My head hurts. It all started innocently enough, too. An article in January's Aviation Safety magazine covered the issue of airspeed and when the plane you're flying considers that it's "going too fast." Here, "going too fast" is defined by the physics of the situation, not the regs. This is the part in the article that grabbed my attention:
"However, it is very important to understand VNE is not an indicated airspeed. Instead, VNE is a true airspeed. Of course, true airspeed and indicated airspeed are going to be the same value only at sea level density altitude with standard atmospheric conditions. At much higher density altitudes, flying at an indicated airspeed, even in the green arc below VNO, can result in a true airspeed substantially exceeding the VNE redline. That's bad."
The article went on to describe how glider pilots flying high can run into problems with aerodynamic (or aeroelastic) flutter because they failed to reduce their VNE as their TAS went up with (density) altitude. Which got me thinking about what I fly and how I fly it. What are the potential consequences of my having a never-exceed speed masquerading as an IAS on my IAS dial? And if VNE really does change with TAS, at what density altitude do I need to become concerned at the true versus indicated airspeed difference? Twenty feet? Two thousand feet? Flight level 200...? Judging by the responses to the Aviation Safety article the following month, as well as some confused questions I saw on internet bulletin boards, I wasn't alone in my concern.
I'm not a glider pilot, why should I care?
Now, I'm an aerobatic pilot who likes to practice nice 'n' high - around 6,500 MSL for safety (which is why it would be ironic as hell if I've managed to reduce one safety concern while introducing another) - and I'm flying in California where it's quite common to see temperatures on the ground near 40 C during the summer months. Assuming standard lapse rate, that means it's a balmy 27 C and a density altitude of 9,000 feet where I'm merrily punching holes in the sky.
Until that Aviation Safety article I hadn't been unduly concerned about density altitude for my aerobatic practice sessions. Unless the plan for the day calls for multiple spins I don't much care about my climb performance. The aerobatic practice area is far enough from the field (KLVK) that I'm at 6,500 MSL by the time I arrive, no matter how hot it is.
Likewise, climb performance is a long way down on my list of priorities after I botch an aerobatic maneuver. If I end up heading downhill in a hurry, it's power to idle and get a few Gs load on the wings, then glance apprehensively at the ASI as the needle rapidly winds up. So long as a decent load is achieved quickly (you want three or four Gs immediately, not a couple of Gs for the time being and let's see how she goes...) it's not that difficult to stay in the yellow arc. Then, once I'm rubbery side down and the airspeed needle is going counter-clockwise I'll zoom climb with whatever momentum I have left and attempt to get my stomach contents to settle, red line averted, wings still attached, control surfaces still controlling, engine still running, oil pressure and temperature green, puking averted (for now)... It's all good, right? Right?
Whoa. Back up just a second there, sonny boy. "Red line averted." Am I sure? According to that article the red line might not be where I think it is. Gulp. Could I be exceeding red line and not even know it? And if I am, what are the consequences of this so-called flutter? Am I going to get my control surfaces, or worse, departing the aircraft? Time for some research.
Indicated versus true airspeed, redux
Let's begin with a quick review. We understand that the force on the leading edge of the wings and in the Pitot tube is the same when the ASI reads 100 KIAS, whether we are at sea level on a standard day or we're at 3,000 MSL and it's 35 C on the ground. We recognize that while the indicated airspeed is the same, the plane is actually moving through the thinner air much faster (its true airspeed) at 3,000 MSL on a hot day than it is at sea level on a cooler day. In other words, the velocity of the aircraft is compensating for the reduced air density to provide the same indicated airspeed and the same amount of lift. (This is also why the stall speed stays at the same IAS value whether you're at 3,000 or sea level.) Thus, everybody remembers from primary training that the airspeed indicator is a pressure gauge, plain and simple. All good so far.
Regime change: not always a good thing, wherever you happen to live
Enter VNE. During primary training in light aircraft we are taught that the colored arcs on the ASI apply at all altitudes achievable by our FAA-certified aircraft. We learn that the ASI reports stall speed at all altitudes, for example.
