Stop calling it the impossible turn

Pilots love a good debate, and some topics seem to come in and out of fashion like bell bottoms. Right now the wars over lean of peak and angle of attack indicators have cooled (thankfully), but the war over “the impossible turn” seems to be heating up. In the last few months I’ve seen multiple articles, videos, and forum threads on the subject. It’s fun to debate, but what problem are we trying to solve here?

As my 8th grade English teacher always said, we first have to define our terms, and there we find the start of trouble. The impossible turn is actually quite a vague phrase. It generally refers to a low altitude, 180-degree turn back to the departure airport after experiencing an engine failure on takeoff. The classic scenario involves a sudden power loss below 1000 ft. AGL, where the pilot feels an overwhelming urge to crank and bank their way back to the safety of a runway. There’s not much margin for error, especially at low altitude, and if it’s not done properly the result can be fatal.

I appreciate the desire to make an impression, and that’s clearly what safety advocates are trying to do with their catchy phrase—if the engine quits on takeoff, resist the urge to turn. It’s impossible. There might be a perfectly good field right in front of the airplane, and landing there is easier and safer than trying to get back to the airport.

But while this may be good advice for a new pilot, very few things in aviation are quite so black and white. Such binary language may discourage a pilot from ever trying a turnback, which is flat out wrong. Sometimes, returning to the airport is the safest option and if you’ve never practiced it or thought about it there’s no way you’ll pull it off successfully.

MDW
Taking off from Midway doesn’t leave you many good options?

Think about some urban airports that are surrounded on all sides by housing or offices. Chicago’s Midway Airport, Atlanta’s Peachtree-DeKalb Airport, and Dallas’s Addison Airport all come to mind. On takeoff from one of these airports, an engine failure a mile off the end of the runway leaves you with no good options. Should you automatically exclude the large airport area with four runways and instead try to shoehorn the airplane into a residential neighborhood? That hardly seems safer.

I believe pilots make the impossible turn successfully every year, but you rarely hear about them because good news rarely makes the headlines. Here’s just one example of a turnback in a single engine piston airplane that worked out quite well. It was done at low altitude, and we can all second guess the decision, but it’s clear that the turnback was successful. The pilot walked away.

Glider pilots practice this maneuver all the time, simulating a rope break on departure. When I earned my glider rating about 15 years ago, I remember being shocked at how low we could make a safe return to the airport in such a situation. In many scenarios, the first step was to turn left, count a few seconds, then turn back around 270 degrees to the right in order to line up on final. Those few seconds felt like an eternity, but with proper planning and technique, it worked out. Granted, gliders have highly efficient wings compared to a Cessna, but it’s clear that the impossible turn isn’t impossible at all. We simply have to know when it’s safe and when it’s not.

This brings up the critical question: how low is too low? The correct answer varies from airplane to airplane and even airport to airport (a large airport with multiple runways means a shorter turn is required to find pavement). About 1000 ft. AGL is a good starting point, but it could be as low as 600 ft. or as high as 2500 ft.

The only way to know for sure is to practice, but make sure to do it from a safe altitude and with an instructor onboard. This is no time for steep turns at 200 ft. AGL. Every year during recurrent training for the Pilatus PC-12 I fly, I practice turnbacks from a variety of altitudes and airplane configurations. I’ve done this both in a full motion simulator at Flight Safety and in the airplane with an experienced PC-12 instructor in the right seat, and I never fail to learn something new.

In the airplane, one option is to simulate a takeoff by flying down the runway at slow speed but from 1000-2000 ft. AGL. Climb out as you normally would, and at a predetermined altitude pull the power back. Make your turn and set a hard deck to create an airport in the sky—if you’re not “on the runway” by 1000 ft. AGL it’s time to abandon the maneuver and go around.

Over the years I’ve come away with some important lessons for this maneuver. First, and most importantly, the initial reaction should almost always be to push. If you’re climbing out at Vy, you will likely be at a high pitch angle and without much excess energy. The human instinct is to pull back, but doing so in this situation can quickly lead to a stall. I tell myself to push forward and reduce the angle of attack before doing anything else. At the very least, this buys time.

Next, it’s critical that you make the turn aggressively. That doesn’t mean dangerously—you’re low to the ground and at slow speed, so you want to be very careful of an accelerated stall—but if you only bank 10 degrees you’ll never make it back. In the Pilatus, at least a 30-degree bank seems to be about right, but that feels like a lot when you’re low to the ground. Don’t let the airplane overbank and don’t pull back to tighten the turn.

Finally, wind matters. This is obvious from a physics standpoint, but it’s often overlooked in the real world. Taking off in calm winds might mean your turnback leaves you far away from the runway, while a 20-knot headwind can leave you right on top of the airport when you roll back out. In some practice scenarios with strong headwinds on takeoff, the biggest issue is getting stopped on the runway after turning back. The headwind keeps you close to the airport as you climb, but then pushes you when you descend. To avoid running off the far end of the runway, full flaps and even a slip may be required. Solving this problem simply takes practice.

