Guys,
Just thinking about a fun ship for next summer was wondering how much separation from the trailing edge does a full flying elevator need to be to effectively act as a turning control as apposed to acting first as a high lift device then adding the turning moment. There was a discussion about why flying wings like the Fierce Arrow don't stunt smoothly. The general consensus was that with the "elevator" on the trailing edge it didn't actually turn the wing, it first acts like a high lift flap then finally pitches the wing over.

So thinking about it, looking back at the Voodoo wing it seemed to turn really good without being touchy with the short tail booms.  How close to the trailing edge could you be and get a true elevator control?

Best,        DennisT

I'm willing to bet it depends entirely on the wing, airfoil, wing loading, etc as no two are alike enough to give a definite answer .

I've been wondering what I can do for easy money.

Get plans and laser cut ribs from Walter Umland, and build an Agile Arrow.  i consider it the best stunt airplane I have ever flown.  I've only flown my own airplanes, so it may not be the best in the world.

Here is the appropriate website. http://www.builtrightflyright.com/New_Web_Pgs/kits/AgileArrow/AgileArrow.htm

I can remember as kids my big brother Gary and I used to take surviving wings from crashed "Veco" stunters like Smoothies and make flying wings out of them by installing motor mounts in the leading edge and using the flaps as "elevators".  They all flew but I've no recollection of the "quality" of their performance.  It is very likely that, at the time, I wasn't aware there was a "quality" quotient!

In my advance, more informed years, I find myself pretty much in agreement with Mr. Rush.

To which I could add...

When as a mini-adult around 1966 I decided to start flying stunt for the big bucks after moving to California I installed flaps on my very nice flying Mathis Coyote because I knew that, to be a "competitive" stunt ship, it needed to have flaps.  Absolutely killed the turn rate...thus my introduction to "adverse" (with respect to desired pitch/turn rate) pitching moments and the birth of my more or less organized pursuit of stunt ship design.

Ted

Seems the flying wing for stunt is fascinating because of the potential lower weight and simpler build. I remember a discussion about the "Arrow" and that's were the idea that the TE elevator didn't achieve the best rotation came from and that separate pitch and lift surfaces were better. But perhaps the layout of the Arrow is not optimum for this type of stunter. Some other layouts like Larry S's Werewolf and Red R's Stuntwing might with some separation of the control surface have better corners.

Best,         DennisT 

I don't think it's as wrong as you're making it out to be, Howard.  It's poorly stated and it implies a chronological relation (first less lift, then pitching) where one doesn't exist, but IMHO it's kinda right.  An upwards-deflected trailing edge flap on a wing is going to both rob the wing of some lift and add some positive lifting moment (or add negative lift and positive lifting moment, if you want to think in engineering terms). 

   I think what it does in the transient case is vaguely interesting, but the problem really is that you are pitching it one way and putting in camber the other way. An extreme example was seen in an airplane Bob Hazle built for VSC, where not only was the trailing edge hinged, it was a compound elevator with 2 segments. It was reported to be nearly unflyable because of the pitch one way, lift the other way effect. As small deflections it seemed as if the controls were reversed, and at large deflections it went the way it was intended.

    I am pretty sure that the offset hinge Fierce Arrow worked about like a typical combat airplane, just with an exceptionally short boom. The version with the conventional hinges definitely had the "wrong way" problem.

    Brett

The version with the conventional hinges definitely had the "wrong way" problem.

Negative direct lift control?  Could be.  I suspect that more likely is the extremely narrow CG range at which such a plane is flyable.  All the more reason I should build a Fierce Arrow.

If the transient negative lift is of the whole dadgum wing, rather than just the elevator, maybe so, but flying wings don't have much pitch moment of inertia, so I'd think the transient wouldn't last long.  This calls for a Fierce Arrow with a barbell that can be used to adjust moment of inertia.

Maybe later.  I should be working on a stunt plane and typing up contest results now.   

    Yes to the first.  The second is a nice-to-have and everybody is glad you do it, but is not going to affect your participation in Poland.

    Brett

Here's a side view of the aforementioned wing, and a front view of the aforementioned Bob Hazle.

