Testing the Impossible Turn

dmccormack

Touchdown! Greaser!
Joined
May 11, 2007
Messages
10,945
Location
Lancaster County, Pennsylvania
Display Name

Display name:
Dan Mc
So I flew a bit this morning -- cool, clear, calm -- perfect tube-and-fabric weather!

Had the airplane started up and rolling down the taxiway by 0715 -- air was smooth and cool enough for nice lift.

I flew around for about 30 minutes, practicing steep turns, turns about a point, slow flight, and low-level flying over the remote hills near the Mon river.

I climbed up to 1000' AGL and did some more airwork on my way back west. I decided to test out altitude loss with power-idle, 180 degree change of course.

I entered each at 60 MPH in a climb (standard Vy climb out speed), reduced power to idle, waited a moment (simulating recognition delay) and then commenced turning.

At 60 MPH in a steep descending turn (about 60 degree bank) there was little G load. I let the speed build after the 90 degree point was reached, then levelled off at 180 opposite heaindg.

Tried it a couple of times -- 300' altitude loss.

Of course we all know it takes more than 180 to head back to the runway, but as far as I am concerned, pointed towards the airport = landing area. There is lots of level grass on the field not found around it (Wal Mart and water tower on 300' hill immediately south, prison to the north, town to west, hills and steep valley with stream to east).

I takeoff from the pavement and land on the grass.

So I thought through the scenarios on the drive to work. When takinig off form 27, If I lose power anywhere before 300' AGL I'll land straight ahead and most likley stay on the runway.

300-500' AGL will mean landing on I-79

>500' AGL and I'll turn back.

When taking off from 9, the options are limited, but a landing somewhat straight ahead will be in a sloped, mowed hillside.

The slow speeds mean the airplane will turn inside itself with a sufficiently steep turn.

It's a worthwhile exercise every pilot should do with each airplane flown.
 
Dan, keep the wind in mind when you make the decision whether to turn around or not. If you are taking off into a strong headwind you at be at your turn around altitude but not very far from where you took off. Now you are looking at a downwind landing to a short field.

Just my $.02 from doing a bunch of "stimulated" rope breaks
 
Dan, keep the wind in mind when you make the decision whether to turn around or not. If you are taking off into a strong headwind you at be at your turn around altitude but not very far from where you took off. Now you are looking at a downwind landing to a short field.

Just my $.02 from doing a bunch of "stimulated" rope breaks

Good point -- though I'm realistically limited to about 12 knots.

Much more than that and it's simply too much like work.

I've flown in 15G20, and it was a rodeo. :eek:

this is a fly-for-fun airplane. :yesnod:
 
According to the paper "The Possible Impossible Turn", the ideal bank angle is 45 degrees. If you bank more than this, you take on a lot of induced drag and start to lose more altitude in the turn.

Note that the altitude loss in the turn is only one part of the equation, you also have to consider whether you have enough altitude to be able to glide back to the airport. In a plane with too little power, your climb will be too shallow, and you'll be too far from the airport to be able to make it back. A headwind helps keep you close to the airport on departure, and also pushes you back to the airport upon return. Turning into a cross-wind can also help keep you from straying too far off the extended centerline, which increases the distance from the airport and the number of degrees you have to turn to get pointed back. And, of course, the longer the runway, the shorter the trip back is.
-harry
 
Dan,

Great, I agree this is a great thing to learn about.

However I think you might have missed part of the situation. Part of the problem is it is more than a 180 degree turn.

You had it right Climb at your normal climb speed to about 1000 feet, but line up on road when you do so as if you had just departed from that road. Wait at least 2 seconds before initiating the turn, it would probably take you at least this long to decide you needed to turn. Also turn downwind (worst case scenario). Your altitude loss will be your altitude about 1-2 seconds after you line up on the road at you normal approach speed. You would need that 1-2 seconds to stabilize the turn and decent enough to round out, flare and land.

Of course don't let your self go below 500 feet AGL.

