More leverage allows for slower and more accurately placed (felt) controls. They're most useful in competition stunt planes doing the tricks for points. Sport planes don't need them, since they fly fast and get flicked about recklessly. At least I like to fly them like that. Great relief to jerk the controls on a sport plane and make it dance. Round appearing circles and 90 degree (looking) corners not an issue.

Ok, I have read about fliers liking 4" bell-cranks but what is the big deal verses 3" ones?

The best thing about 4 in bellcranks is that you lower the forces on the pushrod 33%.  Reduces wear and tear on the controls and give you the option of slowing the controls down for more precise flying.

These are just estimates, bvut here's the principle as I see it:

2" bellcrank with 90 degrees total movement.
3" bellcrank with 70 degrees
4" bellcrank with 50 degrees.

The small bellcrank loses mechanical advantage toward the extremes of throw and the output looses most of its fore-aft vector at the extremes.

The bigger bellcrank operates with the output in a near-linear range.  Futhermore, a given amount of slop (compliance) in the linkage take A LOT of motion away from a tiny bellcrank, but very little from a bigger one.

Of course the thing has to fit in the model without hogging out too much structure.  There have been countless successful mdoels with 3" bellcranks before the 4' fad hit.

The best thing about 4 in bellcranks is that you lower the forces on the pushrod 33%.  Reduces wear and tear on the controls.

Can you explain that?

Bellcrank size has nothing to do with pushrod pressure, only the length of the control horn. Assuming you can get the travel desired, the bellcrank can be any reasonable size. Going from a 3 to 4 inch crank simply reduces the amount of "pull" on either line needed to steer the model.

I thought I read here somewhere that a 4" bellcrank is closer to the grip size of a closed hand, resulting in a movement ratio close to 1:1.  The goal is to reduce sensitivity and afford more precise control.  It makes sense to me.

I don't EVEN want to chime in about expocranks! ;->

L.

"I may have many faults, but being wrong ain't one of them." -Jimmy Hoffa

Another worm...

What about the distance from the BC pivot point to the push rod pivot?


Ward-O 

Thank you, Ward!
Finally someone asks the right question: not all 4" cranks are equal.
There are at least two ways to slice this ... assuming the elevator and flap horn lengths are kept constant for the purposes of the comparison.
1) If the short to long arm length ratio is the same as a classic 3" crank, then you enjoy better linearity, especially when it comes to the difference between UP and DOWN throw curves, but no improvement in sensitivity.
2) If you run a shorter short crank ratio, then you get to fly with wider line spacings at the handle, and you back away from the Netzeband-limit problems. There is no improvement in linearity.

5" cranks anybody?  S?P
Regards,
Dean P.

Fascinating.  I was aware of the sensitivity factor in relation to arm length ratio, but the linearity vs. sensitivity compromise is new to me.

5" cranks sound like a great idea.  With correct control arm lengths, would they allow less compromise than 4-inchers?

Also, Dean, would you like to say a bit more about UP/DOWN throw curves? 

Great stuff.  Thanks.

We tried doing that, and it's too hard.  Somebody should write the formulae for control linkages.  For extra credit, figure out how to do Igor's flap mechansim with a four-bar linkage. 

To really know how this works, dust off your geometry and trig knowledge from high school (you remember -- you were 15, and supposed to be studying the curves on the board, not the curves on the girl next to you).

Tim,
As W.C. Fields would say, "Ahhhh, yes, I remember it well."  Mr. Euclid and his pals were hard put to compete with the songs of Sirens. Wasn't there an early seventies girl-band called The Sirens?  

http://www.netax.sk/hexoft/stunt/ .  Go for it.

I would say the amount of non-linearity for the small amount of control travel we use is insignificant. And of course the controlling entity is also non-linear, plus it possesses a tremendous ability to compensate for such anomalies.
Don

One consideration is not so much linearity or its lack, but asymmetry: the amount of up vs. down with equal handle displacements. For some common  geometries and component placements this can vary some between zero bellcrank rotation and some predetermined deflection, like say +/- 30^o, were you set your flap or elevator displacements equal. I played around with this on a calculator furnished to me by Larry C., who had received it from someone else. I don't remember how large the asymmetry ever got, but it was interesting that it wasn't always worst for our "uncorrected" rod/horn angles. The calculator seemed valid, but did not account for all the adjustments we can make, like tilted bellcranks. I think I posted some graphs on SSWF, but someone who should have made more effort sort of discouraged me from following up. 'didn't need the grief. I'd think that the larger bellcrank might lessen that problem as well as easing the "Netzeband wall" problem. I can say from experience though that finding room in small stunt wings is a challenge. 3.5" is better for my latest (well 'late' has a dual meaning here).

