzled me most about this video was "Positive rate - Flaps up". Is that really the way Mr Cirrus wants his planes to be operated? Do that with every aircraft I have flown so far and you will slam it back onto the runway immediately (at least when rotating at the recommended or calculated Vr).
A bit OT, but I've always wondered about the "it typically takes 4 seconds to react to engine failure" - on the train crash near Santiago de Compostela the other week, the train's data recorder showed there was a 4 second delay between the driver's shout of horror when he saw the curve ahead and brake application (not that it would have made much difference if he'd applied the brake instantly, I suspect)
I used to have the Cirrus Touchtrainer and tried several times to lower the altitude and return to the runway. Then I posted the results on the Cirrus (www.cirruspilots.org) forum. Since an instructor and his student died training the Impossible Turn at low altitude, the policy is not to train for it anymore and not to advocate it. The solution to this was ... pull the chute :-)
The positive climb was with the appropriate airspeed, but it is a video and I can't write that much text with it, can't I.
My personal limits for such a turn would be 1000 feet minimum in the Cirrus in a real case scenario. I would consider stronger cross-winds and the few seconds to realize on what is going on as limiters. In lower cases I would pull the chute on the climb out.
4 seconds sounds about right, probably quicker if it goes "bang" and splatters you windscreen with oil, and a bit slower when it just fades away...
I had one where it failed quietly when reducing power, so it was a bit hidden, it took a few seconds to recognize that speed and sink rate were not behaving as expected. Total time from engine failure to positive lowering of the nose & start of the engine drills - about 5 seconds; total time from failure to landing - 60 seconds...
The problem with the "impossible turn", and why we bang on during training not to turn back, is that the first thing after establishing the aircraft in a glide is to select an appropriate field. Of course that could include the runway, but in practice it rarely does, especially in aircraft where climb gradient is not twice the descent gradient. On a windy day in a SR22 you can be at 1,000ft before the runway ends... but mind the tailwind!
And as far as I am concerned, if the engine fails the aircraft has let me down, it deserves every bit of damage it has coming its way, all I care about is survival without larger injuries, and for that a field will do nicely.
But what puzzled me most about this video was "Positive rate - Flaps up". Is that really the way Mr Cirrus wants his planes to be operated? Do that with every aircraft I have flown so far and you will slam it back onto the runway immediately (at least when rotating at the recommended or calculated Vr).
I was surprised by the same thing. Gear up fine but I would always delay flaps up until I accelerate over 110kts. Versus 85kt Vr. Maybe a Cirrus uses a different procedure given fixed gear....
As said, the flaps up was with the appropriate airspeed and not only at the positive climb. The field to land at as Cobalt suggests might not be such a good option after all as the Cirrus doesn't like to land in the field as its nosewheel will easily collapse and human injury might be the result. The chute would be a better alternative if I must believe the analysis made by COPA on the accident statistics and casualties.
Right now, the Cirrus community is a little shaken as we have now 100 pilots that died in a Cirrus. In the last few weeks there have been quite a number of accidents, half of them with an instructor on board.