172 on short final
Can you make it back?

There’s no doubt that this is a maximum performance maneuver, but then again a catastrophic engine failure is an emergency situation that demands decisive action. With planning and practice, the impossible turn becomes just another tool in the bag, one that can be invaluable if used under the right circumstances. Many open-minded pilots now recognize this, from Captain Brian Schiff (who forcefully advocates for “the possible turn”) to the FAA (who updated their guidance in Advisory Circular 61-83).

The key is to have a plan that you brief before every takeoff (even if it’s just to yourself). You have to know what your options are and be spring-loaded to react properly, given the conditions. For me, this briefing always considers a few essential things:

  • Runway length: when can we just land back on the runway? A 10,000 ft. runway gives you more options straight ahead
  • Airport environment: is there a parallel or intersecting runway that is more convenient to turn towards?
  • Surrounding terrain: are there any good options besides the airport or conversely, are there any obstacles to avoid?
  • Wind: I like to turn into the wind when possible, since this requires a less aggressive turn to be lined up on final
  • Ceiling and visibility: do we need to set up any avionics to find our way back to the runway if it’s low IFR (synthetic vision helps a lot here)?

This doesn’t have to take long, but a typical briefing for me might be, “We’ll be taking off on runway 21L and climbing out on runway heading; below 1500 feet MSL an engine failure means we’ll land straight ahead in that open field; above 1500 feet MSL we can make a right turn into the wind to land on runway 3L; VFR weather means we’ll do it visually.”

Also, don’t get locked into landing straight ahead or turning around. Sometimes there are other options that should be part of your takeoff briefing. For example, Cirrus pilots typically set a minimum altitude for deployment of the whole airplane parachute system (CAPS). In many cases this is a better idea than either a turnback or a forced landing on a road or parking lot. The point is, you should use all the tools at your disposal to give yourself options, but make sure you think about those options from the calm of the taxiway and not at 800 feet with a quiet engine.

Impossible turn? Hardly. Calling it the possible turn is more accurate, but I prefer “the unforgiving turn.” Like an ILS to minimums, it can be done safely with planning, practice, and discipline.

49 Comments

  • I agree with you John. Had to do it in a J-3 once, way back when… Engine went dead silent on the climbout…crickets. It was make it back to the airport or land in tall trees. I liked the idea of the former. It was low, about 400 feet, had a second pilot in the airplane with me. It was a single runway airport. Initial thought was to tear drop back to the concrete. In the course of doing the maneuver it looked like getting all the way back around and properly straightened out was going to be marginal. Elected to land in the mowed grass paralleling the runway. Full control all the way down that way with a solid plan. Took quick action, there’s not a lot of time to run scenarios when its that clutch. You have to have prepared yourself for it beforehand, as you stated above and as I had done prior. Use every trick you’ve got in the bag. Good article. Thanks.

  • You use the conventional label “180° turn” near the start, but later, when you talk about practicing 90-270 turns in glider training, you rightly indicate out that it takes about 360° of turning to get lined up with your departure runway surface (modulo winds). So the real test for pilots practicing up higher is how much altitude they lose in 360° of turning—with a turn-direction reversal somewhere—rather than 180°.

    Obviously, in real life there might be other options that require less turning, like landing on a different-direction runway, on a taxiway, or on the grass in the airport infield, but a 360° turn (or 90-270) really should be the reference point: 180° might give pilots too much false confidence about how low they can pull off the maneuver.

    • A very good point. Really, the biggest lesson I’ve learned is that there’s no such thing as a “standard” turnback. Depending on the airport, the terrain, and the wind, it could be a 90-degree turn or more than a 360-degree turn—I had to do S-turns once because I was too high.

      I know training time is limited, but practicing this maneuver (safely) is the only way to really see how your airplane behaves in different situations. Even if you never use it for an engine failure, I think it’s great for improving overall mastery of the airplane.

      • Thanks John
        Good article. Let me run something by you. On takeoff and after clearing obstacles, what about side stepping your departure toward the downwind side by a small heading. That way if you have an engine failure at an appropriate altitude your turn back could be a simple 180degs as opposed to 270 plus your turning into the wind.
        Thoughts?
        John Arsenault

    • Yes, in the RAF we teach pilots to aim for the airfield on a turn back, not starting below 600ft agl and using 40deg bank maximum. We turn one way, normally into any crosswind, lining up with the runway if height permits. And we practise, a lot, to go around.

  • Thank you for a timely article.

    As you mentioned, there is no data on how often this maneuver is completed successfully, hence no way objectively to evaluate it’s risk – fertile ground for an opinion free-for-all.

    I absolutely agree with your assertion that turns be made aggressively. Time with an instructor practicing steep, low-speed turns while unloading the wing is probably well spent.

    In a full-blown emergency the objective might be to make the airport not the runway.

    • Well John, you forced me into the cockpit!

      In succeeding posts, you’ll see a lot about pushing down the nose to avoid a stall. Excellent advice in many situations, but not all.

