V-Tails ARE Different

[HELModels]

I fly a V-tail and it handles differently than if it were a conventional tail. The wing has very little dihedral and yet it responds very quickly to rudder, rolls into the turn immediately. I believe there is a reason for this which is glossed over every time someone claims there is a difference. I claim the dihedral is in the tail. :) A conventional tail needs more dihedral in the wing, WHY?

[Ollie]

"The Famous V-Tail Debate"
See:
http://www.djaerotech.com/dj_askjd/askdesign.html

[JetPlaneFlyer]

The 'dihedral' in a 'V' tail does work in the same way as dihedral in a wing. Obviously the tail is smaller area than the wing, and the span is less, so the dihedral effect is not as strong, but otherwase it works exactly the same.

The dihedral like effect you notice with your model is therefore exactly what you would expect.

Steve

[Ollie]

There's more:
http://www.charlesriverrc.org/articles_modeldesign.htm
Tail Design and Structure

[HELModels]

Here is a variation on V's which I am about to start building. Do you think the V can be made asymmetrically, or should I tilt an equally sized V to get the asym effect?

Thanks,
Harry

[Ollie]

What is the purpose of your design?

Range of airspeed from slow to high speed?
Maneuverability range?
Size?
What kind of efficiently?
How much of strength and stiffness?
Etc, etc.

[HELModels]

Purpose of design is to fly, ha ha. Alot of these questions have been answered by a more conventional design. The size will be determined by the wing I already have - about 55" span. The draggy extra pod will hold a camera. I know there are other ways to add a camera.

So, Ollie, do you think I can get the asym tail effect by tilting a normal V?
This goes with the discussion on V designs.

I also know that projected area is not enough or best way to design a V and making it work with an asymmetric planform is an added twist, but no pain, no gain.

Harry

[Ollie]

I had a V-tail with a skinny tail boom. The boom twisted with torque because I didn't design enough twisting stiffness. However, I could control the plane just fine. It just looked strange when the boom twisted about plus and minus 20 degrees in flight.

I didn't use an asymmetric planform. I wouldn't advise you using an asymmetric planform for a V-tail.

[Sparky Paul]

This is an elevator only vee-tail. Ailerons for turning.
The position of the tail lags the position of the model significantly in a roll. :)
Needs a much stiffer boom.

[HELModels]

A cracked boom will do the same, obviously. So, with the tail all twisted up in a bunch, did you really notice a difference in the handling or did you just see it twisting? I'd be interested in hearing more about the effect. I've flown a V in the elastic state and it actually seemed to damp some heavy wind gusts. I had to work the rudder like a ratchet. That makes me think you briefly saw a tendency to resist turning and then it would briefly seem normal, but only you can tell us what you noticed, if you would, please.

Harry

[Ollie]

You can't far with just subjective (art) discriptions applied to reasonal objective (science) design. You must understand the aerodynamic equations with measurements data.

Perhaps you don't want the Modeling Science forum.

[Sparky Paul]

On the Ninjas, the tail obviously lags the fuselage when rolling. Once the forces get adjusted in the roll, the tail boom unwinds and the vee looks normal.
On a GL, with that twisty a tail boom and rudder-vators, at speed any rudder input would twist the tail, and the plane would dart around uncontrollably.
That wasn't amusing, so I changed out the tail boom to a much stiffer version.

[HELModels]

I didn't use an asymmetric planform. I wouldn't advise you using an asymmetric planform for a V-tail.

To stay on topic, V tails must be different otherwise an asymmetric planform wouldnt matter, the same principles would apply regardless of tail type. I am capable of designing a V tail (already have) or more conventional tail(did that too). The idea of an asymmetric layout is to compenate for the lateral imbalance by hacking off enough of the horizontal stabilizer. The rudder is a bit oversized on alot of them too. A V combines the horizontal and the vertical. I am going to claim that if a V is no different aerodynamically than a cruciform or even a T, then it could also be designed to work on an asymmetric layout. If youre only reason for saying you wouldnt advise it is because you have made assumptions about my capabilites, then that isnt advise. What exactly is your reason for advising against a V tail on an asymmetric layout?

