I just wanted to check that I am correct in assuming that rotate_extrude has
some specific logic that deals with polygons that touch the x-axis, e.g.
d=1;
p=[[d,-10],[20,-10],[20,10],[d,10]];
rotate_extrude() polygon(p);
The centre hole closes when d=0. Is this OK, no danger of leaving a very
small hole and/or "junk" on the centre line?
Thanks.
--
View this message in context: http://forum.openscad.org/rotate-extrude-with-polygon-touching-x-axis-tp13636.html
Sent from the OpenSCAD mailing list archive at Nabble.com.
I had lots of problems with this while working on this project.
F6 exposed the problem and this approach fixed it.
--
View this message in context: http://forum.openscad.org/rotate-extrude-with-polygon-touching-x-axis-tp13636p13638.html
Sent from the OpenSCAD mailing list archive at Nabble.com.
I had wondered, as you suggest, if leaving a small hole and then filling it
in with a cylinder was a safer option, particularly given that numbers are
held as floating point.
However the OpenSCAD manual does include an example that creates a solid
with no central hole:
http://en.wikibooks.org/wiki/OpenSCAD_User_Manual/Using_the_2D_Subsystem#Extruding_a_Polygon
...?
--
View this message in context: http://forum.openscad.org/rotate-extrude-with-polygon-touching-x-axis-tp13636p13640.html
Sent from the OpenSCAD mailing list archive at Nabble.com.
I agree it should work - and maybe it works with Zero.
But I created my curves in inkscape and exported from there and so mine all
have long floating point values. Top and Bottom are the same but FP.
So I ran into this problem. Its quite likely (given other threads) that its
simply an FP accuracy problem and CGAL should work it perfectly.
So if you can do it with Zero as your major axis endpoint - I'm guessing
it'll probably work.
--
View this message in context: http://forum.openscad.org/rotate-extrude-with-polygon-touching-x-axis-tp13636p13642.html
Sent from the OpenSCAD mailing list archive at Nabble.com.