Originally Posted by
BackPacker
Yep. Nicely done.
So essentially you confirmed the Rogers (?) paper that the lowest loss of altitude, assuming certain stability conditions, is achieved by making a 45-degree banked turn at the edge of the stall. Good.
GtE, I'm still interested in your stab at doing a paper when you remove those stability conditions, and analyze a proper EFATO situation. Mark1234 found out that the best results in that situation are achieved by initially setting a 60 degree bank, pulling to the stall, and reducing the bank angle as you fly the turn and the nose drops.
So am I, and it might be viable to open this up for wider participation.
I'm at the final stages of leading a multi-author paper analysing optimal stall recovery actions for single engined aeroplanes at the moment (5 of us as co-authors, and between us we've tested currently 12 aeroplane types, and that'll shortly be 15) - that should be done in the next couple of months and we can then agree the final wording and analysis and I can get it into the peer review.
We did this on the basis of preparing an outline test plan and analysis route, then opening the project up to participation. Several colleagues have done so - using their own flight resources, and it's been a lot of fun. Also, it's prepared conclusions that are based upon test results on 15 aircraft and brains of 5 variably experienced pilots. The best NASA have ever done on a similar project was I think 7 aeroplanes and 3 researchers.
[Hence our firmly unofficial lab slogan - "if it was easy, NASA would have done it already" !]
So, starting at Rogers' paper, I think one needs to construct algebraically a turnback that looks something like this:
(1) Engine failure, initial deceleration from Vy over a period of seconds while the pilot says "oh f********" slowly, then pitch to target turnback speed.
(2) Roll left to 45 degrees of bank whilst maintaining target turnback speed
(3) Change heading through 45 degrees.
(4) Roll right through 90 degrees to 45 degrees right bank
(5) Maintain 45 degrees right bank and target turnback speed until approaching runway centreline
(6) Reverse again to 45 degrees left bank to intercept centreline on runway track.
(7) Establish 1.3Vs and drop gear and flaps.
Each of those 7 is fairly easily modelled initially, although the model will certainly contain errors. Specifically:
(1) I have a model already for deceleration post engine failure that I published a few years ago and use for planning stall tests in certification programmes. There are standard values for reaction times, but this can also be tried in the simulator with a selection of pilots (in fact I think we might have that data already from another project).
(2) There are standard minimum requirements for achievable roll rate in the certification standards. If we assume that, it's a reasonable worst case.
(3) Standard flight mechanics
(4) As (2)
(5) As (3)
(6) Probably a bit of a guesstimate needed, but do-able.
(7) As (3)
We could then construct a test plan around that model which for a range of airframes at a safe altitude does the following:
(A) Confirm or deny 45 degrees of bank as best for height loss in the turn using a nominal fixed speed.
(B) Determine for that aircraft, the optimal speed for height loss in a turn at the optimum (hopefully 45 degrees) bank angle, and the height lost at that speed and bank angle.
(C) Validate the estimate for that aircraft of deceleration following an initial engine failure with height held for a pre-determined "Oh f**********" time.
(D) Determine height lost and time taken in accelerating to turnback speed following initial deceleration.
(E) Determine height lost and time taken to go from wings level at turnback speed to 1.3Vs with gear and flaps down.
(F) Finally and crucially, assess the actual handling difficulty in handling each part of the manoeuvre, and the manoeuvre overall, with a particular concentration on stall avoidance and runway centreline capture, using standard scales for pilot compensation and workload (Cooper Harper for the first, probably NASA-TLX for the second).
The analysis will take some time, the flying would probably take around 90-120 minutes on a single sortie for an experimental test pilot, nearer 180 minutes over 2-3 sorties for somebody not experienced in flight test ideally working with a flight test engineer in the other seat.
Standard GPS units would do a reasonable job of recording flight tracks, plus voice recorders (a dictaphone plus a tie-clip microphone stuck in the headset earpiece) can help a lot in gathering pilot comments, although I can bring a portable flight data recorder to the party.
If we could get the number of aeroplanes tested into double figures, back it up with some robust analysis, and include a few flying instructors in the authors list to make sure anything recommended is sensible and feasible. Well not only would that go well beyond Rogers' initial analysis, but we'd end up with something that could be a real benchmark in safety practice and training.
Worth doing I think. I think I'll do it and open it up to collaboration (and shared blame or glory as appropriate!); I can certainly bring three aeroplanes (a modern 4-seat low wing, a vintage high wing taildragger, a microlight) to the party and can think of a few fellow flying researchers who'd enjoy joining in. (
Pilot_DAR,
India-Mike, any interest in playing?)
G