how much throw for all-flying stab?

[hul]

how much throw do I need for an all-flying horizontal stabilizer on a 700mm span Hawker Hunter EDF?
I have currently set it at +/-7 degrees, which looks like a lot. But the wing (S3021) probably needs more than 10 degrees to stall.
All advice appreciated.

Thanks, Hans

[fhhuber506771]

+/- 7 deg is a lot for a high speed model... But for some low speed models it may not be as much as you want.

Distance from the center of lift and size of the control surface will affect how much throw you need.

Too many potential variables to say if your case needs more or less. If the plane is ballanced and proportions are reasonable, then +/-7 will be able to fly it... but might be either touchy as the devil or only be enough to make smooth laps around the field.

[hul]

these are the proportions. I have dual rates set at 50% (+/- 3.5 deg) in case it's touchy but was worried about not being able to slow it down if CG is a bit forward.

Thanks, Hans

[fhhuber506771]

calculate the CG... and get it on 30% MAC. the +/-7 deg should be adequate for flight.

Having the dual rates set at 50% of full may be a very good plan...

Tailplanes look proportionately a little small (scale size? Often tailplanes are fudged to a bit larger because of scale effects)

[yoyoML]

deflection = alpha * ( A2*d2 - A1*d1) / ( A2*d2 )

This formula relates elevator deflection to main wing angle of attack, where A1 is main wing area; d1 is distance from CG to main wing's AC; A2 is elevator area; d2 is distance from CG to elevator's AC; deflection is the difference in angle between main wing and elevator.

(A2*d2 is simply the tail volume)

I assume straight-line, symmetric CL curve for both elevator and main wing, with the same slope, and also one wing does not affect another. Since your plane is jet-like and the wings have similar shape, I think the assumptions are good.

The result is quite easy to grasp: main wing alpha is proportional to elevator deflection.

If you know at which alpha the main wing stalls, you can calculate the deflection when it stalls. Try with 100% stall deflection and adjust in relation to that stall deflection.

[HarryC]

You will be surprised how much travel an all moving tailplane needs. It will look huge in comparison to what you would have with a conventional elevator. For example my jet Sabre has an all moving tail, flies very fast, yet has a lot of movement and is not sensitive, and is struggling to hold the nose up to flare on landing. The full size plane I fly has an all moving tail, and it also has what looks like a lot of travel.
AMTs need a lot of travel because they only change angle of attack. An elevator changes the AoA and the camber of the tailplane, and the change of camber alone produces a large change of force.

Do not put your balance point at 30% of MAC, you must do a proper calculation for that model. I have models with balance points ranging from 22% MAC to 50% MAC and if flown at 30% MAC would be unflyable. 30% MAC is an average of what other aircraft use and it tells you nothing about what your plane needs, it is not a scientific rule.

H

[Sparky Paul]

As Hans, and yoyo and Harry really fly these things... pay attention to them.

[hul]

I did a calculation using Gordon Whitehead's formula, CG is at about 23%.

The wing's airfoil stalls at around 11 degrees, the wing probably needs more than that because of sweep before it stalls. Does that mean I need more than 11 degrees of throw to get the wing to stall? Or in other words, is deflection of the stabilizer roughly angle of attack of the wing?

Thanks, Hans

[JetPlaneFlyer]

Does that mean I need more than 11 degrees of throw to get the wing to stall? Or in other words, is deflection of the stabilizer roughly angle of attack of the wing?

Thanks, Hans
There is no simple relationship between stall angle of the wing and optimum deflection of the elevator/stabiliser; in fact the two are not really linked at all. The sensitivity of the elevator is dependant on many variables, CG position and flying speed are two examples but the stall characteristics of the airfoil is not one of them.

[Sparky Paul]

Flying tails use remarkably low amounts of deflection for full control.
With swept wing models, there's also the problem of getting too slow. Swept wings stall at the tips first, which would normally be countered by washout.. but that would work against the plane at speed.
Never let it get slow.
Start with small deflections and keep the speed up.. Stalls... try to never stall.

[hul]

Stalls... try to never stall.
clearly, but minimum speed is close to stall. I thought swept wings behave more like deltas, that don't really stall at all. Built in 2 degrees of washout anyway.

