Chuck and Igor got me thinking.  Some successful West Coast airplanes have tail aspect ratios lower that what one would think they oughta, to wit about 3.4.  Thinking this was because of stiffness, a couple of us made some higher aspect ratio tails with extra-stiff structure. We figured they'd make the airplanes more stable.  The tail I made should have been 6% more effective in stability than the original Impact tail (effectiveness being lift curve slope * tail volume).  It would have been even more if I hadn't put the tips on backward and had to reshape them.  The results were underwhelming.  Putting vortex generators on one of those higher aspect ratio tails improved it so it worked about as well as the lower aspect ratio tail.  Supposing the lower aspect ratio works better, why do you reckon that is? 

Here is the 3D lift curve slope illustrating what Chuck was talking about in another post:

Then again, I found out that anecdotal evidence often trumped the opinions of the smartest, which seemed to be why the oldest ones often built a flying model of a design instead of relying on equations.

    Oh, foolish me, sitting here designing based on experience when I have page after page of calculations I could be doing instead.

    Brett

Well, that wasn't what I meant, but since my detailed explanation just got choked in a compatibility view issue with the software I'll just say never mind.

Methinks Brett was being facetious !!! 
 
Randy

Yep, putting something on the ends of the stab sounds like a good way to get some aspect ratio benefit without the disadvantages.   

Here's the data on some Impact stabs.  I think my 2008 airplane's stab was too small for its flaps.

Yep, putting something on the ends of the stab sounds like a good way to get some aspect ratio benefit without the disadvantages.   

Here's the data on some Impact stabs.  I think my 2008 airplane's stab was too small for its flaps.

The 2008 plane was the Orange one you had through last year?  Now I understand why PW said your flaps were too stiff; you had almost identical tail surfaces to his '92 plane.

 Howard, from the graph you posted it looks like there isn't any great improvement in Cl  once the stab gets up to 4.5 AR.  Not having a plane in front of me, a 3.4 AR tail might have a 23 in. span and elevators about 3.5 in chord.  The wide elevators would tend to give the plane a more solid neutral- you'd be able to feel the stick forces around neutral more than with narrower elevators.

Good luck.

Phil C

In the modern era of carbon fibre how hard is it to get the required stiffness regardless of the aspect ratio?

Not very I would imagine.

Methinks Brett was being facetious !!! 
 
Randy

Of course.

    But "cut and try" and learning from experience is perfectly legitimate as long as you are objective about the results of the "try" and have a basis for doing it. In our case, it's entirely mandatory, since no one has ever determined, in terms of engineering, what makes one airplane fly well, and another very similar airplane not so well. At this point we have better educated guesses than at any time in the past and can pretty easily determine what will *not* work well just by eye. But we are a LOONG way from actually knowing it sufficiently to start from scratch and and accurately predict mathematically what will be a good airplane and what characteristics it will have. Which is OK, because no one can quantify what characteristics we actually want.

    Brett

But we are a LOONG way from actually knowing it sufficiently to start from scratch and and accurately predict mathematically what will be a good airplane and what characteristics it will have. Which is OK, because no one can quantify what characteristics we actually want.

I dimly recall reading a book on sailplane design that was written in the 1970's, and was fairly famous for trying to set out a scientific, "be-all, and-all" approach for designing sailplanes.

What you got if you followed all the science (according to the review), was a sailplane that pretty much followed the trends of 1970's sailplane design, and failed to predict any innovations made since then.

So, yea -- we know where we are now, and maybe this actually is the pinnacle of achievement (or at least maybe this is what's currently in style).  But I suspect that anyone who starts from a "blank sheet of paper" is ether really starting from a decade-long collection of old design ideas, or is about to build a plane that has one or more serious drawbacks compared to current ships.

  With a lower aspect ratio tail the trailing edge of the elevator will be taller than the high aspect ratio tail even at the same angle of deflection.  This gives more leverage to the tall elevator trailing edge.

 Chuck said"Since stall recovery isn't much of an issue in stunt, I have one-to-one aspect ratios on my tails and wings."

Hi Chuck
Really? you have built stuntships with a 1 to 1 AR on the wing?  or am I misreading this?

