Vertical Flap Control Horn vs. 90-Degree Flap Control Horn

[Howard Rush]

Everything you know is wrong, but it doesn't really matter.

Dave Gardner came over to the house yesterday.  He brought some plots of flap control horn angles.  He convinced me that if the bellcrank is in the plane of the wing, it's best to have the flap control horn input arm vertical.  Then I used my cool program to check this, and either the program has an error, or the three-dimensionality of the problem complicates it more than we thought.  Attached is the upchuck.  As you can see from the inputs, it's a 4" bellcrank and 1.5" flap control horn, the bellcrank is straight in neutral, and the leadouts come straight out the wing.  I had Excel put some "trendlines" (the fine, black lines) on the flap plots.  They are the second-order curves that best fit the flap curves.  Trendline equations are next to each flap curve.

[Serge_Krauss]

Howard-

You're doing some fascinating stuff here. I'm just getting a re-look at things now, but am I correct in thinking that your conclusion is based on getting the bellcrank and flap deflections as close as possible?

I was thinking more about symmetry in up versus down flap deflections. The spreadsheet I used, created by a Mr. A.J. Herbon, indicated an advantage in this regard for angled flap horns. I'll post his diagram. In it Beta sub zero is the angle of the horn from vertical, with forward slant being positive. Delta Alpha is the difference between bellcrank rotation angles to achieve equal up and down flap deflections as read on the horizontal-axis scale. The second illustration is for my model, which has a slightly greater than usual upward flap-rod angle. A horn angle of 10.911^o makes the rod and horn perpendicular at neutral. It seems, according to his spreadsheet that somethig less than that angle is the best compromise, and that the *vertical horn is the worst of those illustrated - in this regard. As you say though, they don't seem to make a lot of difference for flap deflections we use.

Anyway, here are the diagram and family of curves...

*Edited  to change "perpendicular" to "vertical" to avoid confusion.

SK

[Howard Rush]

I'd look at the response of controls surfaces to differential control line displacement.  It doesn't matter what the bellcrank does, except for mechanical bearing loads and how it fits in the airplane. 

The number he calls beta0 may have a lot less effect on control response than some of the other things you can vary.  My favorite parameter so far is DeltaBellcrankPushrodArm0, the offset angle of the bellcrank output arm.  It took just a few minutes with the program to get the response linear for my new plane.  You may want neither linearity nor symmetry, though.  I know of one world champ podium finisher that was intentionally nonlinear and another that was way asymmetric.   

[Larry Cunningham]

It helps even more to tilt the bellcrank to match..

Illustration has an Expocrank, but you get the idea.

L.

"If called by a panther Don't anther." -Ogden Nash

[Chuck_Smith]

Everything you know is wrong, but it doesn't really matter.

Dave Gardner came over to the house yesterday.  He brought some plots of flap control horn angles.  He convinced me that if the bellcrank is in the plane of the wing, it's best to have the flap control horn input arm vertical.  Then I used my cool program to check this, and either the program has an error, or the three-dimensionality of the problem complicates it more than we thought.  Attached is the upchuck.  As you can see from the inputs, it's a 4" bellcrank and 1.5" flap control horn, the bellcrank is straight in neutral, and the leadouts come straight out the wing.  I had Excel put some "trendlines" (the fine, black lines) on the flap plots.  They are the second-order curves that best fit the flap curves.  Trendline equations are next to each flap curve.

Pretty cool, but I'd expect that wing centerline, thrust line, and placement of the horizontal tail (all three WRT their vertical realtionship to the airplane's center of gravity play a larger role in control response) and any differences between inside and outside manuvers than the relatively small changes in surface deflections due the control geometry shown.

Let's say the wing is below the entire aircraft's CG. The drag from the wing is causing a nose-down pitching moment. The tail is above the CG, now we have a nose up contribution. If the engine is also above the CG, we get a negative moment from the thrust. There are also pitching moments due to the lift of the wing and the lift of the tail.


Without calculating the stability derivatives and knowing the control derivatives it's all just a crapshoot, and once again, it's the response of an airplane that matters.

Remember, Quantum Physics is for the guys that flunked out of Stability and Control, he, he.

[Howard Rush]

Moments from those drags isn't much.  To wit, drag from a surface is a lot less than lift, and moment arm for the drag is a lot less than moment arm for the lift.  

That stuff (that of it that's not a function of speed) picks the elevator deflection for zero moment.  For my airplane it's about 5 degrees down elevator, which I know in advance, because it's a known airplane design.  Then you can calculate the mechanical control response for flaps and elevator with elevator at +5 degrees for neutral flaps.  I made mine linear and symmetrical, but neither may be the best.  

[Kim Mortimore]

.....I made mine linear and symmetrical, but neither may be the best.

.....You may want neither linearity nor symmetry, though.  I know of one world champ podium finisher that was intentionally nonlinear and another that was way asymmetric.  

Howard, Nonlinearity is not so hard to imagine reasons for (grammar cringe), but it's hard (for me) to imagine reasons for way asymmetric other than as a correction for some opposing asymmetry--really bizarre design, Elephant Man wrist.  I assume you asked the way asymmetrical podiumite why.  Would you be inclined to share that with us?

