What's new
Van's Air Force

Don't miss anything! Register now for full access to the definitive RV support community.

Flutter testing?

TXFlyGuy

Well Known Member
Has flutter testing been accomplished on all RV models? How is the testing done? What are the parameters?
What was the result of the test?
 
Flutter

Freeedom from flutter was tested in a careful scientific way for.the RV-8 after a structural failure accident years ago. This was documented in the RVator at the time. If you can't find it I'll look through old issues. I suspect but don't know that the other older models might have been tested by the control excitation method = i.e. Slap the stick. Anyone (outside Vans)) know how the newer models were tested to be flutter free within the design envelope?
 
Is the technique of "stick slapping", i.e., control excitation mode, an acceptable means of testing? I know of other kit manufacturers who have employed this method. They took the aircraft to altitude, pointed the nose downhill, and applied full power. Then the stick was aggressively slapped, side to side, back and forth. The results were negative in this example.

Just curious how many people actually do this, and the validity of the result.
 
So the Vne speed is derived from this testing? And what is the exact relationship of indicated versus true airspeed with regards to flutter?
 
Don't know.

I don't have an answer to your question. I too remember the retesting of the 8 modal after #2 went in. The others you will have to ask as is recommended above. There is a section in the builders manual of the 8 that touches on this subject. You need to make this call for yourself. Having said this I did this old school flight testing before I would let anyone in the aircraft with me. We did not only the VNE build up slow and in stages but also took the engine and prop up to speed, the for and aft CG to a little over the limit as well a gross weight.
best I can remember 103% for VNE and flutter, 111% for engine and prop speed, 50 Lbs over on weight and 5-8% further forward and aft of CG limit by the book. This is just my nature to test every thing to be sure I can trust it.
I don't think that it is easy to make a statement that every thing is known about every one of these RV's as each is not a carbon copy of the next. In the case of our airframe we can say it did meat with a little margin for our peace of mind. As Dan said it would be nice if you get an answer to post it here so all can see it. Hope this help. Yours, R.E.A. III #80888
 
....They took the aircraft to altitude, pointed the nose downhill, and applied full power. Then the stick was aggressively slapped, side to side, back and forth.....

One issue with this is that unless the airspeed is either stable or decreasing, it's increasing. And if the test is being done as the speed increases, the instant after the hopefully successful test, the plane is going faster and is at an untested airspeed.

Another issue is if the plane is pointed downhill, then as the test is performed, its altitude and therefore the aerodynamic damping is increasing. This also gives results which are both highly variable and difficult to document.

It's a poor test procedure, and I'd be surprised if it was really performed in that manner.

Dave
 
One issue with this is that unless the airspeed is either stable or decreasing, it's increasing. And if the test is being done as the speed increases, the instant after the hopefully successful test, the plane is going faster and is at an untested airspeed.

Another issue is if the plane is pointed downhill, then as the test is performed, its altitude and therefore the aerodynamic damping is increasing. This also gives results which are both highly variable and difficult to document.

It's a poor test procedure, and I'd be surprised if it was really performed in that manner.

Dave

Perhaps I'm in error. The downhill, WOT, was part of the test to check max speed. The flutter testing was probably a different part of the test program.
 
So the Vne speed is derived from this testing? And what is the exact relationship of indicated versus true airspeed with regards to flutter?

There is no one simple answer. "Flutter" refers to different modes: wings twisting, wings flapping. Tail wagging, tail twisting. Some of these modes scale exactly with the square root of density times true air speed (which is indicated airspeed). Other modes scale differently.
 
I have seen data for a few computationally derived flutter boundaries, an experimentally determined one on a rotor, and hand worked through a 2DOF model problem (wing bend+torsion). In each of these cases, the flutter boundary varied with altitude along a line very close to half way between a constant EAS line and a constant TAS line. (EAS is equivalent airspeed, it is the same as IAS if your instrumentation is perfect).

The reason that the flutter boundary tracks this way is that while the loads scale with dynamic pressure (EAS) the damping is dependent on a simple product of density and velocity (rho * V).

So assuming a constant TAS boundary is conservative.

Note that the flutter 'mode' is usually a coalescence of two structural modes, whose natural frequencies change with speed because of the damping, and sometimes because the load distribution changes with speed. Examples would be wing bending + torsion, or fin bending + rudder imbalance oscillation.

