I have noticed that many pilots relate the cause of flutter to aerodynamic loads and believe they are safe to fly at altitude at speeds up to the Vd or Vne, which is generally calibrated for flight under 3,000 metres.Whilst I know that flutter is related to inertia, which is a product of true airspeed, I have not been able to find a simple explanation of how flutter occurs. Are you able to offer a simple explanation?

ANSWER from Steve Smith on May 31, 2000:

This is a great question, and happens to be one of my pet peeves.

There is, in fact, an altitude effect on flutter speed that is different from the altitude effects on the rest of the aerodynamics, but it is not as severe as many people say and think.Many people apply a "constant true airspeed" rule that is much too severe, and can in itself create other safety problems (like unwillingness to fly fast enough to escape sink, to clear a high ridge, etc).

As you may know, most aerodynamic forces are scaled with the dynamic pressure, q = 1/2 (rho) V^2.So as you go up in altitude and the density, rho, decreases, you compensate by adding more V^2 to get the same lift and drag.It also turns out that the pressure that a pitot tube measures is the dynamic pressure, (Pt - Ps = q) , so indicated airspeed (ias) tracks dynamic pressure.Conclusion:Constant indicated airspeed means constant aerodynamic properties.The same lift/per unit area for a given angle of attack.The only minor deviation from this characteristic is the subtle effect of Reynolds number.

But flutter is different, because of the inertial coupling and the damping effect of the air.So flutter speed does not remain a constant indicated airspeed as you increase altitude.The flutter speed (expressed as an ias) decreases slowly as you increase altitude.But nowhere near as fast as the indicated airspeed would decrease if you kept the true airspeed constant and increased altitude.A really good rule of thumb turns out to be half way between the two. Make a plot of indicated airspeed on the horizontal axis and altitude on the vertical axis.A vertical line drawn at the Vne speed (red line) is a constant IAS line.Now, at a few altitudes, compute the indicated airspeed that corresponds to the sea level Vne in true airspeed.Example, Assume the sea level Vne is given as 100. Then, at 15,000 ft, a true airspeed of 100 is an indicated airspeed of 79. Plot this curve on your graph.This is a constant true airspeed line. Now, a good guide for an altitude-corrected Vne line is to draw another line half way in between.For my example, Vne would be at an indicated airspeed of 90.

There are two sources that you can refer to as documentation for this, although only indirectly.First, there was a flaming dialog in the letters to Soaring Magazine several years back, and I ducked out of it because it just got crazy.As part of it, a fellow at McDonnell Douglas Helicopter in Tempe, AZ wrote a letter that had some helicopter flight test data showing the effect of altitude on blade flutter.He did a nice job until the very end, where he confused indicated and true airspeed and thus stated a confusing conclusion.He said, "so, this data shows that the flutter speed increases with altitude, so you don't have to worry about it" and what he said was true of the true airspeed for flutter.It does increase with altitude, but not as fast as the airspeed at a constant indicated airspeed, which also increases with altitude.So, I took his data from his letter to the editor and converted it to indicated airspeed, and sure enough, it matched my "half-way" rule of thumb very nicely.The second source is the new book out of Germany called Fundamentals of Sailplane Design, by Fred somebody, sorry I can't remember his last name.Anyway, itís a very nice book on basic aerodynamics for pilots and amateur aerodynamics hobbyists. I highly recommend it, especially since he cites some of my work!But to the point, he has a plot of sailplane flutter boundary measured at one of the Akafliegs, plotted on a flight envelop, and it shows the same "half-way" rule that I have been suggesting.So I was happy to see that he found that independently.

I would treat it as a rule of thumb, and remember there is nominally a 15% margin for flutter on top of that, which can make up for many sins like instrument error, wear in control surfaces ninges, incomplete mass balance, etc.Thatís another thing.There are many different types of flutter that are possible.The simple models of bending/torsion flutter show the basic trend with altitude because they get the effects on aero loads from damping and inertia in there.But it is always possible for something complex to happen, where a wing torsion mode might cross an aileron flapping mode, and you might get something different at one altitude than another.But also remember this, but don't quote me on it - when we race sailplanes, we push them hard.We are willing to risk a little, especially when there is a long way to the ground.So if a particular glider was going to flutter, you would have heard about it.Have I exceeded Vne in my LS-6?Well, inadvertently, a tiny bit.Do I routinely fly close to Vne...YES.Do I worry about it?No - if it were going to have trouble, it would have happened a long time ago to someone else in an LS-6.Are there cases where sailplanes have had flutter problems? yes.One I'm aware of in particular is the Grob G-102...several of them, after repairs to the stabilizer/elevator, have exhibited elevator flutter well below the Vne. This has been reported, and the factory didn't react, as far as I know. But I know of at least two incidences.It may be as simple as a mass-balancing problem... more gel coat on the elevator without increasing the mass balance, and buzzzz, it flutters sooner.

So, there you are... I hope this is worthwhile and informative.It's a complicated subject and a lot of misperception running around, and the real experts will tell you that it's all so problem-dependent that they are reluctant to make generalizations.So I think I am the middle ground - some generalization to a usable rule of thumb, but with an eye of caution.