      After my comment above, I flew this maneuver 10-20 times at 3000′ AGL in a Maule, reasonably light, trimmed for climb, and with the CG near its forward limit. When power goes to zero in that configuration the airplane’s natural stability drops the nose with alacrity. If the pilot doesn’t counter with aggressive back pressure, seconds after power loss, airspeed and altitude loss grow rapidly in the turn.

      My best results:

      – do nothing for 3 seconds.
      – apply takeoff flaps (decrease stall speed)
      – pitch up to bring airspeed to stall+10kts (minimize turn radius) and initiate a 45 degree banked turn.
      – apply strong back pressure to keep the nose near the horizon during the turn – ease SLIGHTLY if stall warning horn or AOA indicator trips.
      – roll out as airport comes into sight, retract flaps and pitch for best glide.

      This approach was probably optimal because it kept the wing near its max CL and minimized the turn radius. The wing’s Clark airfoil also minimized the risk posed by a stall.

      If I learned anything from the exercise it is how dependent this maneuver is on weight, CG and configuration – and how quickly things can get out of hand when the wing is blindly unloaded without regard for the airplane’s state.

      One thing I can reaffirm from my previous comments is the value of practicing the possible turn. I promise you will learn something.

      • Thanks for sharing the results, Kim. Very interesting. The only universal may be “it depends on the pilot and the airplane.” And I suppose a reminder that practice is always good.

  • Last week I took a day trip up to Greenville, Maine; it was the first time I had been back there since I flew fire patrol during the summer of 1977. Departing runway 32, I suddenly remembered my practice of this turn during that summer, as the forced landing options were, and still are, pretty lousy. I don’t remember the actual numbers, but I had worked out what altitude I needed to successfully make the turn back. Tootling back to Vermont, I had a bit of a chuckle thinking about this, as once you are away from the airport at Greenville, you’re going into the trees anyway. Which raises the question of just how much cut-back of the right-of-way alongside which logging road you would need to land successfully!

    We actually tried this in the MD-80 simulator for a south departure from DFW runway 17R. This has the advantage of the west side runway complex, so we simply cranked the ship around to land on 36L. We could do it, but the importance of unloading the wing cannot be stressed enough. Doing so not only expands the load factor margin in the turn, but reduces induced drag, perhaps briefly improving the glide ratio.

        • Unloading = reduce the load on the wing. In straight and level unaccelerated flight (or sitting at your PC) you experience 1 G. In a level turn at 60 degrees of bank you have 2 Gs. If however you use forward stick/yoke to get that “light in the seat” sensation you are close to Zero G which reduces your induced drag.

  • With winch line breaks in a glider, 500′ allows a mini circuit.

    But the very first reaction after a line break is getting the nose down and waiting until you have approach speed. Until then DO NOT bank. In power aircraft much the same applies except that higher altitudes and speed will be required.

    If any crosswind, turn downwind if returning to the reciprocal runway. That way a 180 will tend to have you lined up reasonably.

    And yes too strong a headwind can preclude a downwind landing.

  • What a wonderful article, John! In a sense, you changed my mind on it. Not that I wouldn’t consider turning back from crosswind, for example, but certainly I’ve been advocating that the simply idea of trying to turn back (high bank angles and low airspeed is an awful combination) is far too dangerous to become a tool – which it might be. In the Cessna 152/172, 1000ft AFE is likely comfortable, and as you point out sharply, the airport surrounding areas are a big part of the equation, and Brazil, where most of my Portuguese written articles readers fly, is prone on having rough terrain and residential areas around airports. Just loved your briefing format, it’s not very far from what we do in the airline I fly for (all type and environment differences considered, of course), and flying professionally has little to do with flying for money: every pilot should seek a standardized operation to remain as far from trouble as feasible, either carrying 400 passengers or your own dog. Having said that, thank you for the amazing perspective you gave us, and the very well drawn exercise with the hard deck with the “virtual” airport. Other than that, the “best glide, look for a field” never gets old! Cheers!

  • I have some thoughts about this. You are right that the situation is never black and white. But for MOST private pilots, and even some commercial ones, the turn back is riskier than landing straight ahead. I teach that it would be better to land in an undesirable spot (on a roof, in the trees, in the water) in a controlled manner than over a runway when stalled. It’s akin to teaching someone to fly into a thunderstorm. What constitutes a thunderstorm? If it’s red on the radar? Pink? What about yellow? I know some pilots that have pushed this and come out fine, but I still avoid anything red or above or yellow and quickly building. On a personal note I have a friend who recently attempted this turn and lived, but with severe injuries that may prevent him from flying ever again. In my local area we have mostly farm fields, trees, and neighborhoods around the airports. I’m always looking for the best places to land in case of engine failure. Even in those congested areas around Midway there’s likely a road that could be wide enough to accomodate a small airplane.

    • I agree that this is a tough subject, but at some point the turn back has to be possible, right? Maybe for a 10,000-hour crop duster it’s 500 feet and for a student pilot it’s 2000 feet, but I just can’t believe it’s never possible.