SK

One consideration is not so much linearity or its lack, but asymmetry: the amount of up vs. down with equal handle displacements. For some common  geometries and component placements this can vary some between zero bellcrank rotation and some predetermined deflection, like say +/- 30^o, were you set your flap or elevator displacements equal. I played around with this on a calculator furnished to me by Larry C., who had received it from someone else. I don't remember how large the asymmetry ever got, but it was interesting that it wasn't always worst for our "uncorrected" rod/horn angles. The calculator seemed valid, but did not account for all the adjustments we can make, like tilted bellcranks. I think I posted some graphs on SSWF, but someone who should have made more effort sort of discouraged me from following up. 'didn't need the grief. I'd think that the larger bellcrank might lessen that problem as well as easing the "Netzeband wall" problem. I can say from experience though that finding room in small stunt wings is a challenge. 3.5" is better for my latest (well 'late' has a dual meaning here).

SK

Hi Serge,
This is where I get lost.  As a descriptive question, rather than a mathematical one, if there is Up/Down asymmetry with a centered bellcrank and 90-degree rod/horn angles, what is the source of the asymmetry.

Thanks.

I want a circular 4 inch bell crank with a gear on top to mate with a gear rack on the end of the pushrod to give linearity at the bell crank.  How come no one has taken up this wonderful idea?

You may not want linearity.  Look at Igor's plane, second in the 2008 world champs.  You may not even want symmetry.  I flew another airplane that came in second at a WC.  It had significantly asymmetric control.  Think of hinge moment and the relationship between loop radius and line tension at constant handle setting.   

What if, rather than raising the bellcrank, which would not be a fun (or maybe even practical) installation, you attached the rod to the bellcrank using a raised post (the Primary Force ARF uses such a post, tho it's not a flapped plane) so that the flap rod is parallel to the fuselage centerline in side view, and the rod/horn angle is 90 degrees with no horn bias?  Fore and aft forces trying to move the post out of vertical might have to be dealt with (the Primary Force's post is fairly wide at the base).

Kim-

I think there might be some problems with the bellcrank being more torqued to twist of rock. Maybe a better approach would be to do something like Brian Hampton did: run a separate arm off of an upward-extended bellcrank center bearing. Keeping it light with strong Materials (CF?) might be a challenge, but it could otherwise be kept simple, if a tube were sleeved over a typical 1/8" bellcrank support shaft, with the bellcrank in the middle and the arm raised. This would restrict any rocking motion of the bellcrank more than usual. Remember though that there still might be a more optimal height than that for a 90^o juncture at neutral.

You know, I'd think that an Excell spreadsheet might be used to compute this without the need to actually solve any equations or combine complicated expressions. I'm pretty burried now wih a couple other things that would also seem a waste of time, but I think this can be done. meanwhile, I've dug up an image of the input diagram fro the software by Mr. Herbron that I used for those graphs scanned to fit the <50Kb limit on SSWF. Maybe it will be legible. Note from the input window that some of these bellcrank rotations are larger than practical to get the really large deflections (It 'only' takes 38^o to get 30^o flap deflection with those dimensions.).

SK

A bunch of far-out and unnecessary ideas here, which basically hover around the real reason for using 4" bellcranks.  It is because we are used to lots of wrist movement in order to get the usual 30 deg. of elevator/flap throw!  And that's it!

So it's really about what we are used to.  If you are used to "quick" controls, you can still fly a smooth pattern, because your muscles are "trained" to move less.

I watch the plane when I fly, not my wrist.  My wrist moves in whatever way required to match the plane's path with the ideal path I have stored in my memory.

Floyd

Having said that, I can't justify it.  I should do some ciphering.

I spent a couple of days on my new plane figuring out this stuff by brute force with Cad.  I got the slanty bellcrank and the slanty slots in the wing laser cut.  I probably should figure it out all over, now that I decided to reverse the sign of engine torque and may shift the CG.  I'd sure like to see that spreadsheet (after it's documented, of course).  I'm not smart enough to make my own. 

