Stabilizer/Elevator Sizes - Flapless Designs

LAST EDITED ON Feb-28-04 AT 01:29 AM (CDT)
 
Hi, All-

I've read everything on SSWF over the past two years and from earlier archives on this subject, but would like any last-minute input from anyone who hasn't posted or whose ideas have evolved further. I'm backing off some on a couple of the more radical features of my little LA .25 powered, flapless, profile stunter, including stab/elevator areas, but am ready to cut more balsa as soon as I get some better contest stock.

Here's what I THINK I understand about this.

1) A tradeoff exists between elevator area and deflection in order to achieve requisite camber for turning.

2) Elevator deflection increases lift across the entire stab/elevator area along the span of the hinge, and further, unless fenced off from any remaining stab area.

3) As tail arm lengthens, longitudinal moment of inertia (square function of c.g. distance) increases faster than pitching moment (linear).

4) As tail arm increases, stabilizing (counter to turn) airflow angle increases at the tail (Zaic's 'circular airflow') during turns.

5) Uncambered flapless wings require no additional stab/elevator area and/or deflection to counter flap-induced negative wing pitching moments in initiating turns.

6) Empenage drag is stabilizing, but positely pitching with a raised stabilizer. A slight, positive stab aoa may be beneficial, depending on the sum of other pitching moments, notably positive-pitching gyroscopic precessions from prop rotation and circular flight and negative moments from l.g.

7) Flapless designs need a more forward (15%-16%) c.g. than flapped designs.

8) As elevator chord increases (hingeline moves forward) for a given deflection, tail aoa increases along with camber.

9) Finally, as stabilizer chord increases ahead of the elevator, more lifting area is provided, but less % camber and aoa are achieved for a given elevator deflection.

What this seems to indicate is that more elevator area/deflection and less stab area are needed for longer tail moments. This makes the TVC coefficient less meaningful, although I gather that values of .40 - .45 are "good" and that increases are not at least immediately counterproductive.

Immediate questions that arise include the following:

1) Apart from questions of elasticity, inertial moments, aerodynamic resistance to turns, and weight, are there any other upper limiting factors on useful TVC's. (Edit: Of course I forgot to mention overdriving and stalling the wing. I am thinking though of large TVC's with limited pitching moments on my 19% thick wing with a somewhat blunt l.e.)

2) Can fenced-off, non-flapped stabilizer area be made to enhance neutral stability more than it hinders turn performance?

3) Besides expocranks, blanketed elevator area, and flaps, are there any other schemes out there to enhance neutral stability or damp pitching around neutral?

Still aiming at a relatively tight turn radius without making a combat plane of it, I'm tentatively interpolating nonlinearly between published extremes, as I lengthen the tail arm a bit. However, I lack data on well-behaved large flapless designs, and I'm kind of in the dark on what happens to the optimum hinge location (camber effects), although the elevators on Streaks and other such models are a smaller percent than on longer models (i.e. at what rate do elevators become a smaller percentage of the horizontal area, as tail arm decreases?). Have I missed anything?

SK

Serge,
I can't answer many of your questions with the sort of data you request. However, I designed this plane with inspiration from flapless design threads by Brett and Ted, a picture of the Hooptee, a Gothcha foam buck and some patterns from a Lightning Streak. It flies VERY well- it tracks beautifully and is capable of very tight turns.
Regarding horizontal tail design parameters, I followed the recipe by Brett and Ted: the total area is 20% of the wing, elevator area is 40% of the total and the aspect ratio is 4.5. This is set on a longish tail moment.
The balance point is presently at 15% MAC and I feel it can be moved further aft and still maintain good directional stability. Your 13% figure may be conservative.
I consider the design a success and there are very few things I would change if I built another. If you follow their recommendations, your design should work well.
Derek Moran

Derek-

Wow, I didn't expect a response so soon! I just restarted the computer to edit an obvious omission from my post and saw your thoughtful response and really nice looking model.

Although mine started out as a rather radical pushing of the standard-configuration "envelope", I have redrawn the moment lengths and horizontal tail to a more conservative departure from those recommendations (I decided to make the plane more likely to succeed). I'll probably still try to push the limits of my experimental wing a bit though, and that's what prompted the post. It's still an experiment, and I just wanted to know whether there are any other things I should be considering.

Thanks for the input.

SK

P.S. My c.g. was computed at 16%.

Derek-

I forgot to ask: what is your wing's aspect ratio, and out of curiosity, how long are your MAC and tail arm? Thanks.

SK

Hi Serge,
Wing span: 52.50 (inches)
Wing MAC: 9.25
Wing area: 480 sq in
Wing AR: 5.67
Thickness: 16% root, 17% tip
CL line is straight (CL=MAC/3)
Tail moment: 19.50 (CL wing to CL stab/elevator)
Engine: OS .25FP
Flying weight: 28 ounces

If you have any more questions, I'll be happy to help.
Derek

Derek-

Thanks much for the data. My wing is the root of my little experiment and as such differs some from yours - although the shapes are quite similar for reasons that will be apparent. The remarkable thing is how otherwise similar our numbers are.

First, my wing is really high in aspect ratio and thicker, with a l.e. radius of about 7/16" and .24-chord points lined up spanwise. Tips are like yours, but raked outward a bit more. Here is how the final measurements compare to yours:

>Wing span: 52.50 (inches)

Mine, as finished: 56"

>Wing MAC: 9.25

9.16 (may increase slightly with t.e. treatment)

>Wing area: 480 sq in

476 (now) - 495 (depending on t.e.)

>Wing AR: 5.67

6.3-6.5

>Thickness: 16% root, 17% tip

19%/20%

>CL line is straight (CL=MAC/3)

Straight, except I use .24 (MAC)

>Tail moment: 19.50 (CL wing to CL stab/elevator)

19.29 between quarter-chords of MAC's

>Engine: OS .25FP

OS .25 LA

>Flying weight: 28 ounces

I wish! The wing minus center sheeting is already 7 3/8 oz.

My horizontal tail is of course not finalized, probably shrinking some more though to save weight and aerodynamic twisting moments, but also in response to interpolations and suggested "numbers" from SSW. Last week's was about 20% plus some unflapped, fenced tip area. So that may end up comparable too.

Again, thanks for showing us your fine example!

SK

I will be interested to learn if you can hear the wing tips in maneuvers. We built 3-4 combat planes with a similar tip. Very easy to do on foam, hand cutting a tip that fit the tip of the main panel and tapered sharply with a 3 in. span, very similar to yours, I think. We could hear the air roaring around the tip in a tight maneuver, over the noise of the unmuffled engine. Must have wasted a bunch of energy, because these planes lost a lot of speed in tight maneuvers. The culprit seemed to be the sharp transition from the wing leading edge to the angled tip.

A squared off straight tip works just fine.

The rounded "Arrowplane" tip seen on many combat planes also works well. If it is smoothly rounded and taperd it has the least drag and seems to add a touch of stability to a flapless plane in maneuvers. Might just be due to the slightly further aft MAC though effectively making the nose a bit longer.

Hmmmm...this would be disappointing. I configured these tips according to some old NACA or NASA data (misplaced) in order to diminish the induced drag - not increase it - by inhibiting flow around the tip. The data indicated that with a certain tip-span to chord ratio, losses mere minimized. When I drew this up, I approached their spec. These are heavier than planned and not my favorite shape. Darn! Radius is something like 1.3+" and, yes, the tip span is just a bit over 3.0". If they do work out that way, I'll post the result and chop them off...assuming they haven't whistled their way into the fairgrounds turf. I might as well proceed with them now, but with a wary eye. Thanks.

