Control Surface Load

[AirX]

Hello All,

A question that came up today concerning the loading of control surfaces and how that translates to the load at the servo. One of the members of our club lost an aileron to flutter but recovered the plane with little are no damage. Basically I need to figure the aerodynamic loading on a surface at a sustained angle of deflection and then to figure the load at the servo. Can anybody help me in this...

Eric

[Sparky Paul]

I've seen sites that do compute surface loads.. but generally if the servo to surface linkage geometry is proper, the surface is well attached with sufficient hinges, and there's not a large unbalance due to excessive material hanging off the hinge, flutter won't happen.
Sometines the gears on the servos are stripped by the flutter or flight loads.. this indicates a stiffening of the entire system is needed.
Ailerons are the most frequent victims.. due to the common use of a single servo and torque rods which can twist under load.
The Goldberg Tiger II is a well-known flutterer.. due to the very long thin aileron supported by few hinges and only the center mounted servo to drive it. A pocket can open up in the meat of the aileron at the end of the torque rod which lets the aileron move independently of the servo position. Switching to 2 servos cures the flutter.

[AirX]

Hi Paul,

Thanks for the help and I am thinking the same way you are that there was too many things that contributed to the problem. The model was a Hangar 9 Ultra Stick. The control horns were plastic and long the servo arms were plastic and also long. the trailing edge of the surfaces were rounded probably not heling the matter.
I have seen the calculator from Multiplex but it does not deal with the moment arms created by the chord length of the aileron and the length of the control horns plus the length of the servo arms. If the lengths of the arms and horns change then the load multiplies by the change in lever length. Just trying to be bette4r informed with the larger models.
Thanks again Paul.

Eric

[Gary Retterbush]

Flutter is IMHO generally caused by poor linkage. (linkage meaning the entire system to include hinges, pushrods etc)

My Euros worth, Gary

[AirX]

Hello Gary,

Thank you for answering my request for help. The problem is going to require a little more engineering than just worrying about the sizing of the control linkage. On the heavier planes and larger airframes the aerodynamic load needs to be figured into the needs of the linkage system such as the control horns, control rods, servo arms, etc. This will require some math and physics to be worked on the aero forces and I am looking to either find a spreadsheet to do the work that is already written or I will end up making it myself. My biggest problem is the figuring the physics to accomplish this for it is not my arena.

Eric

[sguty]

Hi all--
Flutter, in my experience, is rarely a result of the wrong size servo (until and unless the servo starts failing mechanically, as Sparky Paul sez.)
But if you are interested in servo loads, there was an article in RCM a while back (April '90, if I recall correctly) that went through the physics for aerodynamic loads and linkage mech advantages. I pulled that info into a spreadsheet which I think I can attach as a zip--if it doesn't work, send me a PM and I'll send you a copy (and confirmation of the RCM issue, which would require digging around in my stacks of mags.)

[Gary Retterbush]

Originally posted by AirX
Hello Gary,

Thank you for answering my request for help. The problem is going to require a little more engineering than just worrying about the sizing of the control linkage. On the heavier planes and larger airframes the aerodynamic load needs to be figured into the needs of the linkage system such as the control horns, control rods, servo arms, etc. This will require some math and physics to be worked on the aero forces and I am looking to either find a spreadsheet to do the work that is already written or I will end up making it myself. My biggest problem is the figuring the physics to accomplish this for it is not my arena.

Eric

Well, to each his own. I think you are making it a lot harder than it really is; of course, I only fly stuff up to 5 meters. I find if I simply use common sense and good linkage that things just seem to work. As I said, linkage includes ALL components. Now if the entire wing is fluttering, that is something else. Hope you get your problem solved. Gary

[GordonTarling]

Try - http://www.csd.net/~cgadd/eflight/calcs_servo.htm

Found this to work quite well, PLUS it also shows the formulae used.

[AirX]

Hello Guys,

Sguty: thanks for the access to the spreadsheet it is what I was looking to find if not make. Thank you for the help.

Gary Retterbush: I know that on a smaller model it is insignificant but when ailerons get to be 30 square inches in area then there are significant forces at work that need to be figured into the needs of the aircraft. The airplane in question lost 4 servos that day on the flight where flutter occured. The aileron servo on the port wing panel, the flap servo on port wing panel, the flap servo on the starboard wing panel and last but not least the aileron servo on the starboard wing panel was making crunchy gear noises but would still move the surface. These were Futaba 9002 servos with only 1 flight on high rates, with many flights at low rates. So yes your right that flutter on a smaller aircraft would be the cause of spindly linkage but this plane had only the control horns that could qualify as spindly.

GordonTarling: Thank you Gordon for the link to that site it also encompasses what we needed to work out the problems on this aircraft.

Thank you once again guys for the help.

Eric

[Sparky Paul]

30 sq. in>??
That's less than 1/2 the area of one half of the horizontal on my Seniorita.
My fully aerobatic Senior has about 135 sq. in. and has never fluttered.
From your description I'd look at the wing structure. I've seen fluttering ailerons remove wings that weren't stiff enough.
I had a Nordic glider, -no- control surfaces , flutter its way into the ground due to insufficient stiffness in the wing.
.
Sparky Paul
http://www.angelfire.com/indie/aerostuff
PJB's Seriously Aeronautical Stuff
http://www.networkone.net/~pjburke/index.html

[Gary Retterbush]

Hmmmm.... I thought my 5m Discus was sort of large what with ailerons of 74 sq. inches each. Boy, you must have some monster!;)
Gary

[Sparky Paul]

Here's my Stick-it... The Stick-it is notorious for shaking its tail off in flight. Even when power off.
This one doesn't. It has no speed limit, powered or gliding.
Standard servos and hardware...