Keep in mind, I am not looking for a fight. I was only looking to discuss V tails, and in this case V tails on an asymmetric layout. Youre never under any obligation to answer, but if youve got something positive to contribute, then contribute away. Also keep in mind, that were talking about model airplanes here, nothing manrated, nobody's bacon is on the line.

Harry

[Ollie]

The V-tail depends on many design aspects. Paul and I told you about tail boom torque stiffness. The V-tail depends on stability and control in pitch and yaw. Next the pitch stability depends on the wing airfoil and CG position. The V-tail yaw depends on the tail drag. If the tail drag is perfect with the rest of the plane drag then it is not zero yaw. Etc, etc.

My point of view is the V-tail design is integrated with the whole plane configuration.
If you want only the V-tail not any integrated whole plane, I will shut up. Your choise.

[Sparky Paul]

Harry, I frequently don't follow the rules when making planes.
I do look at miserable failures in the past to possibly avoid more of the same in the future, both mine and others.
I'm not into aerodynamic perfection, just something that flies, and if it generates interest and awe, more the better. :)
Build your thingy.

[HELModels]

Back talking Ollie :) wont git er done, so I finished the repairs needed on the test wing in order to do the first test. I didnt use an extra pod, so the weight was uneven. I threw it with motor at full power and it took off into a climbing right turn. I had left trim in and with the prop on the short side, I was expecting it to help balance. I froze up and gave it more right and it crashed, broke the wing again.

The next test will include a balancing weight. I had planned to do that and decided to just throw it - hasty. The goal right now is to get it to fly straight with power. I expected it to act like a twin with one motor out, but the side-to-side CG is much more important.

Harry

[HELModels]

I fixed er again. The first test crash showed I need to counter a hard right turn. I will add weight (and some drag) to put the CG at the center of the wing again and see what the tendency is then. I plan to modify the tail last.

Harry

[HELModels]

I added a draggy pod the left side of the wing without adding any more weight than the weight of foam. I tossed it motor at full and the thingy went left instead of right. I believe the pod balanced the lift/drag as the CG only shifted a 1/4 inch to the left. With the motor off the glide was straight. I adjusted the thrust angle toward the right and at full power it was straighter. There is alot of stuff to balance.

Harry

[Sparky Paul]

You might be zeroing in on why there's so few of these things flying around. :)

[HELModels]

Just might be, but I am going to keep tweaking it. Here is a shot of the actual crash dummy. It serves the purpose since it is pretty durable. It might be time to modify the tail and see what that does. It is flyable right now, but the power on/power off difference is there.

Harry

[HELModels]

I made the modification to the tail. I tilted it to the short side. I need to smooth out the outboard pod and reduce the drag from it. If that doesnt work enough, then I might shift the main pod to the right some more, or move the outboard pod inboard more. It will probably crash, but it is tough and that is what is needed to get through the testing.

[HELModels]

Thingy crashed again, but I wasnt surprised. I removed the pod completely and with the V tilted and mixing removed, I tossed at full power. It did what it did on the first test and that was hard right turn. This convinced me that the drag of the extra pod is definitely a key element to balance. I then took some spray foam and squirted it where I thought the pod should go. I waited for it to dry and gave it another toss. It tracked straight under power. No right turn under power. No real progress either. I've resisted adding weight, but it must be part of it.

Suggestions on lateral CG?

[Ollie]

When you fail Part III, Part IV, Part V, etc, etc. then change your point view. Think differently. Learn from a better master designer like Don Stackhouse.