[Howard Rush]

The asymmetrical podiumite hadn't noticed it, so it may have been my imagination.  Neutral seemed normal when I flew that airplane, but the upside-down wingover pullout was a lot lower than I expected.  Having a lot of leadout travel per degree of elevator in the neighborhood of full down would help overcome hinge moment at the top of the hourglass.  I'd think that would be an advantage once you'd gotten used to flying that airplane.

[Chuck_Smith]

Moments from those drags isn't much.  To wit, drag from a surface is a lot less than lift, and moment arm for the drag is a lot less than moment arm for the lift.  

That stuff (that of it that's not a function of speed) picks the elevator deflection for zero moment.  For my airplane it's about 5 degrees down elevator, which I know in advance, because it's a known airplane design.  Then you can calculate the mechanical control response for flaps and elevator with elevator at +5 degrees for neutral flaps.  I made mine linear and symmetrical, but neither may be the best.  

Hmmm... not sure we're in agreement there.

The distance between the thrust line and the vertical CG is pretty important, would you agree?

If yes then consider the drag equals the thrust. That tells us the net drag vector aka the superposition of all the drag vectors and it's relationship to the vertical CG is equally as important as the thrust. Which it is, because to fly straight the drag moments and thust moments must be eqaul and opposite.
And don't forget, the lift vector is perpendicaulr to the flight path, not the wing. So as the AoA increases you get a forward contribution to the lift that affects the moments. This is why in practie an overloaded airframe's wings can fail forward and not backwards.


I know from computing stability derivatives that drag moments are absolutely important to the control response.

[Howard Rush]

The distance between the thrust line and the vertical CG is pretty important, would you agree?

Only for prop clearance.

I know from computing stability derivatives that drag moments are absolutely important to the control response.

Stunt people who don't compute stability derivatives think they are.  Let's see some computing. 

[Kim Mortimore]

The asymmetrical podiumite hadn't noticed it, so it may have been my imagination.  Neutral seemed normal when I flew that airplane, but the upside-down wingover pullout was a lot lower than I expected.  Having a lot of leadout travel per degree of elevator in the neighborhood of full down would help overcome hinge moment at the top of the hourglass.  I'd think that would be an advantage once you'd gotten used to flying that airplane.

Assuming it wasn't your imagination, and the plane did have increased leadout travel per degree of elevator travel in the neighborhood of full down, getting used to it would be quite a price to pay.  Wouldn't the square eights be pretty wild?

How about increased leadout travel everywhere?  Give up some control surface deflection at the extremes which isn't used anyway, so the full range of leadout travel is applied to a smaller range of deflection, get you a big-a$$ bellcrank, flip the flanges on your hard-point upside down so you can spread them lines out wiiiide, take all that there mechanical advantage, and go to town.  or words to that effect.  notwithstanding.  indubitably.  

Actually, I'm currently experimenting with this on the Imitation, which has the combined advantages of being a high-quality design and providing full access to the flap horn.  A friend of mine machined a horn extension which allowed me to move the flap rod link-to-hinge from 1-1/8" to 2" (better than major surgery to reduce the bellcrank pivot to connector distance).  Results: excellent, at least for me (disclaimer: some of this has to do with feel at the handle which is partially subjective, so it could be blah blah).  A sense of more direct relation between handle motion and aircraft response (especially small adjustments).  A greater sense of confidence and control at low altitude (which is important for some nonexperts like me).  I haven't carefully checked out the difference up high yet, partially because I'm working limited power (LA46) and have more trimming to do.  I could go on here, but need to leave now, so more later if there is any interest.    

[Howard Rush]

Assuming it wasn't your imagination, and the plane did have increased leadout travel per degree of elevator travel in the neighborhood of full down, getting used to it would be quite a price to pay.  Wouldn't the square eights be pretty wild?

How about increased leadout travel everywhere?  Give up some control surface deflection at the extremes which isn't used anyway, so the full range of leadout travel is applied to a smaller range of deflection, get you a big-a$$ bellcrank, flip the flanges on your hard-point upside down so you can spread them lines out wiiiide, take all that there mechanical advantage, and go to town.  or words to that effect.  notwithstanding.  indubitably.   

Actually, I'm currently experimenting with this on the Imitation, which has the combined advantages of being a high-quality design and providing full access to the flap horn.  A friend of mine machined a horn extension which allowed me to move the flap rod link-to-hinge from 1-1/8" to 2" (better than major surgery to reduce the bellcrank pivot to connector distance).  Results: excellent, at least for me (disclaimer: some of this has to do with feel at the handle which is partially subjective, so it could be blah blah).  A sense of more direct relation between handle motion and aircraft response (especially small adjustments).  A greater sense of confidence and control at low altitude (which is important for some nonexperts like me).  I haven't carefully checked out the difference up high yet, partially because I'm working limited power (LA46) and have more trimming to do.  I could go on here, but need to leave now, so more later if there is any interest.     

That airplane's owner got pretty used to it.  He got second with it at a world championships. 

I think you are on the right track with the long control horn if you still have adequate control movement.  It lets you get more bellcrank movement, which is a virtue.