Randy

Since stall recovery isn't much of an issue in stunt, I have one-to-one aspect ratios on my tails and wings. In my mind ( and that can be a scary place ) I can trade off a little TvC for AR and hopefully save a little weight...

That's why I wondered why people use such low aspect ratio tails.  I reckon it might be some aeroelastic thing or something about the laminar-turbulent transition wandering around. 

and then I add to the tail area for good measure anyway so go figure. 

That's because you are an S&C guy.  Have you seen Frank Williams's airplanes?

So, anyhow...

My theory (yes, Howard, it was a theory before I started building low AR tails...not after) was fairly simple.

Tails stabilize and or "direct" changes about the pitch axis of the airplane by producing varying amounts of lift which control the rate at which pitch change takes place.

As has been discussed, lift for a given area is influenced by more that just the area and the amount of camber modification which results from deflecting the movable surface.

Perhaps the most prominent of those "other things" is the aspect ratio of the surface.  High and low aspect ratio surfaces of a given area are capable of producing more or less the same amount of lift--and, as a result, redirect the pitch proportionate to the lift they develop.  Lots of lift, rapid pitch change; lesser lift, slower rate of change.

The thing that is differs between high and low ARs is the angle of attack required to produce the lift necessary for a "desired" rate of change.  By and large, high ARs produce that amount of lift at lower angles of attack (less deflection) and low ARs at higher (greater elevator deflection).  The primary source of the "different" angles of attack is the deflection of the elevator.  The low AR will require more deflection than the high AR.

As has been earlier discussed, the high AR will result in (all other things being equal) the desired rate of change at a lower angle of attack (less deflection, less control input, more sensitive response per unit of input).  It will be more sensitive in response rate in ideal conditions and flying in winds will require more "adaptation" to the changing airspeed in maneuvers and the inputs necessary to adapt will be more sensitive as well.  Put a 10 to 1 AR tail of equal area on your Impact or TP and the response will be dramatically different from the ~4.5 or so to one on the stock bird.  Yeah, you could probably trim the airplane and adjust handle spacing to make it usable but it would always be sitting there waiting to catch you unaware.

Now to the meat and potatoes of my "theory".

There is a big difference between producing the required amount of lift for a given rate of change from a low versus high AR.  That difference is the amount of drag resulting from the production of the same amount of lift.  I believe we can agree that lifting surfaces with a 30 to one AR (sailplanes) produce the lift required for a  given task while developing little additional drag.  They are very efficient and can adapt to significant different missions with very modest changes in the cost (drag) of doing so (and, by the way, ask little in the way of thrust to keep them functioning  efficiently).  Put that same area into an airplane designe for high speed flight and, when slower flight is demanded, for landing, etc. the drag goes up dramatically and lots of power and high angles of attack are necessary to keep in functioning as required.

When these effects of aspect ratio are applied to tails the drag produced by low versus high ARs to produce the lift required for a rate of pitch change varies significantly.  Little drag increase for a very high AR tail and lots of drag for a real low AR.  In the days of Fox .35s increasing drag may have been an issue.  In today's stunt world, much less so.

There is an additional effect from increasing the drag per unit of lift from the tail by lowering the AR.  Doing so moves the Neutral Point (the point at which all of the aerodynamic loads acting on the vehicle are "centered") of the vehicle aft (like adding more feathers to an arrow) which simultaneously increases the static margin (the distance between the CG and the NP making the vehicle "more" stable in pitch.  The greater the pitch rate commanded the greater the lift, the greater the drag from the tail and the greater the aftward movement of the NP.  The result, in my theory, is an airplane that is more stable in pitch response despite an aft CG that allows the aggressive maneuvering required to wing toy airplane "trick" events.  It is much easier to fly into and out of a corner because the airplane's response will be less "twitchy" for a given rate of change than one with a very high AR tail.

In "Ted's theory" there is a certain element of the holy grail in finding an optimum AR to the tail and my feeling is that something lower than the AR of the wing is probably "better".  As in all things stunt, I think extremes must be avoided because we must combine aggressive response with controlled precision.  It's easy to make a stunt plane turn quicker than you can fly it competitively and one of the ways of doing so is to use "very high" AR tails which will produce explosive amounts of lift with very modest input.  Great if you've got Superman's reflexes but not a viable source of trophies for the mortals among us.