Design Philosophy

by Don Stackhouse

There have been some rather heated discussions lately on the R/C Soaring Exchange about the relative merits of computer analysis versus old cut-and-try methods in the aerodynamic design of wings. Similar threads have argued over the shape of the "ideal" planform, the "best" airfoil, the "optimum" aspect ratio, and the validity of modified airfoils or of blending from one root airfoil into a different tip airfoil. All this controversy reminds me of one of my favorite stories:

The Four Blind Men and the Elephant (an old Hindu parable)

One day four blind men encountered an elephant for the first time. They approached it cautiously, but with great curiosity. The first one grabbed hold of the trunk and declared "Aha! An elephant is just like a snake!" The second found an ear and replied "No, an elephant is exactly like a tent." The third bumped into a leg and decided the elephant was just like a tree, and the fourth caught the tail and maintained that the elephant was just like a rope. They all went home arguing, each steadfastly insisting that he was right and the other three were wrong.

The flight of a model sailplane is a complex phenomenon, each portion of the model seeing its own unique set of conditions at any given time, yet still having an influence on all of the other parts of the model at the same time. In addition, we expect our models to perform well at a wide variety of operating points within the overall flight envelope. To help us achieve this aim we have available a large database of experimental data and theoretical analysis tools, plus the insight garnered from all of our own experiences and the experiences of others, and the results of actual tests of the model. The results depend on how well we use all of this information together to reach the final design.

If we get hung-up on one parameter, or one design technique, or one phase of the design process, we automatically give ourselves a case of "tunnel vision". There is no single airfoil, aspect ratio, planform, tail size or type, etc., that is optimum at all flight conditions for even a single model, much less a variety of models.

While it is true that the section at the mid-span of a wing of two wildly different tip and root airfoils may have or may not have any of the characteristics of its parents, it is also just as risky to believe that the airfoil that is optimum for the conditions at the root will be equally appropriate at the tip. Ideally you should study the sections at a variety of points along the wing, as well as local chord, twist, flow characteristics, et cetera.

The effects of what is happening at the tail, along the fuselage, along the span of the wing, all influence each other in different ways at different flight conditions, with corresponding effects on the overall control, stability and performance of the model.

None of our design tools is perfect. None of our data is completely reliable at all conditions, and some of it isn't very reliable at any condition. In my experience the best approach is to use all of the available tools and data to the fullest extent possible, then look for the consensus forming between the different approaches. This way the strong points of the different approaches can compensate for their individual shortcomings. The first '93 Monarch hlg went through 150 hours of computer work, PLUS six fuselages, seven tails and eight wings before we froze the design.

If you try to build an elephant with only a tree, or only a rope, a snake, or a tent, your result will almost certainly fall short of your expectations. Likewise, the model that was designed with only theoretical methods, or only past experience, will probably not be the best design possible. Only by using all of the available tools as cooperative members of a team effort can you achieve a design that is more than just the "sum of its parts".

[HELModels]

No CAD/CAM, no formulas, basically just an R/C chuck glider. I am now using a water bottle and a rubber band for a pod with some lead tape. In the normal configuration this plane flies great. The data I have gathered has been useful and is not the end of it. Basically, I want a plane with an extra pod for carrying a camera, battery, maybe a servo, maybe a GPS.

I have tried a simply asymmetric wing without changing anything else - result noted

I have tried above, but added drag - result noted

I have tried above, but moved the drag - result noted

I have tried above and twisted the tail - result noted

I have taken a perfectly fine flying plane and explored the various ways to screw it up and still have it fly. :D So far, I have only crashed with pure unbalanced lift. I have noticed differences in power and power off. What I think will probably happen is that I will settle on the original V tail with a slightly offset payload pod and the main pod will be offset just enough to maintain lateral CG at the center. The chuck glider stage is to get a better understanding of where the limits are.

I am examining the elephant after altering it. Is it still an elephant?