The idea of a 'stick rap' is that an abrupt, sharp input excites many modes to fairly high frequencies. If you are familiar with a Fourier series, it takes many terms in the series to represent a square wave or spike function. The actual range of modal frequencies that is excited depends on control system free-play, damping, inertia, etc. I would have thought that one could get higher than 6 hz, but I would defer to Carl's experience.
 
Flutter

I think in a group this large a lot of the builders want to know the true answer to the vne question as it keeps coming up and a group funding might be the way to do it with each of the different models being shaken and a study done.
Maybe a college grad study at an aeronautical university?
Bob
 
Bob---thats great, but we're talking experimentals here. Are 2 RV7s built exactly alike so the test data between the 2 would be accurate. Yeah maybe if they were built side by side in an assist center. But mine, and others probably wont be. Like adding 2 pieces of heavier angle across the aft structure for the ADAHRS. Heavier material, stiffer aft structure, so is the harmonic the same in both planes?
I know we're talking about tail flutter here, so does adding Alan's tail mod change the vertical harmonic and the speed that flutter begins?

Tom
 

Carl, what does the GVT and flight test cost $ for a GA? With several inflight failures on the 7, it might be a good candidate.

Tom, it might be more interest in knowing the difference of rudders than build-to-build.
 
Van's does have something to say about flutter testing, it is in Section 15 of the builder's manual.
Flutter testing of factory prototypes has resulted in ... determined through flutter testing at a speed of 20 mph above VNE ...slap-the-stick method...is potentially dangerous and requires a very skilled pilot trained to recognize the subtle control responses...it is suggested that amateur builders do not perform flutter testing in their RVs.
Suggest a few read that section before testing luck by exceeding any printed limitations from Van's.
 
Builder's Manual Points to Ponder

The builder's manual (Section 15) has a flutter discussion as pertains to RV's. In the RV training materials, over on the safety page?Vne discussion begins on page 343. The link to the current draft is in the sticky at the top of the page. You can download a PDF version of the training materials if you'd like. The table of contents is hyperlinked to help with navigation. If you?d like a Word version (or PPT version of the RV aerodynamics briefing), drop me a PM or e-mail. Treating Vne as TAS is a simple technique that will ensure design limits are not exceeded at any altitude. If you haven?t already done so, on the factory web site, there is a good article called ?Flying High and Fast? that has an excellent flight envelope discussion as applies to RV?s. It can be accessed via this link:http://vansaircraft.com/pdf/hp_limts.pdf. If you don?t have a copy of the builder?s manual, it can be obtained with a set of preview plans from Van?s. This is an excellent resource for folks that may not have built their ?new to them? RV. If you are considering the necessity for flutter testing, please read the last paragraph in the builder?s manual carefully. I?ve highlighted it in bold below. One other point to ponder is the effect of aircraft aging and wear and tear (e.g., control bearing wear). Even if accurate flutter testing is properly conducted for a particular airplane, as the airplane ages, flutter characteristics may change, and actual flutter margin may be reduced. Therefore, for airplanes built in accordance with the plans and instructions, respecting the design margins and using the TAS Vne technique are recommended. This is experimental aviation after all, and if the airplane has been modified, then proper testing may be warranted.

For folks that don't have access to the builder's manual, here's what Section 15, Revision 8, published 06-08 says (please check with the factory to determine if there is a subsequent revision to this section):

?Flutter in an aircraft structure results from the interaction of aerodynamic inputs, the elastic properties of the structure the mass or weight distribution of the various elements, and airspeed. The word ?flutter? suggests to most people a flag?s movement as the wind blows across it. In a light breeze the flag waves gently but, as the wind speed increases, the flag?s motion becomes more and more excited. It is easy to see that if something similar happened to an aircraft?s structure the effects would be catastrophic. In fact, the parallel to a flag is quite close.

?Think of a primary surface with a control hinged to it (e.g., aileron). Imagine that the aircraft hits a thermal. The initial response of the wing is to bend upwards relative to the fuselage. If the center of mass of the aileron is not exactly on the hinge line, it will tend to lag behind the wing as it bends upwards.

?In a simple, unbalanced, flap-type hinged aileron, the center of mass will be downward. This will result in the wing momentarily generating more lift, which will increase its upward bending moment and its velocity relative to the fuselage. The inertia of the wing will carry it upwards beyond its equilibrium position to a point where more energy is stored in the deformed structure than can be opposed by the aerodynamic forces acting on it.

