[aeroexperiments]

Why do flex-wing hang gliders need so much low-siding to hold a constant bank angle at low angles-of-attack, with the bar well pulled-in?

I think that we have to appreciate the fact that the tips are making downward (negative) lift, to really appreciate why a glider needs so much low-siding to hold a constant bank angle with the bar well pulled-in, even when the same glider needs a neutral roll input or a high-siding roll input to hold a constant bank angle at a higher angle-of-attack, like near the min. sink angle-of-attack or best L/D angle-of-attack.

Consider-- when we fly with strings trailing from the extreme tips, we can see that the vortex in the immediate vicinity of the tip spins in the "backwards" direction, indicating that the tip is creating negative lift, at airspeeds as low as about 10 mph above min. sink. (Wills Wing Sport 2 with VG off.) When we stuff the bar and rock into a head-down position and pull our knees over the bar, we can make the lower side wires be slack, showing the center of lift of each wing has moved far inboard, and a very large part of the wing is flying at a negative-lift angle-of-attack. (Wills Wing Sport 2 with VG off, and most other flex-wing hang gliders as well, over a fairly wide range of VG settings.) Bearing all this in mind, I think that when we are banked and turning, at a low angle-of-attack, the fact that the outboard part of each wing is creating negative (downward) lift is hugely important to understanding roll stability dynamics at low angles-of-attack. We know that regardless of whether or not we have washout, the wingtip on the outside of the turn is flying at a faster airspeed than the wingtip on the inside of the turn. With no washout or modest washout, this difference in tip speed contributes a rolling-in torque-- the outside tip is flying faster and creating more upward lift than the wingtip on the inside of the turn. But when we have a lot of washout in the wing, and we pull the bar in to place the wing as a whole at a low angle-of-attack, we actually place the tips at a negative-lift angle-of-attack. Now the faster-flying tip on the outside of the turn is creating more DOWNWARD lift than the slower-flying wingtip on the inside of the turn. This contributes a rolling-out torque that tends to roll the glider toward wings-level.

Steve

[aeroexperiments]

Here is some of the background on why this is interesting to me--

I've been thinking about roll stability-- why we tend to often need to high-side at high angles-of-attack (e.g. thermalling) and why we need to low-side at low angles-of-attack (bar well pulled-in). With the bar extremely pulled-in, to hold a constant, steep bank angle, the need to low-side is so strong in some gliders, especially with VG off, that a pilot can constantly "hug" the extreme corner of the control frame without rolling the glider to some extreme bank angle. Try doing that with the bar at the trim position, or pushed out! At high angles-of-attack there is a great chance the glider will eventually roll into a very steeply-banked, high-G turn even if the pilot is just staying near centered on the bar-- as we were reminded in the recent video of the chute deployment in cloud.

At even lower angles-of-attack-- say with the harness unzipped and pilot pulled all the way in and rocked head-down with his knees pulled over the bar-- a glider has an even stronger tendency to roll out of a turn, toward wings-level, especially with the VG off. Pulling the bar "all the way" in in this manner--with pilot rocked head-down and knees pulled over the bar-- may be a reasonable strategy to deal with loss of control in cloud, in some gliders. (Maybe not a very slippery topless glider with minimal sail billow and minimal washout.) In some gliders that are prone to yaw-roll oscillations, I've found that simply pulling the bar all the way while remaining prone and zipped-up in the harness might not be a good strategy for dealing with loss of control in cloud-- with these gliders, the glider does show a basic tendency to roll toward wings-level but this ends up feeding a dynamic yaw-roll oscillation involving some rather extreme changes in bank angle and pitch angle-- probably not a good thing "in the white room". Gliders I've flown that exhibit this behavior include a Wills Wing Spectrum, and Airborne Blade, and a kingposted Laminar R12. Last year I took a high tow in my Spectrum specifically to see if stuffing the bar all the way with harness unzipped, my head rocked down, and my knees pulled over the bar would make the glider tend to roll toward wings-level and then tend to stay near wings-level without going into these dynamic yaw-roll oscillations that the glider is prone to at slightly higher angles-of-attack (bar stuffed but pilot still fully prone). That is exactly what happened.


So it certainly seems that in many hang gliders, especially lower-performance and intermediate-performance gliders, loosening the VG all the way and pulling in the bar all the way-- and perhaps also rocking head-down and pulling the pilot's knees over the bar-- might be a reasonable strategy to deal with loss of control in the cloud. Compare and contrast this situation to a typical light airplane-- the "graveyard spiral" is a well-known common result of untrained pilots venturing into clouds in airplanes. If an airplane is starting to get into a "graveyard spiral", pushing the yoke forward does not tend to level the wings to any significant degree-- it simply ensures that the airplane will go blasting past the redline, likely leading to flutter and failure.

Yet the max lift/ drag ratio of a hang glider is not all that different from the max lift/ drag ratio of a typical light airplane. What accounts for the great differences in the observed dynamics? And specifically, why does a hang glider tend to roll toward toward wings-level at very low angles-of-attack?

