CG on a biplane

[CDNshy_guy]

can anyone tell me how to get the CG on a biplane and is it any different then on a single wing plane?

Norm

[pease1]

If the wings are parallel, not staggered then the CG is the same. If you stagger them then it gets tricky. The CG is the average of the two wings CG, with a bias towards the top wing based on the stagger, how close the wings are, and other factors. Start with 60% bias towards the top wing and you'll be darned close.

[CDNshy_guy]

Thanks
They will be parallel.
I'm building this one from scratch. No kit and no plans.
It's going to be a big 96" wingspan (glow powered).
I know this is a Micro forum but the principal should be the same when it comes to the CG.

I just want to see if I can build one from scratch and make it fly.
I realy like Biplanes so I decided to make it a biplane this morning.

Norm

[XX-Man]

Norm,

Just a reminder…
CG (center of gravity) should not be confused with CL (center of lift). Something we just don't think about that much. I think what you are concerned about is the center of lift on your wings and that depends on the airfoil, sweep and taper of the wings just for starters. Center of lift on air foils can change in flight as air speed and angles of attack change as well. There are several thumb rules that one can apply to finding the center of lift on a wing/airfoil. One I've found to be good enough for my purposes over the years is to measure a zone 25-33 percent of the chord length from the leading edge and that will just about cover it.

What you may want to think about is the relationship between CL and CG. CG should be just in front of the CL or your plane will be unstable maybe unfliable! (A danger in designing your own but don't let that stop you, that's the way you learn). Think about it this way… the center of lift is the pivot point or line that the horizontal stabilizer uses to hinge most of the weight of the airplane about. It is the horizontal stabilizer that "pitches" the airfoil, changes angle of attack, as a result makes your plane go up and down. In a biplane the CL is most likely to be an average of the two wings. Oh well, thank about it and….

Good Luck,
-Jim

[Richard Miller]

There are two ways to establish the CG. One is to consult the plan or the builder or get advice from whomsoever, the other is to work it out in a practical way.

Balance the model til it glides in a reasonable manner. If it is squirrely or erratic it more than likely needs the CG moved forward and decalage added. This means a greater angular difference between the bi planes or some negative incidence in the horizontal surface. Or a bit of both.

Back to gliding in a reasonable manner. Add just a bit of speed to the launch, a slightly faster than normal-gliding-speed velocity. What you want, the indication of suitable trim, is a slight divergence upward from the glide slope leading to a stall, or incipient stall, and recovery.

If there is no deviation from the straightaway, or if the model diverges downward, the cure is as above: CG forward, by increments, with the necessary decalage to hold it there until it responds as in paragraph two.

These are variously neutral [straight ahead], negative [downward] and positive [stability].

-Richard

[Mikerjf]

With the fuse horizontal get the vertical profile that both wings make and just calculate or layout the mac like it's one wing.

[Don Stackhouse]

We have a section on our website:

www.djaerotech.com

called "Ask Joe and Don" that deals with all sorts of questions, including this one. Just go to the AJ&D page and type "C/G of a biplane" or something like that into the search engine, and it will come back with a list of links to which of the over 300 articles refers to that phrase. You might want to browse a bit while you're there, you might find other useful articles on other subjects as well.

C/G involves a number of factors, including the tail moment arm and the tail area. The starting point is the Mean Aerodynamic Chord ("MAC") and Aerodynamic Center ("AC", for an individual flying surface it's located on the MAC of that surface, typically at about 25% back from the leading edge) of the combined wings. From there you find the AC of the whole aircraft (including the tail). You may or may not want to include some correction factors for the effects of aspect ratio and airfoils on the diferent flying surfaces, although in my experience these effects tend to cancel each other out for most models.

So how do you find the AC of a pair of wings? First find the AC's of the individual surfaces, then use a weighted average method to find the AC of the combination.

For example, lets assume we have a biplane (a "sesquiplane" in this case, where the upper wing has twice the area of the lower) with an upper wing of 200 sq.in., and a lower wing of 100 sq. inches (total are of the two is therefore 300 sq.in.).

On a side view, draw a line from the AC of the top wing to the AC of the bottom wing. The AC of the combination will lie on that line, closest to the larger of the two surfaces, at a distance proportional to the ratio of their areas. In this case it will lie 1/3 of the way from the top wing to the bottom wing (that's 100 divided by (100 + 200)).

Once you have the combined AC of all 300 square inches of wing together, you can find the AC of the combination of that plus the tail using the same method. This AC for the whole aircraft (assuming the effects of aspect ratios of the different surfaces, plus the effects of their different airfoils, can be ignored) is going to be the approximate location of the "Neutral Point" ("NP"), the C/G location at which the plane's stability is neutral (i.e. it goes where you point it, with no tendency to come back to the original attitude after it's been disturbed).

If you put the C/G ahead of the neutral point, you will have positive stability. The amount that the C/G is ahead of the NP is called "static margin" and is a measure of how much static stability you have. It's typically expressed as a percent of the MAC. If you calculate the static margin in %MAC of an airplane which has an amount of static stability you like, and use the same static margin on your new model, the static stability of the new model should be similar, assuming the two models are similar enough in general type and arrangement that other complicating details don't get in the way.

