Turnbacks
 

Just to clarify what I mean by "turnback" it's where following an EFATO (pretty much by definition in a single engined aeroplanes) the captain elects to try and return immediately to the take-off runway, landing in the downwind direction.

I've only ever attempted them under (relatively) safe conditions, for my own interest - generally it's left me with a clear view that it's not a manoeuvre I'd attempt under most circumstances. I've however quite a few times had occasion to observe the consequences of a turnback - which invariably seems to have involved destruction of an aircraft (although rarely loss of life).
Working on analysis of yet another case (and yes, yet another destroyed single engined aeroplane), I've been pondering. Has anybody ever done some serious formal research on the turnback? - how best to fly it, go/no-go criteria, relationship to aircraft characteristics? If so, can anybody point me at any publications describing formal turnback investigations?

This is, I should emphasise, pretty much just personal interest in the topic - not any particular case (yes I am working on something at the moment, but no the turnback performance wasn't really relevant to the outcome and anything I learn now will almost certainly have no bearing at-all).

For that matter, is anybody (well, any full time aviation organisatio) anywhere in the world still advocating turnbacks as a standard post-EFATO action?

G

Ahh, I'd forgotten starting a thread on this five years ago (grief, I need to get out more):

http://www.pprune.org/forums/showthread.php?t=66757

However, I don't think it changes my current interest.

G

Transport Canada has been active on this subject, and there is a civil aviation document TP13748E 'An Evaluation of Stall/Spin Accidents in Canada 1999'. It's referred to on the internet in the context of turnbacks, but it's not explicitly available on-line - so it's a suitable challenge for a university library...

I have in my papers somewhere an early 1990s Canadian article on turning back, which contained some numerical analysis and proved to my satisfaction that turnbacks were all but impossible. Now I regard even tree-tops ahead as a better bet.

I shall not rest now until I have found it, and when I have I'll send it to you.

Graham P.

Turboprops
 

I believe the PC 12 operators look more favourably on the turnback situation due to the better rate and angle of climb of these aircraft, giving a better margin. Obviously there must be a minimum turnback altitude, but in some situations it would be less suicidal than it has been in the past.

This a topic that crops up frequently in the GA press, and I well remember writing to Pilot magazine in 1993 in response to a long article that proved it was an impossible manoeuvre!

My only experience of them is on my UAS days flying Bulldogs. It was a standard training exercise for the more advanced students (I was doing them from about 40 hours), and the QFIs were tested on them during their trappers visitis, I seem to recall.

It was only advocated for runways where an EFATO would put you into the spare bedroom of 13 Acacia Avenue! One of the key elements was that it was part of the T/O brief, so was not a spur of the moment decision, and the direction of turn would be briefed as into any crosswind, to minimise the turn radius. The absolute lowest height to attempt it would be 450 ft agl and the aim was to return to the airport environs first, land back on the reciprocal runway, second.

The actual handling of the manouevre required precise and prompt input, and as I recall was something like this (20 years old memory so bear with me!):

• Smartly lower the nose to 'nail' Vy
• Transfer to instruments and roll to 45 degrees of bank
• Maintain an accurate Vy and 45 degrees on instruments for the first 90 degrees of turn
• After 90 degrees of turn, ease bank off and visually assess likely landing site
• When lined up/wings level consider some engine shutdown drills, but concentrate on the landing

I seem to recall there may have been mention of flaps to 'inter', and even 'nibbling the light buffet', but can't swear to it.

One thing is for certain: if an average PPL decided on the spur of the moment, following an EFATO, to attempt the manouevre, without any prior practice or training, it would most probably end in tears!

I imagine the problem it forcing yourself to fly fast enough with the ground rushing up. Should brief what parts of the airfield or surroundings are landable, just because you took off from a paved runway doesn't mean you have to struggle to get back there. I'm only an ex glider pilot from a big field so I usually had options. Important to pull the cable release not the airbrake in the heat of the moment though :ooh:

Damn you guys are good!

That paper by David Rogers is printing as I type - something to read over lunch today I think, and I shall try from my end to get hold of that Transport Canada document.

A note to 212man; the RAF I know used to teach turnbacks as a matter of course. I think that these were finally killed off circa 1994 after a Hawk with a QFI+QWI crew with about 6000 hours between them landed on a wingtip at Valley at the end of a practice turnback. This led to the realisation that HMQ was losing more jets practicing turnbacks than she saved by having crews current in the practice and I believe they were banned. I'm not sure how this related to Bulldog practices (possibly your local OP was based upon minimising risk to third parties?). But, one point you mentioned rings true with my own experiments - I've found that nibbling the stall warning always gave the least poor chance of getting away with it. That seems to be borne out by the conclusions in Rogers' paper too, who advocates turning at 45° of bank and 5% above the stall speed in the turn.