All of the talk about hinge loads are valid, by the way, and militate against excessive use of super low ARs.  Enough line to tension to deflect very low AR, large area elevators against the airloads is a real issue and can't be ignored.  Other things that will effect the "optimum" will be flap size and deflection and the negative pitching moments which result that must be overcome by the tail.

At any rate.  That's more or less the thought process that led me to use lower AR tails.   I've been very comfortable over the years with the TP series which, I feel, provide very predictable and controllable responses under a wide range of flying conditions.  I think the modestly low AR tail plays some role in that.

 Ted

Igor, not sure I agree with all of that.

The slope of the lift curve depends on the Aspect Ratio, and the zero-lift angle of attack depends on the flap deflection.

In other words, we determine the Lift Curve with the flap deflected, and then we adjust for aspect ratio.

The lift slope is simply the Lift Coefficient plotted as a function of angle of attack for whatever configuration ( flap deflection ) of the airfoil we choose.

The Aspect Ratio is the variable we need to convert the 2D flow from a test section to 3D flow on a wing or tail.

We care because it is the dynamic stability, not the static stability that matters to the pilot.

So, *Theoroetically* for a given control response (moment) if we raise the tail's Aspect Ratio we could lower the Tail Volume Coefficient, save a little weight. Another factor is that we would have more stability in turbulence, since any gust that affects the flow field around the wing will create a bigger effect on the aircraft's total moment contribution from the tail (because the wing's downwash angle changes due to the gust and the lift curve is steeper for the higher AR tail.)

But the nice thing is, it doesn't cost millions of dollars for us to build a prototype and test, so as controline pilots we have an enormous database of experimental results to choose from!


<<Edited to fix something I said that I'm not sure was right. >>

Oh Igor, we agree.

All this is nice to think about, but there is no substitute for a well-built model being flown by a skilled pilot.

Chuck

Howard, the chart do a good job of showing the 3-D effects.

Yes, the stick-fixed neutral point will move by the Tail Volume coefficient times the ratio of the lift slopes for the AR's being compared... but that's static stability.

You can balance a marble on the head of a pin and have it be statically stable but not dynamically. Same thing happens with airplanes.

Dynamic stability is what's important, and the nice thing about that is that dynamic stability guarantees static stability.

The higher AR tail should offer more dynamic stability, since the slope of dCm/dAoa  for the whole airplane will be more negative due to the steeper dCl/dAoa for the higher aspect ratio tail.

That means, that when you hit a gust or turbulence the ship with the higher aspect ratio tail will recover faster than one with a lower AR tail - everything else being equal... I think... Need to go have a beer on that and work it out.


<<edited to change "Be more stable" to "recover faster" wrgt to the stabilty and aspect ratio of the tail. >>

Dynamic stability is what's important, and the nice thing about that is that dynamic stability guarantees static stability.

Has anybody tried to estimate the dynamics of a stunter?  I know that guys are really good at feeling the dynamic response and trimming to get what they want.  It would be interesting to see that quantified.  

Maybe what we need is a Cooper-Harper scale for stunt ships.


http://en.wikipedia.org/wiki/File:NASA_diagram_of_Cooper-Harper_rating_scale.jpeg

This was developed the same year I was born and is still the "Gold Standard" for aircraft handling.  I have been associated with Calspan engineers with the TIFFS and VISTA pltaforms, and they still use it.

Maybe we can modify it for stunt ships and create a databse.. of course on a secure server with US access only, LOL!

I propose we call it the Rush-Smith scale.

Funny aside: The VISTA F16 was re-engined from a GE engine to a Pratt and Whitney with a 2D vectoring nozzle. It was the fastest moving "control surface" on the aircraft". I was in the Calspan hangar one night, and they were showing films of the "Cobra " manuever... and suggested that with that nozzle the verical fin and rudder was redundant with the right control system. I suggested they simulate removing the entire asssembly and see how it flew. The answer I got was "What if you lose the engine?" I replied "Feet and knees together, eyes on the horizon."

It was poorly received.

Still, imagine the confused look on some poor Sukhoi driver when a Viper pulled up, flipped and frisbeed around to get a firing solution!

Chuck