SK

Wow! What a beautiful structure, Serge.
It's clear you wanted to prevent warps in the high aspect ratio wing. I took a different approach. I wanted a light weight, quick-building structure. It looks a lot like an Eastern European combat wing, but I used a molded leading edge and a full-depth spar. The covering is MRL polyester tissue (polyspan). It was built straight and I haven't had any trouble with warps. A picture of construction is attached.
My tips are a caricature of Dr. Hoerner's and are commonly used on sailplanes. They are quite different from the tips Phil describes and I am convinced they are efficient (though, as they say, without data, you're just another guy with an opinion).
I was concerned that my wing might be too thin (16% root, 17% tip), but it has proved to work fine. It could probably be thinner. The plane is light and does not require the extra lift and strength a thicker wing could supply.
Derek Moran

Thanks, Derek -

However, I think I overdid it through inexperience. I chose 19% for structural reasons and would probably go to 18% or lower now, after actually seeing the resulting 3-D object in my hands. I wanted it light, but put too much into it; the diagonals, outer bay sheeting, and 20 l.e. stub ribs were afterthoughts. Now the thing is more rigid as-is than any covered wing I've built. It will have to break to twist significantly. I'd try to save the weight back by learning to do plastics, but I at least have to reinforce the center. The spar is "bullet proof", but there are stress concentrators everywhere else. Yours definitely looks like the right approach; although I'm looking forward to flying mine, I may just follow your lead next year. Thanks much for the pictures.

SK

I wanted a light
>weight, quick-building structure. It looks a lot like an
>Eastern European combat wing, but I used a molded leading
>edge and a full-depth spar. The covering is MRL polyester
>tissue (polyspan). It was built straight and I haven't had
>any trouble with warps. A picture of construction is
>attached. Derek Moran.

Derek,

Looks like the molded L.E. and full-depth spar gave you enough strength to get by with so VERY little wood in the ribs and T.E. and end up with the super-light overall weight. Not needing an L.E. capable of supporting flaps certainly doesn't hurt.

How thick is the L.E. wood? A single-layer molding? How about the thickness of the full-depth spar? Is there anything inside the L.E., like say, ribs, for example, or is the molding strong enough all by itself?

Very ingenious wing, and easy on the eyes. I've been drooling over the Hooptee myself, and read in Stunt News a while back that Kilsdonk himself built a replica of the '63 design a few years ago, shortly before his passing, and even with the 1/4" thick "poor man's" I-beam spar, the wing flapped like a bird in the super-tight corners it produced. Looks like your molded L.E./spar unit is giving you a whole lotta bang for the ounce, as well as being easy to build once you have the buck in hand. Thanks for letting us in on it.

Ditto Cyberflyer on the pleasures of constructive group communication!

Oh, how great minds think alike!
I'm right in the middle of my flapless stunt development, which is an extension of my greater body of stunt design experiments.

All this talk of Hooptees, forward swept trailing edges and Derek's construction method seem a little like Deja Vu, as my creation bears various resemblances to each. My new experimental wing method is quite similar in concept to Derek's work above and holds the same combat-esque origins. I'm doing an all balsa one first and then a foam/glass/carbon one to test a number of ideas.

I'm going with 610 sq in wing, perpendicular 1/4 chord line, blunt/forward aerofoil, 5.1:1 Wing AR, 4.6:1 Tail AR, 23+" MAC-MAC tail moment at 20% area with a 60/40 elevator split and probably a 40FP up front. Mine will have a Yatsenko style removeable wing too, which will present a little weight challenge for me to contend with.

A.

Serge, take a good look at the Flite Streak moments. A longer nose, of course, but the areas work. It is a tight turning aircraft and with the right engine and tank, is great doing the pattern. A 15% larger plane with the longer nose, who knows?

LAST EDITED ON Feb-28-04 AT 09:38 AM (CDT)
 
Hi, Ty-

I actually did scale the Streak up to about 500 in^2 at the same time I drew this one up. I almost built that instead, but opted to go out on a limb. I am using it as my short-end standard for comparison. Knowing now how slowly I build (so far), I'm easing up a bit to make this one more useful. It won't be as adaptable as a Streak, but it will still test some fringe ideas.

Serge

Serge:

I didn't think fine arts guys could write like that!

Curt-

They let us out once in a while, when they don't think we'll cause too much trouble!

Serge

>1) A tradeoff exists between elevator area and deflection in
>order to achieve requisite camber for turning.
I have found that when using a stab/elevator the main concern is balancing the control forces for the speed of the model in order to get a "good" control feel. The relative chord of the elevator to the stab and the airspeed affects this alot. WB Netzeband has an procedure in his articles which allows you to calculate the pushrod force for a given tail configuration. It works well. The 40% elevator figure from Brett and Ted works well. Limiting the elevator to about 3 in. chord works well with a 3 in. bellcrank and 1 in. horn. The speed difference between the dive in the wingover and the speed in square loops is enough that it can be a problem.

>3) As tail arm lengthens, longitudinal moment of inertia
>(square function of c.g. distance) increases faster than
>pitching moment (linear).

This doesn't seem to be a problem for any reasonable scale("looks right") tail moment. I am using the following on a very nice flying plane.
Kt= .28
tail moment 20.75 in.
tail area 124 sq.in.(16.7%)
wing area 735
wing chord 12.25
C.G. .51 ahead of MAC (4%)

The following plane, while it flies nicely, is not as stable or pointable:

Wing Area 740 Span 60 in. Avg. Chord 12.25
Stab 108 sq.in. (14.6%)
tail moment 21.5
Kt=.256
balance .61in. ahead of MAC 5%of avg. chord

We'll see if a 20% tail works any better soon.

>4) As tail arm increases, stabilizing (counter to turn)
>airflow angle increases at the tail (Zaic's 'circular
>airflow') during turns.

A gap of at least 2/3 of the average chord is needed between the wing and stab. Shorter and you run into all sorts of erratic control effects. Over 1 avg. chord makes it very difficult to balance the plane with a "nice" looking nose length.

>6) Empenage drag is stabilizing, but positely pitching with
>a raised stabilizer. A slight, positive stab aoa may be
>beneficial, depending on the sum of other pitching moments,
>notably positive-pitching gyroscopic precessions from prop
>rotation and circular flight and negative moments from l.g.

Not enough effect to bother about. Just make sure the stab does not have negative incidence. Trim the turns by shimming the stab incidence after test flights, if you can't get it to fly right by adjusting the elevator travel.
>
>7) Flapless designs need a more forward (15%-16%) c.g. than
>flapped designs.

The above planes are ~ 20% and fly fine. More nose heavy can work too. You just have to balance the control size, chord, and travel to suit. I let several flyers with varying experience fly the planes and they did fine after just a few laps to the point of flying the pattern.

>What this seems to indicate is that more elevator
>area/deflection and less stab area are needed for longer
>tail moments.

This is correct. Lots of combat flyers have found this.

> This makes the TVC coefficient less
>meaningful, although I gather that values of .40 - .45 are
>"good" and that increases are not at least immediately
>counterproductive.

It takes a careful balance of the elevator chord, the TVC, the balance point, and line spacing to get "nice" feeling, well-harmonized controls.

>1) Apart from questions of elasticity, inertial moments,
>aerodynamic resistance to turns, and weigh, are there any
>other upper limiting factors on useful TVC's.

Mainly a big, effective tail will require a very long nose unless you use a very big engine. Take a look at Netzeband's BareCat!!!

>Still aiming at a relatively tight turn radius without
>making a combat plane of it,>

Set the wing span to limit the speed loss in maneuvers that you are after, add chord to get the turn radius, and limit the weight to help both. To get a turn radius comparable to a current good flapped plane you need to keep the wing loading under 9 oz/sq.ft. or use a gonzo 4 stroke for nose weight and pulling power to haul it through turns at a much higher angle of attack.

> I'm tentatively interpolating
>nonlinearly between published extremes, as I lengthen the
>tail arm a bit.....

build a prototype and cut and hack it to suit. The second plane mentioned above, I sliced half an inch off the elevator width right on the field to reduce the control loads, then went home and spliced 2 in. onto the fuse to extend the tail moment.

Thanks much, Phil, for sharing your experience and data - valuable food for thought.