G

GTE, the paper bears out the practice nicely. I last flew one in 1987, so not sure what happened later, but it wasn't a local procedure, as far as I recall.

Bearing in mind that by that stage we were 'spin/aeros' checked and were familiar with sensing stall bounderies (without a warning buzzing away in your ear: disabled) I'm fairly sure now that we were taught to 'nibble the buffet'. Not all students were taught and it wasn't in our syllabus, but certainly the QFIs were expected to remain current.

Quite, one of the reasons for going to instruments for the first half of the turn a) minimise height loss by flying accurately b) minimise risk of stalling!

The intent was not purely to get to the runway, and it would not be contemplated if there were decent areas beyond on which to force land (as there were from some runways.)

An observation there. I'm not sure that I remember the Bulldog's stall warning ever being operable - but it did have fairly clear natural stall warning. That may however not be true for all aircraft.

G

Having started off on the Vigilant (Grob 109) we were taught turnbacks but they were:

A. planned out for each runway and shown during briefing- 2 runways so four directions that were:

B. graded specifically on height and wind speed direction at the point of EFATO e.g. minimum 300' still air, min 500' wind above 10 knots.

C. as an FSC I was once made to carry through a practice turnback and land in a light tailwind to show it is tricky.

D. captain's discretion - you were not expected to slavishly follow them if it was gusty for example.

E. I'll have a look but the FRC's refer to using the MINIMUM angle of bank to minimise height loss something like 10-15*.

F. I found it very hard not to perform an immediate turnback when I started on the PA28 - those drills were still a sharp instinct 8 years on!

G. Feather the Prop in the subsequent actions.

It was disabled, hence my comments about being able to judge stall bounderies without the aid of something whining in your ear.

What an excellent thread, enjoyable read and the pdf document link is excellent too.

I'm part-way through an FI course. The turnback was a demo and practise on my pre-course entry 'test'; and it's been done once more at the end of one of my air exercises both at height, and in the circuit.

My FI instructor wants me to experience the turnback for two reasons: to sharpen my handling skills; and so that I can talk from a position of strength to my students to give weight to what will be my advice of 'don't try it'. Fair enough.

It'll never, ever be used by a PPL with any success. The vast majority I'm sure won't have the handling skills, or presence of mind to do one successfully. More to the point there's no performance data in the flight manuals regarding turnbacks; there's nothing like the CAA Safety Sense leaflets for turnback guidance (these UK CAA leaflets are great for giving you factors to apply in a variety of take off and landing scenarios) which would give info such as increase airspeed by 'x' or never exceed bank angle 'y' or perhaps most importantly, what to do in wind. The above pdf is great for aero engineers like myself or Genghis, or tp's or the very rare flying techno-anorak. But a ppl won't go anywhere near it - nor will most FI's.

The turnback is all-or-nothing. If you don't get back to the field with it you're arguably in a much worse position than a crash straight ahead. Ground speed in any sort of wind will be greater. Eg in a 10 knot wind I'd rather crash straight ahead at a groundspeed of 30 knots (I fly a DHC-1) than turnback, fail to get to the field, but crash at 50 knots -

Most definitely a potential 'killer', one of the major problems being the loss of airspeed going from a headwind to a tail and the tendency to stall in the turn therefore.

Nothing wrong, with a QFI onboard to show you what it is like, but keep the carb heat on and a hand on the throttle:eek:

As D120A and others say, there is normally a better option ahead, +/- 45 deg, unless you are in a high-power machine when you may well have enough altitude. They were always exciting in glider training after a lowish cable break.....................

212man, your recollection of the Bulldog turnback drill isn't quite correct. (I still have my QFI CFS lesson plans and copy this directly from them):

Under the "Stalling in Manoeuvre", Advanced Turns (2).

(6) TURNBACK (n.b. This being the final part of the lesson)

LOWER NOSE TO 80 KT ATTITUDE
SMOOTH AND POSITIVE 45* ANGLE OF BANK
LEVEL WINGS ON RECIPROCAL
FLAP A/R
MIN HT 350' AGL.

I think at Vy you might well come a very nasty cropper, hence this being included in the "Stalling in manoeuvre" sortie.

Yes, we used to practice them and be standardised on them; I used to brief EFATO drills for every takeoff and although the turnback option (or not) was included in the brief, I would be reluctant to do one for real if there was any other option.

BOAC wrote:

There's no loss of airspeed turning out of a headwind to a tailwind. If there were, you couldn't fly a 360 at height without constantly changing attitude to maintain a constant speed, 'cause there's head and tail wind there as well.