    The "pitch around faster crowd" should know - this airplane is barely controllable.

     While we are at it, Ted asked this one a long time ago, but it's relevant here, for the same reason - which way should the flap go on a canard? If you get it wrong, it's worse than having no flap at all.

    Brett

It's certainly been hinted at a lot here, but I don't think it has actually been written out in this thread. So I'll just say it. The "Fierce Arrow" has two disadvantages over the Red Reinhardt and Jim Thomerson designs:

1) It is very short coupled.

2) It is highly swept.

The first requires higher control forces acting over a small distance and gives less c.g. range. The second should, I'd think, present yaw-roll couple problems, more spanwise flow, and perhaps tip stall tendencies.

At VSC 14, there was the then latest iteration of the "Fierce Arrow" and some discussion. I asked Bill why he didn't just reverse things to get the control surfaces further from the a.c. He was quite aware of the advantages, but responded simply, "But then it wouldn't be a 'Fierce Arrow.'"

SK

     While we are at it, Ted asked this one a long time ago, but it's relevant here, for the same reason - which way should the flap go on a canard? If you get it wrong, it's worse than having no flap at all.

    Brett

Ok, I will bite - "which way should the flap go on a a canard?"

Synchronized to move in the same direction as the fore plane or opposed to it?

(By the way, that is a genuine question.)

This is interesting.  An easy way to find out is to build a simple sheet 1/2a where all control options can be set at the flying field and go try it.  I had one 1/2a canard as a kid.  It flew a decent level flight with a noticeable nose up, and was very sluggish to respond to control.  It's pretty hard to keep these from being too nose heavy.  My guess would be that flaps would more take over and function like a regular elevator while overpowering what the canard was doing.  Maybe somebody will try that 1/2a and show us.

Dave

A friend got me to test fly the Dick Sarpolus design that appeared in Flying Models some years ago.  The engine is mounted between booms on the front of a stubby fuselage.  Not a super stunter, but more of a souped up Ringmaster.  We discovered a few things.  As originally built the controls were hooked up as a canard, both flaps moved the same direction to get more lift.  VERY strange.  Under power it was very stable until you moved the controls.  Then it tended to mush and suddenly go into a loop.  As soon as the motor quite it handled and landed predictably and smoothly.  Obviously the prop wash on the flaps over powered the canard, up to a point.

Next, locked the flaps in place.  It flew fine and was stable and controllable.  Much like a Ringmaster, but a little better.

Finally, hooked up the flaps to move opposite the wing.  It flew OK, but didn't turn as tight.

Regards the picture- the combined flap/elevator is about 4-5 times as wide as it needs to be.  If you glue the forward portion solidly in place it will be something like Reinhardt's Stuntwing, or a Wakkerman F2D.  it'll fly better, but the elevator is still about twice as wide as needed.  Might have a CG problem with all the paraphernalia behind the balance point.

I think you can make a canard stunter that has the flaps move the right way (same way as the canard surface) without artificial stability augmentation. 

      Howard tangentially refers to the fact that you want a big canard to overcome the fact that the CG is well ahead of the lift vector from the wing, however, the bigger the canard, the further forward the CG has to go, making the adverse pitch moment worse, meaning you need a bigger canard. This continues until the canard is bigger than the wing, say, 4x as big.

    The solution is to make the flaps move the wrong way (canard down, flap up), however, that puts you in the same situation as Bob Hazle had, and to a lesser extent, the Fierce Arrow - you are trying to do an outside loop in a Piper Cub, with the camber in the wrong direction. You would do better with no flaps.

     You really want the CG to be near the CP, so that creating large amounts of lift does not try to pitch you in the wrong direction, and a canard does not do that. The CG has to be way ahead of the lift vector from the wing to make it stable. The most recent go-around on canards was when it was suggested that if you put the CG at around the CP of the wing, how great it would maneuver. Which it will, even without control input, unless you do something odd.

   Note that there are other unfortunate issues, like, the CG is in the wrong place for the leadouts to come out the wing.

     Brett

     

No, I was thinking of an actual canard where the little wing is in front.  There's one sorta obvious way to do that, and another way that's less pure.  Think of ways to get pitching moment from a canard without Cm_α or lift from flaps without pitching moment.

