SOLVESPACE -- parametric 2d/3d CAD

(you are viewing a thread; or go back to list of threads)

Lathe differences aren't working (by R.Lucchesi)
I create a circle let's say 4.5" in diameter, Extrude it to a height of 2.25".

Then I create a new drawing in plane orthogonal to the extrusion so that I am seeing the extrusion on edge.

The drawing I create is another circle this time 4" in diameter so that 1/2 of the circle is within the cylinder and the other half is outside the cylinder.

I create a control line which is a horizontal diagonal of this circle.

I select this control line and Lathe using it. It creates a dome ontop of the cylinder. But then I check Difference instead of Union.

I thought it should carve half a sphere into the cylinder.

What I get is unpredictable. Sometimes I see the half sphere, sometimes I see a sort of trench into the cylinder that's about 1/4" wide in the form of a half sphere carved into the cylinder. But most often I just get what looks like a circle drawn on the face of the Cylinder.

We've tried this with the Windows and Mac Version (v2.2 for Mac) and the results are unpredictable.

When I get what looks like 1/2 sphere carved into the cylinder I get exceptions and when I use NURBS it generates what looks like a partially cut out cylinder with the 1/2 sphere cut into it.

This is not what I expected.
Sat Nov 12 2016, 17:43:56
(no subject) (by whitequark)
In general, NURBS booleans tend to be fragile when surfaces coincide. This is a known pitfall with the current NURBS boolean implementation.

If you provide a file that demonstrates a problem, I could suggest a workaround.
Sat Nov 12 2016, 23:04:36
sample file (by R.Lucchesi)
So i did this again. When I didn't have NURBS on it didn't show any indentation into the cylinder but when I turned NURBS on it did show a trench (maybe spherical) cut into the Cylinder.

When I exported the NURBS on file to STL it still showed the (spherical trench) but the cylinder looked intact this time.

What I was trying to do was to carve out 1/2 a small sphere out of the cylinder.


I attached the .slvs file to this but will also send the .stl file in the next
Sun Nov 13 2016, 14:10:31, download attachment Sample Lathe-Diff.slvs
STL sample file (by R.Lucchesi)
Here is the .stl file derived from the NURBS on file I just sent
Sun Nov 13 2016, 14:11:17, download attachment Sample Lathe-Diff.stl
(no subject) (by whitequark)
Two tips:
1) You can use the "Show naked edges" command to see whether the mesh is properly watertight. (This should probably be done before export to avoid nasty surprises, I've filed
2) The Revolve command does one *full* revolution. If you revolve a circle when the axis of the revolution lies through the circle, you will in effect subtract a sphere *twice*. This results in coincident surfaces and then broken meshes. Instead, when you want to subtract a sphere, draw an arc that represents a half-circle, as in the attached sketch.
3) Even in the attached sketch, I had to force surfaces to triangle mesh. The ends of the arc coincide with the axis of the cylinder, and as I've said before, this irritates the NURBS backend, causing it to return bogus results.
Sun Nov 13 2016, 17:40:36, download attachment SampleLathe-Diff.slvs
Post a reply to this comment:
Your Name:
Your Email:
(no HTML tags; use plain text, and hit Enter for a line break)
Attached file (if you want, 5 MB max):
© 2008-2018 SolveSpace contributors. Most recent update Nov 22 2018.