However, if you turn back in these circumstances your groundspeed increases markedly (by 2 x windspeed once you've completed 180 degrees). As all this is happening close to the ground, and is thus very noticeable, the natural tendency is to pull off the apparently excessive speed. Result, tears all round.

At height you don't notice the apparent speed change in relation to the ground, and so this doesn't happen.

A bit sweeping and does not apply if you change heading very rapidly. The numbers matter.

JF

I have a couple of links to papers by Rogers on the tech log sticky which may be of interest ..

Ask the RAAF
 

During 2FTS the RAAF used to brief the manouvre for every T/O in the Macchi. I don´t remember the details - maybe someone Dunnunda in Godzone knows. I wonder if the RAAF still do turnbacks in the PC9 or Hawke?

PITHBLOT

Some time ago on the RAF Hawk, we briefed turnbacks at Valley, but only after achieving a minimum of 250kt - and also Valley was fortunate in having an alternate runway not too far round from the main.

Shytorque, without trying to make excuses (apart from the memory loss!) I thought Vy was 80 kts. Now I'm curious: what was/is it?

Loss of airspeed turning out of a headwind
 

I think there is a great deal of discussion that could be had re a loss of airspeed turning out of a headwind into a tailwind. Apart from the obvious discussion of CAS versus EAS and the ramifications of very slow speed aircraft such as a First World War type, there is the issue of slip and skid, whether the aircraft is in or out of balance, whether the aerodynamic forces involved are in equilibrium and, not a insignificant factor, the rate of turn used. Instantaneous effects, regardless of the overriding criteria of 'airmass movement' versus 'ground observation' will show that the airspeed will change during such a turn. If you take the manoeuvre to criticality, i.e approaching the 1G stall speed, in a twin engined aircraft with the critical engine (the engine inop that requires the highest speed for controllablility) shut down; I think you will very well find a decrease in airspeed turning out of a headwind. All thoughts welcomed.

Neville Shute Norway in his autobiography gives an interesting account of F/L George Stainforth, who had been hired to test fly the first Airspeed aircraft, sitting for hours in the cockpit prior to the first flight asking the same questions over and over again and running his hands over the controls repeatedly. The Airspeed group actually became rather uneasy in the situation and thought perhaps they had selected an unsuitable test pilot. On the first flight the engine stopped at less than 500ft after takeoff, F/L Stainforth did a 180 and returned to the field for a perfect downwind landing. Then they figured it out.
After an excellent landing you can use the airplane again!
(and Airspeed did)

In a previous life I used to occaisonally back-seat while the pilots practised turn-backs. I seem to recall that certain combinations of height/speed/wind in the Hawk would provide insufficient energy for a safe return to the field. It was always interesting and sometimes exciting to watch :-)

The RAAF did indeed look into this matter and there is a document on the net somewhere detailing the results of their experiment.

In short 3 of the 5 test pilots on the programme died during the trials.

Turnbacks were (are) routinely practiced by QFI/QPNIs at Linton in the Tucano. You needed to have achieved a speed/height combination (130kts/500ft from memory). Without this you simply did not have the energy to complete the manouvre and still have the energy to safely eject if needed!

Options available included a 'dumb-bell' back on to the reciprocal, or a less agressive turn onto finals for another runway. The former involved a hard turn (45aob), invariably away from any crosswind (to help blow you off the centre-line) prior to a swift reversal onto the the reciprocal. (A bit like a procedural turn!). In the Tucano a top-tip was to take mid-flap ASAP to arm the stick shaker and give a slight increase in performance. Needed plenty of awareness re: speed, height , bank, ROD, seat parameters etc. The aim was simply to see if a safe landing could be made rather than jettisoning the aircraft! Landing ahead in a field was not recommended.

Another option was to simply fly a 180 as discussed by others, but this was rarley practiced due to relative position of the ac v runway. It was only an option when the ac was well upwind or had more energy when you could glide ahead for a few seconds prior to initiating the turn.

Take-off brief included details about the turnback to be flown if required. You needed to think about the planned departure; you were often planning to turn at 500ft anyway, so this needed to be included in the turnback plan. Have heard some people call it a turnback only if you go for the reciprocal rather that an alternate runway.

Finally...

I'm with JF - I don't agree! Wish I could prove it with the numbers, but I can personally testify that, in a light/slow aircraft, a slick turn onto a downwind trk will temporarily reduce your IAS and not instantly increase your gnd speed.