?The ?wing bounces back? and starts to move downward but, as before, the aileron lags behind and is deflected upwards this time. This adds to the aerodynamic down force on the wing, once more driving it beyond its equilibrium position and the cycle repeats.

?At low airspeeds, structural and aerodynamic damping quickly suppresses the motion but, as the airspeed increases, so do the aerodynamic driving forces generated by the aileron. When they are large enough to cancel the damping, the motion becomes continuous. Further small increases in airspeed will produce a divergent, or increasing, oscillation, which can quickly exceed the structural limits of the airframe. Even when flutter is on the verge of becoming catastrophic, it can still be very hard to detect. What makes this so is the high frequency of the oscillation which is typically between 5 and 20 Hz (cycles per second). It will take only a very small increase in speed to remove what little damping remains and the motion will become divergent rapidly.

?Flutter testing of factory prototypes has resulted in establishing a NEVER EXCEED SPEED (Vne) of 210 statute mph for the RV-3,4 and RV-6/6A, 230 statute mph for the RV-7/7A/8/8A and 210 statute mph for the RV-9A. This speed was determined through flutter testing at a speed of 20 mph above Vne. (FAA certification criteria require flutter testing up to Vne plus 10% or about 20 mph) The flutter testing performed consisted of exciting the controls by sharply slapping the control stick at various speed increments up to this level. Under all conditions, the controls immediately returned to equilibrium with no indication of divergent oscillation s indicative of flutter. This testing was performed on factory prototype aircraft, and the flutter free flight operation of subsequent amateur built RV?s has substantiated published Vne.

?The ?slap-the-stick? method of exciting the controls for flutter testing is potentially dangerous and requires a very skilled pilot trained to recognize the subtle control responses which indicate the onset of flutter. For this reason, it is suggested that amateur builders do not perform flutter testing of their RV?s [italics added]. Rather, the airplane should be constructed in strict conformity to the plans with particular attention paid to the control system?trailing edge radii, skin stiffness, control linkage free-play and static balance in particular. Maintaining conformity with the prototype (plans) will provide and adequate level of assurance against control surface flutter. Any design changes to the control surfaces, control system, or primary structure could invalidate the testing which has been done, and require that testing be re-accomplished.?

Fly safe!

Vac
 
Can not say this for all experimental aircraft but I believe all Vans models have been throughly flutter tested.

The result of this flight testing and structural engineering analysis is the basis for published limitations.

It is the builders responsibility to flight test the aircraft in Phase One up to those limits to insure aircraft stability at those limits. If is not necessary to go beyond those limits into the safety pad area, i.e., testing to failure.

If flight testing is done beyond established limits, you are doing more than is prudently required to have a safe airplane.

Builders are not expected to induce in flight flutter, to do so not being professional flight test pilot is crazy.
 
Flutter

Carl, what does the GVT and flight test cost $ for a GA? With several inflight failures on the 7, it might be a good candidate.

Tom, it might be more interest in knowing the difference of rudders than build-to-build.


One only needs to read thru the forums to find why there are problems with the 7 rudder . I believe the plans call for Pro Seal on the AEX between the skins on the trailing edge when it is riveted together. I've read everything from JB Weld -epoxy -silicone and other adhesive being substituted. Then the countersinking and riveting . There is a builder on the forums that just did his " countersinking was not deep enough , I'll fill the gap with epoxy , build on "
Take a look at a 7 rudder , big weight at the top rudder stops on the bottom ,slam that thing against the stops and the trailing edge might "unzip "
 
Flutter

one builder rebuilt his rudder and documented the process , he openly admits that he skipped the Pro Seal procedure . He crashed and died .
 
FFT

What would really be cool is to do design work up on different models and then shake several of same model to see how they differ since no two exp are built the same.
Bob
 
Again, I have no data to back up this supposition, or whether the rudder's torsional mode is even an important player in the flutter characteristics of this design.

Have built one of each of these rudders. I was about to measure torsional stiffness until I read about the shear bucking failure of (the NJ tragedy) HS spar between the skin and fuse. Afterward the SB came out and added the "H" reinforcement to the spar.

I did measure the weight and CG of the two. The riveted TE, (aka 7-tall) is .016 skin, the folded (aka 8) .020 skin. The riveted TE is heavier by 400g and moves the CG from 3.18" to 4.05" from the pivot line. The 7-tall has some wiggle zone torsionally due to the skin looseness. I assume it is due to a combination of thickness and minor stretching from the dimpling and riveting the TE. All w/o glass caps, and with the same, by the print, counterweight.