SK

LAST EDITED ON Feb-28-04 AT 01:53 PM (CDT)
 
>Hi, All-
>
>I've read everything on SSWF over the past two years and
>from earlier archives on this subject, but would like any
>last-minute input from anyone who hasn't posted or whose
>ideas have evolved further. I'm backing off some on a couple
>of the more radical features of my little LA .25 powered,
>flapless, profile stunter, including stab/elevator areas,

I wish that Brett and Ted would weigh in here, as they have a lot more experience on this than some of us. Phil Cartier's comments should probably be carefully considered, as few have more handle time with flapless designs than he does. Here are a couple of additional observations I have noticed:

1. The Flite-Streak is a great ship. It is, however, a little "twitchy." For a flapless ship to "groove", you need a stab/elevator combination that is large enough to do provide the stability needed. Notice the small area of the stab and elevator on the FS. For its intended purpose back when it was designed, the FS was and is wonderful. It is still a good stunt trainer, but can be a challenge to keep "smooth."

2. The bigger the stab, the bigger the elevator needed to turn the ship. Smaller stab, smaller elevator. My personal experience, and this is very, very subjective, is that keeping the total stab/elevator surface at 16-16.5% of the wing area is optimal. I find that keeping the stab/elevator area ratio at 60%-40% to be a good starting point. More area requires either a longer nose, more nose weight, or a shorter tail moment. I have never used Phil's calculation of wing chord as a measuring tool to determine wing/stab distance, but again, he has countless hours of flapless handle time.

My latest flapless profiles use the basic SkyRay 35 airfoil and moments, but with the tail shortened 3/4" and a new larger stab/elevator. Power for this one will be either LA 25 or FP 20. This basic setup showed good promise late last year prior to the weather getting bad.

I am not sure your need an airfoil quite as blunt as on the Doctor or the Medic. These are good foils when run with the right kind of engine, but when the engine quits the glide is a bit challenging.

Let us know how your efforts turn out.

LAST EDITED ON Feb-28-04 AT 01:27 PM (CDT)
 
>>Hi, All-
>>
>>I've read everything on SSWF over the past two years and
>>from earlier archives on this subject, but would like any
>>last-minute input from anyone who hasn't posted or whose
>>ideas have evolved further. I'm backing off some on a couple
>>of the more radical features of my little LA .25 powered,
>>flapless, profile stunter, including stab/elevator areas,
>
>I wish that Brett and Ted would weigh in here, as they have
>a lot more experience on this than some of us. Phil
>Cartier's comments should probably be carefully considered,
>as few have more handle time with flapless designs than he
>does. Here are a couple of additional observations I have
>noticed:
>

I can't speak for Ted, but I think I have pretty well exhausted my knowledge of the topic in numerous previous posts. I'm not actively developing these sorts of models, only occasionally flying them and seeing what I like or don't like about the various designs. In the case of the Skyray, Medic, and Doctor, I have some ideas about each one and how to make them better, but I wouldn't wait around for me to actually do it

I think I like the smaller airplanes better overall than the big old Doctor - but that's probably because I don't think we have ever gotten the engine to run just right, and the Skyray and most of the Medics have had 20's or 25FP, and they ran like watches. Even with that, the smaller (and somewhat denser) little models didn't seem as much like "kites" as Ted's Doctor. Paul Pomposos's Doctor (smaller for reasons I still never grasped), on the other hand, seemed very solid - but the engine (OS 40 Surpass/11.75-4 3-blade (!!!)) also ran perfectly.

Derek Moran's airplane (shown above), in addition to being beautiful and possessing excellent craftsmanship, is also probably at least one generation ahead of my experience and pretty much incorporates most of the ideas I would try.

Just like any other design issue. the "numbers" are useful to get you in the right ballpark. From there's it's a matter of building it, seeing how it performs, and addressing the shortcomings you think you find.

To reduce sensitivity around neutral picture a handle with a squre insert between the lines (3" for example) so that at neutral your line spacing is 3" and as you rotate the handle the spacing increases until at 45 degree of handle tilt you have a line spacing of approx 4.2". Doesn't require any mods to any airplane and then after you fly it you can tweak the shape of the insert to give you any response you want. See the old S/N article about John Wright's round insert handle to get the picture. Then when done testing and decide you don't like the way it works, you can put it in the closet like I did with the one I built years ago!

I simply cannot stand how good this thread is--as a representation of the community memory that SSWF can provide.

Derek, are there plans for your ship available?

Serge, you go!

Cy.

Sorry Cy- no plans. I have a design layout in AutoCAD. It's not highly detailed and it's not pretty, but there's enough data to build a model airplane. If I get a bunch of requests I will generate formal plans.
Derek

Serge, if you compare flapped and flapless model, you have only little differences. It is true that flapless model has no airfoil pitching moment, but you have CG pitching moment – it acts similar if moved to 15%. The tail brings two distinct effects. It gives and maneuverability and stability.

Our usual area and tail length as we use on flapped models loads the tail only little. The lift coefficient of tail in steady radius of corner can be cy=~0.3, it is far less that that cambered airfoil can do. The maneuvering of flapped model is not what is calling for bigger tail. And also AoA of tail is close to its “zero lift” AoA. It will be even better on flapeles model, so also stab/elevator ratio will be similar to flapped model.

The static stability says how far is CG from neutral point. I think that is what makes differences. If you move the CG little front, the same value of stability could be reached by little smaller area. I recalculated one of my “worksheet” models for flapeless model, if we count “modern” value 25% of wing area for flapped models, those mentioned 20% looks like a good value and gives the same amount of static stability if CG is at 15 instead of 23%.

The stab/evenator ratio is critical on flapped model (it gives proper ratio of flap and tail effect), but it is not the case on flapeless model. You can adjust it later anywhere on way from handle to the elevator.

Igor and Brett-

Your perspectives on this are much appreciated.

Don, and Bob-

Thanks for the ideas and for sharing your experiences. I'd forgotten about handle geometry. Lots of good food for thought, even past tomorrows stabilizer!

SK

Hi Serge

A couple of questions.

What does TVC mean?

I am curious about your proposed CG position and the general advice given on this. I would have thought that as a flapless symetrical section has a Cmo ~ 0, that the centre of lift is near the wing aerodynamic centre. I would therefore have gone for a CG of 25% of mean chord and simply adjusted tailplane area to maintain the desired static margin. That way, there is almost no tailplane load in a loop.

This does not seem to be the recommended approach. I am curious as to why a more foward CG than this is so preferred.

In designing a new plane, I decide the wing area, AR and moment arm as a % of MAC, set out the ACw ACtp NP and CG, then design the plane around this geometry. Do you follow a similar proceedure?

>Hi Serge
>
>A couple of questions.
>
>What does TVC mean?
>
>I am curious about your proposed CG position and the general
>advice given on this. I would have thought that as a
>flapless symetrical section has a Cmo ~ 0, that the centre
>of lift is near the wing aerodynamic centre. I would
>therefore have gone for a CG of 25% of mean chord and simply
>adjusted tailplane area to maintain the desired static
>margin. That way, there is almost no tailplane load in a
>loop.
>
>This does not seem to be the recommended approach. I am
>curious as to why a more foward CG than this is so
>preferred.
>
>In designing a new plane, I decide the wing area, AR and
>moment arm as a % of MAC, set out the ACw ACtp NP and CG,
>then design the plane around this geometry. Do you follow a
>similar proceedure?
>
>Regards,
>Ken

An excellent question, Ken. I had answered it essentially as you did when I designed the Doctor and Medic. After flight tests I put my thinking cap back on and came up with the same opinion as did Igor. The CG/AC pitching moment helps provide some feedback to control inputs that pretty much disappear when the CG is at 25% even with a large tail.

After the fact I measured a bunch of old timers (well, at least several) laying around the garage, all of which had done very well at VSC comps over the years and all of which were flapless. Lo and behold, all of these independently trimmed stunt ships had CGs with a fraction of 15% of the MAC.

If anyone is interested this subject a lot more were discussed at great length (probably too great a length for those with short attention spans) in my STUNT NEWS article on the Doctor/Medic.

For Serge's benefit, I haven't been ignoring this thread but felt sure that he was well aware of that article which pretty much reflects my feelings on non-flapped ships.

By the way, re Brett's lack of understanding of why I shrunk the Doctor a small amount from the prototype, he pretty much answered the question in his discussion. I felt it was too big and acted too much like a kite in high winds. Just like he said!