Then try the same maneuver at your best angle speed, (I know, there isn't much difference in the Chief)

I too have done it many times in a glider, my rule of thumb in a power plane is that I usually turn crosswind at 500 feet. If I am already turning when the engine quits I can consider returning the runway. Otherwise it the softest cheapest place I can find to land if front of me.

Brian
 
I've been watching this thread and wondering what ever became of common sense? No instructor should be teaching how to accomplish the impossible turn...the time would be better spent in instilling the automatic reaction to get the nose down...way down, to be sure that the wing is flying...and turn no more than 60 degrees from runway heading. Choose something inexpensive to hit.

My guess is that if we held a "turn back" contest involving those posters who advocate trying the impossible turn, we would lose over half of them. Land straight ahead (or nearly so), under control and you will live to talk to the insurance adjuster.

In the case of an actual engine failure the mind shifts into tunnel-vision mode, and all of these brave statements about bank angle and reaction time go out the window.

Bob Gardner, Pusillanimous Pilot
 
I've been watching this thread and wondering what ever became of common sense? No instructor should be teaching how to accomplish the impossible turn...the time would be better spent in instilling the automatic reaction to get the nose down...way down, to be sure that the wing is flying...and turn no more than 60 degrees from runway heading. Choose something inexpensive to hit.

My guess is that if we held a "turn back" contest involving those posters who advocate trying the impossible turn, we would lose over half of them. Land straight ahead (or nearly so), under control and you will live to talk to the insurance adjuster.

In the case of an actual engine failure the mind shifts into tunnel-vision mode, and all of these brave statements about bank angle and reaction time go out the window.

Bob Gardner, Pusillanimous Pilot

Only if you train at life that way. I just had an engine quit on me a month ago. There was no tunnel vision.
 
I've been watching this thread and wondering what ever became of common sense? No instructor should be teaching how to accomplish the impossible turn...the time would be better spent in instilling the automatic reaction to get the nose down...way down, to be sure that the wing is flying...and turn no more than 60 degrees from runway heading. Choose something inexpensive to hit.

My guess is that if we held a "turn back" contest involving those posters who advocate trying the impossible turn, we would lose over half of them. Land straight ahead (or nearly so), under control and you will live to talk to the insurance adjuster.

In the case of an actual engine failure the mind shifts into tunnel-vision mode, and all of these brave statements about bank angle and reaction time go out the window.

Bob Gardner, Pusillanimous Pilot

So if I'm 3000' AGL I should land straight ahead?

:rolleyes:

There is an altitude which provides sufficient energy to manuever for a landing. The FAA gives us 1000' AGL for 180 power off for the Commercial PTS -- and that's heading away from the landing point when the exercises begins (In other words -- you have to do a gliding 180)

The question is -- what altitude for this airplane changes the emergency response from "Land straight ahead" to "Get back to the airport"?

Take a look at the departure end of 27, you'll see the options are limited.
 
Last edited:
I've been watching this thread and wondering what ever became of common sense? No instructor should be teaching how to accomplish the impossible turn...the time would be better spent in instilling the automatic reaction to get the nose down...way down, to be sure that the wing is flying...(snip)

Looks to me like Mr. McCormack did exactly that. Put the nose down.

At 60 MPH in a steep descending turn (about 60 degree bank) there was little G load. I let the speed build after the 90 degree point was reached, then levelled off at 180 opposite heaindg.

And, IMO, the reason that the so called "impossible turn" is such a bad one is that over and over, we teach pilots to pull on the yoke as they roll into a turn in order to maintain altitude (which is part of the check ride, right?). But in this situation, that automatic reaction which has been drilled into our heads results in dead pilots.

The thing that needs to be taught (again in my own stupid opinion) is not just to turn, not just to find out how much altitude you need, but to put the nose DOWN as you roll into the turn. Practice that over and over. Make the nose down as automatic as pulling during a level turn.
 
I think the 45 degree "limit" factors in expertise as well as drag. You're only in the steep (60 degree) bank for a very small period of time, you reduce the total time required to turn, and you keep the airplane as close to the centerline as possible (slow speed + steep turn = small radius)

As far as low power -- 50 MPH is Vx (which is what I use until I’m 500’ AGL. The reason I’m using Vx for the first few hundred feet is that the steep angle keeps me closer to the airport).