One significant factor is that the wing of the hang glider has a lot more washout than the wing of the airplane. Therefore the hang glider's polar degrades much more rapidly at high airspeeds than does the airplane's polar. Likewise, making a nose-down input to decrease angle-of-attack will tend to cause only a small airspeed increase in the hang glider, and a large airspeed increase in the airplane, relatively speaking. But that doesn't explain the large difference in observed roll dynamics-- why does the hang glider show such a strong tendency to roll toward wings-level at very low angles-of-attack?

One thing that tends to make the bank angle decrease in a steep turn is the fact that a steeply descending turn (high sink rate relative to the airmass) requires a constant rolling-in motion just to hold the bank angle constant. (For an extreme case, imagine a steeply diving turn that is just short of a vertical rolling dive with the nose aimed straight down.) If the aircraft stops rolling toward the inside of the turn, the bank angle will start to decrease. An aircraft naturally has some resistance to rolling--this is called "roll damping" and is based on the fact that the rolling motion tends to increase the angle-of-attack , and the lift force, of the descending wing. "Roll damping" acts upon the entire wingspan no matter how much washout the wing contains. Say for example that the tips are so washed out that they tend to meet the air at the zero-lift angle-of-attack, creating no lift. The descending tip (which in a constant-banked descending turn, is also the inside wingtip) still tends to experience an increase in angle-of-attack (so that it now creates some positive lift) and the rising tip (which in a constant-banked descending turn, is also the outside wingtip) still tends to experience a decrease in angle-of-attack (so that it now creates some negative lift). Clearly this will create a rolling-out torque, which will tend to roll the aircraft toward wings-level. "Roll damping" acts on the entire wingspan, not just the "effective span".

But "roll damping" would seem to operate on a typical light airplane just as strongly as it operates on a hang glider, especially if the light airplane has a relatively high aspect ratio, like a Cessna 152 or 172. This doesn't seem to help us understand the extreme difference in roll dynamics of the two types of aircraft at low angles-of-attack.

We mentioned "effective span" above. With washout, the "effective span" is less than the actual span, because the tips are meeting the air at a low angle-of-attack, and are only contributing a small amount of lift. How does this affect roll dynamics? One thing that tends to make an aircraft roll toward a steeper bank angle, is the fact that the outboard tip is describing a larger turn radius, and is moving through the air faster, and is tending to create more lift, than the inboard wingtip. If we wash out the tips, so that less of the wing's lift is generated at the tips and more of the wing's lift is generated well inboard, then this rolling-in tendency due to the difference in airspeed between the tips becomes less pronounced. When we give a wing washout, we decrease the "effective span", so that the difference in airspeed between the inboard and outboard wingtip contributes less of a rolling-in tendency. The rolling tendency due to the difference in airspeed between the tips is based on the "effective span", not the actual wingspan.

I've long been convinced that this change in "effective span" is the main reason that we see less rolling-in tendency or more rolling-out tendency when we loosen the VG. With the VG tight, the "effective span" is large, and the tips have relatively modest washout, and the difference in airspeed between the inboard and outboard tips creates a significant rolling-in torque. (Say for the sake of discussion, we are talking about angles-of-attack near min. sink or best L/D.) With the VG loose, we increase the tip washout, and the "effective span" becomes smaller, and more of the lift is generated near the root of the wings not the tips, and the difference in airspeed between the inboard and outboard wingtips creates much less rolling-in torque.

I've been aware of these relationships for years but something kind of obvious hit me just a few days ago-- if the wingtips are so washed out that they are meeting the air at a negative-lift angle-of-attack and creating negative lift, then the difference in airspeed between the inboard tip and the outboard tip actually contributes a rolling-out torque, not a rolling-in torque. The outboard tip is moving faster through the air and creating more downward (negative) lift than the inboard wingtip. Now the washout is not simply causing a decrease in rolling-in torque, it is actually causing a rolling-out torque, as far as the portions of the wing that are flying at a negative-lift angle-of-attack are concerned. I think that this is the main source of the strong rolling-out tendency that we see when we pull in the bar of a flex-wing hang glider and place the wing at a low angle-of-attack. In a sense we could say that the "effective span" has become negative! If large parts of the wings are so washed out that they are creating negative (downward) lift, then the difference in airspeed between the inboard tip and the outboard tip will contribute a rolling-out torque, not a rolling-in torque.

Steve

[aeroexperiments]

Caution, flying very fast in turbulence might break the wings due to the fact that the tips are generating a strong downforce and the root a strong upforce. Hitting a strong updraft is probably not a big deal as the lifting part of the wing is concentrated well inboard. But hitting a strong downdraft (or flying out of an updraft into still air) might increase the negative angle-of-attack of the tips, and the download on the tips, to the point that the leading edges fail or the upper rigging or crossbar fails. Everything in this thread is meant as food for thought, not practical advice. Please do fly with a Go Pro or similar if you try any of my ideas in clouds, thanks, and also a freshly packed chute. Steve

[tom emery]

That is one big chunk of "food for thought". I think I get it. I have a question though. What is "polar"? And by "washout" do you mean the tendency to pitch the nose down by creating sort of artificial "flaps" by means of the wash out struts?
Thanks.