[Rotaryphile]

Wind tunnel tests conducted away back in the 20s, when most all self-respecting airplanes had two sets of wings, showed that the center of pressure of a biplane wing combination was at 22-23 percent of mean aerodynamic chord, compared to 25-26 percent for a monoplane wing.

I simply calculate the mean center of pressure of the two wings, applying weighting factors to allow for unequal size wings, and have had no trouble. As a last resort, a useful technique for airplanes of really weird configurations is: Build a cardboard hand-launched glider of the same layout, scaled down as convenient. Flat plate airfoil is just fine. Play around with the little glider, using paper clips or what-have-you to adjust the CG, until you get a nice glide with very little positive incidence difference between wing and stab, (if it has a stab).

It is important not to confuse center of pressure, (center of lift) with CG of an airplane. The CG is usually behind the center of lift, since the horizontal tail normally develops lift. I have flown aerobatic bipes with the CG as far rearward as 50% of the mean chord, with 20% horizontal tail area or more. Calculating the center of pressure of the entire airplane gets more complicated, since you have to allow for downwash over the tail. For commonly used tail moment arms, assuming that stab area is about 50% as effective as an equal amount of wing area is usually not too far off the mark, for a first cut.

[Richard Miller]

Rotaryphile,

Since I'm at work on a biplane of my own I found the information about center of pressure location - 22-23% on biplane wings of interest. Wonder why that is?

You mention a rearward CG and a lifting tail. This surprises me. All the HLG and free-flight models I've built for the last 45 or so years have had lifting tails, but I wasn't aware they were common elsewhere.

Otherwise, you seem to have things nicely nailed down. The critical point, it seems to me, is employing just the amount of decalage you need, no more. More is wasteful. Imparting a little extra speed to the model and observing how it reacts is about the best way I know.

-Richard

[Dave Robelen]

Hello Richard, Rotaryphile,
The term "lifting tail" is easily cofused with the airfoul used on the horizontal tail. All too often just because the tail has a flat bottomed airfoil it is termed "lifting". Models like the Hobby Lobby Telemaster come to mind in this case. For the majority of our R/C models setting up the model so that there is an upload on the tail is an invitation to a sudden "tuck" manuever.
I too have enjoyed the seemingly simple free flight models, especially the balsa HLG type, and even there moving the CG too far aft was an invitation to poor recoveries from the launch. The exception sems to be low ceiling indoor gliders where they are thrown in an overhand "arc" to the ceiling. The trim is super critical, but the results are rewarding.
Do you suppose the slightly more forward center of pressure on those early biplanes is a result of the two downwash fields mixing?
Regards, Dave Robelen

[Rotaryphile]

Hi Dave: Was initially a bit startled by "airfoul" (sic) - maybe the shape of a bird's tail airfoil? All my aerobatic models, whether bipes or monoplanes, have their centers of gravity well aft of their wing's center of pressure, which means that the tail has to be lifting when the airplane is in level flight. For aerobatics, I like to place the CG very close to the aft limit of stability. The test I use to determine whether an RC model is nose heavy is to trim it for upright flight, then roll inverted. If it dives, it is nose heavy. If it climbs, it is tail heavy. Trimming CG so that no down elevator is needed to stay in level flight while inverted needs rather small elevator travel, and usually results in a model that is a bit "nervous" in pitch, so I usually trim CG so that the nose drops about a degree per second while inverted, and in windy weather, trim a little more nose heavy. Placing CG near the aft limit makes smooth aerobatics much easier, I feel. Vertical uplines and downlines can be done with the elevator stick pretty much centered; no "down" elevator is needed to keep the lines straight, and the less work for the pilot means lesser opportunity for error.

I have never seen an explanation for why bipe wings have their center of pressure farther forward than those of monoplanes, but will consult my slew of aero engineering texts dating back to the biplane era. I expect that mutual wing interference is to blame, but don't know exactly how. The lift distribution over biplanes wings is highly peculiar. For example, there is a venturi effect between the wings that tends to suck the two wings together with a force that may exceed the weight of the airplane, in level high speed flight.

I used to worry about flat plate stabs stalling before the wing, but a little checking indicated that for normal layouts, the stab can't develop more than about half the lift coefficient of the wing, so a flat plate stab will not stall until the wing is into very deep stall. I normally use about 9% symmetrical section stabs for structural reasons only, since flat plate stabs tend to need draggy external bracing.

[Richard Miller]

Y'all,

When I think of lifting tails my mind goes back to the Starduster, C.1955. Everybody should build one of these. Wait a minute. First I think about birds and how they twist their tails to turn. We modelers probably got our ideas about using a tilting stab to turn a model in gliding flight from that. Then I think about the Starduster.

Thing is, with a free flight - 1/2A gas - if the tail's going to lift, make it biggish. The one on the Starduster is maybe 40%.

It is known that a couple of ships anchored a beam width apart, will tend to draw together. The Venturi flow. BUT, in the case of an a pair of airfoils you have the increased dynamic pressure on the under side of the upper surface and decreased pressure on the upper side of the lower. Separation is critical of course, which is like saying you've got to move your jaw to chew gum.