Harry

[Ollie]

You will kill the elephant sooner or later. Poor elephant. :)

[HELModels]

Dont worry Ollie, I put the elephant back together using the good wing and then went out tonight and had a great flight. I assure you the anesthesia worked, it had no memory of all the cuttin. I'll test again, but I wanted to fly nice tonight. :)

Harry

[sarmoby]

Hi all-

In my group of slope gliders I must have half a dozen v-tails and from building them and flying them I can observe the following:
1. Many are undersized for the needs of the plane, especially at low air speed and near stall. Don't design for the cruise environment, design for the transitionary regime.
2. Keep the control surface away from the tip of the fixed section. This causes the effect of turning the v-tail into a shovel with uneven airflow that stalls more readily.
3. They typically need some form of rudder differential for a clean yaw effect.( rudder = more up than down or similar mix)
4. I suspect a higher aspect ratio v-tail performs cleaner than a shorter v-tail of the same area though the structural loads may be higher.
5. When possible, set the control linkage to pull for up( ask me how I know).

Back top the debate
Scott

[ghoti]

I heartly approve of Harry's cut and try approach (design by iteration) while being respectful of Ollie's insightful comments. Pigs can't fly but perhaps elephants can.

The best tools for this, IMO, are tall grass to crash in and EPP plastic foam with C rod across the LE of the wing. This makes for crashing with impuniity.

Is there a flight simulator that allows such changes as these? Such might streamline the data flow.
Bill

[reylf_gnijieB]

Harry's having fun. Go, Harry!

My planes don't fly as well as the high-perfomance tweaked designs that I can buy. But, for a couple of dollars in wood, I put something in the air that is fun, and most important to me, different. I haven't built a V-tail yet, but that's next. Thanks to all for what I can learn by reading your arguments, Special thanks to guys like Harry who get Ollie and Paul to share their knowledge by stirring the pot.

[HELModels]

I've been too busy flying the V-tail in the normal configuration to mess with the asymmetric configuration. As you can see, the base wing is from a aerobird challenger. My version handles great. I can fly in the backyard and turn around at ground level + 10 feet, between the trees, straighten out, come down the hill where I stand, hit full power, climb to about 30 feet, cut power, turn around above the little trees in the neighbors, shoot the gap, 1/2 throttle and go back up hill, then repeat. I admit this type of flying usually ends after about 5 laps when I clip a tree branch and spin to the ground and need to glue something. So far the weakest link is the front of the pod, as that is balsa and the removeable cowl doesnt really add strength, just energy absorbing. In a crash, it flies off and the balsa might break. The other way I fly it is to take it to a larger field and get it up in some lift and try to stay up. These flights usually end with a nice landing, unless the wind is really strong and then landing is tricky no matter what you fly. Control surface authority and maneuverability are excellent, which makes all the difference in wind. I've flown it in 25-30 mph winds in the winter. I wouldnt want to try any of this with the stock aerobird, as they are not very responsive and heavy. The creases in the wing are from flying in high winds.

Harry

[slowmofo357]

I have one for sale...

http://www.rcgroups.com/forums/showthread.php?t=717740#post7858480

[HELModels]

What you are selling is an aerobird challenger, mods or no mods, it includes a lousy radio, brushed motor, heavy battery, etc.

What I have been building only uses the wing. The motor is brushless. The pod is molded. The V-tail is redesigned. The electronics are first rate and the battery is lipo. I get 10-15 minute flights on a battery which is 1/2 the size, weight, capacity.

Here is a picture of the pod in mid build:

[macboffin]

Vee tails do have a dihedral effect.Believe it or not, having built very many aircraft with vee tails, the optimum angle is 90 degrees between.
Some of these airraft were fully wind-tunnel validated on account of being for serious military UAV use. Currently I have a AP plane with this set-up, camera in thickness of taill fin as illustrated.

[HELModels]

Interesting that you notice the dihedral effect. I'm surprised that you would say 90 degrees is optimum. Is this optimum for shorter moments like the planes you pictured? I am using about 110-120 and it handles great - no bad quirks, plently of authority.

