The Impossible Turn - with in-cockpit video footage

TangoWhiskey

TangoWhiskey

Touchdown! Greaser!
Joined
Feb 23, 2005
Messages
14,210
Location
Midlothian, TX
Display Name
Display name:
3Green
http://flash.aopa.org/asf/pilotstories/impossibleturn/?WT.mc_id=091204epilot&WT.mc_sect=sap

When his Mooney’s engine failed soon after takeoff, Dave Keller needed to land immediately. But rather than seek a landing spot straight ahead or off to the side, he chose to execute the “impossible turn” back to the airport, a maneuver no pilot should ever attempt. With aerial footage captured by Keller’s own panel-mounted video camera, the AOPA Air Safety Foundation’s latest Real Pilot Story: The Impossible Turn puts you in the right seat as the event develops.

P.S.--AOPA's web guys left the page title as "Pinch Hitting a King Air" from that guy in Florida who landed the King Air with his family aboard after the pilot died. But it's a different story...
 
I tried to watch that earlier today and it kept locking up. I close it twice when I saw the incorrect title. Guess I'll try again later.

Best,

Dave
 
TangoWhiskey said:
How about a left turn (shorter) back to the much longer runway 30? Check out the airport diagram; he was departing 18:
Click to expand...

That was my first reaction as well. In addition he used way to little bank in the turn (45-50° is ideal) and let his speed decay too much immediately after the failure. I'm not saying I would have done any better, just that even though his performance was fairly sub-optimal, he completed the maneuver successfully. I think that says that (at least in a Mooney) the impossible turn isn't all that difficult from 500 AGL.
 
gismo said:
That was my first reaction as well. In addition he used way to little bank in the turn (45-50° is ideal) and let his speed decay too much immediately after the failure. I'm not saying I would have done any better, just that even though his performance was fairly sub-optimal, he completed the maneuver successfully. I think that says that (at least in a Mooney) the impossible turn isn't all that difficult from 500 AGL.
Click to expand...

He also overshot the turn back to the landing runway (36), necessitating additional maneuvering and turns back to the left then right to align himself. Of course, the shorter runway 18-36, doing any more gradual or early of a turn towards the runway may have very well left him with too little runway to stop... and you can see on final approach all the highway traffic zipping by left to right. Making it back to the runway, then going off then end and onto the highway would have been bad.
 
That was sheer luck on the pilot's part. The turn back to airport put him over very hostile terrain (i.e. trees).
Yes, a left turn may have been a better choice, but when everything's going wrong it's a lot easier to be sitting on the sofa watching the video that be in the cockpit trying to deal with a rapidly developing emergency at low altitude.

Yes, his airspeed decayed after the failure, but that's the difference between real life and training. It's easy to keep airspeed when you're doing your BFR and know the failure's coming.
When I practice engine outs, I keep the yoke locked for two seconds after the throttle pull to simulate the OODA loop from the failure. Then you're dealing with the airspeed you will have after the "is this really happening" and "oh shoot!" reaction.

It's also easy to say he didn't bank enough, but I think that's a bad analysis.
The downside of not banking enough is you don't make it to the runway and have to land out. The downside of banking too much is stalling with do power at low altitude (i.e. death)

If you really want to see the downside of a low altitude engine-out stall, do what I do every so often (and which is almost never taught).
Go out to the practice area and pull the power in clear configuration. Hold altitude until the stall, and then recover without adding power. It's surprising how far down you have to push the nose and how long it takes to recover (both timewise and altitude wise) without power.

I'll keep my banks shallow, thanks.
 
Yes, his airspeed decayed after the failure...
Click to expand...
Not necessarily.

There are two warning horns that you hear in that video. Just before touchdown you hear a higher-pitched squeal, which it seems safe to assume is the stall warning horn.

Just after the engine fails, and then again for a while as he's clearing the trees, you hear a lower-pitched alarm which the AOPA program explains is the gear warning horn, which they say is triggered by having the gear up with a low manifold pressure.
-harry
 
Before every takeoff... once complete with preflight run-up... I decide my course of action should the engine fail in the first 1000 feet.

