Stability Questions

[HELModels]

Alright, I built it and I crashed and crashed and crashed. Each crash was a
dive. The airframe and all has survived unharmed except for props.
I used an RAF airfoil and modified the camber and thickness until it ended
up thin and with somewhat concave underside. I was trying to avoid making
a single surface airfoil like every other parkflyer. This modified airfoil
has a very close resemblence to the hobie hawk airfoil. I couldnt quite
figure where to balance it - Xfoil said at reynolds of 100,000 the Moment
was -0.15!!! when producing coefficient of lift of 0. This was a type 2 analysiswhich I think holds the chord length constant or something. I found that hardto believe or understand, so I just balanced it at 25% chord and figuredworst case was forward CG.

Here is how it flew. Climbed with no elevator, but climbed better with
a touch of elevator. Flew very fast at a wing loading of about 7 oz./sq ft.
and full throttle - fast not really what I wanted.
The last crash happened when trying to turn into the wind while at full
throttle - was attempting to fly a square pattern and remain within receiver's
range. what I noticed was the turns were very sharp and aggressive and the
last turn ended in a dive for the ground. There was another transmitter on
my frequency in the impound area, so part of me thinks someone could have been giving me "TRAINER INPUTS" to keep my hunk of junk from the primary airspace. That was my intention anyway and I didnt need extra help to insure that. Funny that the dive occurred short of the "pro" airspace, just as it looked like I might infringe. Dont fly at flyins even if it is the end of the day and most of the hotdogs have gone home. I only broke a prop, but I smelled a rat and turned in my frequency flag and radio to the
impound area. When I requested my radio back before going home the head rat chided me, thinking he was cute, suggesting I had to wait till the flyin was completey over before I could get the transmitter back. Sniffffff, smell the rats?

Anyway, the control surfaces on this thing are huge and too big for a trainer with such a small span, 33".
It might have been a stall and it might have been a loose tail section, so I tightened the tail section up and it is nice and firm. I also replaced the wing with one that has polyhedral(same span, chord) and am considering trimming off some of the control surfaces. The polyhedral ruins the hint of scale as would trimming the surfaces.


My questions are this:

Is a stall followed by a severe dive, an indication of a Forward CG?
Can a rudder cause a stall if it is too big?
What can cause unexplained dives?

[Sail 'n Soar]

Is a stall followed by a severe dive, an indication of a Forward CG?
No. But it could be an indicator of out of trim, that is, too much incidence difference between the wing and horizontal, or too small a horizontal

Can a rudder cause a stall if it is too big?
The large rudder wont' cause a stall, but could cause your aircraft to roll into a dive in a turn.

What can cause unexplained dives?
From your description you plane has too much up trim and is stall-prone to begin with. In addition to the "too big ruddere" comment, stalling in a turn can result in a spin, which you probably wouldn't be able to distinguish from a dive if it occured near the ground.

[HELModels]

Sail n Soar,
I tried again with the polyhedral wing and watched for hints that the incidence was incorrect. It would slightly pitch up, slow up a bit, speed up a bit, repeat. This was with the landing gear attached and so didnt climb very high. Too much drag to overcome?

Turns with polyhedral were tight yet gentle. I didnt get into any trouble until I turned real tight(big rudder again) to avoid a larch tree and then "The Dive" came. This dive broke the motor mount loose, so will have to fix that.

I now see that a pusher configuration might be best for learning and protecting motors/gears/props and no landing gear is aerodynamically cleaner.

If I could get the wing loading down to 4 oz, do you think "The Dive" would go away?

[Sail 'n Soar]

If I could get the wing loading down to 4 oz, do you think "The Dive" would go away?

Lowering the wing loading, assuming you slow down as well, will not cure the problem, but things will happen a bit slower. It sounds like you have too much vertical stab and rudder. The polyhedral wing you are using seems to have more effective dihedral than your original wing. I'm basing that on the fact that the problem got much better, but was not cured completely. The traditional fix would be to add more dihedral or to reduce your vertical stab & rudder area.

