da Rock

And servos with a gear tooth missing. Sometimes it's out where it's trying to hold the up you need to do the loop.

Servos that aren't up to the job can get overpowered too.

Franco2fly

I probably should have posted this sooner, but here's the update now.

Glassing the rear and bottom of the stab, coupled with the struts going 2/3s the way out and attaching to the leading edge on the stab made the entire tail strong enough.  The plane was capable of tight loops, screaming dives and decent rolls.  Unfortunately, the Fuel flow and engine problems hit towards the end of the last check flight.  As I was entering the downwind leg at 50 to 70 feet, and in a bank, the engine froze, I mean stopped quit, no warning.  It had been running at mid-range when this happened and it was so sudden that I could not establish any semblance of a glide.  The nose dropped and the entire front of the airplane was destroyed.  ABSOLUTE POS.

KKKKFL

Rodney

Flutter is a very complex subject. Everything (yes EVERYTHING) will flutter if given the proper input (energy at it's resonant frequency). The only thing that will not flutter is something that is infinitely stiff and has no massan impossibility. Our choice is to change the resonant frequency of whatever is fluttering so that the input stimulus is not at the resonant frequency. How do you do that? Two ways, change the stiffness or change the mass or some of each. The easiest way in most cases is to increase the stiffness of the structure that is vibrating (fluttering). The solution in most cases is to move the resonant frequency (by increasing the stiffness with sheeting or more rigid components). Perhaps, since you say it is slightly nose heavy now, is to sheet the surfaces with 1/32 or 1/16 sheet balsa to stiffen it up. This will not add much weight and will increase the stiffness a great deal.

Rodney

Flutter is a very complex subject. Everything (yes EVERYTHING) will flutter if given the proper input (energy at it's resonant frequency). The only thing that will not flutter is something that is infinitely stiff and has no mass an impossibility. Our choice is to change the resonant frequency of whatever is fluttering so that the input stimulus is not at the resonant frequency. How do you do that? Two ways, change the stiffness or change the mass or some of each. The easiest way in most cases is to increase the stiffness of the structure that is vibrating (fluttering). The solution in most cases is to move the resonant frequency (by increasing the stiffness with sheeting or more rigid components). Perhaps, since you say it is slightly nose heavy now, is to sheet the surfaces with 1/32 or 1/16 sheet balsa to stiffen it up. This will not add much weight and will increase the stiffness a great deal.

AA5BY

The answer to whether the balance surface ratio is too great somewhat depends upon your flight observations. If the balancing surface is too large a percentage, then the rudder will not stay centered and will swing against whatever looseness exist in the servo, linkage and hinges. When control is given to the rudder so as to swing it thru the center position, it will want to slam to the other side. In short, it will not stay in a neutral straight position. Thus, if you had trouble with it holding a center trim position, it would confirm your suspicion that the balance tab is too large. If it were an elevator... you would know instantly that a problem exist because it would never be able to be trimmed for level flight. It would be slightly harder to sense whether the yaw is not able to be trimmed. Does it have problems with loop tracking?

Keep in mind that the dynamic involved is different that what is normally experienced. In this dynamic, the least bit beyond balance is intolerable and could very well stress the vertical fin... because each time it crosses center, it will violently jerk to the limits of the play in the other direction because the rudder will constantly be wishing to force itself full swing one way or the other. The area percentage varies depending on airflow at various places on the rudder and the kind of foil on the rudder. I'm more familiar with balanced rudders on sailboats where generally 15% area is the max.

da Rock

Keep in mind that aerodynamically balancing the surface won't do anything much to the movements you will experience in a model unless you have poor connections or have inadequate servos. The surfaces of models are held rigidly by design. There won't be any slamming about as might happen on a full scale plane. What AA5BY describes should immediately clue you in to bad connections or bad servos. The surfaces of full scale are held by the forces the pilot exercises against the stick. Aero forces can slam the stick and surfaces about. Our solid rigging to a powerful enough servo won't allow surface movement it's strong enough to control. Unless the servo is weak or it's gearing compromised or the rigging has slop, there won't be any slamming about.

The solid hold of the servo is the only thing that will experience anything and it will simply feel less force while holding or moving the surface. Thanks to the force generated by the aerotab, the force of the surface is countered to some degree.

What should also be considered is that the part of the surface from LE to TE where the tab is acts as an unchanging camber wing. The part of the span that has a hingeline running through it acts as a variable camber wing. The hinged part will flex itself into a stall much later than the part with the aerotab. You can usually expect the aerotab part to go into stall with not much stick input from your TX. The AOA of that aerotab part of the system won't have much angle to move before that unflexing section reaches it's limit.

