Turn Around On Take Off After Engine Out.

[redtail]

I only half agree. A proper short field takeoff includes both holding the brakes until full power develops and flaps. This gets you off the ground sooner but actually results in a slower climb. By 50' AGL any edge gained by the flap's lift is lost due to the added drag.

Thus, when taking off from a typical runway (not a short field) I hold brakes until full power develops without flaps. That way I get to Vy and climb as quickly as possible.

I am a big believer in using the whole runway. My home field has a taxiway going into the middle of the runway with enough room for me to take off to the east. Yet I still back taxi my 182 all the way down to the numbers before I go. It is a little thing, but why tempt fate?

YMMV

Same here. I use it all, and at my experience level, 1000' is my absolute minimum to even consider returning to the field! Also, Vy is what I use on practically every takeoff.

 

[redtail]

I say consider because, with highways around my home field, coming up short would probably be worst than just accepting a controlled crash into trees straight ahead or off to one side!

Just thinking about it reminds me of going to the dentist, unpleasant thought but necessary

 

[azure]

I say consider because, with highways around my home field, coming up short would probably be worst than just accepting a controlled crash into trees straight ahead or off to one side!

In any direction from my home field it's industrial parks and residential areas for miles around. Coming up short, straight ahead, it's mostly all the same. There's no place to put down safely. Still, your chances of surviving are greatest if you're in control of the airplane all the way to stopping. For that reason, I don't think I would try the impossible turn here even if I was proficient at it, unless I was at an altitude (1000 AGL at least) where I was 100% sure of making it back to the runway. If what I see after completing the turn is especially bad, being at a lower altitude at that point gives me even fewer options.

 

[gkainz]

does anyone know where this happened? (the vimeo clip in the original post) ... I'm curious what the straight ahead options looked like on google maps/earth...

 

[alfadog]

does anyone know where this happened? (the vimeo clip in the original post) ... I'm curious what the straight ahead options looked like on google maps/earth...

That's easy. There is a date in the camera:

Full narrative:
http://www.ntsb.gov/aviationquery/brief2.aspx?ev_id=20090301X04523&ntsbno=WPR09FA133&akey=1

Summary:

NTSB Identification: WPR09FA133
14 CFR Part 91: General Aviation
Accident occurred Saturday, February 28, 2009 in Rio Linda, CA
Probable Cause Approval Date: 04/28/2011
Aircraft: MOONEY M20E, registration: N3463X
Injuries: 2 Fatal.
The flight was planned to be a maintenance test flight following an annual inspection, which included the installation of a field overhauled engine by the mechanic/passenger. Witnesses reported that a "popping" sound was heard during the takeoff ground roll. After liftoff, they heard the engine surge followed by a puff of black smoke emanating from the right rear of the engine. A videotape recording obtained from a camera located on a nearby building captured the airplane on the initial climb-out before it made a left turn back to the runway. The left wing and nose dropped as the airplane entered a spin prior to descending out of view of the camera. The airplane came to rest in an open field about 300 feet from the end of the runway and was consumed by a post-crash fire. During the postaccident inspection of the airframe and engine, no mechanical failures or malfunctions were found that would have prevented normal operation. The fuel lines, hoses, and fuel pump were too badly fire damaged to be able to determine if any pre-existing malfunctions affected those components. The National Transportation Safety Board determines the probable cause(s) of this accident as follows:
A partial loss of engine power for undetermined reasons and the pilot's failure to maintain an adequate airspeed while maneuvering during the initial climb, which resulted in a stall/spin.

edit: crash site

 

[redtail]

That's easy. There is a date in the camera:



edit: crash site

From the location of the high school video and this map, looks like his best option would have been a 90* left turn. There's the school on the right and a big highway straight ahead!

Just my thoughts.

It's sad but you can see how tempting it was to make a left 180! Even if you don't make the runway, there's a field off the left side of the airport. Boy, this really makes you think.

RIP

 

[cleared4theoption]

That park straight ahead looked like an ideal spot to me.

 

[redtail]

That park straight ahead looked like an ideal spot to me.

I could be wrong but based on the location of the school, I'd say he was probably over the park when the engine quit. Straight ahead, looks like a four lane major highway!

That area beyond the highway looks cool if you're high enough to glide past the highway!

