Van's Air Force
Don't miss anything! Register now for full access to the definitive RV support community.
What Is Flutter?
- Thread starter RV8JD
- Start date
Steve Barnes
Well Known Member
Flutter demonstration (under laboratory conditions)
Take a piece of regular paper
Cut it 1/2" wide +- and about 4"+ long
Once the strip of paper is cut hold it between your thumb and first finger
Blow hard across the top of the paper
THAT MY FRIEND IS FLUTTER
Steve
Take a piece of regular paper
Cut it 1/2" wide +- and about 4"+ long
Once the strip of paper is cut hold it between your thumb and first finger
Blow hard across the top of the paper
THAT MY FRIEND IS FLUTTER
Steve
N941WR
Legacy Member
Bill, I have seen that video of the plane before, but seeing it again is still an eye opener. Wow. I can't see how those parts didnt break off. Maybe they did.
Neal Trombley
Well Known Member
Nasa did the test and stopped before any damage
Bob Kuykendall
Well Known Member
Thanks, nicely done!
That is a very highly technical definition. I think that could be simplified considerably to make it easier to grasp for the layman. I don't think most pilots would be able to visualize the concept of energy extraction from the airstream. I am not saying this isn't correct (I'm not actually sure) but I think most people would read that and their eyes would glaze over 
Because increasing stiffness in one area can easily increase loads in another weaker area.
stiffness vs. frequency
Now it's clear - this calculator also helped.
http://www.amesweb.info/Vibration/Simply-Supported-Beam-Natural-Frequency-Calculator.aspx
Just to confirm - increasing stiffness automatically increases the natural frequency of the structure? Are strength and stiffness always linked? Could you have a very strong but not so stiff structure? I guess I should remember this from my school days, but I don't!
When people say that aircraft design is all about balancing various compromises, I'm starting to understand what they mean.
Now it's clear - this calculator also helped.
http://www.amesweb.info/Vibration/Simply-Supported-Beam-Natural-Frequency-Calculator.aspx
Just to confirm - increasing stiffness automatically increases the natural frequency of the structure? Are strength and stiffness always linked? Could you have a very strong but not so stiff structure? I guess I should remember this from my school days, but I don't!
When people say that aircraft design is all about balancing various compromises, I'm starting to understand what they mean.
Robert Anglin
Well Known Member
Not quite.
NASA may have stopped the test before failure, but to say there was no damage. How many times can you flex a structure over time before it fails?
How many times have you overloaded your aircraft? How many times have you bored on through ruff air with the throttle pushed up? How many times have you made hard maneuvers at above Vs? How much faster will flutter speed the failure of a structure up over time? Now is that a black arts question?
Hope this helps, Yours, R.E.A. III #80888
NASA may have stopped the test before failure, but to say there was no damage. How many times can you flex a structure over time before it fails?
How many times have you overloaded your aircraft? How many times have you bored on through ruff air with the throttle pushed up? How many times have you made hard maneuvers at above Vs? How much faster will flutter speed the failure of a structure up over time? Now is that a black arts question?
Hope this helps, Yours, R.E.A. III #80888
Bob brings up an excellent point.
I've lost two friends of a result of structural failure in the envelope, and had another step over the side after a manufacturing defect allowed sufficient fatigue damage to cause an in-flight break-up. And those were fighter planes designed for high speed, high G operations. The bottom line is no airplane has been engineered yet that is pilot, maintenance or builder-proof. In everyday life, fatigue damage, builder error or (most likely) pilot handling error (either high speed, high G or high AOA or some combination thereof) is far more likely to cause an RV'er problems than aerodynamic flutter, assuming the airplane is operated in the envelope--with the occasional unintentional excursion that we all have, as Bob mentions.
In addition to following sound building practice and designer recommendations, as airmen we also need to have good grasp of the various regions of the operating envelope and stick and rudder skills necessary to maximize performance without exceeding limits; and for those occasional times we transgress, thanks to the engineers for providing some margin!
In no way no way are these comments intended to detract from this excellent engineering discussion. Excellent briefing, Carl. Thanks much for putting the time and effort into this! This is an outstanding example of the resources we share in our community.
Fly Safe,
Vac
I've lost two friends of a result of structural failure in the envelope, and had another step over the side after a manufacturing defect allowed sufficient fatigue damage to cause an in-flight break-up. And those were fighter planes designed for high speed, high G operations. The bottom line is no airplane has been engineered yet that is pilot, maintenance or builder-proof. In everyday life, fatigue damage, builder error or (most likely) pilot handling error (either high speed, high G or high AOA or some combination thereof) is far more likely to cause an RV'er problems than aerodynamic flutter, assuming the airplane is operated in the envelope--with the occasional unintentional excursion that we all have, as Bob mentions.
In addition to following sound building practice and designer recommendations, as airmen we also need to have good grasp of the various regions of the operating envelope and stick and rudder skills necessary to maximize performance without exceeding limits; and for those occasional times we transgress, thanks to the engineers for providing some margin!