If you think torsional stiffness is a key factor, then I will measure both, including quantification of the no-force wiggle zone. The planned method is fastening the CW end (top) with formed boards, then a weight and dial gage at the tip of the opposite end. Pivots will be included. Before doing this, what definitive value will it provide to Vne?

Suppose the 7-tall has a flop zone of .2-.3" and the slope of the force-deflection
is half of the 8, then what? Is it enough? What does it really tell us?

We should note that the 14 rudder has lots of internal bracing for stiffness. Beautiful design.

I will be doing first flight with the 7-tall, and some spin testing, followed by replacement with the 8 and more spin testing to validate the difference. Yes, I know Vans tested it.

Your experienced opinion is appreciated.
 
Last edited:
Carl, great writing...

As I mentioned earlier, I'm convinced Van's airplanes are flutter free to the their design Vd, if properly built and the control surfaces are mass balanced correctly.

So do I, the operative phrase being "to their design Vd", but that's like saying "Peace exists until the war starts." Most want a DMZ, a margin, a place to posture and give the other guy the finger...or maybe just not get shot for taking a wrong turn in the dark. A low flutter margin is the other guy with a hair trigger.

The -8's rudder drawing is below, and it has a folded TE, compared to the riveted TE of the -7......I'm not sure if the riveted TE or the folded TE would be stiffer in torsion.

Also note there is a considerable difference in CG due to the 7's trailing edge wedge. The Canadian report fingered excessive filler and paint, which exacerbates the issue. Carl, could you share a few educational words on the effect of increasing control surface imbalance?
 
Over-G vs Flutter Point to Ponder

From a purely operational standpoint, with the exception of some early RV-3's, RV's constructed IAW instructions and operated within the design envelope using acceptable handling techniques have proven to be free from flutter and structural failure. If limits are exceeded, then experimenting is occurring, whether intentional or not. As dangerous as flutter can be, so is 14+ G's available in the hands of a pilot that doesn't understand how to properly maneuver about three axis that may find himself in an unusual attitude, regardless of tail design and the amount of Bondo present. I can understand where conflating the two hazards is quite easy once speed limits are exceeded.

The "no man's land" Dan refers to in his metaphor can be effectively maintained in an RV-type by understanding the design envelope, using the Vne expressed as TAS technique, understanding when it's acceptable to exceed Vno (maximum structural cruising speed) and combining this with a handling skill set that allows the pilot to effectively manage energy. This requires the pilot to anticipate acceleration and deceleration and know how to control that with a combination of throttle, G, velocity and lift vector management (i.e., power, pitch and roll). Proper control application is also critical: how the airplane is rolled and how the stick is pulled. Another required skill his how to safely recover from an over-speed condition in any attitude.

The Canadian mishap was a pilot conducting some sort of "chase" formation that looked a lot like LOWAT BFM (low altitude basic fighter maneuvers, or 1 v 1 maneuvering) from the video evidence. The pilot did not appear to be a trained fighter pilot based on information in the mishap report. The ability to maneuver effectively in this type of fluid, three-dimensional environment requires some knowledge of energy maneuverability, aircraft handling characteristics, turn rate and radius management, ability to simultaneously assess and react to the other aircraft's maneuvering, and not allowing either the other aircraft or the ground to present a maneuvering hazard. It typically takes some specialized training to provide a pilot with the requisite skill set for this type of activity. A simple startle reflex at high speed is sufficient to generate more than enough G to bend something or start to remove airframe parts if the pilot suddenly realized that the ground was an immediate hazard. Another thing to consider is that overall G available (i.e., how much you can pull without bending or breaking anything) decreases if the airplane is rolling simultaneously--i.e., the ultimate load limit may be as low as 5.9 G's for aerobatic RV-types; however, Van's does not specify asymmetric G limits for any RV-type.