      To use your radar analogy: I agree you should never fly through red, but I worry that our current teaching on turnbacks is often like telling pilots to never fly through green returns. That’s just needlessly restrictive.

      You’re 100% right that landing the airplane under control—anywhere—is better than stalling it over a nice landing field. That’s the key. Fly it as far into the crash as possible.

      • This. At some point the student has to understand his or her own limitations. Its akin to learning to land in Xwind, in the beginning we teach to be very cautious of winds and gusts, but as we gain more experience we need to push those personal limitations. For single engine student pilots I brief and teach any engine-out below 1000′ we’re straight out, but as we start getting into more advanced ratings we talk about and practice more advanced maneuvers. We need to start understanding that 1. A turn back to the departure runway is *more* than 180* and 2. With parallel runways at DVT we have more options like the other runway or multiple taxiways. It comes down to understanding energy management and being able to successfully trade altitude and airspeed per our needs.

  • Thank you, Mr. Zimmerman. It’s refreshing to see a serious discussion underway regarding the ‘unforgiving’ turn. I fly out of fewer and fewer airports where an off-field landing would not result in injury or loss of life to those on the ground. A 180 begun at 450-500 feet is well within many GA planes’ margin of safety and deserves study, and practice, for those of us who fly out of urban airports.

    Theory
    The altitude lost in a 180° turn is a function of drag and time. Induced drag is fixed once the plane is in the air (weight, wing loading, airspeed). It takes the same amount of energy/lift to reverse course, whether done at 20 degrees of bank or 60. The amount of bank has little effect on the induced drag accrued in a turn; however, bank does determine how long the a/c will be in the turn and subjected to parasite drag.

    Bank angle and TAS are the drivers re how long it takes to complete a 180. The formula for rate of turn is 1,091 x (tangent of the bank angle)/TAS. Because the tangent goes up with angle of bank any increase in bank will increase rate of turn and any increase in speed will reduce it. And since the distance between a dead-stick airplane and the ground can be measured in precious seconds, an expeditious rate of turn is desirable. To point out the obvious, more bank and lower TAS are what you want (with an eye on best glide speed and stalling speed as g’s increase). To illustrate: a 180° turn at 30° of bank at 85 KTAS takes 24 seconds to complete; a 180 at 45° takes 14 seconds; and a 180 at 60° can be completed in only 8.1 seconds. Double the bank from 30 to 60 and cut the time needed to head back to the field from 24 to 8 seconds. The steeper bank has the added advantage that the offset from the runway is less on rollout.

    Yes, we are all aware that steep-banked turns near the ground invite a stall. A 2 g tug on the yoke will stall my 20E at 90 KIAS; coincidentally, 2 gs are required to hold a plane in a stable 60° bank turn. Not much room for error when you slip below Vy and crank it over to 60° of bank.

    Practice
    Several iterations (at altitude) in a Mooney 20E yielded average altitude losses in a 180° turn of:
    360′ lost at 30° bank,
    270′ lost at 45° bank, and
    200′ lost at 60° bank.

    These were entered and flown at 90 KIAS; TAS was 100 knots.

    Caveats
    It’s important to bear in mind that these figures are from a slippery Mooney which suffers famously little parasitic drag. Consider the coefficient of parasite drag for several popular GA planes:

    Aircraft Cdp Flat Plate Area (sq. ft.)
    Mooney 201 0.017 2.81
    Beech Bonanza 0.019 3.47
    Piper Arrow 0.027 4.64
    Cessna 182 0.031 5.27
    Beech Sierra 0.034 5.02
    Piper Warrior 0.034 5.83
    Cessna 172 0.036 6.25
    Cessna 152 0.038 6.14
    Beech Skipper 0.049 6.36
    Piper Tomahawk 0.054 6.64

    The altitude loss in most other GA aircraft is going to be greater than in a Mooney. Run your own experiment to get a sense of how much altitude you would need to consider the ‘impossible turn’. Also, a lighter aircraft (Mooney, again) will not pay as high an induced drag penalty as one with higher wing loadings. My recommendation is to try several power off 180°s in your plane.

    Other considerations
    Obstacles. The runway may be the lowest lying thing around. Do you need more altitude to clear trees, buildings, power lines, etc?

    Alignment. A 180 may not be the end of the turning if you wind up having to make two large corrections to angle toward a runway and then line up with it. You can reduce rollout alignment problems by allowing the plane to drift downwind after liftoff, or, in the absence of a crosswind, by turning a few degrees off runway heading. Or – for many urban airports – a parallel runway or a taxiway may be waiting to receive you on rollout.

    Climb angle. If your experience or conditions (short runway, meager vertical velocity) indicate that Vy would provide such a shallow climb angle that you will be so far from the airport that there’s no point in turning around, consider an initial climb closer to Vx – bearing in mind that a slower rate of climb is inevitable on a hot day and the engine is paying the price.

    Dithering. Don’t! Crank it over past 45° asap. If the engine comes back to life you’ve lost nothing and can level your wings while explaining to ATC, in your best Chuck Yeager voice, that you’ve just dealt with a minor hiccup.

