Pitch-Up Problem

[DLC]

A while back I built a Fokker D VIII using Earl Stahl free-flight plans, scaled down to 36" span. I also took out the dihedral and added ailerons. It was very spirally unstable, and eventually spiraled in and demolished the fuselage, but left the wing undamaged. I liked the looks of the plane, but didn't want to continually battle its spiral instability. So I tried an experiment. To get away from the balancing problems of the D VIII, I moved the wing back to where the MAC/prop distance was about 1 chord length and the MAC to tail MAC was about 2.5 chord lengths. I also lowered the wing to be on the same vertical plane as the horizontal tail. Then I sketched up a fuselage with side area distribution which, per the method in Lennon's book, should give good spiral stability. A picture is attached. On the first flight I tried a gentle climbing turn and saw no signs of spiral instability (the D VIII tried to wind up from even a gentle turn) and I was about ready to celebrate. Then I tried a harder turn and the nose pitched up almost 90 degrees. I got it straightened out, but was getting close to some trees, so had to make another hard climbing turn. Again it pitched up about 90 degrees, this time crashing into the trees. The body was badly damaged, but there was no significant damage to the wing or tail, so I would like to rebuild it, but not until I understand why it pitched up. Any ideas as to what might have caused the pitch-up?

[rebell]

There could be a lot of reasons, but I would say a combination of rearward CG and the very small fin/rudder is the problem. In a steep turn, there is not enough fin to prevent the tail from side slipping.

[rofujiyama]

Hi DLC
Those nose and tail moments bring back memories. I used to us one wing chord to firewall and 2 an 1/4 chord to the horizontal stab, both from the wing itself. The dimensions I chose , came out of a RC Glider Handbook that is long lost now :( . The moments made for a nice flying plane at the time, but the areas I chose for the horizontal tail and elevator, was a little to large for a powered plane, when using that tail moment and i ended up trimming both the stab and elevator area, to get reasonable travel/response from the elevator. The elevator ended up being about 1/4 the stab area visually speaking, or a little less.
I found, that scale type elevators of wwI, 30's, early fortys , with alot of area, generally will be more managable if you use lower rates for takeoff and zipping around, and possibly higher rates of travel when landing, especially since you have a nice radial cowl, that will generate some turbulence around the body at lower speeds( I think,lol)
So, basically I think you have a lot of extra area in you horizontal stab/elev. for you tail moment,that will tend to overpower your plane when you apply elevator, at combat speed ,lol :D
Looks very cool , tho :cool:

[JetPlaneFlyer]

I'd hazard a guess that the downwash from the wing (which increases when you pull high 'G' turns) is effecting the horizontal stab. The downwash acting on the stab will tend to push the tail down. The model 'could' be prone to this because the gap between the wing TE and the horizontal stab LE is quite short and both surfaces are in line horizontally. The solution may be to relocate the stabiliser to the bottom of the fuselage, or move it upward into a 'T' tail position to get it out of the wing wash and /or lengthen the fuselage.

Steve

[MarkusN]

...The solution may be to relocate the stabiliser to the bottom of the fuselage, or move it upward into a 'T' tail position to get it out of the wing wash ...
Up- / down relocations of the tail are not very effective in reducing downwash interference. Downwash is a flow field, and it decreases only slowly over height.

[JetPlaneFlyer]

Up- / down relocations of the tail are not very effective in reducing downwash interference. Downwash is a flow field, and it decreases only slowly over height.

Unless you have a better suggestion it's maybe worth a try?... I know relocating tails in the vertical axis is a standard way to deal with wing wash interference on full size aircraft (it's the reason 'T' tails axist in the first place) so no reason for it not to work on a model, if done correctly.

Steve

[MarkusN]

I know relocating tails in the vertical axis is a standard way to deal with wing wash interference on full size aircraft (it's the reason 'T' tails axist in the first place) so no reason for it not to work on a model, if done correctly.
That's not so much about downwash (you can correct that simply by changing decalage) but about getting the tail out of turbulent air of the slipstream (and, sometimes, out of the dead air of the stalled wing.)

Things are different for extreme geometries like fleas, but for a more or less standard tail, these are the effects.

In the case here I wonder how much the delta shape of the tail plays into the equation, together with interference from the propeller and wing slipstream.

Speaking of that: It might be an interference between wing and tail that indeed can be corrected by lowering or lifthing the tail: the interference vortex starting at the wing root when the aileron is deflected. If that leads to erratic flow conditions at the low AR tail, that might indeed be corrected by changing the vertical arrangement.

What might also help (and retain some of the scale look): shortening the ailerons, to move that vortex out. Many of those WW I planes had wing warping for roll control, which makes for a very smooth slipstream.

[kcaldwel]

First thing to check is the CG position. Do a calculation to find the neutral point, and then put the CG at least 10% of the mean wing chord ahead of that. 15 to 20% would likely be better for an easy handling aircraft. Re-trim for the new CG, and see how it is. You'll have to take into account the plane between the wheels in the neutral point calc - it is destabilizing.

http://www.geistware.com/rcmodeling/cg_super_calc.htm

Seems to me that model isn't that much different than 10,000 other ones for tail and nose moments, and stab to wing location. I can't see why it can't be made to behave normally.