At my home airport, I have pre-identified areas to put it down if a turn back is not feasible. With the development in the last 10 years around 3CK (Lake in the Hills, IL) - this is becoming more difficult than it used to be.

Anyway, I think this is something that is not taught enough in our training. All of our CFIs gave us guidance not turn back under 500 AGL, but few teach that an engine-failed landing plan should be part of the pretakeoff checklist. At a minimum, we should be making note of the MSL altitude for 500 AGL, and calling that out as we climb through 500AGL.

For landing areas, Google Maps and Earth can help with this too, especially if the airport is an unfamiliar field.

I have had one instance of an intermittent/blown mag shortly after takeoff. The bad mag was firing at odd times, causing rough operation and a dramatic RPM drop. Fortunately, the engine still was making about 2000 RPM. I made the 180 back to the runway, and other than heavy braking because of the downwind landing, it went well. That said, there was a moment of hesitation before I turned. I learned from this experience that there is a definitely a difference between a real emergency and a practice emergency. All the more reason to have a well-thought out plan ahead of time.


PilotAlan said:
That was sheer luck on the pilot's part. The turn back to airport put him over very hostile terrain (i.e. trees).
Yes, a left turn may have been a better choice, but when everything's going wrong it's a lot easier to be sitting on the sofa watching the video that be in the cockpit trying to deal with a rapidly developing emergency at low altitude.
Click to expand...
 
nyoung said:
Before every takeoff... once complete with preflight run-up... I decide my course of action should the engine fail in the first 1000 feet.

At my home airport, I have pre-identified areas to put it down if a turn back is not feasible. With the development in the last 10 years around 3CK (Lake in the Hills, IL) - this is becoming more difficult than it used to be.

Anyway, I think this is something that is not taught enough in our training. All of our CFIs gave us guidance not turn back under 500 AGL, but few teach that an engine-failed landing plan should be part of the pretakeoff checklist. At a minimum, we should be making note of the MSL altitude for 500 AGL, and calling that out as we climb through 500AGL.
Click to expand...

I don't know which "few" those might be, but this one was taught -- and continues the tradition -- of a pre-takeoff brief to include "No turn back until 1000' AGL."

That 1000 AGL is the figure unless I know the airplane and have practiced the maneuver in that particular airplane.

I suggest to students not to read/say the pre-takeoff emergency brief aloud with passengers (unless the pax are pilots).

This maneuver is not a PTS requirement so may be the reason it's not taught as much as we would like.
 
dmccormack said:
I don't know which "few" those might be, but this one was taught -- and continues the tradition -- of a pre-takeoff brief to include "No turn back until 1000' AGL."

That 1000 AGL is the figure unless I know the airplane and have practiced the maneuver in that particular airplane.

I suggest to students not to read/say the pre-takeoff emergency brief aloud with passengers (unless the pax are pilots).

This maneuver is not a PTS requirement so may be the reason it's not taught as much as we would like.
Click to expand...

My instructors suggested the following:

0-50 agl - back onto the runway / straight ahead
50-500 - pick a landing area +/- 30 degrees either side
500-1000 - pick a landing area +/- 90 degrees either side
>1000 - turn back to the airport if it's the safest alternative
>10,000 AGL - Pick the best airport for a safe landing. ;)
 
Jack Spectre said:
My instructors suggested the following:

0-50 agl - back onto the runway / straight ahead
50-500 - pick a landing area +/- 30 degrees either side
500-1000 - pick a landing area +/- 90 degrees either side
>1000 - turn back to the airport if it's the safest alternative
>10,000 AGL - Pick the best airport for a safe landing. ;)
Click to expand...

All interesting options, but I think "Straight ahead" (meaning "Minimal maneuvering) under a certain altitude is what you'll remember when the feces strikes the air circulation device.
 
I haven't heard anyone say anything about pulling the prop back to get a better glide. Too low? It seems as if he had time to me.
This isn't a fixed pitch prop is it?

Best,

Dave
 
wasnt the engine still running?
long discussion about this on the purple board.
 