Assumptions: All this assumes no wind. A headwind will make it all easier. I don’t take off in tailwinds.

@300 FPM climb = about 2 minutes to 500' AGL
At 50 MPH I've travelled 8800 feet from the takeoff point (approximately 1000’ from the far end of the runway).
The paved runway is 3500’ long.
From that point, I flew over 2500’ of runway.
So I’m 6300’ from the near end of the runway at 500’ AGL.
I need to turn 180 degrees and remain aloft long enough to touchdown on the airport grounds (remember, the runway isn’t my goal – a flat, unobstructed place will do).
Glide ratio of 8:1 means I’ll travel 4500’ horizontally from 500’ AGL.

1800’ short of the runway. :mad:

I’ve been turning downwind when I reach 500’ AGL, and am established on downwind about 650’ or so AGL (we want to remain good neighbors and an extended departure leg would put me over town).

So, while I know I can turn 180 degrees with a 300' altitude loss, the distance from the airport means a "straight ahead" is the best choice until I complete my turn to downwind. Once etsbalished there, i know I can make it to the edge of the grass.

A good exercise: mental and otherwise! :yesnod:
 
Looks to me like Mr. McCormack did exactly that. Put the nose down.

I have no choice if I want this Chief to continue flying!

And, IMO, the reason that the so called "impossible turn" is such a bad one is that over and over, we teach pilots to pull on the yoke as they roll into a turn in order to maintain altitude (which is part of the check ride, right?). But in this situation, that automatic reaction which has been drilled into our heads results in dead pilots.

The thing that needs to be taught (again in my own stupid opinion) is not just to turn, not just to find out how much altitude you need, but to put the nose DOWN as you roll into the turn. Practice that over and over. Make the nose down as automatic as pulling during a level turn.

Exactly right.
 
So if I'm 3000' AGL I should land straight ahead?


You can get to 3000 agl in a reasonable amount of time?!?! Last time I flew the Champ I had a passenger and just over 1/4 tank of gas and thought I'd never make it to 3500 msl
 
From Rich Stowell:
1. Remember the number "8" -- This is the increase in the risk factor for death or serious injury when attempting a turnback following an engine failure during takeoff vs. landing straight ahead (or nearly so).
Source: Transport Canada, "An Evaluation of Stall/Spin Accidents in Canada," TP 13748E, 1999.

2. Remember the number "62" -- This is the overall success rate (percent) for turnbacks in a turnback-specific study using a simulator and techniques ranging from "pilot's choice" to "optimum turnaround bank/speed/g-load."
Source: Jett, Brent W., "The Feasibility of Turnback from a Low Altitude Engine Failure During the Take-off Climb-out Phase," AIAA-82-0406, January 1982.

3. Remember the number "68" -- This is the success rate (percent) for turnbacks in the turnback-specific study AFTER subjects were given specific instructions on how to execute the aerodynamic optimum bank/speed/g-load.
Source: Jett, Brent W., "The Feasibility of Turnback from a Low Altitude Engine Failure During the Take-off Climb-out Phase," AIAA-82-0406, January 1982.

4. Remember the number "Zero" -- This is the margin for error when attempting a turnback at the aerodynamic optimum combination of bank, speed, and g-load specified in the paper referenced earlier in this thread. If the turnback happens to work, the pilot will be the "Ace of the Base" for a day; if the turnback fails, there will be a burning hole in the ground that consumes both the occupants and the airplane. It's one of the two extremes, with no middle ground.

