Best practice: fuselage + wing

[soaring]

Hello,

i've spent some time in FreeCad for dry/run fuselage/wings with standard workbenchs and some additional python scripts.
i've spent also some time on forum and read than many people did the same, with some results / hack / tools that differs but did the job.

As hobby, i do some RC Gliding and use FreeCad as a day to day basis for 3D printing. I think it is possible to extract some best practices from users experience, tools, scripts, dry/run/fail/retry/success and share them to the community.

It's the reason why i starting a dedicated thread, maybe some articles, new tools, workbench will emerge from it, i hope !

For dealing in the past with nurbs, i feel that curves workbench (addon) can help a lot, as i read it offers gui tools available only through API. Microelly did also a fantastic job on nurbs.

Thread is Open ! i will start with a first image of my personal work

welcomed and hoped contributors :

Chris_G | https://forum.freecadweb.org/memberlist ... ile&u=2561
Stéphane Combet | https://forum.freecadweb.org/memberlist ... le&u=18468
microelly2 | https://forum.freecadweb.org/memberlist ... ile&u=2364
hammax | https://forum.freecadweb.org/memberlist ... le&u=12483
GlouGlou | https://forum.freecadweb.org/memberlist ... ile&u=5772
saso | https://forum.freecadweb.org/memberlist ... ile&u=3305
...

--------------------------------------------------------------------------------------------------------------------------------------------------------
Github repository for working on some Templates : https://github.com/soaring-community/freecad-pattern
--------------------------------------------------------------------------------------------------------------------------------------------------------

[soaring]

.

[soaring]

Right here, some links on the forum regarding fuselage / wing / foil, could be used as an index (and evolve) :

Help on designing Fuselage | https://forum.freecadweb.org/viewtopic.php?t=7492
Glider wing / fuselage molds | https://forum.freecadweb.org/viewtopic.php?t=12657
From surface to solid | https://forum.freecadweb.org/viewtopic.php?f=3&t=5717
Linking several fuselage of aeroplane | https://forum.freecadweb.org/viewtopic.php?t=14979
Loft in python script | https://forum.freecadweb.org/viewtopic.php?t=3227
Boolean operations | https://forum.freecadweb.org/viewtopic.php?t=32753
Converting plan for 3D printing | https://forum.freecadweb.org/viewtopic.php?f=3&t=28084
Curves workbench | https://forum.freecadweb.org/viewtopic.php?f=8&t=22675
RC Glider | https://forum.freecadweb.org/viewtopic. ... 7&start=20
Yakovlev | https://forum.freecadweb.org/viewtopic.php?f=24&t=23228
canoe kayak | https://forum.freecadweb.org/viewtopic.php?f=12&t=33108
Gordon surface (1/2) | https://forum.freecadweb.org/viewtopic.php?t=29632
Gordon surface (2/2) | https://forum.freecadweb.org/viewtopic.php?t=32335
car modeling | https://forum.freecadweb.org/viewtopic.php?f=12&t=33189
curvesWB | https://forum.freecadweb.org/viewtopic. ... &start=310
...

[chrisb]

I changed the subject capitalization. Using all caps is considered to shout at us.

[soaring]

Thanks chrisb, i will pay more attention about rules, don't hesitate to PM me to fix issues if necessary !

[hammax]

… the thread title is "Best Practise".
This post is simply "some practise" by an example of experiments with ChrisG's CurvesWB - here Sweep by 2 Rails.
Its only a trial to model a bird's wing. Massive Leading Edge, thin Trailing Edge, no turbulence problems,
because feathers damp them away.
I like seagulls, they are the most elegant allweather flyers.
For comparison to "Sweep by 2 Rails" there is also a loft out of the wing profiles.
The profiles are open to get points for the wires of the ruled surface.
I also experimented with scaled clones of a master profile.
The Pipe profiles are another usable feature for a loft e.g.

Wing_6.PNG (95.77 KiB) Viewed 7239 times

Wing_5.PNG (39.84 KiB) Viewed 7239 times

[soaring]

Awesome ! this seagull wing contains all concepts for a wing (diedrals, arraow, foil sizing / change), perfect !
My current freecad install is currently broken, when it fixed i gonna give it a try !

Thanks for your work ! this sample is a very good starting point to play with !

[soaring]

So, Freecad is back ! My first observation is regarding the Trailing edge.

with standard Loft/Sweep, we can see some broken / "rasterisation" on the edge, with "sweep by 2 rail method", it's ok.
it was for me a limitation for wing design... good news !

[soaring]

2nd observation :

before the final design workflow step (approximation), we've got the "sweep 2 rails" step.

Question 1 : what is the color code used for drawing point ? Is there some information regarding color ?

Question 2 : It seems that naturally some foil where design with plots, but there is only 4 "sketches foil "...what is the method behind ?

Question 3 : there are more "plot foils" where the diehdral is important, is it automatic ? Im feeling answer is settings in property panel regarding knots or approximation algorithm, etc...if all is automatic (following an algo / method), it's opening the door to simplify complex approach existing today outside Freecad, in the real world (ex : moving naturally between 2 different foils, creating by the way intermediate foils on the go !

Question 4 : as i understand (but maybe i'm wrong", foil must be opened for "sweep2rails" compute be ok. Is there a simple way to design or transform the result to obtain a solid wing in the final workflow step ? Idea behind is to allow boolean operations. As is, it could be possible to construct ribs by substraction, one of the simplest approach when you have complex/twisted wing, i guess...

[Chris_G]

So, Freecad is back ! My first observation is regarding the Trailing edge.
with standard Loft/Sweep, we can see some broken / "rasterisation" on the edge, with "sweep by 2 rail method", it's ok.
it was for me a limitation for wing design... good news !

This is only a tesselation pb. It doesn't say much about the quality of the underlying geometry.
The GeomInfo tool (icon is a blue circle with "i" inside) gives a better "feeling" of the quality of a curve or surface

before the final design workflow step (approximation), we've got the "sweep 2 rails" step.

Question 4 ...

I would recommend to use the Gordon surface method if possible, it is more precise.
I think it is possible to build up watertight models with it, while the "sweep 2 rails" way cannot.

Question 1 : what is the color code used for drawing point ? Is there some information regarding color ?

There is no color code. They used to be all black with Py2/Qt4, but they get random colors with Py3/Qt5. I don't know why ...