In another mishap, the pilot was at 6500 feet and 130 knots (150 MPH) and began a vertical maneuver of approximately 3500 feet which resulted in an airspeed of about 220 knots (253 MPH). While the exact maneuver was not described in the mishap report, if a pure vertical pull through (split-S) was attempted from these starting parameters and power and G were applied to control airspeed, a turn diameter of less than 1000' would result. Even with idle power, it may have been impossible to properly modulate airspeed without exceeding G limits as the maneuver was started at a speed above corner velocity (maneuvering speed). If G limits were honored, or insufficient G applied, then the turn circle would open up and the airplane will accelerate, perhaps accounting for the 3500 feet referenced in the report, which also correlates with the 220 knots mentioned. At 220 knots, the mishap aircraft was capable of generating in excess of 15 G's. Since "g available" is really how much lift the wing can generate, recall from pilot training that the amount of lift increases as the square of airspeed. The ultimate load limit for a properly loaded aerobatic RV-type is 9 G's. The mishap report indicated structural failure was a result of static overload, and no evidence of multiple dynamic loading fractures was found. No pre-existing fatigue damage was found.

As a point to ponder, consider that nothing really good happens during high-speed flight when design limits are exceeded in any airplane. It is good to have a healthy engineering discussion regarding flutter, structural integrity and design, but ultimately all of these mishaps share a common thread: handling error. The historical exceptions may be some early RV-3 mishaps involving the rear spar attachments.

It is incumbent upon the pilot to understand the limits and utilize appropriate handling techniques to either stay within them or recover if they are unintentionally exceeded. Hopefully, the information in the builder's manual was sufficient to answer the original poster's questions, and this post isn't off the beaten track too far--now I'll turn the discussion back to the engineers, pitch out of the fight and continue to read and learn!

Fly safe and have fun!

Vac
 
Last edited:
Thanks

I really appreciate the time that Carl and others have shared in this discussion.
 
Last edited:
RV14 Rudder

Bill,

Thanks for the great info.

Questions:

1) Just to confirm, are the mass balance weights the same for the -7 and -8 rudders by the drawing?

2) Are the -7 and -8 rudders the same planform size?

I did not know that the -7's rudder used 0.016" skins (versus the -8's rudder with 0.020" skins). That is really interesting. The -7's rudder will be less stiff in torsion compared to the -8's rudder because of this (if they are the same size). Also, the -7 rudder's torsional frequency will be lower because of this plus the heavier weight and aft cg of the -7's rudder (again, if they are the same size).

If you wanted to measure the torsional stiffness of both, that would be interesting information. But at this point I do not know whether the rudder's torsional mode is even an important player in the flutter characteristics of this design. Therefore, I don't know how we would asses whether the flutter margin is less with the torsionally softer -7 rudder. I don't want you to do a lot of work that we may not be able to use.

Good info on the difference in cg between the -8 and -7 rudders too. ("The riveted TE is heavier by 400g and moves the CG from 3.18" to 4.05" from the pivot line.") DanH had a question about that also, and I'll respond to his query in another post after this one.

Also, you made an interesting comment about the -14's rudder design with respect to stiffness. I'll order a set of CD plans for the -14, but in the meantime could you email me a drawing of the rudder, or point me to a drawing somewhere?



Here's a link to the RV14 rudder plans. It would be interesting to hear the differences to other RV models.
https://www.vansaircraft.com/pdf/revisions/RV-14/RV-14_07.pdf
 
Some important differences with regard to the RV-3B rudder!

The RV-3B rudder:

Has a bent trailing edge, not the strip.

Doesn't have the shear clips.

Doesn't have the rudder horn brace.

Doesn't have the mass balance.

There are probably some smaller differences which I've missed.

The RV-3B has a Vne of 210 mph and .016 control surface skins.

Dave
 
Carl,

You're correct regarding the prototype RV-8 wing spar modification. I reviewed Van's write-up, which I should have done prior to posting (no excuse!). I edited the post and annotated the reason for edit appropriately; so hopefully it doesn't distract from the outstanding conversation going on in this thread.

Cheers,

Vac
 
The 7-tall has some wiggle zone torsionally due to the skin looseness. I assume it is due to a combination of thickness and minor stretching from the dimpling and riveting the TE.

I'm kind of alarmed by Bill's comment here, that there may be a sloppy fit to parts that allows some relative motion with very low force. That sounds really bad to me, if for no other reason than service life/durability.

Ironically, the friction associated with slop in a structure like that may add enough damping to offset the reduced effective stiffness. Its not clear what that might do to flutter margin. But it would be a concern for how that margin might change with time if there is actual working of joints in the assembly.

If it is, instead, just a very low stiffness because the light skin is buckling on one side, thats not so good either. Bill, can you shed a little bit more light on this question of the 'flop zone'?
 