5. Remember the numbers "5, 8, 2" -- In a one year period alone (October 2005-2006), there were a total of 5 accidents in the NTSB database wherein the flight profiles included the intentional practice of simulated engine failures with attempted turnbacks to the runway close to the ground. In each of the five cases, the cockpit contained two pilots: one of the two on board was either a CFI or an FAA Inspector. The results: 5 destroyed airplanes; 8 dead pilots; 2 injured pilots. Interestingly, during that same one-year period, there was only one accident in the database where an instructor and a student were killed during intentional spin training. Yet some in this Forum persist in perpetuating the myth that "spin training" is dangerous (and in general, it is dangerous if conducted by the average flight instructor; not so if conducted by those who specialize). What then does that make the practice of turnbacks close to the ground?
Source: NTSB database search

6. Remember the number "43" -- In the five accidents referenced in item 5 above, some of the narratives described repeated attempts at turnbacks prior to the accidents. In those cases, the success rate was calculated: 43 percent of the turnback attempts were successful, with the last attempt of course being a smoking hole in the ground.
Source: NTSB database search

7. Remember the number "100" -- This is the success rate (percent) for all of the "straight aheads" attempted in the simulator study of turnbacks. Remember, too, that none of the turnback techniques looked at in that turnback-specific study could match the 100 percent success rate logged when the test subjects chose to proceed straight ahead.
Source: Jett, Brent W., "The Feasibility of Turnback from a Low Altitude Engine Failure During the Take-off Climb-out Phase," AIAA-82-0406, January 1982
 
I think there's no harm in this, but only to learn what it takes to do it in your airplane should it be the best option. In the rare event that it is the best option, it might help to have simulated such a turn.

Still, assuming you have enough altitude,the big question is: which will hurt more, potentially- going straight ahead to that golf course, pasture, etc., or landing on the runway, landing long, maybe... with a tailwind, and maybe somebody in position for takeoff, or rolling? That's assuming that you actually pull off the turn itself!

Obviously, the "impossible turn" should not even be contemplated unless it looks far better than impossible. What makes it "impossible" in most cases where a crash results is that these pilots try to turn when they simply don't have sufficient altitude. But we all know that most aircraft can do it below TPA, so there is such a thing as a "possible" turnaround during initial climb, in some cases.

Of course you should practice it with lots of extra altitude... unless you are a glider pilot, in which case you have to practice it with... not so much extra altitude. :D
Even with a glider, there is a minimum altitude, and a very strict sequence that must be followed. For us in the 2-33, it's: 200 AGL minimum, nose below horizon immediately, 45 bank immediately, well-coordinated, spoilers or slip if necessary once you have your spot "made".

Step 1, the nose down thing, is very very important. If your turn is too wide, or you come about too far downwind, that is bad, but not as bad as racking it over when you are near stall speed low to the ground. This mistake is another deal-breaker (and neck-breaker) in such scenarios, aside from trying to turn around without sufficient altitude. Not surprisingly, "nose down!!" is also Step 1 if you decide to land straight ahead. Wishful thinking for more altitude with up elevator can kill you. this point is also worth remembering if you have determined your airplane needs 300 AGL, and the fan quits at 250. Time to forget the turn and find some decent place in front of you.

Another tip for those who would try this: in practice or in earnest, if there is a crosswind to any extent, initiate your turn into the wind. Saves you a little extra turning to get lined up... which in turn saves you some altitude. Could make all the difference, if the wind is strong enough. One might know from simulating this at altitude that they can do the 180 in xxx feet, but what about 10-20 degrees more? and what happens if you overshoot, or drift downwind, and have to turn yet again, the other way? It's obviously not a good primary option, even from your "known" minimum altitude.
 
And, IMO, the reason that the so called "impossible turn" is such a bad one is that over and over, we teach pilots to pull on the yoke as they roll into a turn in order to maintain altitude (which is part of the check ride, right?). But in this situation, that automatic reaction which has been drilled into our heads results in dead pilots.
I'm really not sure this is drilled into us as an "automatic reaction", any more than pulling up when you want to go up (or go down less quickly) is an automatic reaction. If I'm getting a little slow on base I push the nose down a little in the turn to final. If everything is just right I won't push but I definitely don't pull up unless I'm fast, and depending on altitude I might push the throttle in a little or pull it out. It's all part of energy management and as in flying in general, tailoring your actions to the particular situation.

As far as the Impossible Turn is concerned, my first Cardinal checkout instructor 6 years ago wanted to teach me about it but I was skeptical and never took him up on it. Today with my new bird, and considering where I'm based (there are really very few options if you lose an engine on climbout from KVLL), it's definitely something I want to explore to come up with a solid idea of under what altitude and conditions it would be the maneuver of choice.
 