I try to always do a pre-takeoff briefing, especially since I don't always fly the same airplane. Most of the aircraft I fly have somewhat similar glide characteristics, and so behave about the way I would expect.

Two weeks ago, I got to fly a friend's early 1990 KA C90. Suffice to say that my attempt at a power-off 180 didn't go well at all, and my usual turnback altitude was completely useless. It only takes 15 seconds to brief turnback altitudes based on airplane and weight (and things like DA).

One thing that could be helpful might be to accelerate in ground affect before starting the climb. Besides other benefits, you would have a higher stall margin when the engine does quit.

-Felix
 
Last edited:
sba55 said:
I try to always do a pre-takeoff briefing, especially since I don't always fly the same airplane. Most of the aircraft I fly have somewhat similar glide characteristics, and so behave about the way I would expect.

Two weeks ago, I got to fly a friend's early 1990 KA C90. Suffice to say that my attempt at a power-off 180 didn't go well at all, and my usual turnback altitude was completely useless. It only takes 15 seconds to brief turnback altitudes based on airplane and weight (and things like DA).

One thing that could be helpful might be to accelerate in ground affect before starting the climb. Besides other benefits, you would have a higher stall margin when the engine does quit.

-Felix
Click to expand...

As Felix knows. that's what I do when there is enough runway; accelerate in ground effect until blue line if possible. Cruise climb also if no obstacles.

Best,

Dave
 
Couple of thoughts from a Mooney guy. 1. Since the failed cylinder didn't seize the engine, the other 3 were probably making some power. If so, that would make a huge difference in the outcome of this turn. 2. He forgot the flaps; he didn't pull out the prop back. I am not surprised. After the failure his brain was 100% in survival mode. Those "details" never crossed his mind.
 
The population density in some places (like here) is such that often there is no suitable emergency landing area "slightly to the left or right." In such a situation, I think I'd be tempted to try the 180 and hope to land somewhere on the airport, even if not the runway.

I also was taught to accelerate in ground effect, by the way. Now I understand why.

-Rich
 
Lance F said:
Couple of thoughts from a Mooney guy. 1. Since the failed cylinder didn't seize the engine, the other 3 were probably making some power. If so, that would make a huge difference in the outcome of this turn.
Click to expand...

If he left the throttle open with one cylinder cracked in half, wouldn't he be risking a major fire by spraying unburned fuel out the side of the cracked cylinder (and making a spark around it from the still-firing plug)?

Chris
 
mantakos said:
Not necessarily.

There are two warning horns that you hear in that video. Just before touchdown you hear a higher-pitched squeal, which it seems safe to assume is the stall warning horn.

Just after the engine fails, and then again for a while as he's clearing the trees, you hear a lower-pitched alarm which the AOPA program explains is the gear warning horn, which they say is triggered by having the gear up with a low manifold pressure.
-harry
Click to expand...

I think the gear warning horn is activated by a switch on the throttle control not directly by manifold pressure. Either way the pilot would have to have pulled the throttle closed to make that horn sound. Some Mooney's had a system that dropped the gear when the IAS was below a certain speed and it's possible that this also triggered a warning horn. My guess is that he closed the throttle in an attempt to quell the nasty noises from the engine bay and that activated the gear horn.

That said, the ideal maneuver for returning to the airport is a 45 degree bank with the stall warning barely activated until it's time fly wings level.
 
Last edited:
Dave Siciliano said:
I haven't heard anyone say anything about pulling the prop back to get a better glide. Too low? It seems as if he had time to me.
This isn't a fixed pitch prop is it?

Best,

Dave
Click to expand...

The pilot mentioned that he should have done that but in the heat of the moment he never thought of it.
 
dmccormack said:
I don't know which "few" those might be, but this one was taught -- and continues the tradition -- of a pre-takeoff brief to include "No turn back until 1000' AGL."

That 1000 AGL is the figure unless I know the airplane and have practiced the maneuver in that particular airplane.

I suggest to students not to read/say the pre-takeoff emergency brief aloud with passengers (unless the pax are pilots).

This maneuver is not a PTS requirement so may be the reason it's not taught as much as we would like.
Click to expand...

