Walden
Member
Coincidently, I am utilizing this process on the current design I am working on: a rubber band launched 1 sheet, 2 piece origami glider GEV.
Last edited:
Folded wing design assistant
- Thread starter clolsonus
- Start date
Coincidently, I am utilizing this process on the current design I am working on: a rubber band launched 1 sheet, 2 piece origami glider GEV.
clolsonus
Well-known member
One more quick pre-breakfast hack/addition. None of this is super useful unless it can be output to scale. Python has an svg library, so I can actually generate an svg 'cutfile' directly with blue/red lines. This could be imported directly into inkscape, lightburn, coral draw, or whatever you use and theoretically cut out. Minor note: I generate "1px" wide lines which don't get imported as the correct line thickness for laser cutting, so you have to select all, and change the line thickness to the proper value. Here's a quick shot of how the svg file looks imported into inkscape (with correct dimension). This was generated with the command:
bash$ ./ft_wing2.py 230 160 250 23
I know that's a bit cryptic, but the numbers are: <root chord mm> <tip chord mm> <half span mm> <leading edge offset sweep mm>
So now I need to figure out how to get this to my 8.5 x 11" printer (tiled?) or something ... using open-source/linux tools. Any ideas for converting to pdf and tiling using open-source tools? My wing is lacking it's wing tips still!
bash$ ./ft_wing2.py 230 160 250 23
I know that's a bit cryptic, but the numbers are: <root chord mm> <tip chord mm> <half span mm> <leading edge offset sweep mm>
So now I need to figure out how to get this to my 8.5 x 11" printer (tiled?) or something ... using open-source/linux tools. Any ideas for converting to pdf and tiling using open-source tools? My wing is lacking it's wing tips still!
clolsonus
Well-known member
Quick update. I found "pdfposter" which seems to work. So the process I have (under linux) is:
1. Run my python script to generate .svg
2. Launch inkscape and load the .svg
3. Print to pdf
4. Run pdfposter -m letter -s 0.9375 input.pdf output.pdf
5 ...
6. Profit!
Since I'm screwing around with all this anyway, I added support for specifying a dihedral in degrees (positive or negative) and the wing plan is adjusted for the desired dihedral. And you can specify a root dihedral and tip dihedral in case you are doing a polyhedral wing.
And for sure you can do all this with interactive tools if you know how to wrangle them, but I'm a python geek so that's how I think.
So I think I have everything I need now to design tapered polyhedral wing tip extensions for my dumb little test project, and output tiled pdf plans for the tricky parts.
Edit: why do projects always end up way more involved than I initially expected????
1. Run my python script to generate .svg
2. Launch inkscape and load the .svg
3. Print to pdf
4. Run pdfposter -m letter -s 0.9375 input.pdf output.pdf
5 ...
6. Profit!
Since I'm screwing around with all this anyway, I added support for specifying a dihedral in degrees (positive or negative) and the wing plan is adjusted for the desired dihedral. And you can specify a root dihedral and tip dihedral in case you are doing a polyhedral wing.
And for sure you can do all this with interactive tools if you know how to wrangle them, but I'm a python geek so that's how I think.
So I think I have everything I need now to design tapered polyhedral wing tip extensions for my dumb little test project, and output tiled pdf plans for the tricky parts.
Edit: why do projects always end up way more involved than I initially expected????
clolsonus
Well-known member
I'm pretty careful to generate the svg in the correct units, and inkscape seems to honor that and print to pdf correctly. I've done similar things in the past for wing rib shapes and have been able to directly laser cut from the generates svg. I had a little weirdness with pdfposter, but I think I figured out the correct scale factor to make the tiled pages come out right.
This evening I went through the whole process end to end and cut out a tapered wing section and was super pleased with how it all came out. Now I need to redo because I added some code to scallop out the top side for dihedral. I know I could carve and make it work, but it's so much neater if you can get a nice clean exact cut on the first try.
Last edited:
clolsonus
Well-known member
Update: Hi Jason, this was a much better question than I originally realized. I had copied svg writing code from a different part of the project where I was generating svg cut files for ribs with notches and such for stringers and all ... and that had been fine. In my new folded wing work, I sort of had "two wrongs make a right". But loading this into inkscape and printing to pdf straight from there was out of scale. I think I tracked down the issue and fixed it. SVG isn't super definite about units, but you can kind of shoe horn it into what you want by creating a "viewbox" relative to the size of your "Drawing". It's a little weird how that works, but I got inkscape to now tell me my document scaling is 1.0 and it prints directly to the correct size, so hopefully all good now.
Here's a tiled printout for a tapered wing tip. Inner chord is 230mm, tip chord is 150mm, segment span is 250mm (with a little nibbled out for a 10 degree dihedral, since I want to angle these up pretty agressively ... this will be the last 250mm (on each side) of an otherwise straight 2m wing.
Oh, and I also am now generating outlines for the tapered spar as well (spar ends also are fit to honor any dihedral request.)