Last edited:
I think the 45 degree "limit" factors in expertise as well as drag.

Here's the article mentioned that advocates for a 45 degree bank:

http://jeremy.zawodny.com/flying/turnback.pdf

It also quotes considerably better results from test using this technique than would be suggested by the post quoting Rich Stowell.

But, debating bank angles and altitudes is moot if you don't have some game in engine out landings to a spot. What Dan did seems to me to be a pretty decent way to learn your airplane and get a feel for what he might do if confronted with this problem. You'll never convince me to "just land straight ahead" because some rule of thumb said so. You're PIC. Make command decisions and make them work because you've developed some skill at doing so.
 
From Rich Stowell:
1. Remember the number "8" -- This is the increase in the risk factor for death or serious injury when attempting a turnback following an engine failure during takeoff vs. landing straight ahead (or nearly so).
Source: Transport Canada, "An Evaluation of Stall/Spin Accidents in Canada," TP 13748E, 1999.
I've never been able to track down this technical report, so I can't give much credence to the quote. I have to be skeptical that they had enough data to work with to come up with such statistics.

But if pilots are performing a maneuver that they've never been instructed in, and attempting it in cases where success is not possible, and performing it via incorrect maneuvers, then that could be an argument for training as much as it is an argument for not doing that maneuver.
2. Remember the number "62" -- This is the overall success rate (percent) for turnbacks in a turnback-specific study using a simulator and techniques ranging from "pilot's choice" to "optimum turnaround bank/speed/g-load."
Source: Jett, Brent W., "The Feasibility of Turnback from a Low Altitude Engine Failure During the Take-off Climb-out Phase," AIAA-82-0406, January 1982.
The number "62" appears nowhere in that report.
3. Remember the number "68" -- This is the success rate (percent) for turnbacks in the turnback-specific study AFTER subjects were given specific instructions on how to execute the aerodynamic optimum bank/speed/g-load.
Source: Jett, Brent W., "The Feasibility of Turnback from a Low Altitude Engine Failure During the Take-off Climb-out Phase," AIAA-82-0406, January 1982.
Neither does the number "68".
4. Remember the number "Zero" -- This is the margin for error when attempting a turnback at the aerodynamic optimum combination of bank, speed, and g-load specified in the paper referenced earlier in this thread...
That's not true.
5. Remember the numbers "5, 8, 2" -- In a one year period alone (October 2005-2006), there were a total of 5 accidents in the NTSB database wherein the flight profiles included the intentional practice of simulated engine failures with attempted turnbacks to the runway close to the ground...
Which is a good argument for not practicing this close to the ground.
7. Remember the number "100" -- This is the success rate (percent) for all of the "straight aheads" attempted in the simulator study...
Sure, because the simulator didn't simulate any trees or buildings to crash into. If the world ahead of you is a computer-generated runway of infinite width and length, then the runway behind you has nothing to offer that improves upon what's ahead of you, so by all means do land straight ahead.
-harry
 
So if I'm 3000' AGL I should land straight ahead?

If you're at 3000 AGL you have choices available to you, but returning to the runway may not be one of them if you departed straight out. In the absence of sufficient headwind, many general aviation aircraft glide at a steeper angle than they climb, in which case additional altitude hurts instead of helping if your goal is to return to the runway.
 
If you're at 3000 AGL you have choices available to you, but returning to the runway may not be one of them if you departed straight out. In the absence of sufficient headwind, many general aviation aircraft glide at a steeper angle than they climb, in which case additional altitude hurts instead of helping if your goal is to return to the runway.

Give me one hour of fam flight in any SEL GA airplane.

Then place me 3000' AGL over near, next to a runway and I will land on the runway.

As should any commercial-rated pilot.
 
But if pilots are performing a maneuver that they've never been instructed in, and attempting it in cases where success is not possible, and performing it via incorrect maneuvers, then that could be an argument for training as much as it is an argument for not doing that maneuver.

