anelok: CAD (was Re: anelok: first prototype design)

Werner Almesberger werner at
Tue Oct 1 17:34:53 EDT 2013

Cristian Paul Peñaranda Rojas wrote:
> Looks nice, where is the CAD file for the case? :-)

Now we get to the tricky bits :)

Regarding tools, I currently see two choices: FreeCAD + ? and
fped + cameo.

FreeCAD would be a "real" 3D CAD system. I'm not sure whether the
graphical user interface could handle the task, but it's scriptable
in Python, so I assume one could express everything necessary that

Then I'd have the problem of turning the CAD design into toolpaths.
I don't currently have a means to generate toolpaths from FreeCAD
or STL, though such tools probably exist in a usable form (maybe in
the FDM world).

I once generated toolpaths from a "real" 3D CAD. That was for the
Ben counterweight mold and I used (scripted) HeeksCAD. That worked
well, although the toolpath generation process was unpleasantly
GUI-heavy. Alas, HeeksCAD seems to be dead now.

The various libraries HeeksCAD used for toolpath generation could
be turned into a new tool, but that may be a bit of work.

The other approach would be fped + cameo. fped can be used as a
parametric 2D CAD. By breaking down the design into layers one
can produce 2.5D models, which can then be turned into toolpaths
with cameo.

When used via the GUI, fped doesn't have good macro capabilities
for things like rounded corners or such, but when scripting the
design (in fped's own language), one has access to all the power of
cpp and can also write comments and such.

fped can also do the usual 2D measurements. Since fped doesn't know
of the Z axis, it can't do side views, though. I also don't have a
processing path from fped to STL, although that should be relatively
easy to make by feeding the toolpath polygons into a 3D CAD,
extruding the slices along the Z axis, and then applying boolean
operations on the resulting collection of 3D slices.

I've used the fped + cameo process before for a similar case, so
I know that it works in real life:

- Werner

More information about the discussion mailing list