As long as you guys are modelling and anaylizing, any comment on the relative stiffness of the heim joint hinges? I've never been a fan of a cantelevered rod end used as a hinge on a primary flight control. It would also be interesting to see the results with the jam nuts loose, as we often see with flying airplanes.
 
What is the relationship, if any, between G load rating, and the propensity to develop flutter?

That is, would an airframe rated for +9 / -6 G's be more or less apt to develop flutter than an airframe rated at +5 / -3?
 
one builder rebuilt his rudder and documented the process , he openly admits that he skipped the Pro Seal procedure . He crashed and died .

Can someone elaborate on this comment? Did his rudder fail and cause the crash? If so, was the rudder failure due to the lack of proseal?
 
Can someone elaborate on this comment? Did his rudder fail and cause the crash? If so, was the rudder failure due to the lack of proseal?

It would be best if the report was read to draw ones own conclusions. Be sure to access and read all the docket documents and pictures.

With respect, this is not the thread for this discussion.
 
If it is, instead, just a very low stiffness because the light skin is buckling on one side, thats not so good either. Bill, can you shed a little bit more light on this question of the 'flop zone'?

Steve, there is nothing loose in the joints or rivets. It takes little motion to engage the skin. It does appear it is slight buckling is the culprit. I am not real sure how to properly secure the rudder to do a test. I was thinking about cutting the counterweight profile in a 2" thick board and clamping it lightly, then do the same for the opposite end. Then hang a weight at a distance and measure the deflection at the tip. The light force movement zone will be most difficult. It could be bracketed with a slight torque in either direction just to indicate where zero might be, then plot the deflection curve and plot an intercept. It might be a while before I get the actual data, so there is time to get the test procedure right.
 
I am not real sure how to properly secure the rudder to do a test.

Ok, let's spitball it.

Top view, looking down at the benchtop. Three pieces of 3" x 0.25" steel angle.

One bolts to the counterweight rib, just like the counterweight. It also bolts to the benchtop.

One gets hinge tabs and also bolts to the bench.

Third just bolts to the bench. It's a support for the dial indicator.

Known weight hangs from the trailing edge at the point near the dial indicator.

 
Ok, let's spitball it.

Top view, looking down at the benchtop. Three pieces of 3" x 0.25" steel angle.

One bolts to the counterweight rib, just like the counterweight. It also bolts to the benchtop.

One gets hinge tabs and also bolts to the bench.

Third just bolts to the bench. It's a support for the dial indicator.

Known weight hangs from the trailing edge at the point near the dial indicator.


Steve, Dan, or Bill. Perhaps I missed something. Are you saying that torsional stiffness directly correlates to flutter potential?
 
Foam

If the foam was bulging the skin outward, especially at the trailing edge, that would cause the extremely sensitive rudder.
Many years ago a hangar neighbor with a then new RV4 experienced the extreme rudder sensitivity. He squeezed the trailing edge to remove the bulge and after that the rudder behaved normally.
Several years ago Barnaby Wainfan did an article in Kitplanes about how trailing edge shape affects control pressures.
 
Ok, let's spitball it.

Top view, looking down at the benchtop. Three pieces of 3" x 0.25" steel angle.

One bolts to the counterweight rib, just like the counterweight. It also bolts to the benchtop.

One gets hinge tabs and also bolts to the bench.

Third just bolts to the bench. It's a support for the dial indicator.

Known weight hangs from the trailing edge at the point near the dial indicator.
* Image clipped to save space*

I was considering holding the entire weighted end rib stationary, but if it is loaded this way, it will be more representative of flex against the counterweight. Maybe a couple of dial indicators, the second on the tip of the stationary end. This will determine if there is any flex in that rib. If so, it can be subtracted from the opposite end to get frame only twist.

I considered if the torsion should be pure torque, but air loading is not pure torque, it is a side load, so a point load (creating some shear loads) should suffice for this stage.

Steve, Dan, or Bill. Perhaps I missed something. Are you saying that torsional stiffness directly correlates to flutter potential?
The more I read about flutter, the more I need to learn. So - while generally there is a connection, as part of the system, quantifying that is beyond me.
 
Last edited:
It appears that the current RV-7 rudder is identical both aerodynamically and structurally to the RV-9 rudder (someone please correct me if that is not right).

(BillL, I assume this is for the completed rudder (less paint), including the balance weights, correct?)