 

[TangoWhiskey]

That park straight ahead looked like an ideal spot to me.

At 9am on a Saturday (2/28/2009), he may have felt otherwise based on the view out the windshield. Park may have been occupied... I see baseball and soccer fields. Yes, February, but not in Minnesota...

Zooming in on the park til you see "fly by" views shows each ball park is separated by chain link fence, too... uggh.

 

[TangoWhiskey]

Gene Benson discussed this accident here:

http://www.genebenson.com/newsletter/newsletter_aug11.htm#complex

 

[redtail]

Gene Benson discussed this accident here:

http://www.genebenson.com/newsletter/newsletter_aug11.htm#complex

Thanks, I'm going to read it but just taking a quick look at the diagram shows he was a lot closer to the runway than I thought!

Looks like if he had just turned a little left perhaps 90*, and leveled out, they would have made it! Sad

 

[alfadog]

Thanks, I'm going to read it but just taking a quick look at the diagram shows he was a lot closer to the runway than I thought!

Looks like if he had just turned a little left perhaps 90*, and leveled out, they would have made it! Sad

If he had done just about anything other than spin it in, they would have had at least a chance. Fly the airplane, fly it as far into the crash as possible - to paraphrase Bob Hoover.

 

[redtail]

Edit: after getting the nose down!

 

[Jack Spectre]

The trouble with a rule like that is that altitude alone doesn't tell you whether it is possible to glide back to the airport.

Agreed. It was taught with the caveat of a 180 turn being an option...if it was the best one. I took it to mean "you'll be able to point your plane back at the airport without hitting the ground".

 

[alfadog]

Agreed. It was taught with the caveat of a 180 turn being an option...if it was the best one. I took it to mean "you'll be able to point your plane back at the airport without hitting the ground".

AOPA says the rule of thumb is you should have 2/3 of your decision altitude for turn-around made by the end of the runway. So for the 172, I should be at 400'. I will have to see. Obviously, it varies by wind and runway factors. In my successful attempt last weekend in the Luscombe, I doubt I was 200' at the end of the 3000' runway with a 15kt headwind.

 

[Jay Honeck]

Also, for the new pilots; A good practice, when arriving at unfamiliar airports, is to check out the departure end of your landing runway for possible emergency landing areas. That way when you depart, you will already have some idea of the choices you will have if the engine quits.

95% of the time, we take off to the South from our island base, Mustang Beach Airport. (KRAS)

If our engine quits, we have two choices: Land on the beach, or ditch in the Gulf. Since the beach is often loaded with people, my plan is to make a shallow 90 degree turn and ditch just off-shore. The water is very shallow, and we should survive.

I have NO idea what will happen when we do that, however. My guess is that our fixed-gear Pathfinder might flip, especially if I don't nail the approach speed just above stall.

It's always on my mind, as the wheels start to feel light....

 

[ebetancourt]

One thing that I never see covered in these discussions is how to land straight ahead. By that I mean, are you going to use best glide? Or a massive slip? Or something in between. So much in these discussions depends on the airplane, the wind, and the pilot's mental preparation that there really are very few absolutes. Climb at Vx or Vy? Flaps? No flaps?

In the Citabria, I have come to realize that short field T/O, Vx climb, means that straight ahead will be the only option to around 1,000 ft, because there won't be much runway behind me. The turn can easily be made in 500 ft, but... On the other hand a nose low "big" slip and I will have access to the runway and overrun for a long time. As far as the turn back from 500 or 1000 or? the wind will make a great difference and turning back may not be worth it. Any discussion of turn backs should include a discussion of downwind landings.

 

[azure]

One thing that I never see covered in these discussions is how to land straight ahead. By that I mean, are you going to use best glide? Or a massive slip? Or something in between. So much in these discussions depends on the airplane, the wind, and the pilot's mental preparation that there really are very few absolutes. Climb at Vx or Vy? Flaps? No flaps?

FWIW I would

1. IMMEDIATELY establish best glide;

2. QUICKLY choose a suitable landing spot (or the least UNsuitable landing spot that catches my eye);

3. use whatever combination of flaps, gear (maybe), aggressive slipping to get the airplane to the target spot at minimum controllable airspeed.