Paul's very nice flying Doctor was built straight from the PAMPA plans and is a very pleasant flying airplane. You can do the pattern almost in your sleep with it (which, given the events of this past week-end ... don't ask ... isn't such a good idea with old familiar pond scum type airplanes).

LAST EDITED ON Mar-01-04 AT 11:49 PM (CDT)
 
Thanks, Ted-

>For Serge's benefit, I haven't been ignoring this thread but felt sure that he was well aware
>of that article which pretty much reflects my feelings on non-flapped ships.

I haven't read the article on the doctor, but intend to do so. I do know your philosophy on flapped planes from the "Imitation" articles and your SSW posts. I think I know some of your take on flapless designs from SSW forum posts: somewhat blunt l.e., forward 'high point' (18%, now probably back to 22%-24% chord), longish tail arms, lower-aspect-ratio (~4.5), 60/40 horizontal tails of no more than 20% of wing area,... there may be something I've missed (perhaps <10 oz/ft^2 wing loading??). I don't know just where you stand on useful aspect ratios. I think you were revising your thoughts on ideal wing thickness - or was that Mikey? I know his increased for his bigger flapless design, but yours may have diminished a tad? 'hope these are close to accurate. Anyway, I'm always happy to hear your thoughts for planes with various ...goals! Thanks.

SK

P.S. (Edit) I forgot "not" to ask. I HOPE this was minor!

>
>P.S. (Edit) I forgot "not" to ask. I HOPE this was minor!
Serge

Let's just put it this way. If I could sell pieces of a purple pond scum two time Nats winner for five bucks apiece I wouldn't have to worry about Social Security! It was ugly .

LAST EDITED ON Mar-03-04 AT 02:30 PM (CDT)
 
Hi, Ken-

>What does TVC mean?

I used the nomenclature from some previous posts on the subject. "TVC", then, is "Tail Volume Coefficient", an indication of the effectiveness of the horizontal stabilizer. There were two different measures, one of which factored in the tail length twice, for a squared term influence (See Frank Williams' post in "Horiz. Stab Area", a thread from 10/2002). For comparison, I have just been using the simpler one that uses each relevant quantity once. So for the numbers I quoted,

TVC = Lt(At)/{Aw (Cw)}, where

Lt is the tail arm measered either from the c.g. or quarter chord or the wing MAC (Ted uses Cw/3, I think) to the quarter-chord point of the horiz. tail;

At is the horizontal tail area (stab/elevator combined);

Aw is the wing area; and

Cw is the length of the wing's Mean Aerodynamic Chord (or MGC) as computed from the definition (as in our previous discussion). See Palos Verdes R/C MAC page for trapezoidal half spans.

I saw these grouped this way as a coefficient in expressions for static stability in an NACA report, but don't see it nearby (piled in!). However, it is called Hunsaker's "th." in TR #293, where this particular expression is used to compute the horizontal tail area, once its necessary value (the number I am calling the "TVC") is computed from several other factors. i.e At = (TVC)(AW)(CW)/Lt.

>I am curious about your proposed CG position and the general
>advice given on this. I would have thought that as a
>flapless symetrical section has a Cmo ~ 0, that the centre
>of lift is near the wing aerodynamic centre. I would
>therefore have gone for a CG of 25% of mean chord and simply
>adjusted tailplane area to maintain the desired static
>margin. That way, there is almost no tailplane load in a
>loop.
>
>This does not seem to be the recommended approach. I am
>curious as to why a more foward CG than this is so
>preferred.

I guess Ted pretty much summed it up in his post. I essentially asked the same question about 16 months ago on SSW, and several (Igor, Ted, Brett, I believe) said that without the negative pitching moment from flaps, this much static margin was required for reasonable feel. It appears to be empirical, although Igor may have somecomputed reasons. There have been several such threads, and this one was one of the most interesting for me.

>In designing a new plane, I decide the wing area, AR and
>moment arm as a % of MAC, set out the ACw ACtp NP and CG,
>then design the plane around this geometry. Do you follow a
>similar proceedure?
>

Yes, but I'm not yet finalized on it. I choose area based on engine size/type and whether or not the wing is flapped. The engine/wing area corrolation is empirical, basically judging from what has worked with certain engine types. It would be easier if engines were run at rated power, but when I asked a year or so ago about power used, I drew a blank (there have been a couple quantitative posts, one on various Fox .35's and a couple .40's, although they did not specify useful power or torque). The corrolation takes into account the complex interrelationships among induced/profile drags and weight (in itself dependent on engine and plane size). Then I choose aspect ratio and, based partly on that and where I want the AC, the taper ratio or approximations to elliptical forms. I haven't been using the AC of the entire plane, since the static margin has proved to be necessarily empirical anyway. These geometrical decisions can be based strictly on algebra and later adjusted to suit aesthetics, RN, etc. I choose quarter-chord sweep now too to determine AC of the wing; it can drive the taper ratio, and minor trade offs can be made to aspect ratio.

My order usually goes like this: engine -> area (or engine -> weight, and then weight and wing loading -> Area); flight characteristics -> A/R and taper; area and A/R -> span. Area, span, taper ratio, and tip span -> root and tip chords (As posted a couple weeks ago, aspect ratio and position of MAC -> taper ratio too). Root and tip chords and hinge line (desired yaw stability, tank space, 'etc.') -> wing sweep. Sweep, root and tip chords (and span) -> MAC and longitudinal wing MAC position (and spanwise MAC position). MAC position -> NP. NP and static margin -> c.g. But this turns out to be ahead of where I would have guessed. Then I get all confused about tail moments. Anyway, TVC -> moment arm x tail area. As stated, I havent been using the aircraft NP, since this seems overriden by unknown factors to reduce c.g. to about 15-16% MAC.

It's the next part that prompted this thread: how to choose the trade-off in horizontal tail area vs. tail arm and in all this to place the hinge advantageously. I use TVC to juggle these, but am not yet sure about the static stability vs. pitching moment thing. However, I have some good real-life data in this thread to think about. Choosing the moment arm in terms of wing MAC length seems reasonable, but I think there is a range that depends on desired final weight and probably on control quickness, neutral stability, and devices used to alter control deflection rates.

I just took an hour to do some necessary things here, so if any more comments have come in - or I had brain fade - I may have to edit this post later.

Lots to mull over.

SK

Edit: Most recent Edit concerned sweep in the design "flow-chart".

I am not sure if the 15% is only to make more feedback. In any case, if I take some my “virtual” model with CG at 22%, which gives feedback (difference in line tension in corner) 50% of line tension in level flight, and if I make larger area instead of flaps and I move CG to 15% then the feedback will be at ~20% instead of few if CG is at 22-25%. So that idea looks be proofed, by I have one exception. My model from WC in Sebnitz which has logarithmic unit on flaps to limit the hinge moment transfer to bellcrank and thus the feedback in corner needs also CG at 15%. The device limits the feedback, but not the flap pitching moment. So here is some disproportion and I afraid to make some definitive relations here.

BTW Serge, do you have my spreadsheet? If yes, you can model an airframe also without flaps. It allows only straight TE, but is enough to give hints or compare two different ways.

Igor-

>I am not sure if the 15% is only to make more feedback. In any case, if I take some my
>“virtual” model with CG at 22%, which gives feedback (difference in line tension in corner)
>50% of line tension in level flight, and if I make larger area instead of flaps and I move CG to
>15% then the feedback will be at ~20% instead of few if CG is at 22-25%. So that idea
>looks be proofed

Your virtual stunter data is interesting. However, I'd like make sure I'm correctly interpreting what you have written. Do you mean that with flaps and the c.g. at 22% the difference between the tensions of the two lines is half of the total line tension of this model in level flight. If so then I understand that it drops to 20% of the total level flight tension when flaps are replaced by greater tail area and c.g. is moved to 15%, but that is higher than it would have been with the c.g. still at 22%. Is this correct?

My hard drive was replaced last year with some loss. I do not have your spreadsheet, but am interested. Thanks.