Personally, if the area ahead doesn't look too good for a landing one could walk from, I'd try for as much turn as needed to point the plane in a more favorable direction. I see nothing wrong or dangerous about starting a turn as long as you maintain sufficient airspeed and recognize that more speed is needed when turning. This is one place where an AoA indicator would be a great benefit. The danger lies in letting the plane stall or continuing the turn so low that a tip drags. IMO either can be avoided without maintaining runway heading even if you're only a couple hundred feet AGL. Also I see little point in establishing hard altitudes for specific turn limits. The potential for turning varies significantly with the airplane and the pilot's skill. I have practiced a 270/90 reversal in my taildragger and found that I can make it with less than 200 ft of altitude loss. It has no stall warning but (at least when solo) there's not enough elevator authority to make it stall in a 45 degree bank (for reasons I don't understand the bank limits the stall potential).
 
PilotAlan said:
That was sheer luck on the pilot's part. The turn back to airport put him over very hostile terrain (i.e. trees).
Yes, a left turn may have been a better choice, but when everything's going wrong it's a lot easier to be sitting on the sofa watching the video that be in the cockpit trying to deal with a rapidly developing emergency at low altitude.

Yes, his airspeed decayed after the failure, but that's the difference between real life and training. It's easy to keep airspeed when you're doing your BFR and know the failure's coming.
When I practice engine outs, I keep the yoke locked for two seconds after the throttle pull to simulate the OODA loop from the failure. Then you're dealing with the airspeed you will have after the "is this really happening" and "oh shoot!" reaction.

It's also easy to say he didn't bank enough, but I think that's a bad analysis.
The downside of not banking enough is you don't make it to the runway and have to land out. The downside of banking too much is stalling with do power at low altitude (i.e. death)

If you really want to see the downside of a low altitude engine-out stall, do what I do every so often (and which is almost never taught).
Go out to the practice area and pull the power in clear configuration. Hold altitude until the stall, and then recover without adding power. It's surprising how far down you have to push the nose and how long it takes to recover (both timewise and altitude wise) without power.

I'll keep my banks shallow, thanks.
Click to expand...

No question, a stall must be avoided at all costs. Even if recovery were possible you'd lose all chance of aiming for a suitable landing area and the chances for any recovery when stalling close to the ground are not good. That said, I see no advantage in keeping the bank angle small. This significantly reduces your available landing areas and it's actually more difficult to stall most airplanes in a steep bank than in level flight. In addition stall recovery in a steep bank can be accomplished by simply reducing the bank angle. I suggest you try the same exercise you mentioned in a 45-50 degree bank and see for yourself what the recovery potential is. Of course the most important thing is to keep the plane coordinated. If you rudder the turn, a stall is likely to be a lot more difficult to terminate.
 
gismo said:
Also I see little point in establishing hard altitudes for specific turn limits. The potential for turning varies significantly with the airplane and the pilot's skill.
Click to expand...


...which is why I mentioned that 1000 AGL was a guideline until the pilot has assessed his/her skill in the particular airplane.
 
gismo said:
No question, a stall must be avoided at all costs. Even if recovery were possible you'd lose all chance of aiming for a suitable landing area and the chances for any recovery when stalling close to the ground are not good. That said, I see no advantage in keeping the bank angle small. This significantly reduces your available landing areas and it's actually more difficult to stall most airplanes in a steep bank than in level flight. In addition stall recovery in a steep bank can be accomplished by simply reducing the bank angle. I suggest you try the same exercise you mentioned in a 45-50 degree bank and see for yourself what the recovery potential is. Of course the most important thing is to keep the plane coordinated. If you rudder the turn, a stall is likely to be a lot more difficult to terminate.
Click to expand...

The issue with steep banks close to the ground power off for most pilots is the results of a poorly coordinated slow-speed steep turn would be very bad.

Better to plow straight ahead into some trees than drop in from 50' AGL in a spin.
 
wasnt the engine still running?
Click to expand...
The pilot says that he pulled the mixture to try to prevent further engine damage, though I don't think it's clear at what exact moment he does this.
-harry
 
sba55 said:
One thing that could be helpful might be to accelerate in ground affect before starting the climb. Besides other benefits, you would have a higher stall margin when the engine does quit.