    Restart? There’s not much you can do. In my E the fuel selector is almost inaccessible and your boost pump is already on, right? (But, doesn’t hurt to check it.)

    Planning. Three items to consider before you take the active: a) Decide if you want an initial climb at Vy, or closer to Vx, b) how do you plan to obtain some lateral distance from the extended runway centerline to improve alignment after completing the unforgiving turn, and c) are there obstacles that would factor into your decision altitude or direction of turnback.

    The take-away
    For what it’s worth, this research and in-flight experimentation led me to the following:
    1) Initial climb at Vy – or Vx if necessary.
    2) If a cross-wind, allow plane to drift off centerline (if no parallel runway) or turn 10° to gain lateral distance.
    3) I call 300′ as the altitude at which the ‘impossible turn’ becomes feasible (in my Mooney! Determine your own plane’s performance.)
    4) Engine failure, rack plane into 45 – 60° banked turn, heading into the crosswind if there is one. The nose drops automatically in the steeper bank – if you’ll allow it.
    5) Pull prop to feather. (Increases the Mooney’s glide ratio almost 20 percent; Bonanza reportedly similar.)
    6) Maintain Vy.
    7) Notify tower to break out the ticker tape or marshmallows.

    I would not challenge the 1000′ rule but for the location of the airport (PDK, in my case). Rural airports? Forgot all this and glide into the nearest pasture ahead of you. But if, as for many departures, there’s nothing but innocent civilians in your path, know the altitude at which the ‘impossible turn’ is possible. Then practice it aloft and think it through before each departure.

    • Hi Mike,
      In your details of the altitude loss in the 180 degree turn in the Mooney 20E, you show the 60 degree bank as having the lowest altitude loss in the 180 degree turn. You also indicate the KIAS was 90 knots for 3 bank angles you flew. In order for the experiment to be valid, you must keep all the independent parameters constant except the one your trying to obtain information on. In the gliding turn, the independent parameters are the aircraft wing loading, air density, bank angle, and angle-of-attack. Thus, to draw the correct conclusion on the optimum bank angle, one must fly the identical angle-of-attack at the different bank angles. If you keep the airspeed the same and vary the bank angle, the angle-of-attack will be different. One can accomplish this by flying the aircraft at an airspeed which is a fixed percentage above the accelerated stall speed corresponding to the to different bank angles. If you do this, you will find the optimum bank angle is between 45 and 46 degrees. For example in a C-172 it is 45.4 degrees. However, what is more interesting is the fact that the bank angle can be flow between 40 and 50 degrees with a minimum impact on the altitude loss in the gliding turn.
      If you are interested, I wrote a White Paper on the turnback maneuver and it has all the details about this issue.

  • With the new AC-83J, Appendix A 11.4, published in Sept 2018, there has been a renewed interest in the turnback maneuver. There has always been varied opinions on whether or not to attempt a turnback maneuver. Even among those who advocate the turnback maneuver, there are many conflicting view points on the question of ” How much altitude above the runway does one need before attempting a turnback maneuver?, or even what bank angle and airspeed to use.

    As an aerodynamicist for over 5 decades and a CFI for over 4 decades, the reason there are conflicting statements is due to the true lack of understanding of the aerodynamics of the turnback maneuver. It is important to understand that the knowledge of the aerodynamics is at least as important as acquiring the stick and rudder skills in executing the turnback maneuver. In fact, it is the aerodynamics that tell the Pilot how to fly the turnback maneuver. I will say that understanding the aerodynamics of the turnback maneuver is in no way “Rocket Science”. As Pilots, we should all understand what is in the Handbook of Aeronautical Knowledge, FAA-8080-23B, Chapter 5, on the subject of aerodynamics. In fact, that is all that is necessary to understand the aerodynamics of the turnback maneuver.
    In an attempt to convey to the Pilot community, the aerodynamics of the turnback maneuver, I have written a White Paper, “Single-Engine Failure after Take-off: The Anatomy of a Turnback Maneuver”
    The paper outlines how every Pilot can determine the “Required Minimum Runway Length and corresponding height of the aircraft above the departure end of the runway, which will allow for a potentially successful turnback maneuver in the particular aircraft they fly. I demonstrate it for a C-172. The paper shows how the Schiff method is overly conservative in that it dismisses potentially successful turnback maneuvers closer in to the departure end of the runway.
    If you have an interest in receiving a copy, send me an email to lgtech@roadrunner.com and I will provide a copy.

  • Good run through, a couple things implied…get to know your airplane and your capabilities in it, if you’re not satisfied fly more…go out with your CFI and learn what an accelerated stall feels like enough times that you recognize it through control feedback without any instrument reference…this will give you more time outside working the problem with a heightened sense of how much margin remains throughout your maneuver.

    The other is communicating your intent…this is not a negotiation, you are informing so others can move, or be moved, out of your way. The first response may not align with your plan, if not, be assertive, use “emergency” to get attention…at the end of the day, though I seriously doubt anyone will argue at this point, no one outside your cockpit can fly your airplane and I doubt any tapes played back with “emergency” are going be a problem, but if they are, better to be wrong and alive than correct and dead.