[AirX]

Hello Guys,

You figure it out, The plane is a 120 ic stik that weighs over 12 pounds. Flies at over 70mph in level flight. has barn door ailerons and flaps. It lost 3 servos due to stripping gears on its last flight and it lost an aileron. It is an ARF from Horizon Hobbies. It is not mine because I dont like big planes. I know that the control horns before the flight were stiff and relatively strong, the control rods were 4/40 threaded rod, the servo arms were long plastick arms the same length as the control horns. After the flight where flutter occurred one of the ailerons departed the airframe. The flap servos stripped and the other airleron servo barely worked due to eminent shreading of gears. The servo that was in the aileron that departed was stripped. So it apears to me to be a series of problems that caused the overall problem. The owner tried to fly with high rates at relatively high speeds.

Eric

[Sparky Paul]

I've seen a couple of those ARFs.. none fluttered, but they weren't flown at real high speeds.
There's no way to know how well an ARF has been constructed.. material or glue.

I've seen and flown the .60 sized ARF in the series. This particular one was well built -the transparent covering helps here- and flew nicely. But it's not ever going to be pressed to the speed limit.. unless it's sold to someone else by its owner.

[green66]

To accurately calculate air load on a control surface at various deflection angles probably requires coefficients obtained from actual wind-tunnel testing. Although the configuration is simple, the fluid dynamics isn't - As the deflection angle increases, flow on the convex side (top side of a surface deflecting downward) of the control surface gets increasingly separated resulting in a growing wake, while flow on the concave side tends toward stagnation, both situations increasing the air load.

Since you ultimately want to know the worst-case load, I'd ignore the load at various angles, and instead consider the control surface deflected 90 degrees, e.g. fully drooped flaps, the high-load case.

A flat plate held at 90 degrees to the airstream has a drag coefficient ranging roughly from 1 to 2, depending on its aspect ratio. Since flaps, ailerons, spoilers, etc are relatively slender, and since the confinement of flow along one edge of the control surface by the wing surface will increase Cd some more, I'd use a Cd of 2.0.

Not sure where the center of pressure would be, given the asymmetic flow around only one edge of the panel, but don’t think it would be that far off-center; taking it at the center (half-chord) of the control surface is probably a decent assumption.

So how about a simple equation for servo torque based on the above? Stepping through (in?) it. . . .

Required servo torque = pushrod force x servo moment arm
Pushrod force = hinge moment / control horn moment arm
Hinge moment = drag force on control surface x half-chord of control surface
Drag force = dynamic pressure x projected control surface area x drag coefficient. Since the control surface is at 90 degrees, projected area equals actual area
Dynamic pressure = 1/2 x air density x airspeed squared
Air density (sea level) = .00238 slug/cu ft

Combining all the above, and working the unit conversion factor, we get:

Required Servo Torque (oz-in) = V squared x A x Cd x w x Ls / (15130 x Lh)

Where:
V = model speed in feet per second*
A = control surface area in square inches
Cd = drag coefficient (2.0), unitless
w = width (chord) of control surface in inches
Ls = servo moment arm (inches) from pushrod attachment point to axis of rotation, measured at a right angle to the pushrod
Lh = control horn moment arm (inches) from pushrod attachment point to the hinge axis, measured at a right angle to the pushrod

*To use V in miles per hour, substitute 7030 for 15130 in the equation.

For the model in question - 30 sq in control surface at 70 mph, and guessing a servo arm and control horn length of 3/8" and 3/4" respectively, and control surface chord of 2” (15" long), plug into the equation:

Required Servo Torque = 70 x 70 x 30 x 2 x 2 x .375 / (7030 x .75) = 42 oz-in

But this is for level cruise. I'd use servos with an ample safety margin over the torque based on normal flight speeds, and also consider the application - ailerons don't deflect nearly as much, so values from the above calc might be conservative as is, whereas for a flap servo, I'd add about 50%.

If you want to truly know the air loads you’re dealing with, you could rig a digital scale or force transducer to the pushrod, mount your wing on struts to your car and drive around in a big parking lot, noting the air loads at various speeds & deflections. But this is the point where the whole thing becomes a research project more than a hobby.

Hope this helps some, or at least provides some add'l insight.

[Sparky Paul]

Looks good, green..
I've save this one..
42 oz-in is the nominal value for the standard servo torque.
As it would be almost impossible* to deflect the flaps at 90 degrees at 70 mph.. or even 35.. this is a good limiting case.
* The flaps would tear off.. the plane would pitch up uncontrollably..the horn would be ripped from the surface..
Something in the structure or the control system is letting the aileron flutter.. as mentioned, either the control system isn't stiff enough or the wing structure isn't.
All planes have a flutter speed. The trick is to find it, and do what is required to move it beyond the speed capability of the aircraft.
That's what the "keel" on fuselage on my Stick-It (note on the photo) did to the tail end flutter.. and to keep the ailerons from fluttering what I did was add a shear web to the trailing edge sheeting, making the wing a double-D-tube assembly. It CAN NOT flex within the speed range the TT 36 PRO can take the plane at full power.
.
Sparky Paul
http://www.angelfire.com/indie/aerostuff
PJB's Seriously Aeronautical Stuff
http://www.networkone.net/~pjburke/index.html

[AirX]

Thank you Green66, very good explanation. Thank you for the help, I am sure we can figur out the problems that have given headaches to one of my local club members.

Eric Barnett