Hi everyone. I'm new here, but have been observing for months. I keep seeing threads of servo flutter and how it destroyed peoples' planes, but I can not find the definition of flutter anywhere. What exactly is it? I have heard that having underpowered servos can cause it, and result in a total destruction of the plane, but how can an underpowered servo cause a plane to come apart in midair? Can someone please explain this to me?


Thanks!
Johnny

G'day Mate,
It's actually control surface flutter, not a lot to do with servos, unless they are very weak, the control surface whether it is elevator, aileron or rudder, oscillates, back & forth, at a very high speed caused by many things, such as loose linkages, which can allow movement, or large gaps in the hinge line.
I have even had the experience of aileron flutter caused by a sloppy aileron torque rod, where it was positioned in the aileron, it allowed the aileron to move up & down at a very high frequency, when the plane was at full throttle, luckily I throttled back & slowed the plane & landed safely, but it can cause the destruction of the tail or wing very quickly, if allowed to continue.
And planes don't fly very well without a tail, or wing half.

Alan is on the money, you can hear the flutter, sounds like a hum and will occur at high speed. If you know what it is you can save your plane. Ive seen it once on a video, once in person.....and it happened to me once on an older airplane, was doing a desending left turn for a high speed pass down the runway as i rolled level and added power that hum I was talking about was loud and i could see the Rt aileron was fluttering. cut power did a desending 360 to touch down. (dual aileron servo aircraft) This was caused by the gears stripping on the RT aileron servo. Changed servo and was flying her again the next week.
I'm telling you this cause fore warned is fore armed, dont worry about it though if your aircraft is set up properly it wont happen unless you lose a servo like I did.

When the hinge line is ahead of the control surface's center of mass, there is a positive feedback loop. The wing acceleration caused by the control surface deflection acts on the control surface's center of mass to increase the control surface's deflection until the wing has deflected enough to resist the force and then, just like a penulum, returns which causes the control surface to reinforce the motion in the other direction. This oscillation continues until something breaks.
Full scale aircraft usually have control surface counterweights called "mass balancers" that are adjusted so that the control surface's center of mass is on the hingeline. The importance of this mass balance is one reason that even something as simple as a paint job needs to be done by a certified A&P on full scale planes.
Mass balance can be used in models also.
At high enough speeds, even wings without control surfaces can flutter, i.e. a Gentle Lady glider in a long steep dive. Even propeller blades have been known to do it in high rpm pylon racing engines. The prevention of flutter is why a lot of helicopter blades have lead wire glued in the leading edge.

Make the control surfaces as light as possible.
The linkage should be as stiff as possible.
Consider the inclusion of control surface counterweights on high speed models.

A fair amount of the time, when you here flutter or someone else does and yells at you, it's too late. Sort of goes along with the phrase "hey! Look Out!!" or and all time classic "Watch this!"

Here is a link to a video of a crash caused by flutter. In the slow motion shots of the crash, you can clearly see the wings fluttering and it shows just how violent wing flutter can become.

http://www.youtube.com/watch?v=ENcIZ...elated&search=

Yes, a weak servo can allow the surface to flutter, but the majority of flutters are caused by sloppy connections. If your pushrod is weak or the connections aren't snug, the surface won't be held firmly. Once the surface starts to flutter, it generates the forces that tear the wing up, or break the horizontal stab off the airplane. Sometimes the flutter destroys the servo's gears, allowing even more surface freedom to flutter even more.

What basically happens is that the surface is moved by aerodynamic forces in one direction. When the surface winds up maxed in that direction, aerodynamic forces have built up in the opposite direction and the surface is driven back the other way. This cycles and builds up greater forces with each cycle. It might wreck the servo gearing and get more freedom of movement, and that gives it even more power and speed. And depending on the strength of the wing or stab, the force of the flutter can tear the supporting structure apart.

Whenever a surface is deflected, it generates forces that try to twist the supporting structure. The ailerons try to twist the wing when they deflect for example, even when there is no slop in the aileron connections. Add some slop and oscillation, and bad things happen.

The discussion of flutter ought to include a lesson or two for beginners....................

Slop in connections is bad. OK...... Obvious lesson.

What hasn't been mentioned is the slightly bad advice a bunch of beginners get from some of their more experienced flying buddies. Quite often, you're told that "it's easier to adjust the servos with that radio, you paid for an expensive computer radio, after all." You had no idea what holes to snap the connectors into and you've wound up with an elevator that works like a barn door and ailerons that go straight up and straight down. And it's not on a 3D, but on your trainer. You're actually in a world of hurt and don't know why. And the field expert has just shoveled some dirt out of a grave he's digging for your airplane.