:)

Having practised turnbacks as a routine exercise as a QFI in the RAF, I have always had the thought as a possibility in later life, but practices (at a safe height) in draggy light aeroplanes convinced me that it wasn't a goer. However, for the sake of interest, I investigated the technique again as a part owner of a Europa. I found that in this slippery aeroplane, it was perfectly viable given certain criteria as mentioned by previous guys above.
You must pre brief before takeoff as to the lower and upper height limits and the direction of turn. These vary considerably depending on the wind but roughly in the region of 600 - 1200 ft for this type.
Looking out in the initial stages of the turn is not good news due to the apparent slip and acceleration over the ground. Once round the problem is nearly always a tendency to overshoot rather than drop short.

Do you know- I can't remember it and can't find that figure in my notes - which are now some 18 years old! Having said that, turning at 45 degree AOB increases the stalling speed by a factor of approximately 1.2 hence the need for an increase over normal glidespeed.

I don't think "Vy" is the correct term to use for a glide because it refers to best climb speed. :)

This may be simular to an EFATO turn back and what may be encountered -

When I do fixed wing or Heli cattle mustering, I encounter what I think of as a rolling turbulence (Roller) that travels with the prevailing wind.
In the Super cub, when I do a come back turn to keep the pressure on an animal,[ once I have the aircraft pionted down hill ] I sometimes get the down side of this roller and see the effect of a rapid increase in ground speed and desent, yet a fixed airspeed (Due to the position of the main herd and other facters, a musterer can not always operate into wind)
What I do in windy conditions to compensate for rollers and wind shear, is to manuver at a faster airspeed, then if all is right when pointed down hill, I do a full slip to slow down if needed.

I used to believe myself turning from headwind to tailwind would result in a marked decrease in IAS, but this is a common mistake. The only loss in IAS results from increased induced drag during the turn.

No matter how rapid your heading change is when turning head- to tailwind there will be no decrease in IAS. Your airspeed doesn't change in relation to the cube of air you are in, only to the ground beneath you.

In order to understand this fact (and I still doubt myself sometimes) I imagine the cube of air being still and the ground map moving like on a rollerfloor below me. No matter what my heading is the air around me doesn't move and IAS remains the same. On the other hand, my speed in relation to the ground does vary with my heading.

I72 driver

Sorry – not very fair of me to set you up like that. Well not you particularly but just anybody who came out with that sort of standard comment.

I said it depended on how quickly you turned and I stick by that. If your mount is a 172 you are quite correct. However some other aircraft have very different capabilities which can change the situation.

Take a Harrier on a heading of north doing 60 kt IAS on a flat calm day. Now change you heading to south in three to four seconds (easily done with full rudder I can assure you). I think you will probably be able to see for yourself that you will end up not with less than about 45 kt of backwards velocity (a bit more than a ‘reduction’ in IAS) because the drag that acts for that brief period of sideways flight is fighting the momentum of eight tons of aircraft doing 60 kt and does not represent much of a 'brake'.

Do this same downwind turn manoeuvre into an original headwind of say 30 kt and you might even die should you let the nose drop a tad so that the backwards IAS can get under your tailplane and blow it over your head.

BOAC knows ‘cos he was a Harrier pilot too.

Regards

JF

Rogers' Paper
 

We are dissuaded from turning back following EFATO because experience has shown this all too often ends with stall/spin fatalities.

Am I alone in thinking that, in advising this maneuver be flown using a 45deg banked turn at 5% above the stalling speed, Rogers must therefore be completely barking?

Whilst some pilots may remain calm and collected following an actual EFATO, thereby fitting the profile of the simulator test group that supported the feasibility of his recommendation, I suspect the vast majority would exhibit varying levels of panic and be unlikely to fly with a very high degree of precision. Is it a terribly good idea, then, for pilots to be taught to turn back flying a profile where, using numbers appropriate to Rogers' Bonanza example, a stall would follow from them getting the speed low by 4 kts or from overbanking by 5deg?

I do believe that a turn back plan is good to have in mind ... but one where the numbers are based on Eckalbar's 45deg bank at 1.3x Vs45deg (which Rogers rubbishes as "from the popular aviation press" and "demonstrably incorrect") strikes me as infinitely preferably to the truly dangerous approach proposed by Rogers!

JF wrote:
Not that I was Harrier pilot (although I was a design engineer at Hawker's during your era, JF!), but do you not think your momentum example applies equally in turning upwind? Just posing the question!

Regards
Islander2

JF, now I see what you mean but then we aren't talking about the wind anymore, are we?

I'll try to turn around quickly next time I ride my Cessna... maybe if I pull really hard? :ok:

So guys (specifically the increase groundspeed - don't lose airspeed - when you turn downwind brigade). Tell me why you think you feel any turbulence in a gusty wind? If your little box of air keeps moving about then surely your aeroplane inside does???