2 would follow 1 so quickly if the fan quit at low altitude that I might not even have time to establish best glide, if the best landing spot was close by. The important thing would be not to second-guess myself, and concentrate on flying the airplane to the scene of the crash (sorry, I keep thinking VLL, and that's the best I could hope for there).

 

[gismo]

One thing that I never see covered in these discussions is how to land straight ahead. By that I mean, are you going to use best glide? Or a massive slip? Or something in between. So much in these discussions depends on the airplane, the wind, and the pilot's mental preparation that there really are very few absolutes. Climb at Vx or Vy? Flaps? No flaps?

It really depends on your altitude and airspeed once you recognize the failure. If you're only 100 AGL there may not be time for anything but dropping the nose briefly and flaring. Up a little higher and you should have time decide whether to glide as far as possible (flaps up till the last 50 ft of descent) or go to full flaps and slip aggressively so as to land closer in. Whatever the case the two most important actions are to maintain flying speed (no slower than min sink and preferably a bit faster) until touching down with minimum forward speed and to avoid plowing directly into any solid objects.

In the Citabria, I have come to realize that short field T/O, Vx climb, means that straight ahead will be the only option to around 1,000 ft, because there won't be much runway behind me. The turn can easily be made in 500 ft, but... On the other hand a nose low "big" slip and I will have access to the runway and overrun for a long time. As far as the turn back from 500 or 1000 or? the wind will make a great difference and turning back may not be worth it. Any discussion of turn backs should include a discussion of downwind landings.[/QUOTE]

 

[Palmpilot]

You need to try it in your airplane and find out for yourself. I know in a 172 at 600', I turn back.

According to the analysis I've seen, whether you make it back to the airport depends on the runway length and the headwind component. At gross weight with no wind and a runway shorter than about 3500 feet, a pre-1996 172 won't make it regardless of the turnaround height, because the glide angle exceeds the climb angle. (I don't have data for the newer 172s.)

Tried the return in the Luscombe last weekend. The Luscombe is 65 hp so climb is 500 fpm max and the glide ratio ain't so hot.

Try # 1. After some practice down the runway, I clear the departure end low, climbed to 500', pulled the power and hiked it around at best glide and 45 degrees. No way I was going to make the field because the late departure and crappy climb had put me pretty far out.

Try # 2. Normal takeoff, same climb, pulled power, same return, made it back no problem.

I've been taught that practicing this at low altitude is a bad idea.

Lesson learned: If you use all available runway and make every take-off a short field takeoff, you maximize your chance of making it back from your target altitude.

Unless you calculate the rate of climb needed, based on the headwind component and the runway length, you're trusting to luck as to whether a turn around would be successful. And if the engine fails at a higher altitude than your predetermined turn around height, the calculation becomes more complicated, because if your glide angle is greater than your climb angle, then the additional altitude hurts rather than helping.

 

[Palmpilot]

Thus, when taking off from a typical runway (not a short field) I hold brakes until full power develops without flaps. That way I get to Vy and climb as quickly as possible.

Vx is better than Vy for turn around purposes, because angle of climb, not rate of climb, is what determines success or failure.

 

[Palmpilot]

AOPA says the rule of thumb is you should have 2/3 of your decision altitude for turn-around made by the end of the runway.

That seems rather arbitrary. If you are lucky enough to have a climb angle equal to your descent angle, your point of landing would be near the point at which you had the height required to make the turn.

Can you point me to where to they say this? I'd like to find out what they're basing it on.

 

[gismo]

Vx is better than Vy for turn around purposes, because angle of climb, not rate of climb, is what determines success or failure.

And unfortunately Vy let alone Vx puts you in a very difficult situation if the engine quits when your deck angle is that high in many high performance airplanes. In a 172 it's probably not as big a deal.

 

[TMetzinger]

That's a total BS made up statistic. Engine quits and you DO make it back, how is that recorded? It's not, unless you've got a tower controller taking notes.

The Impossible Turn is indeed possible. The minimum altitude depends on the wind, the airplane, the technique. It is a high-performance maneuver, and the penalty for not getting it right is quite severe. But it can, and has, been done successfully many times.

All true. I'll say however that which the exception of folks who are actively flying all the time and at the top of their game, the safer bet is that unless you know you have enough altitude to make it back, you should make limited turns at no more than medium bank angles and come back to earth under full control and minimal energy, even if it means hitting trees or cows or something else besides other people.