If the ARF instructions don't tell you what holes in the servo arm or aileron/elevator/rudder horns, then do this:

Use the hole in the servo arm that's closest to the servo and the hole in the horn that's the farthest from the surface. And if that doesn't give you the amount of surface deflection the ARF instructions have suggested, then move the servo connection out one hole and try again. Still not enough? Move the connection on the horn IN one hole. Still not enough? Move the servo connection out one, etc etc Or you can simply try each hole in the servo from inside out. Both plans work.

The idea is to give the servo help with leverage. Or at least to not hurt it. And once you've got the mechanics worked out to the advantage of the servo's and the model, THEN you can fine tune it with your expensive computer radio. But don't screw things up with the radio until you've gotten the physical connections in the ballpark. If you've got bad leverages to begin with, "correcting" the surface deflection with the radio winds up making the servo's work even harder. You'll actually get less force out of the servo and probably get less accuracy as well. And to defeat flutter before it begins, you need to insure that the connector rigging is helping, not hurting the servo's leverage.

And learn yet another even more subtle lesson.......... Next time you find an expert with that advice about using the radio first, put him at the bottom of your list of people to ask for advice.

Good point daRock! 'tis one of the reasons that throttle control and/or a speed limiting large diameter low pitch prop must be used on planes designed for 3-D aerobatics.

Another thing to avoid is a round trailing edge on a control surface. Sharp pointed wedge is ideal but impractical because it gets damaged so easily. Almost a good but a lot more practical is a wedge shape with a squared off trailing edge.

Mass balancers that are easy, effective and work every time for aircraft that have demonstrated a tendency for flutter.

John

ouch that was not a nice thing to witness. Very expensive problem that was. And to a beautiful jet too I cant bear the thought!

http://www.geocities.com/mgd3/flying/flutter
http://www.giantscaleplanes.com/flutter.htm
http://history.nasa.gov/SP-4302/ch1.10.htm

Some interesting IMO reading on the subject.

Just Google: "Aileron flutter" for more

Everyone has pretty muched described it on the money. If you ever hear a buzz or vibration sound at high speed, especially in a dive, IMMEDIATELY cut the throttle! I ave a GP U-Can-do 46 that will flutter the ailerons if you don't manage the throttle in dives. Luckily the worst thing that has happened was some cracked balsa in the aileron itself. I've added stiffer linkages and done a few other things and the problem is a lot better.

Wow. thanks for all the insightful comments guys. This helps out a bunch. I never knew that the forces on a plane were so great. I'll have to triple check my control linkages. Also, I had another question. Do fiberglass control rods flex? Because I'm looking for a pushrod that's still like metal, but is made out of some other material that's lighter and non-conductive.

Thanks again,
Johnny

Flutter is not only confined to movable surfaces. A wing, a fuselage, a fin or stab can go into flutter without the moveable surfaces fluttering. Everything will flutter given the proper stimulus, even an iron bar. The most common cause of flutter is flexable surfaces, flexable hinge points, soft (not rigid or stiff) structure. The poor or sloppy hinges etc. are not the cause of the flutter, just the means for inputing the stimulus that puts the member into oscillation or flutter. Yes, eliminating the inputs does help prevent the flutter but the only sure cure is to have the offending member stiff enough or rigid enough to put it's mode of oscillation outside the frequency range of the stimulus.

Try carbon fiber. Slightly pricey but great.

Bruce,
Can carbon fiber be bent to make Z or L-bends? The only CF setup I can find is from Central hobbies, and their setup is very complicated and uses ball joints and clevises. I've heard from many people that Z and L bends are the most reliable setup.

Thanks,
Johnny

I've taken .060" ID CF tube and slid it over 2-56 metal pushrods. They're lighter than larger metal rods but very stiff and you still have the metal for Z-bends, etc.

No carbon fibre rod cannot be bent at all. I use it on most of my competitive as well as sport airplanes and will typically use the 1/16 th size and the 2-56 threaded rod fits perfectly, It is incerted about an inch and secured with epoxie. A clevis is used on one end And usually I will use a piece of rod on the other end with the S bend already made up, every thing is fitted and then marked the rod is removed and that last end is epoxied. Perfect length every time.

Actually the the one thing that would improve linkages on many airplanes and particularly many arfs is not necessarily changing push rods but eliminating all bends. By carefully planing exit holes in the fuselage rods in most cases can be totally straight. This may not be the location indicated by arf manfacturers.

John

Also careful laying out the path of the rods during assembly and providing sufficient support midway will reduce flex of standard ny-rods too. Some complain of thermal expansion and such on these but it is minimal on sport planes and can always be trimmed out.