Originally Posted by maguro

...I'm tired from lack of sleep. If I missed anything important, please let me know.

Roger

Uh, having fun?

You're doing fine. Keep up the good work mate.

You might be the only guy that ever experienced something like that and is blaming it on the KFm2 wing and it's step height.

I'm not trying to start an argument and would like to talk about it politely, please read no disrespect or challenge into my words.

I have seen a similar issue when I had a combination of an wing and hstab incidence problem (not at zero degrees to each other) combined with a tail heavy plane with one of the KFm wing (either KFm2 or KFm3).

The glowing praise you heap on your yardarm and longarm builds on your blog pages don't even provide a hint as to the instability you mention here unless I missed it...

http://www.rcgroups.com/forums/showt...8#post12736738

http://www.rcgroups.com/forums/showt...3#post12545198

Jack

Neither the SE5a nor the test plane ever showed any instability at all. Admittedly the SE5a was never a candidate for high speed problems, but it flew beautifully as does the test ship. Both have zero incidence between wing and stab. I've used zero incidence on all my KF builds, and never had a problem.

Roger

Great work roger, a very thorough 'tufting job' you have done on the wing.. I look forward to the video and flight log data.

I think you were looking for input on testing.... I'd expect the biggest quantifiable differences between airfoils to show up in tests such as:

• Top speed (test for drag).
• Duration and/or power required to maintain level flight (test for efficiency / L/D ratio)
• Stall speed (test for Cl max).

I'd be surprised if the airfoil made any significant difference to stuff like roll rate but you may as well test those sorts of issues too if you have time and inclination.

Steve

Originally Posted by A Useless Geek

The symptom is the plane will want to suddenly pitch down in level flight at higher speeds.

Geek,
I'm with Jack on this one. I suspect that this is most likely nothing to do with the KF airfoil. The usual cause of sudden pitching down at speed is structural flexibility. If the wings are slightly lacking in stiffness the wing will twist at high speed, also the fuselage or stabiliser may bend .. This can occur very suddenly when the aerodynamic forces on the plane reach some critical point.

If it is wing twisting then changing the airfoil would probably have some effect, not because of the airfoil as such, but because of the increase/decrease in wing stiffness and/or pitching coefficient.

The other thing I'd be suspicious about is the extreme short coupled nature of the plane layout, and considering the short coupling a very small tail.. These things will naturally make the plane pitch sensitive, rather like a 'plank' flying wing.

Perhaps I was not clear enough in my previous post.

Forget the Longarm; that was an experiment gone bad in a number of ways. The Longhorn doesn't use a KF. I tried it and it slowed the plane down too much. Some of the Smarty Pants have shown a slight tendency towards pitch instability at relatively high speed, but that could be the result of other factors. When I say "high speed," I mean high for a Slow Stick or something along those lines.

The Yardarm has suffered pitch stability problems every time I tried to make the step height more than a few millimeters. I tried all kinds of combinations to get the KF working with a "proper" step height. I moved the wing back, forward, up off the boom, down below the boom, upside down (KFm1), without dihedral, with some dihedral, with polyhedral wingtips, with tip plates above and below the wing surface, with some positive incidence, and pretty much every other stupid idea I could come up with. These experiments were conducted on at least three different airframes (maybe more, I can't remember), all of which were plenty stiff enough to avoid boom flex as the culprit in instability problems.

Nothing worked.

The only thing that worked right was the original layout with a single layer of FFF as the step and a total dihedral in the 10-20 degree range. I have made a pile of these planes. At least five other flyboys that I know have made one or more copies of this plane and have reported good results with all of them. Additionally, one guy took this layout down to Florida over the winter a couple of years ago and now there are a dozen or so of 'em flying around the south of Florida.

This plane is the only failure I have had with a KF. I've tried KFs on tons of other planes and had nothing but good results with all.

Originally Posted by A Useless Geek

This plane is the only failure I have had with a KF. I've tried KFs on tons of other planes and had nothing but good results with all.

