.

[jancellor]

Recently I have been working on a fluid-structure (or aero-elastic) computer model of flex wing hang gliders. I have made a simplified website demo of this model. Partly this was for me to practise setting up a public-facing website (my next job will likely be my first in web development), but I also thought it would be a good way of generating discussion on the feasibility and usefulness of this kind of thing.

The address of the site is http://18.218.120.168/hgmodel. Please have a play with it. When you load the page, the model will run with the given parameters. You can tweak those parameters and re-run the model. Be aware that it's quite slow and the less you change the parameters the faster it will run. You may wish to, for example, set the VG to -3 degrees to see the extra billow, or set the roll rate to 30 degrees per second to see the passive billow shift.

The model considers both fluid (aerodynamic) and structural effects. On the fluid side, it's a vortex lattice method for those of you who know what that is. Structurally, it considers the effects of the sail stretching, the leading edge bending, the battens and the floating keel. In this simplified demo you can just see the resulting shape of the wing, but it is also possible to read the lift, drag, sideforce and torques on the wing which give an indication of performance and handling. I believe the model currently captures a large number of the significant factors in the design of a flex wing. It's debatable whether it captures all of them accurately or whether it could be adapted to do so.

Incidentally, the reason why I started making this model is because I personally believe that it is likely to be possible that we can get a much better trade off between performance and handling in designs similar to current flex wings. Crudely speaking, having more sail tension reduces twist and hence induced drag, but it also reduces passive billow shift and hence handling. So there's a compromise, but there's no strict relationship that says for a given level of performance there is a particular limit on the handling. It depends on the exact shape and stiffness of the sail and frame, which is difficult to reason about analytically and to explore through trial-and-error. I believe a computer model could be the best way of finding a better design. An optimiser can explore a huge design space, even if the model is limited in its accuracy.

[blindrodie]

Cool and fun to mess with...

[red]

Jancellor,

Okay, I suppose we need to start somewhere . . . I give you real credit for the simulation, and I hope it can help with our merging into the sky. It seems to me though, it will be a very complex task to get things right in a simulation.

I realize that a lot of work went into your creation, but realistically, there is a lot to know about our "slow" flight characteristics. Any computer model will work from the inputs given, but I am skeptical that we even begin to know the accurate numbers and the interactions of the forces we use in flight. Here is a very common design, but given just one minor change. Consider (for discussion) this little "history report," if you will:

viewtopic.php?f=16&t=36286&p=405928#p405928

[jancellor]

The better address is http://jchancellor.uk/hgmodel. If anyone can edit that into my original post, please do.

[red]

Jancellor,

Speaking of the Comet, glider handling, and ancient history . . .
This history comes to me second-hand, but from pilots that I trust: There was a time when the Comet prototype was still being developed. It was obvious that the basic Comet design was a sea change in hang glider performance, but it was heavier than the traditional hang glider of its' day. It was a bear to turn, in the air. They tried many different ideas, but the basic problem remained. Turning was very stiff.

Back in the day, my glider and all other HGs were cabled with zero slack, and most top cables had tensioning devices (such as turnbuckles, or for the Comet, a flip-over-and-hook gadget) to allow easy assembly of the glider, and tensioned rigging in flight. One day, one of the Comet test pilots landed shouting that whatever had been done for the glider this time, it was the right answer, and now the Comet was turning nicely. The last changes to the glider had been rather minor, and actually inconsequential. What made all of the difference was the test pilot, who had forgotten to tension the upper rigging for this last flight. His pre-flight inspection had been somewhat deficient, as well. He accidentally invented the slack-wired hang glider, when nobody had even conceived of such a remedy. A lot of HG designs incorporated the idea of slack cable rigging, soon after, for easy handling in the air.

Huge debates started in the HG community, which I watched, concerning the safety of such an option. Cables can withstand constant loads nicely, sure, but cables may soon fail due to shock-loads (work hardening). Slack cable rigging meant that every weightless event in flight would cause a shock-load, as the glider hit normal air again. How much cable slack is permissible with our HGs? How long will the cables be safe? We still do not have absolute answers to these questions, and so we change out old cables on a regular schedule, to be safe now. The advantage of easier turning requires some gliders to have slack in the rigging, and we replace cables often (maybe too often) to be safe with that choice.