Serge

>>>I'd like make sure I'm correctly interpreting what you have written. Do you mean that with flaps and the c.g. at 22% the difference between the tensions of the two lines is half of the total line tension of this model in level flight. If so then I understand that it drops to 20% of the total level flight tension when flaps are replaced by greater tail area and c.g. is moved to 15%, but that is higher than it would have been with the c.g. still at 22%. Is this correct? <<<
Yes that is that feedback “induced by CG position” as we spoke before. I would like only note, that it does not depend on tail area (or at least minimally). It comes from the lift on tail which is balancing that CG pitching moment and stab to elevator ratio – you just support the lift by pushrod, that is all, the area does not change it – half area makes the same lift at higher AoA, but hinge moment is the same – twice lift, half area = the same force.

But I wanted to say that there is exception – that model which HAS pitching moment of airfoil by flaps, but STILL needs CG at 15%. The only difference to usual flapped model is, that there is very limited flap hinge moment (and thus feedback to handle). I do not know how to interpret it at present time. It can be that it has something to do with fact that it was relatively small and light model, so may be the percentual feedback (to the line tension) is high, but absolute value could be still small, I really do not know now, may be I need load my shaking hands by some absolute value or what

Thanks Surge and Ted

I am comming from an RC glider background where feedback is not so important. In gliders, it is common to put the CG on the centre of lift for best Cl/Cd trim. That means that trimming for min sink, a higher Cl (up trim), the centre of lift moves forward of CG (cambered wing. When equilibrium is reestablished, the the tail plane load reverses to an up load. With a fully flying tailplane hinged on the ACtp, feedback to the servo is negligable.

I Guess that if that ever happened in CL, your feedback would be really messed up and any slop in control linkages would make for a wandering flight path. So point taken to ensure CG is always 10% or so ahead of centre of lift.

For calculating neutral point or tailplane area, I have just been using that the wing area moment about the NP is equal to the effective tailplane area moment about NP. Effective tailplane area, taken from a nomogram of wing and tailplane aspect ratios and wake allowance. Perhaps this is an outdated method and the wake allowance has been pretty subjective to say the least. I saw one site where this was being done, thith no regard for tailplane efficency. I am interested in using a more calculatable approach.

Can you blokes point me to some good URLS on the subject of static margin and stability?

Also from your pics, it looks like most modern stunters have a moment arm (distance between ACw and ACtp) of around 300% of mean chord.
I am using 200% based on stunters like the spector form years ago.

What moment arms % of mean chord is about right these days?
Do you see moment arm to be a function of line length as well,
shorter lines requiring tighter manouvers?

Ken-

I am exploring the same areas. Aside from juggling TVC factors, I haven't gotten further (and I have to build now!). This quote from Phil above is relevant:

>A gap of at least 2/3 of the average chord is needed between the wing and stab. Shorter
>and you run into all sorts of erratic control effects. Over 1 avg. chord makes it very difficult
>to balance

In a series of posts from early November of 2002, we talked about a "Pitch-Damping Factor". It was defined by one of the Randys as TVC x Lt/Cw (same defs as in earlier post). This simplifies to (At/Aw) x (Lt/Cw)^2, that is the ratio of tail and wing areas times the square of the tail moment expressed in wing mean aerodynamic chords. This expression could be evaluated for significant models to see what "works", but again I believe we'll find a wide range of models whose idiosyncracies may be only appparent to the experienced fliers who have flown a variety of them.

Frank Williams also wrote an interesting description of the math that goes into stability derivatives. I think it's worth an effort to find through a SSWF search. Incidentally, the analysis in NACA TR 293 (mentioned in a post above) by the excellent researcher and author Walter Diehl is unfortunately couched in terms appropriate to certain classes of full-sized aircraft as they existed in 1928; if not obsolete, they require interpretation in terms of the contrasting desired performance of our different sized models - I suppose again from empirical data. These are among the things I try to assemble and align in my own mental model. FWIW.

SK

>
>Can you blokes point me to some good URLS on the subject of
>static margin and stability?
>
>Also from your pics, it looks like most modern stunters have
>a moment arm (distance between ACw and ACtp) of around 300%
>of mean chord.
>I am using 200% based on stunters like the spector form
>years ago.
>
>What moment arms % of mean chord is about right these days?
>Do you see moment arm to be a function of line length as
>well,
>shorter lines requiring tighter manouvers?

Ken,

Haven't researced the web for neutral point info. Best source I've found is Martin Simon's book called, I believe, Model Aircraft Aerodynamics. It is one of the few places I've seen the subject discussed in a fashion understandable by math challenged people such as myself.

Most of the aerodynamic pubs in the general press seem to focus on flight dynamics for the benefit of the pilot rather than design considerations. Thus, his book looks into alternative configurations like canards, etc. and has to deal with the development of hypothetical neutral point development to determine static margins and the like.

Popular books like the Aerodynamics for naval aviators tend to be very pragmatic and discuss generic CG considerations in terms of Weight and Balance, etc. they speak in general terms of problems with aft CGs and so forth but don't generally involve themselves with the esoteric facets the designers had to consider when developing the airframe the aviators are learning to fly.

Like that dude in the Flight of the Phoenix, us modelers are head and shoulders above the average "mere" pilot when it comes to our "need to know".

Let's all pat ourselves on the back now while I go back to work.

Hi Ted I purchased Martins book many years ago and it certainly has been the basis of most of my aerodynamics knowledge. The method I use for NP and static margin calculation is based on the empirical one described in his book. The problem is that his method relies on an estimation of tailplane efficiency from a nomogram in the book. Unfortunatly he only gives wake allowance vaules for 2 extreme cases.
A T tail and a tailplane placed in the wing wake close to the TE.

There is no guidance for values in between. A reliable calculated solution requires something more definitave like a table of values, graph or algorithm that relate wake allowance to moment arm and hight above the wing. From tabulated 2D data or graphs, I can define an algorithm as a polynomial curve fit or regression curve to use in a program.

>Hi Ted I purchased Martins book many years ago and it
>certainly has been the basis of most of my aerodynamics
>knowledge. The method I use for NP and static margin
>calculation is based on the empirical one described in his
>book. The problem is that his method relies on an
>estimation of tailplane efficiency from a nomogram in the
>book. Unfortunatly he only gives wake allowance vaules for
>2 extreme cases.
>A T tail and a tailplane placed in the wing wake close to
>the TE.
>
>There is no guidance for values in between. A reliable
>calculated solution requires something more definitave like
>a table of values, graph or algorithm that relate wake
>allowance to moment arm and hight above the wing. From
>tabulated 2D data or graphs, I can define an algorithm as a
>polynomial curve fit or regression curve to use in a
>program.
>
>
>Regards,
>Ken

Ken,

I agree that Martin's "paper cutouts on a stick" with a guessed at tail plane efficiency isn't exactly the sort of diagnostic tool on which you'd want to stake your life on a trip to Mars or something. On the otherhand, for a guy like myself (and, I think, most modelers) the understanding of concepts and relationships of understood values is the value of Martin's book.

Simply understanding that stability is the relationship of the CG to the Neutral Point (or Aerodynamic Center of the whole vehicle) is an illuminating revelation when it comes to designing and understanding the dynamics of our stunters. Numbers are great ... but they don't fly. Utilizing the concepts the numbers describe to our benefit is kind of the holy grail of progress in this stuff we call stunt.

The whole neutral point/static margin concept that Martin describes so well was the driving force behind whatever impact I might have had on the history of stunt design when I started to advocate larger tails and aft CGs ... now more or less ubiquitos in modern designs.

I'm not sure what the thread was called but there was a healthy discussion of the concept a year or two ago in which we discussed the progresssion of the realtionships of CG and Neutral point from a conventional planform with a small tail volume; through a planform with two identical wings fore and aft with the CG somewhat behind the forward wing; all the way back to a canard with a tiny foreplane. In each case the neutral point and CG of course progressively further aft to maintain the positive static margin. Howard even commented he had once made a tablet flip card illustrating the progress like a movie.

Ted,

I remember an article by Bill Netzband where he is describing the 35-40 average stunter. In that article he recommend using a much larger stab/elevator volumne. I think it was in a 1967 Aviation Modeler magazine or other magazine such as that. Does that mean that you advocated larger "tail" volumnes before then??