I'd try taking the plane out from between the wheels, get the CG at 15% like that, and see how it flies. Add the plane back in, move the CG forward 5% to compensate, and see if the behavoir returns.

Kevin

[rebell]

I agree with Kevin. That is what I said before, rearward CG is causing the problem. The horizontal stabilizer is big enough to keep the plane level in straight flight by carrying some of the planes weight. (That is supposed to be the mainplane's job.) The moment you turn/bank, the plane depends on the fin/rudder to keep track by letting the fin carry some of the weight. (Rearward CG) Because the fin is to small to carry weight when banked, it slips downward, letting the nose pitch up. Once it is in that attitude, it is difficult to recover.

The solution is like Kevin said. Move the CG forward.

[kcaldwel]

The horizontal stabilizer is big enough to keep the plane level in straight flight by carrying some of the planes weight. (That is supposed to be the mainplane's job.) The moment you turn/bank, the plane depends on the fin/rudder to keep track by letting the fin carry some of the weight. (Rearward CG) Because the fin is to small to carry weight when banked, it slips downward, letting the nose pitch up. Once it is in that attitude, it is difficult to recover.

Well, we agree for different reasons! The CG position has only a small influence on yaw stability. And the horizontal tail may have a down load or up load, and that does not indicate whether any given aircraft has any particular level of pitch stability.

I guess this begs the question though: is it really a pitch up (purely pitch axis), or a yaw effect you are seeing when banked?

Kevin

Still check the CG.

[rofujiyama]

Oops, :( I used a different reference ,, point for my tail moment, as compared to Kevins calculator program. I was wondering about it myself.
That being said , I still feel , that the elevator/stab in scale planes, like yours , will give you alot of over control , if you use, standard 30 deflections for control throws, as your starting point, to trim out your plane.

[kcaldwel]

I vote for a yaw problem. The fixed fin is virtually non-existent, and the total vertical stab area is small. The fuselage area ahead of the CG is reasonably large. I know that is how the WWI planes were, but they all had bad yaw instability issues.

Any linkage play or a weak servo on that rudder is going to be a problem. If it was designed from a book that says you need certain fuselage side areas for spiral stability, I'm very suspicious. That simply isn't true. I think it needs lots more fin area. And the CG in the right place.

[markschaffin]

Do you have a pitch up problem when pulling up from straight level flight?

If you moved your wing back, your CG should have moved forward relative to the wing. Did you need to add some up elevator to keep it trimmed?

Are you sure it didn't tip stall and snap on you? I'd check your elevator throw and CG and try it again, this time making sure you don't slow it down too much.

Mark

[DLC]

Thanks for the help everybody. I am still digesting it. It was definitely not a tip stall and snap. Also, it seemed to be trimmed well in straight flight, both climbing and level and in gentle turns. One thing I have found since my first post was that I had accidentally left out a spacer I used to adjust the decalage, and the decalage was on the order of 6 to 7 degrees, maybe even a little higher, during the flight. The C.G. was at 25 percent of the chord.

[kcaldwel]

If it was trimmed for level flight, then the decalage doesn't matter. Decalage and elevator trim are equivalent until the surface deflection gets quite large.

That plane between the wheels may move the neutral point far enough forward that 25% is not pitch stable enough.

Kevin

[vintage1]

Well, we agree for different reasons! The CG position has only a small influence on yaw stability.



With respect, it has precsiely the same influence on yaw as it does on pitch.

The dynamic are exactly the same..whether the side area presented to the airflow in a slip, will tend to have a summed force ahead of, or behind, the CG. If ahead, the plane will be yaw unstable.

The trick is to adjust the fin to match the pitch stable CG location, and the fuselage side area. People will keep 'increasing tailplane area for stability' which is rubbish. Just move the CG forward!. If you increae tailplane area without increasing the fin, you may end up with a model that drags its tail on turns or wrse.


And the horizontal tail may have a down load or up load,

WILL have a down load AND upload, depending on the speed the model is flying. Its quite rare to have it all one way.


and that does not indicate whether any given aircraft has any particular level of pitch stability.

I guess this begs the question though: is it really a pitch up (purely pitch axis), or a yaw effect you are seeing when banked?

Kevin

Still check the CG.

[rebell]

Well, we agree for different reasons! The CG position has only a small influence on yaw stability. And the horizontal tail may have a down load or up load, and that does not indicate whether any given aircraft has any particular level of pitch stability.

I guess this begs the question though: is it really a pitch up (purely pitch axis), or a yaw effect you are seeing when banked?

Kevin

Still check the CG.

Yes, it is a yaw effect and not pitching. In the banked attitude the end result is still that the nose is pointing up.

I had a model that did the same as DLC described. I had to move the CG way to the front, to about 20% of the MAC. After that it was a pleasure to fly.