From the RVAtor "Tests on the RV-7 showed that, while the airplane met the spin recovery standards for a Normal category airplane, spin recovery was enhanced by installing the larger RV-9/ 9A rudder. On May 20, [2002] Van's made the decision to make this rudder a part of the RV-7/7A empennage."

Yes, my measurements are with the counterweight attached, new, clean, no paint, but without the end caps. The end caps will aggravate the issues.

The 7-tall (9) rudder is 936 in^2, the 8 (7 short) is 718 in^2.
 
Thanks for the confirmation of those two items. My Post #59 stands as written then.

It is correct.

The RV-7A I built was shipped with the short rudder and later received free the RV-9 tall rudder. Vans was changing to tall rudder based on spin test results. I never installed the tall rudder, am quite certain after RV-7 spin tests, Vans reinstalled the short rudder.

The 8 rudder, like 7 short rudder, has bent trailing edge. The tall rudder has wedge. I built one of them before 7 build thinking I wanted a 9 but changed my mind.

Finally got it right building the 8. :)
 
Two cents.

I have come back to this thread a few times to read the good write-ups on this subject. I was in number 2 about three weeks before she went down. So we paid close attention to the re-testing and changes that were made after that. Looking back on it, I don't think I can say that if you build it spec and do a good job of it, then fly within reason and the guidelines that are set out for your aircraft, that you will have a problem. We made one small change to the ruder and the elevators at the time we put them together with just this concern in mind. At the time we had the option of installing the stiffeners as we wishes so we put the front edge of them as close as we could to the spar, then put the first rivet as close to that stiffeners front edge as the book allowed. Then we change the rivet spacing from 1 1/2 inches to 1 inch. We also made sure we put good "RTV" between the two stiffeners at the trailing end. This seams to make the control elements a little stiffer and a little more resistant to torsional flexing without adding more than a few rivets more of weight. I did not know tell this tread that some of the bigger rudders were made with .016 rather that .020 skins. I would personally feel better with the thinker skins. I also like the dash one changes to the wing, but don't have a concern with the older type as long as you adhere to the weight limits set out for it. I have been lucky enough to have good factory test pilots and PAR 23 people as sounding boards to draw from. I think if you build good and respect your envelope you are going to be just fine. Hope this helps, Yours, R.E.A. III #80888
 
Last edited:
One more thing, it would very interesting to know which rudder the RV-7's that broke up in flight had installed, the original RV-8 rudder or the RV-9/-7 rudder.

G-GNDY and N174BK were both -9 style, riveted trailing edge. Both failed into three sections, (1) counterweight and CW ribs, (2) a central section, and (3) a lower section. The trailing edge unzipped. In both cases the VS separated from the aircraft in flight.

This from G-GNDY's report:

Factors that can contribute to the onset of flutter include high speed, a reduction in stiffness and a change in mass distribution. The kit manufacturer had conducted flight testing of the RV-7A prototype. No indications of flutter were encountered at a speed of 217 knots. Additional theoretical flutter analysis was done where the flutter speed was calculated to be 300 knots for the baseline design. The addition of weight, however, can decrease flutter speed by 50 knots or greater. Any imbalance, such as paint and filler, which increases the weight of the rudder aft of the hinge line, has an adverse effect on flutter speed.

It is unclear if the prototype was tested with an early rudder or the -9 rudder.
 
This from G-GNDY's report:

Factors that can contribute to the onset of flutter include high speed, a reduction in stiffness and a change in mass distribution. The kit manufacturer had conducted flight testing of the RV-7A prototype. No indications of flutter were encountered at a speed of 217 knots. Additional theoretical flutter analysis was done where the flutter speed was calculated to be 300 knots for the baseline design. The addition of weight, however, can decrease flutter speed by 50 knots or greater. Any imbalance, such as paint and filler, which increases the weight of the rudder aft of the hinge line, has an adverse effect on flutter speed.

This is the first time I've seen any actual numerical figures given for Vans flutter testing in all the time I've been on this forum. That data is jealously guarded, and for good reason, but it's good to see some actual numerical data rather than "You should be fine."
 
This is the first time I've seen any actual numerical figures given for Vans flutter testing in all the time I've been on this forum. That data is jealously guarded, and for good reason, but it's good to see some actual numerical data rather than "You should be fine."

During flight testing of 7A, I accidentally, stupidly, took it to at least 220 TAS before appreciating Vans limitation was set at TAS rather than IAS.

The test started at 8500' WOT and pushed nose over until 200 IAS, which did not take very long.