When you design in an aero balance tab, you're basically tying a limited AOA part to a part that stalls at quite a significantly larger movement. You can think of it as throwing away some efficiently working area for less force needed by the joystick. We don't have a joystick, we have a TX stick that absolutely controls the surface rigidly. Why do we design models with those tabs? Beats me.

da Rock

Theoretically, aerobalancing would allow a modeler to use less powerful servos. Less powerful servos often have less robust gear trains. Nowhere in that is there anything about reducing the mass of the movable surface. In fact adding the aerotab to the movable surface usually adds more movable surface and increases the mass the servo and linkage must deal with.

da Rock

The Corsair was a design that included aero tabs on the elevator. It's interesting to note that not long into the design process they added robust boost tabs. They chose not to increase the area of the aero tabs, and incorporated the far more effective boost tabs.

BTW, the area of the Corsair's aero tab is about 12-13% of the area of the rudder aft of the hingeline. The planform of the rudder would have forced them to move the hingeline aft or increase the depth of the tab to get a larger percentage of area into the tab.

I wonder if there is an example of a full scale airplane that has a larger percentage of tab area than that percentaqe?

AA5BY

As offered, my experience with balanced rudders comes from sailboats and attempts to minimize helm effort needed for control. I can attest that once one crosses the balance point... that whatever play exist in a system (and there is always a slight amount) the rudder will upon crossing the center position, be pushed hard over against what play exist. And... if the autopilot is engaged, it will drive it literally bonkers as it can never trim the rudder for a straight course.

I'm also thinking that an autopilot is not that much unlike a servo trying to hold center stick.

It is therefore possible that instead of a harmonic oscillation that Rodney refers to, the issue may have been servo induced oscillation... aka - hunting.

jester_s1

Neat theory, but there's not a servo on the market that can hunt fast enough to shake the tail loose from a plane. This was a good old fashioned case of flutter caused by exceeding the speed limit of that particular control surface. The counterbalancers may have contributed by making it easier for the surface to deflect, but probably were not factor in the flutter happening in the first place. The problem was most likely too heavy an elevator mated to too weak a stabilizer, with a little bit of too weak a joint between the stab and fuselage thrown in.

[email protected]

My Feedback: (1)

hobby king is not the best place to buy anything get a american made plane

AA5BY

It wouldn't be a case of the servo being fast enough or strong enough to shake the stab loose... it would rather be the control surface is caught by the airflow and forced to the limits of the play then the servo seeks to recenter the surface and pulls it back across center just enough to be forced to the limits of play in the other direction.

jester_s1

That's still a very slow oscillation compared to flutter. Not that the scenario you describe couldn't break an elevator, but the OP described hearing the noise of flutter shortly before the tail departed the aircraft. The design phase is going to test counterbalances to see if they are too big, so unless the OP modified the plane to have bigger ones I doubt even Hobby King would be selling a product that is that far out of whack.

AA5BY

Granted... such a flutter oscillation would be considerably slower than a harmonic oscillation and the observer should be able to sense the difference. When approaching a near balance point, the normal resistance to harmonic flutter would also diminish.

flycatch

My Feedback: (26)

Seal the hingeline gaps on the vertical and horizontal stablizers.

jester_s1

That's a common myth that sealing the gaps prevents flutter. Sealing the gaps improves control authority and reduces drag so one should do it, but it doesn't prevent flutter. Flutter is caused by turbulence hitting the back edge of the control surface, which has nothing to do with the hinge gap.

Rodney

Very true. However, the act of putting tape over the hinge line does tend to stiffen up the surface a tiny bit so that fact may change the flutter characteristics a bit (shifts the resonant frequency to a higher number) so that the stimulating energy may no longer be at the new resonant frequency giving one a false sense that closing the gap stopped the flutter.

da Rock

Also, sealing a leaking hingeline will usually change the airflow/turbulence the hinged surface will experience.

mithrandir

My Feedback: (2)

indeed... mass balance

freakingfast

My Feedback: (3)

#1 I would sheet the stab with plywood or stiffen with carbon fiber on both sides of the horizontal stab.
# 2 Tight, tight, tight! No slop or springiness in the controls! Loose hinges, servo backlash, hole slop in the control horn or weak push-rod must be addressed.
#3 Gap seal ! This keeps air from leaking thru and upsetting the boundary air as well as stiffening the control surface hinge and flex. A clear piece of tape or covering material in & out of the gaps on the bottom only works fine.
#4 Take the elevator out of resonance! If you were to look at a fluttering elevator in slow motion, you would see both sides moving up and down in unison like a bird in flight. Attach a piece of lead 1/4 to 1/8 oz. to the the bottom (out of sight) ON ONE SIDE ONLY near the tip of the elevator. It must be affixed to the wood, covering is not good enough. This worked like a champ on a Kadet LT 40!
#5 One important thing I've found is having the elevator/rudder the same thickness or thicker at least at the leading edge where it meets the stab. It should also be well centered. What happens is when the boundary layer air comes off the stab and does not contact the control surface, it creates a small vacuum in this little void. This draws the control surface back into the airstream, but the other side is then pulled away from the airstream thus a vacuum pulls it back the other way. This vicious cycle repeats rapidly and creates one cause of flutter. I have solved flutter issue on three planes that were pushed well passed the intended speed by addressing control surface thickness when the other possible causes were corrected and had little effect.