[vintage1]

..or make the fin bigger.
The early planes were mainly flown on rudder anyway.

I have the pilots notes from a Blenhem - circa 1938 or so. They found it necessary to make the point that 'this aircraft flies on banked turns using the aileron and elevator: The rudder should be only used in take off and landing, and for controlling adverse yaw in turns.' or somesuch.

I have a scale tiger moth. If I fly it on ailerons only. the nose pitches up a bit turns. On rudder, it pitches down a bit. The secret is to use BOTH..

[DLC]

Thanks again everybody. After looking at the fuselage closely, I found that the damage was not nearly as severe as I first thought. So, I am going to repair it as is and hope to get it high enough flying straight to be able to experiment with the effects of various parameters on turns without disaster.

[kcaldwel]

Vintage1,

Just as a thought experiment, consider a slope soaring saiplane with no dihedral. The only side areas are the fuselage pod well forward of the CG, and the tail fin at the end of a skinny boom. Moving the CG forward an back 5 or 10% of the MAC will have only small effects on the moment arms to the major side areas. The tail at the end of the long boom predominates hugely, and changing a 20" moment arm by even 1/2" makes only a small change in the yaw stability terms.

Even on more "normal" powered airplanes, the same holds true - moving the CG within bounds that the pitch stability will alow, makes only very small changes to the yaw stability.

That said, this model may well have a yaw stability issue.

Kevin

Edit: Mark Drela had some interesting comments on the old FF theories of fuselage areas and yaw stability awhile ago. I don't have time to search for them right now.

[macboffin]

If it was trimmed for level flight, then the decalage doesn't matter. Decalage and elevator trim are equivalent until the surface deflection gets quite large.

That plane between the wheels may move the neutral point far enough forward that 25% is not pitch stable enough.

Kevin I agree ; that plane between the wheels is lower aspect ratio than the wing, and will keep lifting when the wing is stalled, thus lifting the nose. Plus, usual thing is for small versions of full-size airplanes to increase proportions of the tail fins relative to wing size, due to Reynolds numbers effects. You might also check the incidence angle of that tween- wheels surface ; it's possible that at slow speeds and higher angle of attack you are effectively flying a tail-heavy canard, on account of the tail fins having gone out of effective business!

[vintage1]

Vintage1,

Just as a thought experiment, consider a slope soaring saiplane with no dihedral. The only side areas are the fuselage pod well forward of the CG, and the tail fin at the end of a skinny boom. Moving the CG forward an back 5 or 10% of the MAC will have only small effects on the moment arms to the major side areas. The tail at the end of the long boom predominates hugely, and changing a 20" moment arm by even 1/2" makes only a small change in the yaw stability terms.

Even on more "normal" powered airplanes, the same holds true - moving the CG within bounds that the pitch stability will alow, makes only very small changes to the yaw stability.

That said, this model may well have a yaw stability issue.

Kevin

Edit: Mark Drela had some interesting comments on the old FF theories of fuselage areas and yaw stability awhile ago. I don't have time to search for them right now.


Yup, but that is an example of a model with huge yaw stability anyway, although at a given AofA the wing will not have an inconsiderable 'yaw resistance'

In a more typical model there is enough side area to knife edge at reasonably low angles of (fuselage) attack: At that point a model in full knife edge has precisely the same criteria for what is now pitch stability, as it did when upright.

In the end consider an arrow: you can have one with flights, or a weighted tip, or both. But not neither. CG IS (just as ) important when on the slenderest of fuselages with NO wings ;)

[kcaldwel]

Yup, but that is an example of a model with huge yaw stability anyway, although at a given AofA the wing will not have an inconsiderable 'yaw resistance'

In a more typical model there is enough side area to knife edge at reasonably low angles of (fuselage) attack: At that point a model in full knife edge has precisely the same criteria for what is now pitch stability, as it did when upright.

In the end consider an arrow: you can have one with flights, or a weighted tip, or both. But not neither. CG IS (just as ) important when on the slenderest of fuselages with NO wings ;)


Actually, slope soarers typically have no more yaw stability than any other sailplane - too much drag and handling issues with large yaw stability. I just used them as a simple example of yaw stability.

Even in knife edge, you have a wing chord that is now the length of the fuselage, which is many time the wing chord. Even if yaw stability was as critical as pitch stability, 20% of the fuselage length may be 100% of the wing MAC so the allowable CG position would be huge compared to the allowable range in pitch.

I think you will find an arrow works quite well as long as the CG is ahead of the centre of pressure - the distance isn't very critical. 3" ahead works about the same as 1". This is easily done with fins or weight as you mentioned. If you try it, you'll see that the fin size and/or nose weight can be varied hugely, and the arrow still works. This is not true for pitch stability on an aircraft. Aircraft will be yaw stable as long as the CG is behind the lateral centre of pressure as well.

On an airplane, it is us the yaw/lateral stability coupling and relative magnitudes of the yaw and lateral stability that may cause issues such as dutch roll, not the yaw stability itself.

Kevin