 

[Palmpilot]

And unfortunately Vy let alone Vx puts you in a very difficult situation if the engine quits when your deck angle is that high in many high performance airplanes. In a 172 it's probably not as big a deal.

In a 172, Vx has at least a 12 knot margin above stall speed with the wings level, so that should give the pilot a little time to react correctly. I don't know about other models.

If the plane is trimmed for whatever climb speed is being used, won't the nose lower itself if the pilot does not apply back pressure?

 

[gismo]

In a 172, Vx has at least a 12 knot margin above stall speed with the wings level, so that should give the pilot a little time to react correctly. I don't know about other models.

It's not the stall margin that's at issue here. It's the difference in deck angles between a Vx climb and a best glide descent along with the higher inertia of a larger airplane that causes the problem. In an airplane that climbs "briskly" and has a high glide speed that difference can easily exceed 20-30° and while it's quite possible to avoid a stall if you react reasonably soon after the engine quits (typical reaction times are in the 2-5 second range if you're not expecting the failure) but you are likely to find yourself without enough speed and altitude to successfully flare by the time you get close to the ground if you were fairly nose high when the fan stops.

If the plane is trimmed for whatever climb speed is being used, won't the nose lower itself if the pilot does not apply back pressure

Yes but that's not sufficient to build enough kinetic energy for flaring.

Like I said, this probably isn't much of an issue in a 172 but it might be interesting to experience (at a safe altitude) anyway. Set up a climb at Vx, pull the mixture, count off 3 seconds and then see how much altitude is required to reach a normal no flap approach speed and then level off.

IIRC in a Bonanza at MGW it takes something like 100-150 ft, in my Baron it's more than 200 ft so if you lost one at 100 AGL you'd be underground by the time you "landed"

 

[alfadog]

Practiced the return today in my Arrow and made a separate post of the vid.

 

[Red Air Rambo]

If he had done just about anything other than spin it in, they would have had at least a chance. Fly the airplane, fly it as far into the crash as possible - to paraphrase Bob Hoover.

I think Jimmy Doolittle said it 1st.

 

[Dan Thomas]

I think that what you're seeing as "overbanking" is actually the beginning of the spin. Right around 0:18 you can tell that a constant bank (looks like about 45°) was held while the turn rate continues to increase and immediately after that the nose drops along with the left wing. This is exactly what a skidding turn to spin looks like. If the plane were coordinated when the stall occurred the nose would drop away from the camera.

In a coordinated descending turn, the inside wing is at a higher angle of attack than the outside wing and will stall first. And this is in a coordinated turn. A skid exacerbates the difference, a slip minimizes it. A tight turn like the one in the accident video makes the AoA difference rather large and even if he had the ball centered the left wing would have fallen off into the spin simply because he stalled the airplane.

In a descending turn the whole airplane is descending at the same rate, of course, but the angles at which its parts descend at are different. The inside wing is describing a tighter helix so it's travelling downward at a higher angle than the outer wing, which is describing a larger helix. This is exactly the same reason why propellers have large pitch angles near the hub and smaller angles at the tips. Trigonometry. It can kill.

Dan

 

[Palmpilot]

It's not the stall margin that's at issue here. It's the difference in deck angles between a Vx climb and a best glide descent along with the higher inertia of a larger airplane that causes the problem. In an airplane that climbs "briskly" and has a high glide speed that difference can easily exceed 20-30° and while it's quite possible to avoid a stall if you react reasonably soon after the engine quits (typical reaction times are in the 2-5 second range if you're not expecting the failure) but you are likely to find yourself without enough speed and altitude to successfully flare by the time you get close to the ground if you were fairly nose high when the fan stops.

If you avoid a stall, how can you not have enough airspeed to flare?

Have you tried this? If so, how quickly did the airspeed bleed off to stall speed?

Like I said, this probably isn't much of an issue in a 172 but it might be interesting to experience (at a safe altitude) anyway. Set up a climb at Vx, pull the mixture, count off 3 seconds and then see how much altitude is required to reach a normal no flap approach speed and then level off

An instructor once pulled the throttle to idle on me without warning in a 172 while there was still enough runway left to land on. I had no difficulty flaring. That was a Vy climb, but from the way that went, I have a hard time imagining that a Vx climb would have produced any particular difficulty.