Surge,
Well there are a number of ways you can approach the flapless issue. Wild Bill's idea was a very low aspect wing with a thick airfoil. This required the use of a huge stabilizer and elevator to over come all the drag the wing created in the turns. I have flown Bill's model and it flies quite well. The turn rate was good but not what I would call great but it did lock onto a bottom and was quite smooth through out the flight. All and all a very nice flying model.

My approach was different. I wanted a snappy feel that was light and easy to fly for engines around the .25 to .35. Using these size engines rules out the thick wing from the start. The P Force has a 16% thick section at the root with an 18% section at the tip and 500 square inches of wing area. Just about the perfect size for engines in this range. The first prototype had a tail area of 22% of the wing area and even with the C.G. moved way forward at 12% mark proved to be too much elevator authority. The P Force was more like a combat model than a smooth stunter. I reduced the area of the stab & elevator to 20% and tried it again and it was better but still too much. Reducing it further to 18% was even better allowing the P Force to fly a sharp crisp corner with solid bottoms. Now I moved the C.G to the rear at 16 to 18% of the MAC to get the feel I wanted. In addition, this moved the C.G. closer to the high point (25%) of the airfoil.

The other thing about the flapless designs is they have to be light to make the crisp turns. If they are over weight they just can't make the turn with out banging the controls and upsetting the smooth look of the pattern. One of the things that I found was moving the C.G. back to 18% of the MAC and reducing the handle spacing gave me the best overall pattern that was relaxing and a joy to fly. I did fly it at 20% of the MAC but at this location the P Force seemed a little too sensitive and the overall appearance of the pattern suffered.

I agree with one of the statements where Brett said that the "formula's and equations get you in the ball park". But I believe hard work, meticulous testing, and changing is what really makes the end model excel.

Mikey

>The first prototype had a tail area of 22% of the
>wing area and even with the C.G. moved way forward at 12%
>mark proved to be too much elevator authority. The P Force
>was more like a combat model than a smooth stunter. I
>reduced the area of the stab & elevator to 20% and tried it
>again and it was better but still too much. Reducing it
>further to 18% was even better allowing the P Force to fly a
>sharp crisp corner with solid bottoms. Now I moved the C.G
>to the rear at 16 to 18% of the MAC to get the feel I
>wanted....

>One of the things that I found was moving the C.G. back to 18% of the MAC and reducing
>the handle spacing gave me the best overall pattern that was relaxing and a joy to fly.

Mikey, and others-

Thanks for the review. Did you experiment with changing the percent elevator chord or its span relative to the stab?

I wonder whether, for instance, a smaller elevator chord might equate to narrowing the handle spacing, requiring more deflection to achieve the same horizontal-tail camber (I think it was Bob who hinted at proportioning). I am also wondering about the relative effects on pitch damping and control authority of spreading a smaller chord elevator across a greater span, either maintaining the same overal area or increasing it. This would be a bit complex because of altered aspect ratio and Reynolds Number effects. A sort of small-angle, linear approximation for a trigonometric relationship: area-wise, 40% of an 18% tail ~ 33% of a 22% tail...?

Finally, I was also thinking about what happens if some unflapped tip area is retained for damping purposes or to create some possibly stabilizing interference drag when the elevator is deflected (may require in-line wing/stab).

Has anyone tried any of this?

Serge

Surge,
Yes, I did try a little of that sort of. I trimmed off the elevators first and test flew the model. I tried to keep them at a 60/40 but it would up at 55/45 or so. Like the Skyray and the Flight Streak the smaller elevator did just what you are asking (more deflection with a smaller area). I may be wrong, but I don't think that is a good road to follow. Mainly because the smaller elevator has to be deflected more for given turn rate. My opinion is, it is harder to hit a consistent flat bottom with the smaller control surfaces because of the large amount of control deflection needed.

But that does raise another interesting question. What would a counter balance at the end of each elevator do for a flapless stunt ship???? Hmmmmmm!

Mikey

All the fussing with the size of the elevator, chord, etc. is to get the "feel" of the controls where you like it. Adding fixed stab area at the tips could be used as a fix on an already built plane, but starting from scratch it would be simpler and easier just to make the elevator slightly less chord and run it out to the tips. Running it out to the tips also doesn't introduce another vortex on the stab. You get one automatically at the tip. Cutting the elevator short would add another one, which I cannot see being a benefit.

The Spit I posted a picture of a few days ago is now flying well. It has the scale counterbalance. Can't see that it makes much difference. The biggest thing is the stab is 1/16 in. thicker than the elevator. This has made a big difference in the level flight. Much easier to make small corrections without making a noticeable change in the plane's attitude. The other thing, the Spit elevator is 3.5 in. wide. This is too much. The controls stiffen up when the airplane speed gets over 50 mph(5.5 sec lap on 63 ft. lines). They also stiffen up significantly if the plane picks up speed coming downhill in the wingover, or if the wind picks up when the plane is headed down. I may just get out my french curve and slice a half an inch off the chord and to hell with how it looks!!

If you really want to do it right, use a stabilator. With the right balance point, and some counterbalance to keep the control forces in check, a stabilator gives a lot more latitude in the balance point and very flexible trimming. I just haven't been able to figure out an easy, cheap way to build a stunt size(100+ sq.in.) stabilator and keep the hinge from being too sloppy.

Phil,

Ever tried a closed loop control system?

These are sometimes used on RC sailplanes for rudder where a large deflection with little slop and accurate centering is required.

Basically, use a double control horn and take a wire from each horn back to both arms on the bellcrank. You would probably need a different kind of bellcrank "X" rather than "T". I see no reason why it could not be adapted to CL. The wire tension is critical to minimise slop.

I wasn't concerned about slop in the pushrod on a large stabilator. It just increases the dead band about neutral a bit. The problem is keeping the whole stab square. Say the fuse is 1 inch wide. .005 in. slop in the bearing would translate into the stab rocking in the bearing about 1/4 in.(actually +- .05 in. on a 20 in. stab). The rocking wouldn't necessarily be smooth or predictable. Probably have to angle the outboard tip forward(heh, heh!).

Making the bearing wider means you have to mount the bearing in a small stabilizer with a short chord, so it isn't very sturdy and gets cracked easily. Longer tube bearings are draggy and heavy and spew black goop all over. I've tried K&S tubing. They control the size quite well. 11/32 tubing and 3/8 tubing make a good bearing with .005 slop. It weighs a full ounce. Nothing like a built in one ounce tip weight in a plane that tends to tail heavy anyway. 8 Mylar hinges for a conventional stab weigh a gram.

All this is for a combat style stab that is one piece. Plug on stabs are even sloppier and heavier. Some of the pattern guys are using steel tube for the stabilator hinge.

All in all, while I'd like to try a stabilator, a stab/elevator is just so much lighter and easier. Just have to live with the narrower trimming range.

>>>This would be a bit complex because of altered aspect ratio and Reynolds Number effects.<<<

I think the elevator makes only little lift, so it will not be sensitive too much, but in any case, if you use thin flat stab under 3%, it is NOT sensitive to RE number, and it is also highly recommended for such small models.

Hi Serge

I'm not sure if I understand what you are getting at here.
I think you are trying to dampen stabilising forces.

I think of the the tailplane (stab/elev) as one unit.
I put the elev hinge line on the end of the fuselage and fin TE.
That allows the elevator to run right across the TP span the with no rudder notch. That means no disruption to the TP lift distribution curve.

If the TP planform is optimum (eliptical) and the local chord stab/elev break is constant across the TP span, than you have the same camber for the whole span and no aerodynamic AOA twist.
In short, the whole TP is at the same CL, more or less.

Is this a good thing? I think that would mean the tailplane would respond to pitch disturbances very quickly.

On the other hand, If the stab chord were constant but the elev tapered towards the tips, there would be a reduction in camber and AOA towards the tips so starting a turn, the whole TP would load but as the turn progressed, the load at the TP tips would reduce, possibly some increasing part of the tip may go negative load untill
a turning moment equilibrion is reached. In short the total TP load would reduce as pitch rate increased.