  • This is a maneuver I have practiced with my instructor in an my SR 22. We start out at 3000 ft., and simulate a take off climb off climb.
    At 4000 feet, pull the power back to idle, maintain nose up for 5 seconds (about the time it might take to recognize a problem). immediatly achieve glide speed, 30-40° bank into the wind, and try and line up with the runway. Glide speed is CRITICAL.
    So far, I have been able to do it in 700 feet, but believe I can get it down to about 600 with practice.
    Obviously hope to never use it, but it is a real skill and confidence builder because the airport I fly out of does not give me many options for an off airport landing.

  • Having experienced a total engine failure in a single engine aircraft, I appreciate the information presented. All good advice. However, I would add one additional item to consider. When you practice the turn back maneuver, you will do so with the engine at idle. That means the propeller is presenting little resistance to the airflow across the plane. If you develop a feel for how much altitude you will lose under those conditions, you may develop a false sense of security. With the engine totally dead and a windmilling propeller, the drag will be considerably greater, resulting in more altitude lost in a given time period. So, if you feel confident that you can make the turn back in, say, 700 feet, the real altitude lost could be significantly higher. I don’t advocate practicing with the engine shut off, but just add some more altitude to your decision on whether or not you can make it back to the runway. Some people say that you can slow the plane enough to stop the propeller from turning, which will reduce its drag, but in a minimum altitude situation, there is little time to do that. Plus, that is something to do in straight flight, not in a 45-60 degree bank. I was fortunate that my engine quit in cruise flight and I was well within range of an airport. And, I had done some engine out practice with my instructor shortly before the incident. But, I was still surprised at how fast the plane was descending and how much wind buffet there was on the way down due to the dead engine. Just FYI.

  • This is a great article. When I’m instructing we brief every takeoff – modifying altitude requirements based on terrain, airports, and wind.

    I instruct in the Sling 2 LSA, and we can generally make the turn at 500′ AGL. We practice it also. Ask the tower and if it’s not crowded they generally allow it.

  • Also, IMHO, practicing the turn at a higher than actual altitude isn’t a very effective training tool. You can make the turn in 300′ at altitude. Things are just different when you are low, and your brain is kicking in impulses that you shouldn’t listen to. You have to account for the delay in your reaction, and the fight to lower the nose towards the ground when you don’t want to hit said ground.

    • You are exactly right. There used to be a concept in the FAA Flight Training Handbook called ground shyness, which is a natural reluctance to bank the airplane at low altitude. This could lead to bad technique with disastrous consequences even on any base to final turn if the pilot tried to increase turning performance by skidding or pulling on the control wheel. I agree it is important to train an adequate amount at the low altitude appropriate for the make and model.

  • John, I am quite familiar with the scenario in the video you reference. I have flown into and out of that airport. Obviously the pilot did some things right, but he was fortunate to have been flying a Mooney that allowed him to have a favorable outcome. With my CFI shortly after this event we analyzed the video and noted a few things that likely would have made that flight easier. First, rather than turning left, which would have been favored by any turning of the prop, he turned right which forced his flight to be over a wooded area. Second, the departure runway was 18, and turning left would have brought him around to the longer crossing runway 30 without making a full 270 degree turn followed by an opposite over 90 degree turn to align with the 36. In addition a left turn would have been over open fields and got him over the airport sooner if he had run out of altitude. I credit the pilot for making the best of an unfortunate event, but in hindsight I feel a left turn would have been a much safer option. My CFI said not to attempt something like that in a Cherokee from that altitude. What it did prove that given the circumstances and the aircraft this pilot could return to a single runway from that low altitude. What I’m trying to say is that there are a few options on every takeoff to consider should this type of event happen. We pilots best consider those options before we apply takeoff power. They would include takeoff roll, length of runway, altitude, terrain, and aircraft when should things go downhill in a hurry. Also, with some airports it is better to arrive back on the airport somewhere than off airport where there are few if any favorable landing options. This has been a very good effort on your part and others to shed further light on this important subject.

  • What’s not being talked about much here is the crucial element of the climb angle, the rate of climb, from takeoff to that point where the engine fails, whether it be 500 ft or a 1000 ft, or 3000 ft. What matters in whether the turn back to a safe landing is possible (meaning with a realistic chance of a good outcome) is this question: was your climb angle steeper than your descent path will be? If so, you’re golden.

    Take this example: a C-182, cool day, pilot and one pax. You’ll climb out at 1200 fpm +. Maybe 1400 fpm. Climbing at 90 kts, that equates to 800-900 feet per mile (calm wind). More altitude gained per mile of climb, than you lose in a power-off glide, where you lose about 600 feet per NM.