That statement alone should tell you that the problem is almost certainly not the airfoil but some other feature specific to this plane.

Most likely culprits IMHO would still be the extreme short coupling and very small tail (considering it's moment arm) and/or twisting of the wing. In your experiments did you try a much longer tail moment arm and larger tail?.. Did you do anything to stiffen the wing torsionally?

Boom flex on the current design is unlikely due to the extreme shortness of the boom..

Steve

I've done a couple planes with control surfaces on the small side (relatively speaking for me it's mostly just TLAR) and I can get the condition where I move a stick, don't see the response, and continue to move it more, then get way more response than I originally wanted. I then first think that the model is twitchy, but once I really think about it, realize that it's just control effectiveness. With a large or even oversize control surface small movements give quick, effective responses and we get used to it. With a scale or undersize control surface, we get a weaker or not obvious response and overcontrol.

Most of us get into some kind of conditioned response or even bad habit that doesn't work with all out planes. When I first flew the 96" wing span glider, I was all over with the sticks at first trying to catch up with it and keep it going straight.

Then I realized that for small corrections a small stick input and waiting 3 or 4 seconds was what that plane best responds. Now I like it best on high rates and just a small stick movement and a some patience will give me the big flat turns I want for thermal hunting. That is RET and has polyhedral tips that are augmented by the way the wing tips flex upward in flight.

Jack

Jack,

Large rudder/elevator gliders generally dont have very 'snappy' response to rudder but if you are literally having to wait for seconds then that's a bit extreme, especially as 96" span is only medium size by sailplane standards so turning should be reasonably crispy.. Usually really sluggish response to rudder indicates inadequate rudder area and/or not enough dihedral. If the turns are especially flat then that also points toward inadequate dihedral.

For rudder area, within sensible limits, it's a case of bigger the better. For normal 'V' dihedral somewhere between 9 and 12 degrees on each panel is the general rule, nearer 12 being better. For polyhedral if you put the tips where the V dihedral tips would have been you will be erring on the safe side.. I do have a spreadsheet for working out the exact equivilent polyhedral to V dihedral if you wanted it.

Steve

Who among us does not get a little snippy when we are accused -- regardless of how mildly or politely -- of sloppy science? I did try lengthening the boom, as I stated before. Also, all the wings I tried were plenty stiff, with no torsion, longitudinal, or lateral flex. I hand tested that, and was satisfied that the taller steps on the wings in fact produced greater stiffness than the original wing.

The pitch change was happening "by itself," with no correction input from me at all. (And let's not forget that I actually do know how to fly, boys. I'm not some fumble-fingered newbie who is over-correcting on his first flight.)

I still have no idea whatever was happening with this plane. After weeks of testing, with bottles of CA and sticks of hot glue down the drain, I gave up on it. I put my Yardarms back to the original configuration and just enjoyed the way they flew.

My point is that there is a potential for some weirdnesses associated with the KF for which we need to keep a lookout. Just because this pitch instability happened to me doesn't mean it'll happen to you, but we're here to share what we've discovered and to provide warnings when appropriate.

Geek,

Sorry, I didn't see any reference to lengthening the boom in your previous posts.. i just checked and still cant see it... maybe i didn't look hard enough

Anyway.. what you can be sure of is that if a plane shows some unusual flying behavior it's because of some physical aspect of the plane. If you compare the plane that has this unusual behavior with one that doesn't then one or more of the physical differences between the planes MUST be causing the different behavior.

Looking at your model there is only one obvious unusual design feature that would set it aside from 'well behaved' planes; that is the short coupled small tail. Add to this process of logical deduction the fact that it's well known that short coupled planes are more pitch sensitive and draw your own conclusions.

If we rule out the 'floppy' wing then I'm pretty sure that making the plane more 'normally' proportioned will fix the problem.. For 'normal proportions, say, tail about 20% of wing area, distance between TE of wing and LE of tail at about twice average wing chord.