Turning a slow Harrier, or helicopter downwind is still flying. The 'problem' is that they can both achieve high rates of turn and fly at a low IAS at the same time, so the problems of losing IAS when turning downwind are far more pronounced.

:ugh::ugh::ugh:

Surely if you are trying to return to the runway, you want to minimise the turn radius and therefore the number degrees of turn required to return. In a utopic world you would turn about the centre axis and point back at the reciprocal heading.

H Peacock wrote:
You were doing fine, until you got to that bit. Maybe the head-banging, though, suggests you won't want to re-consider!

There is no difference between turning down upwind and turning downwind at a given IAS. In both cases, a high rate of turn will result in loss of airspeed due to inertia and/or increased induced drag (the amounts of each depending on how the turn is achieved), and the loss will have exactly the same magnitude in both cases.

For conventional fixed wing aircraft flown according to IAS, there is no difference between upwind and downwind turns in steady wind conditions.

Harriers and helicopters are different! A hover stationary with respect to the ground is the extreme example. Turning infinitely rapidly from a 30 kts headwind to a 30 kts tailwind produces a 60 kts loss of airspeed. Vice versa produces a 60 kts increase in airspeed. This is the effect JF got very accustomed to and very much needed to concern himself about with respect to the wind getting under the tail. BUT, the IAS is completely different in each case. Turning downwind, you're starting with an IAS of +30kts. Turning upwind, you're starting with an IAS of -30kts. Try this now without keeping the Harrier (or helicopter) stationary above a point on the ground, and arrange for it to have a +30 kts IAS in both cases, and the loss of airspeed will now be 60 kts whether turning downwind or upwind.

I was reading through it yesterday, and it did strike me as a nice bit of theory mercifully untainted by too much contact with the real world.
Whilst without doubt, it's a useful start to any analysis of the problem, and that you'd not really want to go below 1.05Vs - one's ability to fly this would depend a lot upon the aeroplane. I dont know the Bonanza personally, but if taking (say) a PA28 or Hawk you'd probably get away with it - the combination of high stick forces, clear audible stall warning at about that margin, and benign stalling characteristics should allow you to nibble the stall effectively. On the other hand, there are aeroplanes with poor stall warning, low pitch forces less than glaring stall warning, and less than benign stalling characteristics that I suspect would offer much too high a probability of crossing over into the stall - most likely fatally.

Oh yes, and how accurately can somebody hold a precise bank angle? At low level, under stress, and whilst keeping looking out of the window most of the time.

I certainly hope that if I'd been asked to referee that paper (I wasn't), I'd have had the sense to ask for inclusion of a few test points in support of the theoretical analysis. Without that, the drawing of conclusions is arguably a bit over-confident.

G

212man, I take your point, but the aim here is to use the crosswind to help in a different way. A jink away from the crosswind will quickly displace the aircraft from the centreline such that a tight 180 (plus a bit) will line you up back on the reciprocal. If there is no crosswind and you fly a turn you will clearly be displaced by the diameter of the turn from the centreline. If you had a very strong crosswind you could dispense with the jink away and just turn towards it, this would again reduce most of this displacement. Wish I could draw it to make things clearer.


Islander2 wrote
Think I follow that and agree; your IAS will change when turning upwind/downwind.

Not true. A slow aeroplane/helo does not need lots of 'g' to achieve high turn-rates, and therefore doesn't suffer from a big increase in induced drag.
In the Puma I could fly downwind with a low (+ve) IAS, but when I flew a balanced turn in to a strong headwind the IAS would rise dramatically before steadily reducing back to the steady state condition. You gain IAS turning in to the wind, lose it turning downwind - is that what you're saying?

:)

H Peacock wrote:
Have another look at my post: it doesn't say that, does it? What it says is: "a high rate of turn will result in loss of airspeed due to inertia and/or increased induced drag (the amounts of each depending on how the turn is achieved)"

Also:
Categorically not, 'cos it isn't necessarily true! What I am saying is: 1) that the impact on IAS is identical whether you're turning upwind or downwind for the same pre-turn IAS, and 2) that whether or not the turn results in an increase or decrease in IAS is dependant solely on whether the IAS is initially greater or less than zero, and is nothing to do with whether the turn is upwind or downwind. If the pre-turn IAS is positive ahead of a very rapid turn, you'll temporarily lose an equal amount of airspeed whether turning downwind or upwind; if it's negative (obviously not a usual fixed wing scenario, but certainly one for Harrier pilots), you'll temporarily gain an equal amount.

Edited to add an intepretation note: all the above assumes a steady wind.