Would this provide any more damping than the first case?
Not sure the overall effect (apart from drag) would be any different.

What do you think?
Would the CL/alpha become curved rather than the usual straight line?

Interesting thoughts, Ken.

>I think you are trying to dampen stabilising forces.

I'm looking for ways to damp motions away from equilibrium, when the elevator is minimally deflected, but to minimize penalties to pitching moment from intended use of the elevator. Igor aptly stated a while back that I need an expocrank. Perhaps I want to "have my cake and eat it too." "All wet"?

>I think of the the tailplane (stab/elev) as one unit.

Me too, but I'm thinking of what happens when one separates spanwise areas where there are both from areas where there might only be stabilizer. I don't know the magnitude of interference drag between such areas, but aside from wake effects from vortex formation, there might be advantages to the tail drag caused at initial elevator deflection.

>I put the elev hinge line on the end of the fuselage and fin TE.
>That allows the elevator to run right across the TP span the with no rudder notch.

That's my present configuration.

>If the TP planform is optimum (eliptical) and the local chord stab/elev break is constant
>across the TP span, than you have the same camber for the whole span and no
>aerodynamic AOA twist. In short, the whole TP is at the same CL, more or less.

>Is this a good thing? I think that would mean the tailplane would respond to pitch
>disturbances very quickly.

Elliptical: Best L/D, but not necessarily most lift possible for given structure - as I understand it. Due to lower stabilizing drag, it might be close to maximum, but as I understand things, at least one top designer deliberately looks for higher induced drag and lower sensitivity with low-Aspect-ratio horizontal tails.

>as the turn progressed, the load at the TP tips would reduce, possibly some increasing part
>of the tip may go negative load untill a turning moment equilibrion is reached.
>In short the total TP load would reduce as pitch rate increased.
>
>Would this provide any more damping than the first case?

I don't think the basic lift curve characteristic would change, but the tail would probably reach equilibrium at a lower turn rate. I think that this damping would be counterproductive, since it's already inherant in the increased angle (against turn) with which the stab meets the relative wind, as turn radius decreases. Generally, damping around neutral is desired, so that plane doesn't hunt and comes cleanly out of turns in the pattern.

>Would the CL/alpha become curved rather than the usual straight line?

For any given deflection, I don't think so, but interactions in turns seem more complex anyway. I'll have to think on it when I feel better. As usual, you have provided food for thought.

SK

>Elliptical: Best L/D, but not necessarily most lift possible
>for given structure - as I understand it. Due to lower
>stabilizing drag, it might be close to maximum, but as I
>understand things, at least one top designer deliberately
>looks for higher induced drag and lower sensitivity with
>low-Aspect-ratio horizontal tails.

This is best L/D because Cl is the same right across the span,
you don't get local chords contributing only drag with little lift
as is the case with a constant chord. In Cdi = Cl^2/(Pi * AR)*k,
the plan form correction factor "k" in minimised. You also don't get overloaded local chords that reach Cl crit early. With the correct compensation for local Re crit by thickening shorter tip chords or in the case of cambered wings, increasing camber while aligning beta for all chords, The whole structure can reach Cl crit at the same time. That means max lift from can be achieved only with an eliptical plan form though a good straight taper planform will go close enough.

That said, with a fixed flap C/L, the tailplane will not reach Cl crit unless you also rotate the stab. I believe the stab opposes the elevator particularly as the turn progresses. This is one of the reasons I favour Flapped wings, there is incidence produced between wing and stab.

Someone said, they make the elevator thinner than the stab so small deflections in level flight have little effect. Seems a good idea to me. Is what you are trying to overcome a stability issue or flutter?
Do you mass ballance the elevator? If you aerodynamically ballance the elevator by protrusions ahead of the hinge, I expect feedback will be reduced. That might lead to a more forward CG, reduced manouverability probably not good.

If there is hunting, I suspect the same airflow interacting with 2 or more surfaces. Is the TP low in the wing wake? The kind of thing I am thinking of is, airfow attaches to a surfaces, the resulting deflection causes opposing pressure that detaches the flow and starts the whole thing again. Can you imagine anything like that happening?

What section do you like using?
I like Eppler 472 and related sections. Seem to give high Cl crit and ample opportunity for flow reattachment.

I don't have Xfoil thougj I note you do, you could try some of the ideas out on it. I am thinking of getting Xfoil, does it modle Cmo?

LAST EDITED ON Mar-09-04 AT 02:19 PM (CDT)
 
>Someone said, they make the elevator thinner than the stab
>so small deflections in level flight have little effect.
>Seems a good idea to me. Is what you are trying to overcome
>a stability issue or flutter?

Stability. Past the experimental planform and wing section, the whole thing's been about stability. I don't see any problem with lift or aoa. No flutter concerns yet, since I've never encountered it in my models.

>Do you mass ballance the elevator?

No.

>If you aerodynamically
>ballance the elevator by protrusions ahead of the hinge, I
>expect feedback will be reduced. That might lead to a more
>forward CG, reduced manouverability probably not good.

I'd rather reduce control forces by increasing control arm effectiveness and not destabilize the aerodynamics. I'm intrigued by the "exponential", "self-centering", and servo type mechanisms, several of which seem clever, since they reproportion forces for varied deflections.

>If there is hunting, I suspect the same airflow interacting
>with 2 or more surfaces. Is the TP low in the wing wake?
>The kind of thing I am thinking of is, airfow attaches to a
>surfaces, the resulting deflection causes opposing pressure
>that detaches the flow and starts the whole thing again.
>Can you imagine anything like that happening?

Yes, perhaps - sort of a resonant like thing? It might be between wing and stab, but thrust line is part of the alignment thing too. In prevous threads, negative stab incidence has been blamed by more than one person. Positive thrustline has also been the culprit, and that's consistent with negative stab incidence. Since positive wing incidence is consonant with both negative TL and negative stab incidence, it sounds as though the relationship between thrust line and stab incidence (and all other moments?) outweighs the wing in this. Of course usual configurations (stab above T.L above wing), may dictate a different solution than in-line configurations. I have given this less thought than some, but do not see most stunters as symmetrical; therefore I don't see symmetry arguments of upright vs. inverted flight as relevant to every hunting issue that is brought up.

>What section do you like using?

I listed and drew it in the other thread. Reasons below...

>I like Eppler 472 and related sections. Seem to give high
>Cl crit and ample opportunity for flow reattachment.

I'm playing. I made up one that is quite similar to one favored by John Miller. It has a l.e. radius and transition that allows the high point to be rather forward, with what looks to me to be better tolerance for high a.o.a. and perhaps (!) also better wind "penetration" than blunter airfoils previously used on flapless designs. I would expect it to land longer (not fall as quickly) and transition more gracefully from downwind to upwind - that place being where maneuvers are performed. I hope it works well high. This was all "eyeball engineering", but the limited X-Foil analysis liked it above the NACA 00XX section of the same thickness with 6% further aft high point (BTW, the NACA and 63A derivatives have very similar l.e. radii when viewed as overlapped drawings). That's all on my other (almost) current thread.

The E 472 has been recommended more than once on this forum, but I haven't compared it in any program. You should do a SSWF search; there are at least a couple good related threads mentioning this section. My preliminary opinion is that it isn't a bad approach, except that based on previous discussion (especially from Phil C.) I expect it to "groove" less well in a flapless application due to its aft surface camber. It might make a good flapped wing, but Al and Igor have shown what I think are better aft cambers for flapped planes. If you haven't looked at Igor's site, you will enjoy it and get some good ideas. Al's airfoils are in certain of his articles, one of which I THINK is on Dave Day's site.

>I don't have Xfoil thougj I note you do, you could try some
>of the ideas out on it. I am thinking of getting Xfoil,
>does it modle Cmo?

EDIT: I forgot to answer the last question. Sorry. XFoil graphs Cm Vs alpha (also draws polars, pressure distributions, and graphs Cl/Cd, among additions to what I've shown on the other thread).

I'm so short of time as I regain strength from this stupid bug, that I don't really have the time to play with this much more for now. However, I will surely return to exploring, after I've gotten a couple priorities back on track. My XFoil comes with Profili II and is very convenient to use with its large airfoil library (as I stated before, apparently largely from Lednicer/Selig UIUC) and ability to generate new ones and combine variously.