    Now take a little C-150, warm day, and fully loaded with 2 people. Climbing at 60 kts, you’ll see 350-400 feet per minute, and the same feet per mile. But in a glide, it won’t do any better than the 182, and may do worse. So you’ll still burn through 600 feet per mile as you try to make it back to the airport, and so even if you turned around at 3000 ft AGL, you won’t make it back, but of course you should find some other place to land that little bird from that altitude, considering it took you 8-10 miles to get that high.

    As to the turn at low altitude, it’d sure help to be at a place like my home airport, with crossing runways, just 30 degrees apart. Taking off on runway 34, engine quits at 500 ft, especially with a headwind, turn left and land on 31 opposite direction of course, 13.

  • I’ve done this a bunch of times in my 182. A couple of things I’ve noticed: first, with the fan turning at idle, it glides like a set of car keys. 10:1? Nonsense. My estimate is about 4:1. We are so used to landing with at least some power – at least enough to negate the prop drag – that your first landing at idle power will take your breath away. Second, figure out what is your best angle of turn; it makes tremendous difference. Above 45° and I lost too much altitude to make it. Above 45° and it took too long to turn and I ended up too far away to make it. 45 was it. This will be different for every airplane. Third: actually CHECK these things. It will horrify you how quickly you get behind the airplane if you are not ready.

    • No way it’s anything close to 4:1. The glide ratio in your POH is reliable information. And there is no reason to think it gets worse with age of the aircraft or engine hours – you’re gliding after all. With the constant speed prop, you can try setting the prop for course pitch and that will help. Go up to 6000 ft AGL sometime (1NM), bring it back to idle, and without a wind in front or behind you, note on the GPS how many miles you glide towards your favorite field. It will be 8 NM minimum, likely more, if you are at best glide and not faster. If it’s not, please post a video for proof, so we can see why.

  • A windmilling prop is like having a drag parachute. The glide ratio of an aircraft with a working endine at idle is significantly better than it is when you have a ‘real’ engine failure. Its all very well saying it works in the sim, or a glider, or a PC12 (where you can feather the prop) – try it with the mags off and I believe things could be very different. Then add in the ‘startle factor’ – the physical and psychological effects of extreme stress are not conducive to flying ‘on the edge’ and often, I believe, the root cause of inappropriate control inputs leading to disaster. I remain unconvinced that contemplating a turn-back in piston-engined GA types is a good idea.. A good, thought provoking article 🙂

    • Gerry, are you suggesting in a “real” engine failure the prop stops turning? Because this is usually not the case. Otherwise, why would the 2 glide ratios be dramatically different?

  • I’m a high time pilot and flight instructor that has practiced this scenario many times, successfully. I will never teach my students this practice though. In the average GA aircraft you’d be lucky to be at 500 AGL at the departure end of the runway and an attempt to return would be disastrous. The skill set is just not there folks. An engine failure at low altitude requires mitigating a sound decision and with not much time. As for me it’s better to crash land under control than in a stall spin situation. This is my humble opinion and I will never teach it.

    • Mandy, I understand your point, but consider there may be times where just getting back to the airport property – such as grass, empty taxiway, etc., let alone a runway, is a much better option, if you have the altitude, than an urban jungle you could be presented with at many airports. Every takeoff is different, and it’s it good idea to talk about what you would do if the engine quit between 150 and 300 feet, the worst place for it to quit, especially taking off from a runway less than 3000 ft. And yes, at those altitudes, it’s a little left or right, no turning back. As CFI’s we need to go over all possible scenarios with students.

  • I would advocate practicing this on take off after having done it at altitude several times. You can see and hear what it really looks like. At Van Nuys early in the morning with no other traffic, the tower would allow me to practice teardrop returns to the airport and would clear me to land on both of the runways available. Climb to 1000 AGL, cut the power to idle, slow count to five and then begins the turn. Bank to 45 degrees drop the nose for 65 KTS and you are headed back to the airport in just a few seconds. Center the ball. If you end up short on final, admit defeat and add power! More often than not, I would need a slip to land to get it back down. The tower would critique my landings knowing what I was trying to do. I would never attempt this under real conditions if I had never practiced it during a real take off. You can watch it here: https://youtu.be/m94I5yQ-QPQ

  • The amazing pilots among us can do it, but many (most) of us can’t. There are so many examples of this not working which makes it border on irresponsible to get this into people’s heads. Google N4166Z for a recent example.

    True, there are airports where the only “good” option is trying to turn back. Be sure to brief that you are going to land in an alley or on a side street or in some trees. If you go out and practice this in your aircraft, and understand the limitations, and the exact airport environment, you might have a chance. If you do this, you are already in the 99th percentile of pilots.

    My default game plan is get the nose down and land wherever I can avoid hitting anyone or anything solid.

  • Very nicely written and this is easy to teach in motorgliders. It’s a requirement before I let anyone solo. I was once checking out an FAA inspector and pulled the engine at 250 feet on her first takeoff. She failed to lower the nose promptly so I lowered it for her. She said she learned a lot from that experience.

    A key point at uncontrolled airports is to make a 30 degree turn off runway heading after liftoff and clear of obstacles. When the engine quits you only have to make 210 degrees of turn to line up on final. I proposed this idea to Barry Schiff and he agreed, but this is not commonly taught.