I'll give you $10, but I wouldn't know what to do with it.

Why wouldn't it work to just bend the pushrod to go vertically up to the horn height and then horizontal to the horn?  Brace the 90 degree bends to avoid flex.  Or do I not understand the problem?   

                        ___________________________
                        |     
                        |
                        |
       ___________|                 

Russell,

The crucial relationship in the control system layout is the straight line from eye to eye: e.g., from the bellcrank's flap pushrod hole to the flap horn's pushrod hole. If you mounted a 'transfer shaft' to raise the flap pushrod to the height of the flap horn hole you create a new pushrod driving 'eye' in a better location for linear response. You have to be careful doing that  - it has to be strong enough to carry the loads and stay in the right position and "shape."
 
Whatever the shape of pushrod, the force acts from eye to eye. Think of an "in-line" model: a flapped model with the wing and stab chord lines at the same height...  The cables used as pushrods in RC don't have any easily understood eye-to-eye pieces, or perhaps just the locations where the cable enters the sheath, then where it leaves it, relative to the servo at one end and the control horn at the other...

* the bellcrank is typically flat, on the rib centerline.

* the flap horn is above the rib centerline, where the hingline is.

* the elevator horn is below the stab/elev centerline, where its hingeline is.

The flap pushrod has to tilt up to meet the flap horn's pushrod hole. The elevator pushrod is most often driven from the flap horn, above the flap hingeline. The elev pushrod has to slant down to enter the elev horn hole.

From the side view, the flap and elev horns rotation is visible. The bellcrank's is not. Viewed from above, the bellcrank's rotation is visible, but flap and elev horn rotation is not.

Angling the flap horn 'point-to-point'  - bellcrank pushrod horn placed where the line from it to the flap hingeline is at right angles at neutral flap - helps quite a bit. The 'golf-club' or 'cobra' shaped horns provide a rough approximation of putting the pushrod tangent to flap horn pushrod hole tangent to the arc it traces as flaps are deflected.

Similarly, where the elev pushrod hole in the flap horn is placed should be angled to fit the slant down from there to the elev horn's hole. And the elev horn's hole should be angled so that this pushrod is tangent to the arcs they trace at neutral elev. These arcs follow pushrod travel pretty closely around neutral, with a deviation growing modestly for 10° to 20° each way from neutral. The discrepancy gets significantly worse from a bit under 30° either way, on up.

The bellcrank output hole motion causes that end of the pushrod to move sideways across the fuselage; the flap and elev horns don't shift in that direction.  The bellcrank is usually mounted flat - in line with, or maybe parallel to, the rib centerline. That affects the 'linearity' of pushrod  motion; the geometry when the pushrod hole moves forward is diffferent from when it moves back. Not much, particularly near neutral... Which is why I suggested tilting the bellcrank to be parallel to the (straight) pushrod eye-to-eye line at neutral. That makes the error almost disappear.

Now, the thought of shifting bellcrank pivot, and/or flap horn position across the fuselage width... If the pushrod, at neutral, is slanted - seen from above - that will also introduce some non-linearity of response. Further, if the bellcrank output hole is tangent to bellcrank rotation, and inline with the flap pushrod there, ALL bellcrank rotation adds an error, slight but increasing with rotation angle.

Since the error is very small around neutral, I suggest moving the bellcrank so that the pushrod hole "slices" an arc 5° to 10° 'below' its arc.

What'd-he say??? 

Try it this way: mount the bellcrank so the flap pushrod is in-line with the flap horn when the bellcrank is rotated that 5° or 10° either side of neutral. I think this extends the "very linear" rotation range. The sideways motion is very slight (Cosine of angle * arm radius) at small angles, and this shift of the bellcrank put part of the error on the 'other' side at neutral, then it crosses 'just-right' before continuing to grow with more rotation angle.

Similarly, it should be a good idea to do the same with the flap and elevator pushrod holes...

Why wouldn't it work to just bend the pushrod to go vertically up to the horn height and then horizontal to the horn?  Brace the 90 degree bends to avoid flex.  Or do I not understand the problem?   

                        ___________________________
                        |     
                        |
                        |
       ___________|                 

This might work if the bottom two legs are fixed to the bellcrank and the joint between the vertical leg and the one to the right has a pin joint, but it might not be the lightest way to do it.