If, as you said earlier, both Vx and Vy can put you in "a very difficult situation," what airspeed do you recommend for takeoff in a single?

 

[alfadog]

If you avoid a stall, how can you not have enough airspeed to flare?

Have you tried this? If so, how quickly did the airspeed bleed off to stall speed?

He is referring to heavier aircraft that perhaps have lift issues that would not arrest a descent in that case. You also get into issues of insufficient elevator authority. If you watch the video I posted of doing this manuever in my Arrow, you see I "save a little speed" for the round-out. That allows me to arrest the descent with no power. In my previous two tries, I had "cheated" without thinking about it and added some power in the round out.

 

[Palmpilot]

He is referring to heavier aircraft that perhaps have lift issues that would not arrest a descent in that case. You also get into issues of insufficient elevator authority.

Once you're in ground effect, doesn't that reduce the stall speed?

If you watch the video I posted of doing this manuever in my Arrow, you see I "save a little speed" for the round-out. That allows me to arrest the descent with no power. In my previous two tries, I had "cheated" without thinking about it and added some power in the round out.

It's been a while since I have flown an Arrow, but what you're saying does not surprise me.

I would think the flare issue would have more to do with avoiding damage to the airplane than with saving your life. The former is a worthwhile goal when you're just practicing, but if the engine quits for real, then saving your life becomes the first priority.

 

[alfadog]

Once you're in ground effect, doesn't that reduce the stall speed?



It's been a while since I have flown an Arrow, but what you're saying does not surprise me.

I would think the flare issue would have more to do with avoiding damage to the airplane than with saving your life. The former is a worthwhile goal when you're just practicing, but if the engine quits for real, then saving your life becomes the first priority.

Well, what ground effect does is, to put it simply, increase lift and reduce induced drag on the wing so you create more lift at any AOA. CsubL is higher for any AOA. Therefore you can fly slower at a given AOA so yes, I guess it would reduce your stall speed because you would be flying slower at stall AOA.

As far as damage, obviously. If leaving the gear up in the Arrow means making a better (more survivable) landing then the gear stays up.

 

[Dan Thomas]

If you avoid a stall, how can you not have enough airspeed to flare?

Some airplanes will start sinking rapidly well before stall speed is reached. This is particularly true of short-winged aircraft, and sometimes of airplanes that have gained weight as the models matured. My Jodel stalls at 40 MPH but I sure don't go making approaches at 50 or 55; it will be sinking so badly and generating so much drag at that speed that a successful flare is unlikely.

Dan

 

[zaitcev]

If you avoid a stall, how can you not have enough airspeed to flare?

Very easy, try it in a Mooney M20E. If airspeed in that thing is below 80, you won't be able to flare. It just plops down and may hit prop. But it stalls quite a bit lower and flies below 80 just fine. I received an object lesson in it when I tried to check out in an M20E and the instructor confirmed it (I didn't wheelbarrow it badly though, even checked out after another couple of landings).

 

[gcd89]

Both of my main airports have either houses and trees straight ahead, or water...and then trees...then what?

 

[Morne]

Both of my main airports have either houses and trees straight ahead, or water...and then trees...then what?

How much runway? What is your max gross takeoff roll on a standard day with no wind?

 

[steingar]

Again, you brief before you take off. If there are no good options, you take the least bad (i.e. the one that doesn't hit any people). Been in the situation many times where an engine-out would have presented few good options. But I had already determined what those were, and would make a quick decision if such were needed. If you only have seconds and need seconds to figure out what to do, you are far more likely to do the wrong thing.

 

[Palmpilot]

There have been a couple of interesting discussions on this subject recently on the BeechTalk Forum:

http://www.beechtalk.com/forums/viewtopic.php?f=7&t=0&p=721497

Apparently you have to register to view the following thread. (It's free.)

http://www.beechtalk.com/forums/viewtopic.php?f=41&t=65762

 

[bbchien]

All those "one-up" on the forums-
"I did it in 600 feet!"
"I did it in 500!"

Whadda buncha garbage. The average GA joe needs 1,000 and he can still f-u really really good.