-harry

Exactly right. :yesnod:
 
Best glide in my airplane is 60 MPH.

Why would I fly at 45 MPH?
Best glide in a 45 degree bank is actually higher than best glide wings level, but in this case you're balancing conflicting goals. Best glide gives you the most distance per altitude loss, but the slower you go the tighter your turn radius is, and the less distance you need to travel. And note that the "optimum" is stall speed for a 45 degree bank, not wings-level stall speed. And, of course, a sensible person would allow margin above that.
-harry
 
Best glide in a 45 degree bank is actually higher than best glide wings level, but in this case you're balancing conflicting goals. Best glide gives you the most distance per altitude loss, but the slower you go the tighter your turn radius is, and the less distance you need to travel. And note that the "optimum" is stall speed for a 45 degree bank, not wings-level stall speed. And, of course, a sensible person would allow margin above that.
-harry

Right -- but -- the turn radius difference between 45 and 60 MPH (to say nothing about the increased comfort, familiarity, proficiency during a high stress scenario -- hardly seems worth the risk...?
 
I think the 45 degree "limit" factors in expertise as well as drag. You're only in the steep (60 degree) bank for a very small period of time, you reduce the total time required to turn, and you keep the airplane as close to the centerline as possible (slow speed + steep turn = small radius)

No, expertise did not enter into the calculation. The 45 degrees was not a limit, it was the bank angle that gave the least altitude loss during the turn. The paper I cited in post #20 describes the assumptions made in deriving that number.
 
No, expertise did not enter into the calculation. The 45 degrees was not a limit, it was the bank angle that gave the least altitude loss during the turn. The paper I cited in post #20 describes the assumptions made in deriving that number.

Did they posit 45 degrees because 60 imposed a load factor (and requisite speed increase)?

Doesn't have to in a descending turn....
 
TV Time Out

----------------------

I'm not advocating turning back in every circumstance.

I am suggesting we test ourselves and our airplanes and figure out how high we should be before we attempt a turnback.

Part of my pre-takeoff brief includes "Don't turn back until [feet MSL]" and I point to that altitude on the Altimeter.

A Chief is a lightweight, low powered airplane that takes 1/2 a day to reach 1000' AGL. So it was important to me that I test and come up with a more reasonable target altitude.

Your Mileage May Vary.

----------------------------------

End of TV Time Out
 
Last edited:
If you are REAL good, you go nose up, put in 1/2 a turn of a spin , nose down, then flare...Wheeeeeeeeeeeeeeeeeeeeeee!

I've only done this successfully....never.
 
From Rich Stowell:
1. Remember the number "8" -- [etc.]

When Rich posted that on the AOPA board, I asked for comments from Steve Philipson, a local instructor who has studied the turnback issue extensively, and I posted them there.

http://forums.aopa.org/showpost.php?p=752607&postcount=72

Here they are:

>To say that the gist of this thread is "extremely dangerous" would be an understatement.

I don't know what was going on in this discussion, but more info is generally better than less.

>1. Remember the number "8" -- This is the increase in the risk factor for death or serious injury when attempting a turnback following an engine failure during takeoff vs. landing straight ahead (or nearly so).

>Source: Transport Canada, "An Evaluation of Stall/Spin Accidents in Canada," TP 13748E, 1999.

TC doesn't mention the Rogers/Jett work. Their conclusion is based on the performance of pilots who did not know the optimal technique nor when it could be performed successfully. Thus for the general case, their conclusion is correct.

>2. Remember the number "62" -- This is the overall success rate (percent) for turnbacks in a turnback-specific study using a simulator and techniques ranging from "pilot's choice" to "optimum turnaround bank/speed/g-load."

Sounds right. The success rate is going to be pretty low when anything other than the optimum technique is used AND one doesn't have metrics as to whether the maneuver can be done under a particular set of circumstances. All this means is that the maneuver has a high failure rate when you don't know what you're doing.

>3. Remember the number "68" -- This is the success rate (percent) for turnbacks in the turnback-specific study AFTER subjects were given specific instructions on how to execute the aerodynamic optimum bank/speed/g-load.