Steve

Originally Posted by JetPlaneFlyer

Jack,

Large rudder/elevator gliders generally dont have very 'snappy' response to rudder but if you are literally having to wait for seconds then that's a bit extreme, especially as 96" span is only medium size by sailplane standards so turning should be reasonably crispy.. Usually really sluggish response to rudder indicates inadequate rudder area and/or not enough dihedral. If the turns are especially flat then that also points toward inadequate dihedral.

For rudder area, within sensible limits, it's a case of bigger the better. For normal 'V' dihedral somewhere between 9 and 12 degrees on each panel is the general rule, nearer 12 being better. For polyhedral if you put the tips where the V dihedral tips would have been you will be erring on the safe side.. I do have a spreadsheet for working out the exact equivilent polyhedral to V dihedral if you wanted it.

Steve

I took a semi thoughful approach to that build and looked at several of the glider related spreadsheets. In the end I used Sailplane Calc, got everything entered in that and went with it's recommendations. You can see the results I got in this thread over on the Electric Sailplanes forum where I posted a question about Sailplane Calc:

http://www.rcgroups.com/forums/showthread.php?t=1469092

Got some good replies including one from Chris Suter, the author of Sailplane Calc.

Some of my numbers are above and below the range of values recommended but I went with what I had and it has flown well. Curtis also made a couple of suggestions along the line of what you are saying:

"..Just by looking at the model it lacks dihedral and would turn much better with added dihedral. It makes a lot of difference if you get the EDA values correct. Also you'll find a higher aspect horizontal tail to be more effective..."

But it looks like it is defying logic a little and flying better than the numbers say it will. Must be the KF airfoil, it fixes everything! The wing is pretty flexible, the tips lift in a gentle curve as it flies, and I think that the amount of dihedral I have is not too far off the mark. When I looked at the getting the polyhedral tips long enough and up as high to meet the recommend EDA range I decided to build with the lower EDA and fix it later with ailerons if I had to. But now I don't think I need them.

I have a build thread for it here:

http://www.rcgroups.com/forums/showthread.php?t=1470211

And there are several videos of it flying posted there too. If you'd like to look at those there and comment on the Sailplane Calc results or what you see in the videos, feel free to do that. It will put it where people that might be building or thinking of building the plane can see it.

The more I've flown it the more I like the rudder response I'm getting now, it seems OK and fast enough. But if I do build another tail assembly (it has a detachable tail and a wing made in three sections for better transportability) I'll probably build a taller tail just to see the difference. Maybe I'll build a V tail too...

Jack

jack,
Sailplane calc is a great tool, Curtis is a very clever guy. The design ranges included in sailplane calc come from Mark Drella, and the formulas used behind the scenes are, if i recall correctly, from Martin Simons' book Model Aircraft Aerodynamics.. two better sources would be hard to find.

Lazy control response is ok for relaxed flying in calm conditions but you may one day find yourself in trouble if you hit turbulence low down and don't have enough control authority to recover. The overall vertical tail area itself looks fine but you should find that you improve response a bit simply by making the moving part (the actual rudder) larger and (if you want to keep the area the same) the fixed part smaller.. As Curtis says, a higher aspect ratio surface also gives more response from smaller control movement.... These things should help a bit but i still think more dihedral would be needed to really get it turning quicker.

Anyway, it's a nice glider and looks great in the air..

Steve

I was not able to get the two videos edited together with the software I have. The external video (which should be in the small window) is Blue Ray HD, and the video of the wing is 720x480. My editing software put the small window where I wanted it, but by the time the video got to the end, the little box expanded to cover 90% of the screen. So here are the videos from the wing and the ground. If you start the ground video first, you should be able to start the wing video so the two are in reasonable sync. It was just a proof of concept, but worth a look.

Roger

KF Wing Flight Test (5 min 48 sec)

Flight of KF Test Aircraft (5 min 51 sec)