-Felix
Click to expand...

In a twin, sure. I'm not convinced that would really be useful in a single, though.


Trapper John
 
sba55 said:
Why not? It worked well for me in the Bonanza.
Click to expand...

I think I see John's POV... in a single, you don't have a blue line speed you're trying to accelerate to. Instead, altitude is your best friend.

I'd think (haven't tested this) that one would gain more altitude by climbing out at Vy or Vx as soon as you hit that target airspeed than you'll get by "zoom climbing" after accelerating to 100 knots in ground effect, if we have an engine out at the same mm:ss point into the flight.
 
TangoWhiskey said:
I think I see John's POV... in a single, you don't have a blue line speed you're trying to accelerate to. Instead, altitude is your best friend.

I'd think (haven't tested this) that one would gain more altitude by climbing out at Vy or Vx as soon as you hit that target airspeed than you'll get by "zoom climbing" after accelerating to 100 knots in ground effect, if we have an engine out at the same mm:ss point into the flight.
Click to expand...
If the only thing you're concerned about is altitude over time, then you would want to always climb at Vy. Realistically, however, that isn't the only concern, and accelerating in ground affect addresses some of the other concerns, such as speed margin above stall, engine temperatures, and the ease of putting down the plane straight ahead if something goes wrong (it's much easier if you're at 20' than at at 250' - try it in a sim if you get a chance).

So, sure - flying ME, there's yet another reason to do this. But even SE, it can make a lot of sense if you're concerned about the above.

-Felix
 
TangoWhiskey said:
I think I see John's POV... in a single, you don't have a blue line speed you're trying to accelerate to. Instead, altitude is your best friend.

I'd think (haven't tested this) that one would gain more altitude by climbing out at Vy or Vx as soon as you hit that target airspeed than you'll get by "zoom climbing" after accelerating to 100 knots in ground effect, if we have an engine out at the same mm:ss point into the flight.
Click to expand...

I did the same thing in my single. Arguments both ways; I prefer speed for a couple reasons. First, if there is a long runway (which is the only time this applies), if I lose an engine, I want to be close to that runway when I try to land. I can get down from a couple feet up pretty fast and get the brakes engaged. I see folks nose up with thousands of feet of runway left at low speeds. Second, if I lose an engine after accelerating to a higher speed, an engine loss can be dealt with with less change in attitude; at slow speed one better get the nose down RIGHT NOW or stall. Visibility is better.

Now, if one is uncomfortable low, they may not like this. Back when I flew in the Army, we had to stay low when the bad guys started shooting missiles. If we lost an engine, one could pop up slightly, look left or right maybe 30 degrees and put the plane down.

Higher certainly gives a better view to see where you want to go after the abrupt attitude change. The energy is a bit higher according to some folks with fancy degrees that have commented on this, but not significantly more. One can have potential or Kinetic. I lean toward the second and know how to use it.

In the twin, we have compared methods and found one comes out pretty close to the same after about 45 to 60 seconds of climbing on one. In the climb at Vy case, one loses a lot of energy when the engine failure takes place and onemakes an abrupt attitude change. Many make the case where the plane says nose high with the pilot figures out what the heck is going on before he/she reacts. In the second case, one is lower longer but the pitch change is almost a none event. Remind me sometimes and I'll relate the story to you of when I was in the sim with a fella that liked Vy.

John Deakin wrote an excellent article in Pelican's Perch about it; worth a read.

I don't try to convert folks over to the dark side, but it's what I do and it's been acceptable to every CFI and DPE I've flown with once briefed. Now, if one is taking a check ride or currency training where there's a PTS or other FAA requirements, you might want to discuss it before varying from what the FAA professes.

Best,

Dave
 
sba55 said:
Why not? It worked well for me in the Bonanza.
Click to expand...