As a potential third solution, I had a similar problem with a constant cord/constant foil wing that should have had a very forgiving stall, but would seem to snap into a dive whenever I hit rudder. I added a little of what should have been unnecessary washout and the plane became very responsive, but lost all tendencies to shap into a dive or spin. I went back and calculated vertical tail volume and discovered that I must have goofed on my calculations when I designed the plane. The washout reduces any tip stall tendency. It also reduces the effective incidence slightly. I'm not sure what problem the washout cured, but I am pleased with the results.

[HELModels]

The fuselage is scrap now after the final encounter with the larch tree. I threw tennis balls to dislodge it and one good throw went right through the balsa portion of the fuse. I am now working on a pusher configuration to save the motor/prop/gears.

What I am now wondering is when an airfoil program says the Moment Coefficient is -(negative), is that 25% chord minus whatever the moment value? IOW, If the moment is -.15, then the CG is located at .10 of the root chord length, for constant chord? :confused: Or is it .25 - -.15 = .40?

If the .10 is right, that would mean that the CG has been rearward even at .25, for the airfoil I used. Would a rearward CG and Oversized Stab/Rudder help explain the difficulty in turning this ship without spinning?

[Sail 'n Soar]

The fuselage is scrap now after the final encounter with the larch tree. I threw tennis balls to dislodge it and one good throw went right through the balsa portion of the fuse. I am now working on a pusher configuration to save the motor/prop/gears.

What I am now wondering is when an airfoil program says the Moment Coefficient is -(negative), is that 25% chord minus whatever the moment value? IOW, If the moment is -.15, then the CG is located at .10 of the root chord length, for constant chord? :confused: Or is it .25 - -.15 = .40?

If the .10 is right, that would mean that the CG has been rearward even at .25, for the airfoil I used. Would a rearward CG and Oversized Stab/Rudder help explain the difficulty in turning this ship without spinning?


The negative CM means that with the plane balanced at the 25% cord the wiing will try to rotate the nose down, tail up. i.e., you will have to have an offsetting downward force on the horizontal stabilizer to counterbalance the nose tuck tendency and hold trim. The actual wing moment about the quarter cord is M = .5*rho*V^2*Cm*C*A, C = mean cord, rho = air density, A = wing area, V = flight speed.

[HELModels]

The negative CM means that with the plane balanced at the 25% cord the wiing will try to rotate the nose down, tail up. i.e., you will have to have an offsetting downward force on the horizontal stabilizer to counterbalance the nose tuck tendency and hold trim. The actual wing moment about the quarter cord is M = .5*rho*V^2*Cm*C*A, C = mean cord, rho = air density, A = wing area, V = flight speed.

Thanks for the formula Sail 'n Soar. For my wing area and airfoil I get a small negative number: -.09 This is using Slugs/Ft^3 at sea level 59F, but I end up with units left over as in Slugs/Sec^2 :o

Assuming -.09 is ballpark, if I were to put the CG at .34 chord(.25 - -.09), is that a reasonable place to start?

I have some formulas for setting the incidence of the stab, when the distance to the CG is known and so is Zero Lift Angle of the wing.

[Sail 'n Soar]

Thanks for the formula Sail 'n Soar. For my wing area and airfoil I get a small negative number: -.09 This is using Slugs/Ft^3 at sea level 59F, but I end up with units left over as in Slugs/Sec^2 :o

Assuming -.09 is ballpark, if I were to put the CG at .34 chord(.25 - -.09), is that a reasonable place to start?

I have some formulas for setting the incidence of the stab, when the distance to the CG is known and so is Zero Lift Angle of the wing.

I put the CG at ~ 30% and don't worry about it. Key thing is the value of:

Ah*l/(A*C)

Where Ah = horizontal stab including elevator area, l = distance from wing to stab quarter cord, A = wing area an C = mean cord. I usually size the tail to achieve a value of ~.5 with the CG at ~ 30%. A larger value makes it more freeflight/early RC stable, less than .5 makes it less stable/more pitch sensitive. For most ships I think the tail comes out looking about the right size with a value of .5.