We also interpret V-G (or V-N) diagrams that show where stall speeds reside, why VA changes with load, where gust factors are used to determine the structural limits and so forth. One thing that gets short shrift, however, is the particular origin of the point(s) on the V-G diagram labeled VNE. We assume that its origin is the same as the rest of the graph. And it ain't necessarily so. Take a refresher look at the V-G diagram below. Do you see VNE marked with a single number, like power off stall speed? Nor me. Instead, there's a big red area with "Danger, danger!" marked all over it.
Back in the cockpit, on the airspeed indicator, VNE is reported as if it is an indicated airspeed when it is actually based on true airspeed. This is where the subtle shift of definition seems to insert itself, leading to confusion in amateur pilots such as myself. When we're looking at the V-G diagram we are, not unreasonably, considering the structural limits of the aircraft, how much lift (and load) the wings can support, etc. It is perfectly reasonable for us non-aeroengineering types to continue to interpret VNE as if structural load is the only consideration, because that's the focus when we're interpreting the rest of the V-G diagram. However, it turns out that wing loading is only one consideration when the clever aeroeng people determine where VNE needs to be. There are other ways your plane can tear itself apart if mishandled, and they don't have anything to do with the amount of lift being produced. Thus, they have nothing to do with that V-G diagram per se! Once I'd figured this snippet out I could see that I had some new learning to do. I clearly didn't understand V-G diagrams, loading, structural limits and so forth as well as I thought I did.
Fortunately for me (and you) there are lots of smart people with oodles of experience who post on the web. So what I've done below is collate some of the best information I could find. If you have the time, go read the various threads and blogs in their entirety.
Here is Tailspin's Tales take on flutter:
"Flutter is a so-called aeroelastic effect that results from the springiness (elasticity of the airfoil and the surrounding air), aerodynamic forces, and inertia.
A wing has mass and when it's tweaked - when a force is applied, say by turbulence or a control input - it overcomes the inertia of the mass, and the wing responds by flexing (even a metal wing). Usually, that acceleration is absorbed or damped by the wing, the structure it's attached to, and the surrounding air and it returns to its original position. But if that tweak is suddenly imposed or is energetic enough as the result of high speed, you can make the structure resonate like a tuning fork.
Why does VNE go down with increasing altitude? Because there's less air to act as a shock absorber. Because there's less resistance to the wing (or other structure) flexing, you have to reduce the forces on that structure. The only way to avoid flutter (without redesigning the wing) is to go slower. "
The post goes on to discuss how VNE is set during the design certification and testing process, so that the issues of flutter and structural failure can be safely put into a single number that can be placed on the ASI (in IAS units). The details, while interesting, aren't needed to understand when and how flutter can ruin your airplane (and your day), so let's stay on the phenomenon itself and come back to avoidance later.
We are starting to see why thin air moving quickly over a wing isn't the same as thicker air moving more slowly - even when the dynamic pressure being experienced on the wing's leading edge (and in the Pitot tube) is constant. The air that's above and below the wing, perpendicular to the direction of flight, is partly responsible for damping bad vibrations out of the wing. Take away that damping effect, e.g. by flying the wing very fast in very thin air, and the effects of the vibration become paramount to the structural integrity of that lifting structure.
Now that we've had an introduction to the aerodynamics, we can go see what the very talented people in the Van's Aircraft home-built community have to say on this subject. The entire thread is really interesting, here are just a couple of highlights. Bill Wightman states the issues thus:
"The airplane's structure must react all primary forces acting on it. The structure is designed to handle only so much force in any direction, with a safety margin applied over that. Loads applied to the structure create stress levels that must be controlled to remain under safe levels for the materials used."
In the case of flutter, it's the forces perpendicular to the wing's cord that are of primary interest.
"While most normally aspirated aircraft can safely enough use IAS for VNE flying generally below 10,000 feet, flutter becomes a more important concern at higher altitudes due to reduced damping and higher TAS. Forced induction engines allow much higher altitudes to be used routinely and most WW2 aircraft for example had limiting dive speeds published in IAS for the pilots to refer to at different altitudes but actually based on TAS."
while scsmith adds some detail to the reason why the altitude dependence creeps in:
"The classical flutter equations contain forcing terms that are proportional to dynamic pressure ( q = 1/2 rho V^2) and damping terms that are proportional to just density and velocity ( rho V). This produces the altitude dependence, BUT it is not strictly related to TAS either"
I found similar information on the jetcareers.com forum:
"BTW, air density plays a role in flutter suppression, which is why the tendency rises with altitude. While the forward velocity compensates for air density in the horizontal dimension, for things that flap up and down the air is still thin."