Seems reasonable, but I don't recall whether there was any actual flight training for that maneuver. In any case, it really doesn't matter -- my position is that you shouldn't attempt the maneuver if you have a viable alternative, i.e. a good landing site ahead. If the landing alternatives are poor (solid trees, buildings, cold water) then a 68% chance of success is superior to a straight ahead landing attempt.

>4. Remember the number "Zero" -- This is the margin for error when attempting a turnback at the aerodynamic optimum combination of bank, speed, and g-load specified in the paper referenced earlier in this thread.

That's factually incorrect. There are margins in stall speed, altitude and runway length for the maneuver, and if you use pre-takeoff performance metrics you know whether any or how much margin exists.

>If the turnback happens to work, the pilot will be the "Ace of the Base" for a day; if the turnback fails, there will be a burning hole in the ground that consumes both the occupants and the airplane. It's one of the two extremes, with no middle ground.

Not true. There are at least two other types of outcomes. One can attempt to land straight ahead under unfavorable conditions and also be injured or killed. Also, there are cases (including one at RHV a few years back) where the aircraft reached the field after a turnback and tore off a wing, but the pilots were not injured. Even if they were, that might be a considered a preferred outcome to hitting a house and injuring/killing people on the ground.

>5. Remember the numbers "5, 8, 2" -- In a one year period alone (October 2005-2006), there were a total of 5 accidents in the NTSB database wherein the flight profiles included the intentional practice of simulated engine failures with attempted turnbacks to the runway close to the ground. In each of the five cases, the cockpit contained two pilots: one of the two on board was either a CFI or an FAA Inspector.

No doubt low altitude practice of this maneuver is stupid. The maneuver should be practiced at an altitude appropriate for aerobatic training. I've talked to at least two pilots who had accidents while practicing from takeoff to touchdown. Neither had a clue as to the optimal technique and both made serious errors that caused the maneuver to be unsuccessful. Both thought they were knowledgeable in the maneuver, but neither knew of the Rogers/Jett work or had seem my presentation on this subject. Another point though is that there are no statistics on the numbers/percentages of successful attempts of this maneuver. I've used it successfully twice during partial engine failures, and have read numerous anecdotes of similar successes. Those events don't make the news or statistics.

>The results: 5 destroyed airplanes; 8 dead pilots; 2 injured pilots. Interestingly, during that same one-year period, there was only one accident in the database where an instructor and a student were killed during intentional spin training. Yet some in this Forum persist in perpetuating the myth that "spin training" is dangerous (and in general, it is dangerous if conducted by the average flight instructor; not so if conducted by those who specialize). What then does that make the practice of turnbacks close to the ground?

The accident record on spin training reflects the relatively low rate of that maneuver being performed. When spin training was mandatory there was a much higher incidence of fatal spin training accidents. That was the main reason that the FAA switched from required spin training to stall/spin awareness training. As for safety, specialization is not what counts. There are plenty of specialized instruction programs that teach sub-optimal (or even dangerous) techniques. And while I agree that training/practice of turnback maneuvers close to the ground is reckless, training at altitude is not. Most pilots to whom I've given this training conclude that it's too difficult a maneuver for them to perform reliably. There's considerable value for them in making that realization. For others, it becomes a technique they can use.

>6. Remember the number "43" -- In the five accidents referenced in item 5 above, some of the narratives described repeated attempts at turnbacks prior to the accidents. In those cases, the success rate was calculated: 43 percent of the turnback attempts were successful, with the last attempt of course being a smoking hole in the ground.

I'm actually surprised that the rate of success was that high. Very few pilots are familiar with the Rogers/Jett (and dare I say Philipson) research. This is meaningless though. Yes, this is incredibly dangerous if 1) you don't know what you're doing, and 2) you practice it from/to the ground.

>7. Remember the number "100" -- This is the success rate (percent) for all of the "straight aheads" attempted in the simulator study of turnbacks.