In no way no way are these comments intended to detract from this excellent engineering discussion. Excellent briefing, Carl. Thanks much for putting the time and effort into this! This is an outstanding example of the resources we share in our community.
Fly Safe,
Vac
Last edited:
logansc
Well Known Member
The fatigue life of any aircraft is finite. You can pull maximum "g's" only so many times. Recording g-meters on tactical jets go back to the F-4 (a '50's design) at least. Once the airplane has surpassed the calculated number of 4, 5, 6, 7 and up "g" excursions, either it will normally be subjected to a service life extension analysis and/or upgrade. Usually, you'd be talking about a lot of lower "g" excursions and fewer and fewer (but still a lot) of higher "g" departures. There won't be very many serious over-g departures in the tank.
When the Navy bought several 9-g F-16's for TopGun, I believe I am correct that they only lasted 5 or 6 years. At TopGun, it is likely that they were subjected to 9 "g" on virtually every tactical flight. They don't last long that way.
Lee...
When the Navy bought several 9-g F-16's for TopGun, I believe I am correct that they only lasted 5 or 6 years. At TopGun, it is likely that they were subjected to 9 "g" on virtually every tactical flight. They don't last long that way.
Lee...
David Paule
Well Known Member
1. "Just to confirm - increasing stiffness automatically increases the natural frequency of the structure?"
Generally but not always. Adding stiffness increases the mass of the structure and added mass reduces the natural frequency. Done carefully, adding stiffness should increase natural frequency. Done ignorantly, it might not.
2. "Are strength and stiffness always linked?"
Usually but not always. Typically, the designer has a particular goal in mind and optimizes towards that. Sometimes a designer will want to tailor the natural frequency and mode shape in a particular manner and may chose an approach where the strength isn't significantly improved. Also, the "strength" is determined by many, many things.
Here's an example: if you have a C-channel in bending, one flange is in compression. You can increase its bending stiffness by widening the flanges. The flange on the compression side will buckle at a lower stress than before, and that might result in a lower strength (might not, too, since the moment of inertia is increasing as well). But adding thickness instead of flange width will increase both the stiffness and the buckling strength, and probably improve natural frequency in spite of the additional mass.
3. "Could you have a very strong but not so stiff structure?"
Sure, no problem. Here's a very easy example. Titanium has a Young's modulus of 16 million psi and a yield strength of 120,000 psi. 4130 steel has a Young's modulus of about 29 million psi and a yield strength of 75,000 psi.
For an identical part of both materials, the titanium one is stronger and more flexible than the steel one.
Using the same material for both of these, you can imagine an I beam compared to a square bar, equally long and equally strong. The square beam will be more flexible in bending. Shape matters.
Dave
Generally but not always. Adding stiffness increases the mass of the structure and added mass reduces the natural frequency. Done carefully, adding stiffness should increase natural frequency. Done ignorantly, it might not.
2. "Are strength and stiffness always linked?"
Usually but not always. Typically, the designer has a particular goal in mind and optimizes towards that. Sometimes a designer will want to tailor the natural frequency and mode shape in a particular manner and may chose an approach where the strength isn't significantly improved. Also, the "strength" is determined by many, many things.
Here's an example: if you have a C-channel in bending, one flange is in compression. You can increase its bending stiffness by widening the flanges. The flange on the compression side will buckle at a lower stress than before, and that might result in a lower strength (might not, too, since the moment of inertia is increasing as well). But adding thickness instead of flange width will increase both the stiffness and the buckling strength, and probably improve natural frequency in spite of the additional mass.
3. "Could you have a very strong but not so stiff structure?"
Sure, no problem. Here's a very easy example. Titanium has a Young's modulus of 16 million psi and a yield strength of 120,000 psi. 4130 steel has a Young's modulus of about 29 million psi and a yield strength of 75,000 psi.
For an identical part of both materials, the titanium one is stronger and more flexible than the steel one.
Using the same material for both of these, you can imagine an I beam compared to a square bar, equally long and equally strong. The square beam will be more flexible in bending. Shape matters.
Dave
That was me....19 years ago
Carl,
That particular F-117 video was actually me back in Sep 1997. Quite the "E" ticket ride. Sudden and rapid doesn't begin to explain what that felt like in the cockpit. The high rate of vibration made me first think that it was an engine coming apart. Aviation Week did an interesting breakdown on the event and it turns out it all happened inside of a second from beginning of flutter to wing separation.
After that the airplane did a 90 degree break turn and went from somewhere near 400 knots to nearly zero...and the Gs were in excess of 10 causing the gear uplocks to break and the landing gear to extend. It also shed some other parts including the nose... Somewhere in there I decided I would walk the rest of the way and the jet and I parted ways.
The cause of the flutter turned out to be an improper left elevon repair. The aircraft had flown over a year like that and the engineers tell me that conditions just happened to be "right" for the "event" to happen on my flight...at an airshow no less.
Luckily I haven't experienced anything like that in my RV-8 after 10 years of flying it.