Anyway, right now I am at an impasse esthetically. The aerobird wing is kinda sharp and modern in planform while the tail I designed is more old timer looking and matches the pod nicely. What I need to do is use roughly the planform of the aerobird wing and round the tips to match the tail. This is purely esthetic, but for this type of model it counts.

Then there is the aerobird airfoil which works great for this plane. I have examined it by taking a slice and I might be able to build something close using a very thin and high camber NACA.

Harry

[macboffin]

90 degrees does appear to be optimum for all types. The "Bonanza" came as conventional tail or Vee tail, with 120 degrees stting ; it had a disconcerting "galloping" tail twitch at higher speeds, they figured too much angle, but since the Vee tail version wasn't selling so well they didn't do a different version. Regarding tail moment, for the high aspect ratio AP plane, CG is at junction of fuselage and leading edge of wing, due to swept forward wing, so not such a short moment arm ; the Twinwings, ( another pic to illustrate) the CG is just ahead of the trailing edge of the front wing / fuselage junction, so again, not that short a tail moment arm.

[macboffin]

Poor pics, will repeat

[macboffin]

Swept forward wing Vee tail, and Twinwing Vee tail

[HELModels]

Mac,
You build some nice planes, rugged looking. I hear you when you say 90 degrees is optimum, but I figure optimum is relative to the plane and flight mode, maybe? Mine is a smallish plane and not really very fast, it will turn on a dime and is stable in pitch and yaw/roll. Hands off the stick and it returns to straight and level. So, maybe optimum for this plane is 110 - 120.

All the advice on sizing a V, says pick the size of the semi spans based on formulas(some opinion on which formulas) which consider the original vertical and horizontal sizes and some chosen angle for the V. I've also read where if you want more vertical influence, make the angle closer to 90, more horizontal influence, shallow it some up to 120. What I'm hinting at is that the angle shouldnt matter much as long as the result is enough stability.

What makes 90 degrees optimum in your experience?

Harry

[macboffin]

Mac,
You build some nice planes, rugged looking. I hear you when you say 90 degrees is optimum, but I figure optimum is relative to the plane and flight mode, maybe? Mine is a smallish plane and not really very fast, it will turn on a dime and is stable in pitch and yaw/roll. Hands off the stick and it returns to straight and level. So, maybe optimum for this plane is 110 - 120.

All the advice on sizing a V, says pick the size of the semi spans based on formulas(some opinion on which formulas) which consider the original vertical and horizontal sizes and some chosen angle for the V. I've also read where if you want more vertical influence, make the angle closer to 90, more horizontal influence, shallow it some up to 120. What I'm hinting at is that the angle shouldnt matter much as long as the result is enough stability.

What makes 90 degrees optimum in your experience?

Harry My experience especially with UAVs, but also smaller types.
As you say, more pitch authority, widen them out, more rudder, close them in. Reason for popularity with UAVs is that most have safety parachutes for when it all goes wrong.You chuck the 'chute out of a dorsal hatch and it passes between the fins before deploying.Not so easy if there is a vertical fin in the way!

[HELModels]

Mac, I did read where you said wind tunnel validated. I still have to wonder about the why on 90 degrees and that maybe it isnt the angle but the combination of angle and sizing for that particular plane. I'm having trouble thinking that only one angle is best.

Harry

[macboffin]

Mac, I did read where you said wind tunnel validated. I still have to wonder about the why on 90 degrees and that maybe it isnt the angle but the combination of angle and sizing for that particular plane. I'm having trouble thinking that only one angle is best.