Now I have to quit stalling and build the actual wing ... I have all the parts cut out and it looks like they fold correctly, so I just need to commit to gluing everything together.
Last edited:
clolsonus
Well-known member
Pictures of folding together the main wing body, and then attaching the first tapered wing tip. I think it's going to look really sharp when it's finished, but a tiny bit worried about folding it up on the first flight ... definitely not designed for heavy duty aerobatics:
clolsonus
Well-known member
Here is the mostly completed wing (still need to add servos/ailerons) next to the original 1900mm wing. The new wing is about 2020mm and I shaped the wing tips differently:
Also, lol, mental note, comb my hair at least once a day (and ask my kids to teach me how to take selfies) ...
Original wing is molded foam ~230mm chord, but extra thick 37mm with significant undercamber. New wing follows the proportions of clarky so is only 27mm thick at the same 230mm chord. The tapered wing tips (and tapered spars) were designed with the python script mentioned earlier in this thread.
Also, lol, mental note, comb my hair at least once a day (and ask my kids to teach me how to take selfies) ...
Original wing is molded foam ~230mm chord, but extra thick 37mm with significant undercamber. New wing follows the proportions of clarky so is only 27mm thick at the same 230mm chord. The tapered wing tips (and tapered spars) were designed with the python script mentioned earlier in this thread.
clolsonus
Well-known member
Quick update: I cut in the control surfaces and installed the servos, then did a bit of extra work to beef up the trailing edge near the center so the rubber bands don't cut into the foam too much. Here's a picture of the new wing attached (next to the original wing). The main wing has zero dihedral (as per the original) but the wing tip sections tip up substantially (maybe 10 degrees) and taper (thanks to my python script.) It's a little hard to see that on the new wing due to camera perspective, but it will have a nice look in flight. The fuselage of the old skywalker is completely beat up too along with the original wing (> 100 flights and 46 hours in the air) so I was planning to retire the whole thing, but it can live a little longer (maybe!) as a test bed for building larger foamboard wings. There is a lot I like about the original skywalker, the main thing I don't like is that the fuselage/tail is pretty floppy and it wiggles a lot in flight. That affects the autopilot because any little disturbance gets the whole structure wiggling and flopping around.
clolsonus
Well-known member
I finally caught a break in the weather / and my schedule today so I test flew my folded foamboard replacement wing. I was super concerned it wouldn't hold up well structurally. The original wing had a big carbon tube for strength. I did add two 3/8" x 1 meter sticks glued in with the center spar on this foamboard wing, so it's not pure foam & tape. It flew majestically and gloriously! It held up to some pretty turbulent winds. It is much thinner than the original wing and the orignal had a huge amount of undercamber in the airfoil. The pitch characteristics seem to have changed a fair bit. I think I need to move the CG forward a chunk. It flew great under autopilot. The autoland was going well until I noticed I had never updated the 'home' location. I had powered the beast up in our shelter and that's where it wanted to touch down, so I had to quick take over manual control, go around and land manually.
It's a pretty simple wing, but the tapered wing tips are way over engineered and required about 2 weeks of python hacking. They fit perfect with almost zero trimming, so I can't complain.
It's a pretty simple wing, but the tapered wing tips are way over engineered and required about 2 weeks of python hacking. They fit perfect with almost zero trimming, so I can't complain.
Hey Curt,
I'm glad I came across all of this through your friendly response on the Flight Test Fans FB page and it makes one imagine all the limitless possibilities including something for fuselage builds with perhaps a set of nose cones/overall shapes to select from.
However, the application is completely inaccessible! I regard myself as being more knowledgeable than most regarding computing but did not succeed in installing the required components to use your clever application. I tried again last night with the help of my go-to mate when it comes to things a little too difficult for me in this same regard and together we failed (pip install polygon3 - came up with loads of errors concerning compatibility with Python3).
It's a pity to have such a useful looking tool which would normally make an already rather simple building process even more sure and fun, become a deep dark rabbit hole instead.
What are the chances of developing this into a more user-friendly package even if just for the installation process. I'd certainly look forward to something like this, even a commercial product - and as I mentioned the further possibilities are endless with all sorts of components for a whole aircraft.
So all in all your program looks great and I have a huge respect for the amount of work you've evidently done but I honestly think most of us others can only read your own comments about it without being able to truly experience it.
clolsonus
Well-known member
All good points. One thing to keep in mind is that all the stuff related to FT style wings is in a subfolder so you shouldn't need to go through the complicated install process for the madesigner gui tool.
Jump straight to this link: https://github.com/clolsonus/madesigner/tree/master/sandbox/
The subdir is poorly named, but this was a 1 hour evening hack that has grown into something more involved. I do need to circle back around and clean it up. If I was a web genius (which I'm not) this might be a good candidate for a web-based online design tool, but sadly when you are cherry picking an hour or two of time once in a while, you hit the limits of all you can learn and do.
So you shouldn't need polygon3 packages to run the ft wing scripts, sorry if that wasn't made clear at the top.
Best regards,
Curt.
Similar threads