I have no idea how such information can be handled in a computer simulation of a hang glider in flight, but I would cheer for anybody who can succeed there.

[USHPA7]

Even though I wasn't flying from 1979 to 2015 I never lost contact with hang gliding but didn't pay close attention to the developing glider details. When i started flying again, in 2015, I noticed that the gliders weren't rigged with tight upper and lower cables anymore and wondered why. THANK YOU RED, FOR ANSWERING THE QUESTION I DIDN'T ASK!

Now, I know to rig the Puffin loose, I had been thinking about how I would rig it, tight or loose?

Since this Basic Trainer is going to experience landing shock loads much more than hard rising loads I'll pay close attention to the upper cables, maybe even shock spring the top of the king post. Anybody know if any manufacturer has done that and the results?

Frank Colver`

[jancellor]

Argh, just wrote a reply that disappeared Thanks for your comments Red. You've been around a lot longer than me in the hang gliding world and I'm not surprised if you have more examples of small changes having big effects. I'm also skeptical about the ability of a model to reflect reality. I'm not convinced it's impossible or impractically difficult.

Re the leading edge thickness example, you can't change the stiffness in the web demo but you can change the sail cut (eg tipCurvecut = 0.05) and see the difference on the leading edge curvature. Yes, I'd be surprised if 20% extra thickness went from flyable to unflyable, but could another factor have been significant -- sail mount hole position for example?

Re the Comet and frame flex, I just tried an example on my own version of the model where you can fix the floating keel. With 2 degrees slacker VG from the reference configuration above and a 10 degrees per second roll rate, the roll torque (roll damping) went up 30% with a fixed keel and the yaw torque (adverse yaw) 15%. The starting configuration may or may not be realistic but it shows that a model is capable of showing this effect.

[red]

Re the leading edge thickness example, you can't change the stiffness in the web demo but you can change the sail cut (eg tipCurvecut = 0.05) and see the difference on the leading edge curvature. Yes, I'd be surprised if 20% extra thickness went from flyable to unflyable, but could another factor have been significant -- sail mount hole position for example?

Jancellor,

I used the term "unflyable" in a broad sense. When the owner launched his "repaired" glider from the bunny hill, it did fly very well. Then he attempted to turn left, then right, several times. When the glider did not turn after strong pilot inputs, he quit trying, with the valid viewpoint that if he managed to get a turn started, he might not be able to correct or reverse the turn. He flew the glider straight ahead, and landed long, but normally. Except for a long walk back, we thought the flight looked okay. Turns out, the pilot had been scared spitless by his non-turning glider. He really had intended to fly the glider first from a 3500' (1km) launch. Even without any problems, that straight flight would have left him with a day-long desert hike back to a road. He actually thanked the crew for hammering some sense into him, on the issue of test-flying from the bunny hill.

As said earlier, the extra stiffness of the new (thick-walled) aft spars prevented the billow shift needed to turn the glider. There were no viable options on the sail mounting point; it had to be where the sail ended.

I will add another factor in turning here, that being the extra mass of the thick-walled tubing. Adding any mass at the wingtips can inhibit turning, by the extra inertia. Back when video cameras were the mass of two or three quarts (liters) of milk, I cautioned a new pilot about adding that much mass to one end of a crossbar. He listened, somewhat; he went home and cobbled up an equal mass of metal, to install on the opposite end of the glider crossbar. The weights were exactly equal. When he launched, he found out that his normally responsive glider now turned with the ponderous grace of a large yacht: slow to turn, and slow to level out, needing a lot of pilot effort. He abandoned his video plans then, and went for a good landing in the LZ.

The Comet repaired with the thick-walled tubing had both factors working together against the pilot: extra stiffness, and extra mass at the wingtips. The pilot reported that his Comet felt like the wingtips were set in concrete in flight; the glider seemed to be ignoring his control efforts.