 |
|
|
|
|
|
| |
| |
|
|
|
|
| |
| |
|
|
I'm working on a csg model that has a lot of
bevelled edges. I'm posting an example
of the problem (not the model I'm working on)
to p.b.i with the same subj.
Basically I can bevel convex and concave
edges by a combination of spheres and
cylinders, but it all breaks down when
a convex edge meets a concave one (coloured
red in the example picture). The bevelling
should flow round the corners and fade out.
Anyone's math up to this? Mine isn't. I
know the solution should have only a single
variable, the bevel radius, but I'm stuck
beyond that point.
The pic I'm working on is much more complex
than the pic posted, so isosurfaces are not
an option in general, though an isosurface
plug to fill the corner might work.
--
Bill Hails
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Hi Bill,
I had to think a little, but the solution is to use torus. I've posted a
picture to illustrate it. It's quite simple in terms of math. You need to
bevel the corners that are sharp at the moment. By using clipped torus (so
it's only 1/4 torus) the corners will be bevelled to fit the inversed
cylinders.
After doing this, you will probably need 3 more cylinders to .. eeh how can
I explain it.. oh, you will easily see and fix that little problem.. oh it's
difficult to explain with words. ;o)
Regards,
Hugo
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Hugo Asm wrote:
> Hi Bill,
>
> I had to think a little, but the solution is to use torus. I've posted a
> picture to illustrate it. It's quite simple in terms of math. You need to
> bevel the corners that are sharp at the moment. By using clipped torus (so
> it's only 1/4 torus) the corners will be bevelled to fit the inversed
> cylinders.
>
> After doing this, you will probably need 3 more cylinders to .. eeh how
> can I explain it.. oh, you will easily see and fix that little problem..
> oh it's
> difficult to explain with words. ;o)
>
> Regards,
> Hugo
Thanks Hugo, It worked perfectly, see p.b.i
I had to union/intersect/difference with a box, cylinder and torus
to get the shape but the result is spot on.
--
Bill Hails
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
|
|