Depends on your goal, which probably should change with the specific circumstances of the area you're taking off towards. If you want to reach a an altitude sufficient to make returning to the airport as feasible as possible in the shortest time, the only time you should spend in ground effect would be before reaching Vy (Vx if obstacles are involved) and in most singles that's not much time and you'd often better off just staying on the ground until the appropriate speed is reached. But if you want to make the transition to gliding flight easier and the terrain favors landing nearly straight ahead, staying in GE while accelerating to several knots above Vy and climbing at that higher speed would be the best option.
 
Dave Siciliano said:
The energy is a bit higher according to some folks with fancy degrees that have commented on this, but not significantly more. One can have potential or Kinetic. I lean toward the second and know how to use it.
Click to expand...

Dave, the only real difference is that when climbing above Vy you are spending more of your "energy" overcoming drag. I agree that a 10% increase over Vy isn't significant in that regard though, and I definitely agree that unless there's a compelling reason to go for altitude as rapidly as possible (e.g. when launching over terrain that doesn't offer much survivability if you lose an engine in a single), the advantages of a shallower, higher speed climb outweigh any advantages of a steep climb at Vy. I think it's also worth pointing out that this is more true in overpowered airplanes than in underpowered ones where the climb attitude isn't all that steep at Vy and the climb rate is marginal at best.
 
sba55 said:
I

One thing that could be helpful might be to accelerate in ground affect before starting the climb. Besides other benefits, you would have a higher stall margin when the engine does quit.

-Felix
Click to expand...


But would you have the altitude?
 
Not to beat this to death, and changing to a twin, a good friend that flies for Fed Ex and I ran all the numbers to do a Vx climb in my Baron and went out to a 4,000 foot runway and practiced it; it scared the snot out of me.

I accelerated to a few knots under Vx where I had to yank the plane off the ground in about a 10 to 12 degree up attitude in order for momentum of the plane to keep the speed at Vx. John, in the right seat, hit the garmin to register coordinates when at 100 feet. We did three takeoffs and punched in the coordinates. On the first one, we pitched up a little late and, as a consequence, went a bit faster than Vx. On the second, we adjust from the first one and stayed right on Vx in the climb. Vx was 72 ot 73 knots IIRC--I have the figures somewhere. On the third departure, we accelerated to 90 knots (wheels on the ground in each case) and pitched up to about 15 degrees to 100 feet.
When done, we went back and compared all three; there was less than 100 feet horizontal difference in where we reached 100 feet. So, the takeoff procedure didn't make that much difference, but the consequence, should one lose an engine was significant. Going at 72 knots and losing an engine would definitely have me below Vmc and a twin would roll. If I rapidly pitched the nose down to keep speed up below 100 feet, I'd smash it hard into the runway.

So, I'm a take off from a longer runway fella; accelerate in ground effect or wheels on the ground longer. If there are obstacles, get more speed, pull up over them, then lower the nose right after. I don't want to be 50 to 100 feet, nose pointed up just above stall or Vmc.

My 2 cents. The best way to avoid this situation is to use a longer runway which I do.

What I see a lot of folks do is exactly what I wouldn't do; pull the plane up into a nose high attitude with a bunch of runway left that could be used if lower. They are in a very vulnerable position that first couple hundred feet if they don't react immediately to an engine out and they can't see a thing in front of them as far as traffic if there is any.

I think John Deakin did an excellent job of explaining it. If you go through the drill I did with my Fed Ex friend, you may appreciate the difference more.

Best,

Dave
 
Last edited:
gismo said:
Depends on your goal, which probably should change with the specific circumstances of the area you're taking off towards. If you want to reach a an altitude sufficient to make returning to the airport as feasible as possible in the shortest time, the only time you should spend in ground effect would be before reaching Vy (Vx if obstacles are involved) and in most singles that's not much time and you'd often better off just staying on the ground until the appropriate speed is reached. But if you want to make the transition to gliding flight easier and the terrain favors landing nearly straight ahead, staying in GE while accelerating to several knots above Vy and climbing at that higher speed would be the best option.
Click to expand...
Exactly - like I said, if altitude is your primary concern, then climb at Vy.

AdamZ said:
But would you have the altitude?
Click to expand...
You'd have slightly less than if you climbed at Vy. If your only concern is altitude, climb at Vy.....
 
You must log in or register to reply here.
Back
Top