[aeroexperiments]

The polar is the graph of airspeed versus sink rate.

Washout is the twist in the wing, the way that the angle-of-incidence decreases as we go from root to tip. In other words the way that the wing chord, or crudely speaking, the airfoil, gets twisted to point more and more nose-down as we move from root to tip. In hang gliders, washout is created largely by the way that the trailing edge billows upwards as we go from root to tip, plus the fact that the wing (chord) gets narrower as we move from root to tip.

The washout struts don't have much to do with it; my experience is that they don't touch the sail normally. Even with the bar way pulled in. You're on the right track, but it just so happens that other aspects of the sail geometry are the main cause of washout, normally. The washout struts are just their for insurance in case of a tuck or tumble.

I lost one of the washout struts for my Falcon and have been flying it with just one. I tried stuffing the bar to see what would happen but it didn't make it turn. I don't remember if the test included pulling my knees over the bar; I think it did.

Steve

S

[Suneagle]

Have you got a long version to explain it? (Joke)

Two questions, not meaning to be rude:
1/ What's your point?
2/ Have you got a summary in under 200 words?

[tom emery]

Thanks for the explanation of washout. Still not real clear on "polar". Could be the lack of caffeine.

[aeroexperiments]

[Suneagle]

Okay, I read it all again and I'm still none the wiser.
Call me dumb.
Are you offering a helpful way to fly if you are in cloud?
Are you just asking questions for the fun of theory?
Is this post about acrobatics?
Are you just asking why we highside on some settings and lowside on others?
Are you asking or telling?

What's negative lift? Sink? Are you saying the wingtips are pushing down?

What sort of a turn has the outer wingtip flying much faster than the inner? A spin? I'm pretty sure that in a high banked turn there isn't much difference in airspeed over the wing tips. Am I wrong? Show me some stats. If you are turning at 90 degree bank the tips are in the same plane about a vertical axis.

Please spell it out for me. What's your point/question/theory? I might learn something.

Jerry

[aeroexperiments]

Speaking only for me and my own glider, I would stuff the bar all the way and pull my knees over the bar, vg off, if I lost control of my glider in the clouds and wasn't able to recover in the usual way (reference to instruments etc.) For the reasons given in these posts.

Yes my interests are theoretical. Yes they are also practical.

Emphasis on the practical. The idea that a hang glider has a very strong tendency to fly wings-level with bar fully stuffed, pilot head-down and knees pulled over bar, was first told to me by a champion aerobatic pilot who knew it to be true from practical experience. I was a bit skeptical, in part because I had previously been flying with a harness setup that did not allow me to put the glider at these extreme low angles-of-attack. I git rid of the pitch limiter line and found out that he was right.

Negative lift is downward lift. I am using lift in the sense of the force generated by the wings, or by some part of the wings, not the rising or sinking of the airmass, which is mostly irrelevant to this thread.

You didn't know your wingtips generated downward lift when you stuffed the bar? They do! If you are flying a flex wing hg. See the thread about slack lower side wires at high airspeed / low angle of attack.

I am mostly telling not asking... but what I am telling is what I believe to be the answers to questions that I have long been asking.


Yes part of the question is why we high side at some angles-of-attack and low-side at others...

Every turn has the outboard tip describing a larger radius than the inboard tip. Therefore the outboard tip is flying faster. (The effect does start to become less significant at bank angles above 45 degrees.) Why else would a hang glider, or a sailplane, ever need a high siding input to hold a constant bank angle in smooth air? (I can think of another reason for the hang glider but not the sailplane.)

Before you say "the airspeed difference is trivial, and the difference in lift is trivial, and this cannot act to roll the glider to a steeper bank angle", stop and think what effect will even a very small roll torque have, if it is sustained over multiple circles and nothing opposes it? What effect would holding your body say 1/2 inch off center on the bar have, for circle after circle? We are not talking about a lump of lead sitting on the ground; when we are slipping the surlies and flying through thin air, small roll torques are not trivial.

There is no 90-degree-banked turn... at least not in any kind of steady-state condition, in a hang glider...

If you have more questions I will try to answer...

Steve

[henderthing]

Aeroexperiments is simply proposing a well thought out reason for why the roll stability of flex wings seems to change over different AOAs.

The only bank angle that the two wingtips are travelling at the same speed is 90 degrees. I think it's fair to say that this is not a typical bank for most of us.

Given that our tips have a lower AOA in all states, it's easy to understand that this fact, combined with the faster speed of an outboard tip in a bank, AND a low overall AOA would produce a tendency to "roll out" of the bank-----or "roll stability."