   With or without flaps, or without pitching moments from the wing airfoil alone,  I think I still see ~60 lbs of lift centered on a point 6"+ behind the CG, pitching the airplane the wrong way. You have to make the canard big enough or the long enough to create something like 50-60 ft-lbs of torque, 30 ft-lb just to stay level. That the canard lift helps rather than hurts is the supposed advantage.  But that also makes you move the CG forward to maintain stability.

    If you try putting the CG on the CP of the wing, (also with or without flaps), it won't be passively stable unless you let the canard free-float, and then put in reflex like a flying wing.

    I suppose I do not have sufficient imagination, so you will have to spell it out for me.

 
     I have a rocket glider with a free-floating canard, so it's stable and has no lift or pitching moment in the ballistic boost, and then the canard is locked down when it is supposed to glide. Works pretty well, when the explosive-driven piston that locks the canard doesn't blow the entire front end off.

    Brett

I suppose I do not have sufficient imagination, so you will have to spell it out for me.

Looks like you do.  Floating the canard is the main trick.  Also, if you have a high aspect ratio wing with enough sweep, flaps on part of it are far enough forward not to fight the canard.  I made a hand launched glider that gave me the impression that the flaps would work.  I worked with a guy from Kawasaki who gave me some stuff on the Shinden from WW2.  It looked kinda like a Long-EZE.  The drawings appeared to show a floating canard.  I put a overfloating rudder on the front of an RC glider once.  Its main effect was adding drag.

That's a cool glider.  Who thought it up?

You can control the canard with a tab on the back, but I'd use a stabilator (for some of the same reasons one would use it with a wing).  The mechanism is a little challenging.  The forward rudder on the RC glider was for stability, with no control hookup, and was driven by a weather vane with linkage to drive the rudder in the same direction as the vane, but with a bigger angle.   

Guys,
Just thinking about a fun ship for next summer was wondering how much separation from the trailing edge does a full flying elevator need to be to effectively act as a turning control as apposed to acting first as a high lift device then adding the turning moment. There was a discussion about why flying wings like the Fierce Arrow don't stunt smoothly. The general consensus was that with the "elevator" on the trailing edge it didn't actually turn the wing, it first acts like a high lift flap then finally pitches the wing over.

So thinking about it, looking back at the Voodoo wing it seemed to turn really good without being touchy with the short tail booms.  How close to the trailing edge could you be and get a true elevator control?

Best,        DennisT

Different airfoil. Low PMI.

So thinking about it, looking back at the Voodoo wing it seemed to turn really good without being touchy with the short tail booms.  How close to the trailing edge could you be and get a true elevator control?

Best,        DennisT

The VooDoo turned pretty well if the balance point and the stabilator movement were trimmed properly for a specific speed range.  If you got it trimmed for a plain bearing ST and then bolted on a G-21 it would suddenly turn like a tank- nose heavy and not enough control power to move the stabilator at a higher speed.

So thinking about it, looking back at the Voodoo wing it seemed to turn really good without being touchy with the short tail booms.  How close to the trailing edge could you be and get a true elevator control?

Best,        DennisT

The VooDoo turned pretty well if the balance point and the stabilator movement were trimmed properly for a specific speed range.  If you got it trimmed for a plain bearing ST and then bolted on a G-21 it would suddenly turn like a tank- nose heavy and not enough control power to move the stabilator at a higher speed.

  Thats because the tail volume was very small, so the CG range was correspondingly small. I can easily guess what happened if you got it 1/16" too far aft, too.

    Brett

Interesting stuff but still want to know how far away from the TE does the stabilator need to be 1/4", 1/2"?

Second, does the combo layout with the wide center section and narrow outer wing layout smooth the turn?

Best,          DennisT

      Howard tangentially refers to the fact that you want a big canard to overcome the fact that the CG is well ahead of the lift vector from the wing, however, the bigger the canard, the further forward the CG has to go, making the adverse pitch moment worse, meaning you need a bigger canard. This continues until the canard is bigger than the wing, say, 4x as big.

   
     

so,, what your saying is that my new Impact derivative is actually a Canard with the correct proportioned canard wing,,

look at it this way...