Harry Thats a fair comment. The statement about "90 best" perhaps should have been "usually best" since particular airplane configurations might dictate other angles. Factors such as wing down wash, canards etc can influence tailplane positioning and angles, as well as control mixing.Faced with an unusual configuration I would start at 90 degrees and modify if necessary. One factor about Vee tails often overlooked is that they give a longer tail moment arm than a normal tailplane ( and fin) at the same distance from CG/wing CP. The upward sloping line from CG to centre of Vee surface effort is longer than the (usually straighter) line to a conventional tail group.Also, due to the higher centre of effort, the up-pitching power is slightly greater.(Same applies to T tails of course.) Picture of a T tailer enclosed ; was later converted to Vee tail for comparisons.Result, Vee preffered.

[HELModels]

Here is something which I have been wondering about, Mac. Alot of these RCAP like simple autopilots use FMA copilot for stabilization and rudder controls route- a vtail makes this difficult, I believe. If a Vtail were hooked up to an FMA copilot and a Vertical stab/rudder were added and controlled by an RCAP, would the V still work? It would look like a Vtail with a vertical smack dab in the middle.

[macboffin]

It works ok for the"Predator" UAV, although it's the other way up on that bird.
(The reason why, the central fin is primarily to keep the prop off the ground at rotation/landing) The multiple fuse / fin intersections cause a fair bit of drag, but of course theres a prop just behind sucking hard, so less effect due to that. I'd be inclined to put the central fin underneath, easier structurally, and you'd get a bit of positive roll input as well as rudder input, not a bad thing.Make a little chuck glider and check it out with fin under, and over.(Moveable rudder).

[HELModels]

This is an elevator only vee-tail. Ailerons for turning.
The position of the tail lags the position of the model significantly in a roll. :)
Needs a much stiffer boom.

Well, I cracked the boom on mine recently and it now is very "springy". I found out yesterday after a cartwheel landing the crack is growing. I thought boy this boom doesnt seem so stiff anymore. I didnt let that stop me from charging up the battery and going for a flight in 10-15 mile an hour winds.

What I noticed is that it seemed a little jumpy, but also resistant to gusts. I didnt notice a problem until I turned and it needed more opposite rudder to stop the turn and would overshoot sometimes, going into a turn in the other direction. In steady winds, no intentional turning it would occasionally make lots of little oscillations, but would resist getting into a bank. This seemed contradictory, but is what I noticed in the sense that it seemed both more stable and less stable depending on the situation.

Could this contradiction be another characteristic of a V? Is it the V that provides the gust dampening?

[HELModels]

To clarify, it was almost like exponential stability. In straight and level, it resisted the gusts by absorbing them, but it was divergent in turns, requiring extra control movement to stay ahead until the oscillations subsided.

I dont believe a cruciform tail would be as tolerant to gusts with a springy boom.

[HELModels]

Oh well, so much for studies in aeroelasticity. I couldnt stand it any longer and jammed a dowel into the boom with some PU.

[macboffin]

I should think you are right about that ; but a springy boom doesn't sound too good an idea, on any kind of tail, because unless the tail surfaces are operated by rods straight down the center of the boom,any movement of the tail relative to the servos will result in unwanted surface movements.
A related effect is that caused by bowing, and thus effective shortening of control rods due to G forces when manouvring.Pull a couple of G, the rod shortens, and if the horn is on top of the elevater extra movement results ; if the horn is under, the opposite occurs, less course change than should have resulted from the amount of movement commanded. So long, flexible rods unsupported along their lengths can in the worst case result in the wing tips meeting over the fuselage at the bottom of the loop! (Seen it happen!).

[macboffin]

Oh well, so much for studies in aeroelasticity. I couldnt stand it any longer and jammed a dowel into the boom with some PU. Sounds good. CG checked after open-boom surgery?
Might well save you from an experiment regarding the elasticity of the ground in response to sudden arrival of a plane!

[HELModels]

Sounds good. CG checked after open-boom surgery?
Might well save you from an experiment regarding the elasticity of the ground in response to sudden arrival of a plane!

Yup Mac, CG is good. I've done that experiment with the ground many times. Not so much lately, as in this week :)

Harry