You can get Profili I free and Profili II as a free trial and then permanently for $10.00 sent to its author. Type "Profili" into your search engine. It will come right up. Click on the site for all details, and you will see how things are analyzed. Profili has been convenient for me, since it prints out ribs for linear and elliptical taper with a limited, but useful set of cut-outs for lightening and spars. I just used its XFoil-component analysis for the first time out of curiosity to see what it said about fixed flaps. I do not know its specific limitations, although I believe the "Nurflugel" List (Yahoo) archives from 3-4 years ago lists cogent comments from Lednicer, Bowers (Chief aero at NASA Ames), and a couple other experienced "heavy hitters." When I decide to do this again, I may return there to get a better perspective.

Larry Cunningham is doing some neat work on generating sections for construction and export too!

SK

Serge,

I haven't really been ignoring this thread, just monitoring it to see what I can learn.

It occurred to me, however, that you at one time listed a few things you thought you understood about my take on the subject. Amongst those was the thought that I preferred a low aspect ratio wing.

That isn't exactly correct.

Ideally (all other things being equal ... which they ain't), a max performance flapless ship would have a higher aspect ratio than otherwise. The reason is quite simple, high a/rs provide greater lift with less drag and, equally important, do so at lower angles of attack than otherwise equivilant but lower aspect ratio planforms. Thus, a high a/r planform can produce the lift required for a given level of performance with either less area or the same area at a lower angle attack, the preference depending on the wing loading.

Like everything in stunt, however, one must seek the holy grail with tiny steps and not kangaroo hops. Because high a/rs develop less drag for a given amount of lift when they are in atmospheric conditions which induce wind up, they will do so more readily than lower a/rs. I believe in the DOCTOR article I reminisced about a local combat flyer (actually more of an engineering driven, Rich Porter type than a competition flyer) who loved to build ultra high a/r combat ships with very competitive powerplants. It was not at all unusual to watch him demonstrate his every tightening high speed consecutive loops accelerate to the point of destruction in even modest winds.

Obviously, for stunt use where we must "manage" the elements while maintaining an acceptable level of precision to our maneuvers, such susceptibility to the environment must be militated against. Just a wild guess but I expect six to one is probably a realistic max for stunt work.

Further, I would advocate a slightly higher wing taper than for a flapped wing, say 60 to 70% ratio of tip to root. I know, the DOCTOR is constant chord but (and this is addressed in the article as I recall) this is in response to the expected user, a fairly new flyer whose ability to build straight structures without some sort of jig is at least in question. A constant chord wing can be built straight much easier than a tapered one and, ultimately, a straight wing is more important than one that adapts more closely to the ideal "elliptical" lift distribution about which we and Supermarine wax so elequently from time to time.

LAST EDITED ON Mar-09-04 AT 04:16 PM (CDT)
 
Serge,

By the way, although I wouldn't for a minute discourage investigation and testing of any of the parameters you're discussing, I do feel that any gains to be made in precise determination of tail volumes and airfoils etc. is less likely to be ultimately valuable than is time spent on precise trim of an otherwise slightly imperfect example of the genre. Although the aerodynamics of flight in a tethered environment are complex to the point of irrationality, the fundamentals of the machines we use to employ that environment are pretty straight forward.

Complexities of design really don't become significant until we throw the flaps into the mix. That was precisely the reason for the development of the DOCTOR. If you wanted to build a pretty darn competitive flapless stunt ship you could simply increase the root chord 10%, decrease the tip 10% (WAGs, I haven't sat down to do the math but get in that 65%+/- taper ratio area), Make the tail about 20% (the DOCTOR is larger than necessary) and connect them with a swoopy built up fuselage that meets your aesthetic demands.

Build it straight and reasonably light , put in a good "one speed" engine like a four banger or a high rev/low pitch two stroker; then send it to any "hot shot star" and they'll qualify at the Nats with it. Send it to Walker or Werwage and they will likely be in the top five.

LAST EDITED ON Mar-10-04 AT 11:42 AM (CDT)
 
Ted-

Thanks for the thoughts in this and your next post.

>It occurred to me, however, that you at one time listed a
>few things you thought you understood about my take on the
>subject. Amongst those was the thought that I preferred a
>low aspect ratio wing.
>
>That isn't exactly correct.

'glad I checked in tonight (sick of sanding and fitting). When I saw this I at first thought, "Oh-oh, I must have written something too late at night again, since I remember that you had advocated higher than average (not extreme) A/R in the "Imitation" article. I knew that you had rethought - re-adjusted - some things, and I thought one might have been aspect ratio. What I did remember was your advocacy of reduced aspect-ratio horizontal tails and hoped that I hadn't made a typo. So I checked the posts above and found the quote in #24.

Here's what I said:

"lower-aspect-ratio (~4.5), 60/40 horizontal tails of no more than 20% of wing area,... there may be something I've missed (perhaps <10 oz/ft^2 wing loading??). I don't know just where you stand on useful aspect ratios."

Sorry - confusing punctuation. The "lower-aspect-ratio" referred to "horizontal tails". Only the last part referred to the wing. I hope the other things I said were close...?

>Ideally (all other things being equal ... which they ain't),
>a max performance flapless ship would have a higher aspect
>ratio than otherwise. The reason is quite simple, high a/rs
>provide greater lift with less drag and, equally important,
>do so at lower angles of attack than otherwise equivilant
>but lower aspect ratio planforms. Thus, a high a/r planform
>can produce the lift required for a given level of
>performance with either less area or the same area at a
>lower angle attack, the preference depending on the wing
>loading.

I may have gone a bit high, but that's the wing on my table.

>Just a wild guess but I expect six to
>one is probably a realistic max for stunt work.

Thanks for the estimate. I may revisit this problem again some time and use something closer to that. For now, I've exceeded this a bit looking for tighter maneuvers (at the expense of susceptability to upsets). I have backed off on my short tail moment though - two inches longer now. The plane should last longer with this "pilot" now.

>Further, I would advocate a slightly higher wing taper than
>for a flapped wing, say 60 to 70% ratio of tip to root.

Well, I got THAT right! By the way, no one much noticed it, but I posted a nice way to compute taper to match the spanwise half-span center of lift of a higher aspect ratio wing to that of a more conventional one of equal area. I can't find it without losing this post, but it was about 2-3 weeks ago.

(EDIT: Here it is:

http://www.clstunt.com/cgi-bin/dcforum/dcboard.cgi?az=show_thread&om=7706&forum=DCForumID1&omm=35&viewmode=threaded)

Your last comment and what you say in your next post is appreciated. That's why the "Imitation" is next in line. 'looking forward to flying that one! I've really appreciated the advice and stunt wisdom.

For Ken W., if you're tuned in:

I was looking through my Martin Simons book while on,...er...IN the "reading room" earlier and ran across the Eppler 472. My first thought was that this was not what I remembered when I replied to your post. So I looked back through my stack of wing-section printouts and found the one I remembered. It was one I'd modified by moving the high point BACK to 25% - big difference. So now I'll amend my comment to "I don't know. I haven't thought much about it recently." As thin as it is, but with its blunt airfoil, it might have stall - limited max. lift. It might be really interesting with a light wing loading. However, when I get off this keyboard, I am taking Dirt's implied criticism to heart and finishing my center section tonight, rather than diving into X-Foil.

Sorry to be so long winded, but writing short and concise material takes too long!

Serge

Hi Serge

Found the E472 polars again.
According to NASG polar database:

Real test
E472 Cl max 1.25 Cd min 0.01 @ Cl +- 0.1 Re 300000

Xfoil
E47s CL max 1.15 Cd min 0.06 @ Cl +- 0.55 Re 300000

There is a big difference in Cd between the real test and xfoil
at Re 300000 especially near Cl crit.

This should get my size models (.16 m^2 wing, 2.5cc engine, .5kg mass) turning in a 4.2 m radius without flaps or 2.8 m radius with flaps. Looks good to me for a 12% section.

E474 and 476 are thicker with much the same high point.