Flying tail... how not to...

[Sparky Paul]

Seen at the SAE Aero Design West Competition this weekend.
Open class entry.. geared 1.20 motor.. probably 30 foot span, composites par excellence.. fabulous! up until it flew.
A limit cycle in the horizontal from takeoff to crash.

[Sparky Paul]

Close-up with dimensions (from the image):

[AndyKunz]

Can you explain the "why" please?

Thanks.

Andy

[William A]

Ya, me too Sparky, I'm pretty ignorant on these full surface control setups.

My unexperienced eye's see 2 problems, one being the distance of the pivot point or MAC being to far back.

The second I'm not sure about, but wouldn't it have been better to have the pivot point at the top of the post, next to the control surface ?

[Sparky Paul]

Originally posted by William A
Ya, me too Sparky, I'm pretty ignorant on these full surface control setups.

My unexperienced eye's see 2 problems, one being the distance of the pivot point or MAC being to far back.

The second I'm not sure about, but wouldn't it have been better to have the pivot point at the top of the post, next to the control surface ?
.
William, you got it!
Check out Mark Drela's flying tails on his Allegro etc to see how it's done.
Had the pivot point been -in- the surface, and not where it was, and forward to the 25% mac, the plane would have flown, and probably well, although the geared motor wasn't all that impressive.
The surface demonstrated a continous "limit cycle" in flight, going to the limits in each direction of what it could do. Why the servo case didn't break is a mystery.. the dollar two-ninety eight DuBro ball-links might have had something to do with the longevity, short as it was.
Note the Lockheed Jetstar and U-2R use a tilting horizontal with the vertical mounted to that.. (the white stripe at the leading edge of the vertical on U-2R photos).. but that system is like that which Mark uses; hinge at the surface, seperate actuator attach point forward to the surface.

[Sparky Paul]

Looking at a different view, the hinge for the horizontal is about where it should have been. The wimpy pushrod to the servo, and the wimpy horn weren't up to the loads..

[JRuggiero]

Frankly, I don't see the benefit of all-flying stabilators on model airplanes. There's a lot of stress on that hinge from aerodynamic loads. A conventional stab-elevator, glued to the tailboom, works just as well.

Jim

[Sparky Paul]

A flying tail has the benefit of streamlining to a trimmed position. No offset between the horizontal and the elevator. (The surface won't be cambered, IOW.)
And it can be smaller than a conventional type.
PROPERLY set up the loads on the servo are small.
Looking at the inset image in the photo above is a Drela flying tail installation, improperly implemented.
The hinge axis being outside the surface when it must be on the chord line is the cause for the limit-cycle and crash of this splendid airplane!
This creates a serious aerodynamic moment around the hinge axis which the oh-too-weak horn and pushrod couldn't prevent, permitting the oscillation, which was fed by the propwash.
.
This was a dangerous situation.

[JRuggiero]

"A flying tail has the benefit of streamlining to a trimmed position. No offset between the horizontal and the elevator. (The surface won't be cambered, IOW.)
And it can be smaller than a conventional type.
PROPERLY set up the loads on the servo are small."

Sparky Paul -- thanks a lot. That's a great explanation, and I am enlightened.

Jim

[Salto]

OK, so if what you want as a high priority is absolute minimum drag for different trimmed positions then the all flying tail is optimum.

But one advantage of a conventional stab/elevator is that the cambered airfoil that is produced when the control surfaces deflect can reach a higher lift co-efficient than the all flying design can. So control authority will be higher.

I guess this is why Mark Drela's super efficient sailplanes use an all flying design, but low speed electric 3D models use conventional stab/elevators. Horses for courses!!

Graham.

[Sparky Paul]

Originally posted by Salto
OK, so if what you want as a high priority is absolute minimum drag for different trimmed positions then the all flying tail is optimum.

But one advantage of a conventional stab/elevator is that the cambered airfoil that is produced when the control surfaces deflect can reach a higher lift co-efficient than the all flying design can. So control authority will be higher.

I guess this is why Mark Drela's super efficient sailplanes use an all flying design, but low speed electric 3D models use conventional stab/elevators. Horses for courses!!

Graham.
Absolutely correct!
For least drag a flying tail works best, but you'll notice their absence on manuvering airplanes.
You get higher Cls from a hinged surface.

[FlightofSong]

So how long does a 1.20 run on 2 oz. of fuel?!

[Sparky Paul]

The fuel restriction (4 oz of fuel used regardless of motor size or number of motors, with a penalty for using more) went away after the 2000 contest here in Lancaster when it was determined this requirement for the Open class planes caused accidents.
There wasn't enough fuel to permit a safe flight pattern.
Not being in the business of causing accidents, the restriction has been removed.

[Ray Lowinski]

So on this setup would it have been any better to connect the servo/rod to the TE? It seems like a bit less work for the servo.

[Sparky Paul]

With the hinge axis at 25% mac, the load on the servo is low as symmetrical airfoils have no pitching moment, and will not move with changes in constant airflow, but would respond to the propwash, for instance.
Further forward isn't bad, but aft of 25% is BAD!
The surface will want to depart from a streamlined position to one extreme or the other.
The servo in this case was a high-end metal geared type.
It held the load.
The elasticity in the pushrod and horn permitted the surface to flip up and down and provided a somewhat centering force, keeping the surface from locking hard up or down. It was awesome to see a surface of that size going stop-to-stop!
Videos show the plane pitching in response.

[markdrela]

Pivoting the stab at or ahead of 25% MAC is clearly important to prevent static instability and potentially catastrophic divergence. But this may not be sufficient. For such a large all-moving tail I think it's also important to mass-balance it about the pivot. Otherwise it will be susceptible to *dynamic* instability (aka flutter), just like an unbalanced aileron. The aileron is a more severe case because the wing ahead of it effectively amplifies the airloads resulting from the aileron's wiggling, while the stab is on it's own. Still, balancing the stab, even partially, is certainly prudent.

[Ray Lowinski]

OK, I guess I do have a glider in the building stage that has the flying stab. It is around the 25% mark and operated by snake to a horn.

To prove the point I took an old stab out in the wind (aka house fan) and played with pivot points at different %'s. This works great with a piece of Coroplast with a skewer through the holes. Back at the 40% area it started to "hunt" for a nice position with little adverse input from me.

Keep 'em hummin'

Ray

[Sparky Paul]

There was another plane there with classic flutter. Also a flying tail. And a 48" snake from servo to horizontal. Probably 1 inch of free play up and down.
The tail boom itself although quite wide in diameter was understrength, which contributed to the scary in-flight oscillations which didn't go away when it was reinforced, just changed the speed band where they occured.
I told suggested they change to pull-pull, which they did. The plane then flew well...
I've seen flutter on full-scales remove the surface involved in less than 1/2 a second!

[William A]

"I've seen flutter on full-scales remove the surface involved in less than 1/2 a second!"

Hmm.....that can't be good.