BK
Carl,
That particular F-117 video was actually me back in Sep 1997. Quite the "E" ticket ride. Sudden and rapid doesn't begin to explain what that felt like in the cockpit. The high rate of vibration made me first think that it was an engine coming apart. Aviation Week did an interesting breakdown on the event and it turns out it all happened inside of a second from beginning of flutter to wing separation.
After that the airplane did a 90 degree break turn and went from somewhere near 400 knots to nearly zero...and the Gs were in excess of 10 causing the gear uplocks to break and the landing gear to extend. It also shed some other parts including the nose... Somewhere in there I decided I would walk the rest of the way and the jet and I parted ways.
The cause of the flutter turned out to be an improper left elevon repair. The aircraft had flown over a year like that and the engineers tell me that conditions just happened to be "right" for the "event" to happen on my flight...at an airshow no less.
Luckily I haven't experienced anything like that in my RV-8 after 10 years of flying it.
BK
Ya I was just being honest. And I am an Aero Engineer. I know enough to know that flutter is something that is only well understood by the people who do it full time. We touched on it at school but it was very top level - enough to scare us into staying away from it unless one was going to become deeply involved - a little knowledge is very dangerous.
I have seen airplanes undergoing GVT (ground vibration testing) where all of the static structural modes are identified, and I have been in the control room when flutter testing has been conducted on biz jets and single isle commercial transports, but as a flight dynamics support guy. The flutter guys would be looking at very strange plots and making go-no-go decisions for the next test points but none of us had a clue what their decisions were based on. They all had bow ties and thick glasses and sat in the corner away from everyone else. A lot of work was done up front and they were simply confirming that everything behaved as predicted and the trends were looking good.
I would expect that without all the GVT data and flight test instrumentation, such as on a light airplane, a flutter program has a much higher potential to get "interesting".
Wow - every time I see a video like that - and I've seen yours many times - I hope and pray to see the chute and the pilot talking to the press afterwards. It's really a delight to know that all worked out ok, and that you are flying an RV8. Just amazing. I can't begin to imagine what that ride was like.
And apparently they wear masks so their adoring public won't recognize them.
The paparazzi can be sooooo tiresome
Seriously Carl, great stuff. I learned a lot....mostly how little I know.
In the flight dynamics field it seems like every manager/director thinks they know it better and are quick to second guess. But when it comes to flutter, when those guys say 'we have to stop' nobody in management ever says a word. They just do what the flutter guys say. I envy that sort of absolute power. I guess that's why they get all the chicks!
Caveman
Well Known Member
I'm going to make an admission here that hopefully won't get me roasted too bad. Back during my phase 1, 8 1/2 yrs. ago, I decided if I was going to carry my loved ones in this machine, that I wanted to flutter test it first. I used the dive to a target speed, pull up to a very slight climb and slap the stick method. I started at 1 mph below published Vne. If my memory is correct, I only did this at two different altitudes and increased speed in two mph increments until I reached 5 mph over Vne. So, about three runs at each altitude and I considered it good. I used indicated airspeed after making airspeed calibration runs using Kevin Horton's test pilot spreadsheet. Obviously, after you guys educated a bit more, this was probably insufficient to fully prove the airplane. What I'm now curious about is how different methods of keeping the ball centered in cruise such as plastic trim wedges, in cockpit rudder or aileron trim using springs, and bendable aluminum tabs affect flutter. It would seem to me that all bets are off and the only real safe way would be to change the alignment of the vertical stabilizer... or re-run the tests after each adjustment. It seems a spring system, even on the ailerons, which I have, would always be an unknown depending on how much spring pressure was being used. Is this correct? ****, even bending the aileron trailing edge changes the flutter characteristics, right? I had added the rudder wedge and squeezed the ailerons before my testing, but I'll admit I squeezed the elevators a bit just last year to see if it would lessen the sensitivity while landing. Didn't even think of that affecting flutter margin.
David Paule
Well Known Member
I hope Carl will comment.
My much more limited understanding as a simple structural engineer is that the springiness of a surface and the mass (and mass distribution) of it are what count, not the balance forces and maybe not the aerodynamic moments that are generated. But if you add a tab or a wedge, you're adding mass inertia, that that will have an influence.
The mass inertia and the spring constant determine the natural frequencies. Certainly to a first approximation, the forces generated do not. If they are large, however, they might, but usually it's not necessary to generate new large balance forces -- at least not after the flight test program has progressed far enough for a flutter evaluation.
Dave
My much more limited understanding as a simple structural engineer is that the springiness of a surface and the mass (and mass distribution) of it are what count, not the balance forces and maybe not the aerodynamic moments that are generated. But if you add a tab or a wedge, you're adding mass inertia, that that will have an influence.
The mass inertia and the spring constant determine the natural frequencies. Certainly to a first approximation, the forces generated do not. If they are large, however, they might, but usually it's not necessary to generate new large balance forces -- at least not after the flight test program has progressed far enough for a flutter evaluation.
Dave