Yeah, that's correct, but consider that the criteria for success was (roughly) that the aircraft remained under control and had a relatively low descent rate upon landing. I know of several cases where a straight ahead landing attempt after engine failure was unsuccessful, some of them very high profile (e.g. the fatal crash of the LA traffic reporter "Yellow Thunder" in the 1980s from Fullerton). This past year there was a fatal accident at Cameron Park during a straight ahead abort. In real life cases, straight ahead landings have considerable risk too.

>Cirrus Design Corporation’s Pilot Training Bulletin PTB 02-06-01 doesn’t mince any words: "The ability to successfully complete and execute a return to the airfield maneuver after takeoff is a dangerous maneuver to practice at low altitudes and is therefore not authorized during training."

Sounds reasonable to me. Key words: "to practice."

>The Beechcraft Pilot Proficiency Program used to teach turnbacks after takeoff during their clinics based on the paper previously referenced in this thread -- that is, until an instructor and trainee were killed in the process. They no longer teach that close to the ground.

They never should have taught it close to the ground. Unfortunately we can't seem to stop people from doing unnecessarily dangerous things. In CAP we had a multiple fatality accident during training of canyon turns at low altitude, despite the fact that our training handbook has a notice in bold, capital letters warning that performing that maneuver at low altitude has a high risk of crash/injury/death (I wrote that warning). One crew did it anyway. They confirmed that the warning was correct.

>Bear in mind, too, that an actual engine failure in a fully-loaded airplane with passengers may very well include screaming passengers, smoke in the cockpit, oil covering the windscreen, and a lot more adrenalin than you might think -- how well will you perform under these conditions where your margin for error is zero? With turnbacks, you are playing Russian Roulette. You have been warned in no uncertain terms.

Sure, actual emergencies are stressful and your performance degrades. People do crash and burn on straight ahead landings too.

>Besides, if it were possible to train pilots to turn back with an equal or greater probability of success and survivability as the time-proven straight ahead, I or someone like me who specializes in emergency training would have developed and marketed such a training methodology. And that person would be able to name his/her price, because insurance companies would demand it as recurrent training. In my case, the fee for one day would be a percentage of the combined value of the airplane plus the pilot's life insurance policy. And I would be making more money in aviation than John and Martha King! But then again, I'd just be dreaming…

Well, maybe I'm a lousy businessman. My experience is that people do NOT flock to emergency maneuvering courses in general as they tend to be very expensive. 'Course, I don't claim that the probability of success is greater than a straight ahead landing for all cases, just that in certain circumstances the risk of the maneuver presents a better risk than a straight ahead landing. That training doesn't seem to make pilot's jump to spend even 1% of a $1.5 million loss on training. Perhaps I just need a marketing department.
 
Last edited:
Give me one hour of fam flight in any SEL GA airplane.

Then place me 3000' AGL over near, next to a runway and I will land on the runway.

As should any commercial-rated pilot.

If you're within gliding distance of the runway, I don't doubt it.
 
Did they posit 45 degrees because 60 imposed a load factor (and requisite speed increase)?

Doesn't have to in a descending turn....

They didn't posit 45 degrees. They wrote an equation for the altitude loss and then used it to find out what bank angle produced the smallest loss.
 
Last edited:
BTW, I'm not here to tell anyone they should or should not attempt to turn back, but I do think that people should have correct information when they are making their decisions.
 
Right -- but -- the turn radius difference between 45 and 60 MPH (to say nothing about the increased comfort, familiarity, proficiency during a high stress scenario -- hardly seems worth the risk...?
Which is why I'd say it's perfectly reasonable to add a margin of airspeed.

We have to differentiate between "this is what the numbers say provides optimal performance" and "this is what provides the best likelihood of a successful outcome under real-world conditions".
-harry
 
Another factor to consider.

Downwind landings are seldom practiced and should result in a very flat approach that looks lower and faster than normal. As a result it is very easy to subconsciously climb and slow down to try and make the approach look more normal.

After experiencing this in an off airport landing in a glider I now equate it to the "leans" when instrument flying. As soon as you look away from the airspeed indicator and stop consciously pushing forward on the stick/yoke you will likely start slowing down.

Brian
 
Back
Top