Don't remember when 200 IAS was achieved, but if it happened before 7000' which it probably did, TAS was close to 230.

Airplane was steady as a rock. This was with short rudder.

All other airplanes I had flown before, military and airline, were IAS/Mach limited, not TAS.

I got trapped on that one with Vans limitation at TAS.
 
Vne as TAS Technique

Point to ponder: Observing Vne as TAS in an RV is a technique, not a procedure.

The builder's Manual addresses Vne as IAS. The TAS technique accommodates high altitude, high speed operation (to a certain extent), as explained here: https://www.vansaircraft.com/pdf/hp_limts.pdf. In a certified or conforming design (e.g., military airplane) it is possible to address dynamic limits as IAS by virtue of engineering accommodation and operational limitations. All RV's are different, especially in terms of engine and propeller combinations. If, for example, you were to specify a maximum allowable operational altitude for a specific airplane with a specific engine/prop combination, it would be fairly straight forward to specify Vne as IAS. RV-types do not have a maximum specified operational altitude. They are also relatively low-drag, high-performance sport planes with a wide speed band (ratio of Vmax to Vs). Observing Vne as TAS is simply a conservative handling technique that helps preserve flutter margin at all altitudes.

Vno is a structural, not dynamic, speed limit and is expressed as IAS (top of the green arc). Any time you exceed Vno, gust protection is reduced. RV's are not certified designs, but conventional practice provides 50FPS gust protection at Vno. Above about 10000 feet, it's possible that Vne TAS will be an indicated airspeed BELOW Vno.

So it's necessary to observe structural limits (G's and airspeed) and dynamic limits (airspeed) all while using proper handling techniques to keep all of the big parts attached to the airplane.

Edit Addition

There is nothing precluding the manufacturer (builder) of a specifying a procedure for the operation of that particular airplane. In terms of airspeed, anything equal to or more conservative than designer recommended limits would be a sound basis for developing such a procedure (i.e., something that MUST be done, in this case a specified manner of observing an operational limitation). Exceeding a recommended design limit without proper engineering and test would not be a sound basis for developing a procedure. I included a look-up table in the training materials (Table 2-1, page 91, https://drive.google.com/file/d/0B8EIT6g2n8o_NmJNRmRfcGR3dm8/view?usp=sharing) as an example of how to operationally apply this to a simple, conventional airspeed indication system. Basic "winter" or "summer" conditions are accommodated in the example table. Such a table could be further simplified and even included as a placard in the cockpit, if the manufacturer/builder desires. As stated in further follow-on posts, advanced instrumentation that displays TAS greatly simplifies things for the pilot. Alternatively, a manufacturer/builder could specify a maximum altitude for operation or maximum power settings for different altitudes for a specific airplane to accomodate dynamic airspeed limits.

Service Bulliten 02-6-1 (8 June 2002) addresses the RV-7 rudder evolution: https://www.vansaircraft.com/pdf/sb02-6-1.pdf

Fly safe,

Vac

P.S., David, I recall during my first or second checkout sortie hearing an awful lot of wind noise during a lag roll while maneuvering in chase formation with my IP flying the lead airplane. The noise prompted me to look at the airspeed. I won't repeat the actual IAS as this is a family forum ;) but suffice to say, I learned a bunch about how easy it is to let the airspeed get out of control in an RV when the velocity vector drops below the horizon and the airplane is inverted! Fortunately, I was in relatively thick air, and a smooth unloaded roll allowed me to get the nose back up and decelerate quickly using low G...I've subsequently done checkouts with experienced fighter pilots that get caught off guard by this RV handling characteristic until they've seen it one time. If my students don't accidentally "hamfist" the airplane first, I demonstrate this acelleration characteristic as a rolling, nose-low, high-airspeed recovery.
 
Last edited:
I don't think it is enough for Van's to merely state that Vne is in TAS. Pilot's shouldn't have to think in terms of TAS while flying. I think that Van's should state their Vne (Redline) in terms of IAS (or CAS) as a function of altitude, up to the service ceiling of the airplane. Builders may even want to mark their airspeed indicators accordingly.

I know modern EFIS panels with an OAT sensor input can calculate and display TAS, but Van's (or any designer) should accommodate all levels of instrument sophistication.

+1. The FARs require Vne to be in IAS for normally certified aircraft, as described above. For Vans to not follow this convention is, at the least, confusing to new owners.
 
Back
Top