I'm enchanted by the means you use to refine your trim forces, Rotaryphile. Stuff like that had never occurred to me, but then I was never occupied in flying aerobatic models. By the way, did you ever do a scale analysis, wing area against volume of air affected, of multi-plane wings?

A note on separation. Dynamic pressure of either polarity diminishes as the inverse-square of the distance. And it would be nice to know something definite about just how much air, up and down, an airfoil effects, say as a function of its CL.

That Starduster. I took mine, freshly built, all surfaces and alignments carefully checked, engine - Holland Hornet - ready to run, to the Sheep Meadow in Central Park, NYC. Nobody there but me. Can you imagine. I put a 64th offset in the trim tab on the vertical surface [as I remember it now], did some glides, felt I was ready to fly.

I started the engine, set the timer, launched the model straight up, or very nearly so. It climbed to maybe a hundred feet in the couple of seconds of motor run, made a smart transition to gliding flight, and glid. That was it! Completely check out in one flight and ready for the big time. It was enchanting. Nothin' like that ever before, nor since for that matter. A little golden moment.

-Richard

[Dave Robelen]

Hello Richard,
Re. a "golden moment". My first 1/2A gas experience was with an early Goldberg kit. High pylon, big stab, very easy going, just a little shy on duration with my .049 Wasp. Emboldened by that experience I built the Spacer ( Starduster ancestor) and fitted it with a Holland Hornet. My first flight was going much like yours, except the engine quit with the nose vertical and the model flipped nose down and came all the way to the runway like a spear! That was when I became aware of the potential for fussy trim with the competition free flight birds. From then on it has been HLG and rubber powered projects.
My experience with aerobatics parrallels that of Rotaryphile in the trimming process. The exception being that I worked with monoplanes. Harold Debolt defined a force setup about 1960 with his magnificent Interceptor where the thrustline was high on a low wing machine, definitely above the CG, the stab was large, maybe 30%, Airfoil symetric of course, The wing had some positive incidence, but the stab had 1 deg more. It looked spooky, but with a 40% CG the trim was all but identical upright and inverted.
From flow visualazation work, the wing is affecting the air at least 2 chord lengths above itself, and probably more at very high values of Cl. My background is basically experimental vs theoretical, so you and I are often talking "apples and oranges". Still, I am picking up much from your thoughtful insights, especially the challenge to the "establishment" about the various reasons for lift. Keep up the good stuff.
Regards, Dave Robelen

[Rotaryphile]

Hi Richard: When I was deeply into competition free-flights, it was common to place the CG at the wing trailing edge, and give the wing maybe two degrees, at most, more incidence than the stab, with the same airfoil on both wing and stab. Stab area was around 40% of the wing. This layout was arrived at by patient trial and error by many talented, thoughtful, and above all, patient modelers.

I find that the larger you make the horizontal tail on an aerobatic model, the better it tracks, particularly in windy weather. One problem with bipes is that a bipe with a 40% stab would look just a bit peculiar, with the tail almost as big as one of the wings, so I usually don't go beyond about 25%. If a pattern model were built with the long fuselage currently in fashion, and a 40% stab, I would expect that it would need its CG somewhere near the wing trailing edge for best results. The only type of model that needs its CG close to its wing's center of pressure is a trainer, since trainers need very strong pitch stability. A model that tries to drop its nose automatically when airspeed gets a bit low is far more friendly to a beginner. Later, when the beginner has gained some proficiency, I often suggest trying a more aft CG, which makes the trainer much more responsive in pitch, and vastly better in inverted flight, since it does not need as much "down" elevator to keep its nose level. It will also spin much better, particularly when its control throws are increased.

Getting back to bipes, many of them tend to glide like bricks, and slow down very rapidly in tight turns. This is largely due to mutual wing interference. Away back in the 'teens, it was determined that the cross section of air influenced by a wing is nearly all contained within the smallest diameter circle that can be circumscribed by the wing tips. If a bipe had its wings separated by a distance equal to its span, its wings would perform almost exactly like those of a monoplane. Move the wings closer together, and they begin to interfere with each other very seriously.

Wind tunnel tests that were conducted in the 1920s established that mutual wing interference had exactly the same effect as lowering the aspect ratio. A bipe with an aspect ratio of 6.0 may suffer as much induced drag as a monoplane with a wing aspect ratio of about 3.5 to 4.0. To combat this, I like to use higher than normal aspect ratios on my own design bipes, and with an aspect ratio of about 8.0, they glide about as well as a monoplane with an aspect ratio of about 5.5 - typical of many of today's competition aerobatic monoplanes. Regardless of wing chord, a bipe with typical inter-wing gap needs about 90% of the wingspan of a monoplane to achieve the monoplane's induced drag numbers.

'Nuff said. I covered this stuff, and more, in a series on bipe aerodynamics in M.A.N. a few years ago, and reprinted in Air Age's "How To's", Vol. II, currently available in hobby shops.

[Richard Miller]

This is just to say I composed a thoughtful reply to the above emails and the *&^%$# system done away with it. Curse!

-Richard