    • Hi Russ,
      There is a property of the teardrop geometry that shows that the teardrop angle in a no-wind condition is just a function of the ratio of the distance from the runway that the aircraft turns back divided by the radius of the turn. So for a given TAS and bank angle, the radius of the turn is fixed. Thus, there is only one distance from the departure end of the runway for a turnback that the 30 degree turn will work in your comment. Consider a turn radius of 500 feet and the aircraft turning back at 3500 feet from the runway. One can easily see the aircraft will not be lined up with the runway centerline
      The 30 deg teardrop will work for the case of D/R=3.7, i.e. when the distance from the runway on the 30 deg angle is 3.73 time the turn radius..
      One also runs into this issue in holding patterns using a teardrop entry to the hold. The FAA recommends a 30 degree turn and then fly for one minute, followed by a 210 degree turn. In the one minute outbound/inbound time for the teardrop, the teardrop angle is exactly 35.3 deg in the no wind case. If you fly the entry using 30 degree to the holding radial you will always overshoot the holding radial. In fact with a 90 second inbound leg to the hold, the teardrop angle is 23 degrees.
      Any comment?

  • I think that the best procedure for surviving the turn back is to prevent engine failures. This can be 95% accomplished by making sure that there is fuel available to the engine. Other than that, paying good attention to the engine is also important. If it doesn’t start right, something is wrong, and you shouldn’t launch after just assuming it’s an anomaly. I saw a guy spin on a second takeoff attempt after the engine quit on the first takeoff attempt. Actual catastrophic engine failures do occur on takeoff, but it’s in the noise.

  • A few weeks ago a New Cirrus had engine problems upon take off from the small, lake surrounded airport near my home . For him it was the “impossible turn” as he and his wife lost their lives in a stall induced accident short of the runway they were trying to return to. Two days later, with this tragedy still on my mind I had an engine outage from the same airport. I had the same choice, a 180 degree turn or straight ahead to a water landing. As they say, hindsight is always 20/20 and without a recent example perhaps I would have fell victim to that “irresistible urge” to return to the safety of the tarmac. I swam away that day and live to fly again. For me the choice was clear.

  • John:

    The most important point–the airplane belongs to the insurance company after an engine failure on takeoff climb. The focus should be on saving our butts. If you have practiced the turn, as you have and as glider pilots are required to do, it is a good safety option. If you have not practiced the turn in the type aircraft you are flying, then the turn should be considered impossible. You are asking for a low altitude stall or excessive sink. If you have practiced, then you also need to consider the wind and other options.

    But it is still most important to do the safest thing for the crew and passengers–not for the aircraft.

  • Being a glider tow pilot we are trained to “drift” down wind on takeoff. In the event of a rope break, it gives the glider pilot the advantage of making the turn to the runway into the wind and allowing him to have to make only a 180 degree turn to the runway rather than a 270 degree turn to try to get aligned with the runway. As one of the earlier commenters suggested, it makes sense for powered pilots to do the same thing when it can be safely done. Glider pilots are taught early on to plan on an emergency happening on every take off and decide on the appropriate action ahead of time if it occurs. As others have suggested, take some glider lessons. It will make you a better pilot, plus it is a lot of fun!

  • Using an altitude turn back criterion seems misguided. What you really care about is the slope of the climb vs the slope of the glide. An aircraft without sufficient climb slope compared to its glide ratio will never make it back from any altitude without a stiff headwind. So people should test their airplanes and know if they are in a plane that can usually do it, or not, given the headwind, load and runway length.

    Assuming that living to tell the tale is the primary goal, making the runway is not very important. Making the airfield is slightly more important, vs landing into an obstruction. But the most important thing for longevity of the pilot is to not stall while turning back. Anything that can be done in those few moments after engine failure to minimize the likelihood of a stall and get the airplane pointed toward a clear landing field is what people should be doing.

    • Hi Karl,
      The statement you make about climb angle versus the glide path angle is not completely accurate. This type of situation will usually occur when departing a high density altitude airport in an aircraft with a high power-loadings at gross weight, i.e. C-172.
      I wrote a White Paper ,”Single-Engine Failure after Takeoff: The Anatomy of a Turnback Maneuver”. It describes the issue you raise, and shows that in the case of the climb angle being lower than the glide path angle, a potentially successful turnback maneuver, even without a wind is possible, however, only over a limited range of distances from the departure end of the runway.
      If you are interested in getting a copy, send me an email.
      Les Glatt

  • Sorry trying to turn back below 1000 ‘ AGL is a fools errand. Most that try it will die as they haven’t practiced it recently (or at all) and no timing is given for the “I cant believe this just happened to me” pause that inevitably occurs as well as the airspeed decay.

    I lost a Cardinal engine at 500’ AGL with a student who instinctively wanted to turn back. We ended up going through wires 60 degrees off the runway heading in control and walked away. The airspeed decay and the shock take much more time than you can imagine.

Leave a Reply

Your email address will not be published. Required fields are marked *