I moved the mechanical stops to +/-12 degrees. I'll probably use the full 12 degrees on high rates, about 4 degrees on low.
Yoyo's formula gives alpha = deflection in my case (because CG is close to MAC, therefore A1*d1=0).

First swept wing plane and first all moving tail for me...

Thanks, Hans

[Sparky Paul]

They stall... and frequently that's the -last- manuver they do!

[HarryC]

My English Electric Lightning stalled once, about 3 feet up. Paul is correct, the tips both stall together resulting in a massive nose up pitch that full down elevator can not stop. The Lightning went vertical and then reversed backwards down to the ground.

H

[yoyoML]

clearly, but minimum speed is close to stall. I thought swept wings behave more like deltas, that don't really stall at all. Built in 2 degrees of washout anyway.

I moved the mechanical stops to +/-12 degrees. I'll probably use the full 12 degrees on high rates, about 4 degrees on low.
Yoyo's formula gives alpha = deflection in my case (because CG is close to MAC, therefore A1*d1=0).

First swept wing plane and first all moving tail for me...

Thanks, Hans

I think you mean the CG is close to "AC", as in "aerodynamic center".

Then deflection=alpha is correct for your plane, and for all cases where A1*d1=0. For example a dart which has A1=0.

Ahh... one point: if CG is before main wing AC, d1 is negative.

But for a conventional config, you really should put CG somewhere behind main wing AC, but before neutral point. CG on AC is likely over stable, and requires a bit too much elevator deflection to trim level flight -> a bit more drag. It is a safe starting point nevertheless.

[hul]

CG is ahead of wing's AC because I moved it forward by 5% to compensate for the long nose. I'll move it back if possible after a flight test.

I'll definitely try a stall at a very safe height.

Thanks again, Hans

[BMatthews]

FWOW I found on my gliders that if I used about the same amount of travel at the trailing edge of the all flying as I would if it was a conventional elevator that it worked out nicely.

It sort of makes sense if you think about it. The all flying works by altering the angle of the stabillator. The elevator works by altering the angle of the overall airfoil formed by the stabilizer and elevator. So making it so that both move their centerlines the same should give the same results.

[JetPlaneFlyer]

So making it so that both move their centerlines the same should give the same results.

Not really because a the tailplane with the elevator changes it's camber as well as changing AoA. A cambered surface will produce more lift than an un-cambered one given the same AoA... So a all moving tail 'should' need more movement than one with a conventional elevator.

That’s the theory... I can’t argue with your actual results as I've never done any back to back testing on the same model.... Bear in mind though that as an all moving tail pivots at a point just forward of centre (and not at the LE) then measuring deflection only at the TE is very misleading. Actual change in AoA is much greater because the LE moves up as the TE moves down and visa versa.

[biber]

There is no real difference between an all moving stab
and a conventional one in terms of flight mechanics,
as long as both stay within alphas where flow is fully attached.
And with a particular given tailplane shape all that counts regarding the manouvering
is the state of the tails zero lift angle relative to the wing.
Thats easy for an all moving stab, since the zero lift angle just moves with the whole tail.
For a conventional tail there are some more factors involved in the connection
between the elevator deflection angle and its effect on the zero lift angle of the tail.
But Xfoil or e.g. Profili (wich incorporates the xfoil code) are nice tools to calculate just that.
Keep in mind that the more aft portions of the airfoil have the biggest
influence on the zero lift angle, the nose does count much less.

biber

[Ollie]

In models, the AC is moving, not perfectly fixed, with angle of attack. It shifts with turbulent bubbles. The Cm is not straight with angle of attack. So the pitch stability has to be dependent on shape of the plan plus AC shifting position with angle of attack. Every area ahead the CG is destabling and every area aft the CG is stablizing. Be careful from too much nose area ahead the CG.

[Troy]

To confuse you more, I had built a camber changing full flying stab on one of my slope racers years back... basically an elevator on a full flying stab. I wasn't able to determine if it had any major improvements over a full flying stab in hard pylon turns, but it sure looked cool. However, my full flying stabs with reversed camber airfoils did pull better in the turns with noticeable energy retention.

[hul]

I flew the plane, ended up using +/-6 degrees to make it fly the way I like it (I prefer reasonably high throws).

Thanks again for all your inputs, Hans