When was the last time you saw a canard entered in a full scale aerobatic event...let alone a competitive one??

Ted

p.s. This excludes, of course, those canards the canard surface of which is 4X the area of the "wing".  Mark's canard Impact for instance.

I pretty much agree with everything Brett has said...no big surprise!

For Control Line purposes I think you could probably design a canard stunter that would work very well under a specific set of conditions.  The CG, by definition, must be pretty far ahead of the center of lift of the main plane (the big wing in back).  For that "given" set of more or less ideal conditions you could then compute a canard size and distance from the CL to a provide adequate pitch capability to fly well.

Unfortunately, stunt contest conditions seldom provide a predictably ideal set of conditions.  As conditions change...primarily wind velocity...the extended distance between the CG and the CL will become an issue as the airplane accelerates and decelerates as the wind vector changes throughout the hemisphere of flight.  As the airplane speed up or slows down while maneuvering the distance between the two forces will have a dramatic effect on the the control input necessary to maintain the desired (necessary according to the rule book) rate of pitch (and, therefore, g forces) change.  As the airplane flies faster the moment between the CG and the CL will increase and try to open up the loop/corner radius and vice versa when it slows down.  Control input to maintain concentricity and maneuver size will have to be varied quite dramatically throughout any maneuver that is exposed to acceleration or deceleration due to the winds (or, for that matter, other factors).

One of the primary advances "conventional" stunt ships have made in their 70 or so year history has been the realization by designers that the tail needs to be large enough to allow a CG location as close as possible to the CL of the main plane (the wing).  The closer the two are to one another the less the control inputs to retain a given rate of pitch will change as the airplane speeds up and slows down while maneuvering in modest or greater winds.  The conventional configuration allows the two to be essentially co-located as long as the tail is large enough to allow stability with the CG that far aft.

A canard configuration, on the other hand, hasn't that luxury and, to be stable and flyable, the CG will always be well forward of the CL and increased G forces while maneuvering will always demand greater control deflection during acceleration and vice versa.

I think....

Ted

Chris-

I don't fully understand the question, I guess, but I'll try to contribute. Is your illustration from an R/C Proving Grounds calculator? I think I've seen it before, but I don't have a direct link to it. If so, there are a couple things to look out for. First, their static margin expectations are more attuned to R/C and full-sized aircraft, whose c.g.'s are significantly behind those used in CL. Also, the N.P. locations are based on a simpified analysis that I did not fully "get" and which also actually depends on vertical aft-wing placement and a lot more. I have pretty well given up on reasonable analyses, but there are some out there, including good papers from Stanford aeronautical. Anyway, their N.P. is not necessarily accurate, and for sure, you don't want to locate it as though the aft wing is as efficient as the forward one.

Here's what I think I know about Canard's in general (and agree with Ted for stunt). Not only is the "main" wing less efficient, due to down wash from the canard, but for reasons of longitudinal stability and stall behaviour, the canard must be more heavily loaded. The main wing ends up unable to use its entire lifting capacity. So the canard loses its intuitively apparent advantages and is less efficient in overall function than an aft tail. If computer aided control/stability is employed, then it becomes about as efficient as an aft-tailed plane.

From this cursory "analysis", I'd conclude that you'll want the c.g. further ahead of the plane's NP than shown, but that the NP may be a bit off anyway. These on-line calculators are great for wings, but probably not too sophisticated regarding aircraft NP's. I think they're great for normal stunt configurations, coupled with our normal rules of thumb, but we don't really have that luxury with canards. We already have the wing and tail a.c. approximations to deal with - the assumption that all points on a wing lift equally. We adjust for this, but it's tougher when you couple them together and then add a smaller lifting surface that compromises the larger.

So you may be correct about their proximity in this mathematical model, but the facts that it is an approximation applied to approximations, and that our c.g.'s are placed forward of their R/C predictions, makes the analysis harder. In reality, the NP may indeed be further forward than predicted (especially when predicted by assuming equal lifting ability per unit area, front and rear), but so is the CL c.g. and shifted c.g. needed to load the canard surface more heavily. Some reading and further researching are probably in order!

SK