Okay then, I'm starting to see how how flying fast in thin air is markedly different to flying slower in thick air, even whey the lying dynamic pressure gauge (the ASI) is telling me the same number; there is a difference in velocity dependence between the dynamic pressure and the damping of any flutter in the lifting surfaces.
Now that we have a simple understanding of altitude and velocity dependence as it pertains to flutter, let's get into avoidance. On the Blue Mountain Avionics message board for home-builders I found this, courtesy of ballistic bob:
"If you always fly at low altitude and low airspeeds, then it doesn't much matter which one you use [TAS or IAS for VNE - Flying Donald]. For airplanes with turbines or turbochargers, flight at high altitudes can have surprisingly different IAS and TAS. When flying sailplanes in wave lift, high altitude flight (25K to 40K), it is possible for the stall speed to approach the VNE. Careful speed control is mandatory in those situations."
And Tailspin's Tails offers this:
"Aircraft manufacturers determine VNE by using numerical methods to compute VD, design diving speed. (FAR 23, by the way, requires that VD be not less than 1.4 design cruise speed, so the days of aircraft that climb, cruise and glide at 80 are over.) Flight testing is used to establish VDF, flight test dive speed, which must be lower than or equal to VD. Then VNE is set at 90% of the VD or VDF, whichever is lower, to add a margin for error--theirs and yours."
Having read all the threads above I now understand that the FAA certification process places the red line below the level of risks from structural failure (G limits, as presented on the traditional V-G diagram) as well as from flutter, provided I fly the plane in the manner in which it was intended. I cannot operate the plane - even an FAA-certified one like a Great Lakes or a Super Decathlon - outside of the limits developed during the certification/testing process without becoming a test pilot myself. Thus, if the ceiling for my IO-360-powered aircraft is 17,000 MSL and I drive on up there and then initiate a rapid dive, I'm asking for trouble. This isn't a regime that the certification process considered, because it's bone-headed.
Why is VNE so misunderstood? I think it's because of our needs for simplicity and to have a hard number for when catastrophe is guaranteed to happen. We're human, we tend to take calculated risks up until we know for a fact that disaster will strike. Every one of us tends to drive on the freeway faster than 70 mph; we're bargaining that we can overcome the negative aspects of breaking a law to get to our destination a little earlier. Yet, knowing that momentum is half the mass times the square of velocity, if we crash we're in much deeper poo than we would have been below 70 mph. It's like we're driving in the yellow arc, if you will.
VNE is a peculiar case because it's an approximation, not a definite speed like the power off stall speed. It's as if the manufacturer was told to put a hard number on something that isn't a hard number at all; it depends, like VA, on the circumstances of the flight.
So, where is my red line now?
If I build my own plane and I have options when it comes to putting an engine in it, then I would have to consider very carefully the possibility that the performance of the engine (in terms of its ceiling and ability to produce gobs of thrust way up high in thin air) might easily exceed the ability of the airframe and/or control surfaces to operate at the true airspeeds attainable in cruise flight, or a shallow descent. Alternatively, if I fly a glider then I have to watch out if I get into high density altitudes. And finally, for normally aspirated, FAA-certified planes of the type I generally fly below, say, 10,000 feet density altitude, I am not likely to have an encounter with flutter unless I initiate a prolonged dive with airspeed towards the high end of the yellow arc.
I am now reasonably satisfied that the VNE indicated in my (FAA-certified) aircraft has a considerable safety margin built in, and I am not unduly concerned about a "moving red line" when I'm botching aerobatics at a safe (sic) altitude on a hot summer's day. Staying in the yellow arc is still the key for me. I didn't have any plans to do aerobatics at 10,000+ feet on a hot summer's day anyway, and now I will actively avoid doing so; the risks of botching a maneuver in such thin air may place my plane at risk, one that I don't fully understand. So, provided I have sufficient altitude to safely recover from an inadvertent spin, etc., I'll be happy.