Subject:
|
Re: Accuracy of curved surfaces
|
Newsgroups:
|
lugnet.cad
|
Date:
|
Wed, 7 Apr 2004 13:08:59 GMT
|
Viewed:
|
1123 times
|
| |
|
|
In lugnet.cad, Steve Bliss wrote:
> In lugnet.cad, Niels Karsdorp wrote:
>
> > I am working on 'Electric Train 12V Brick 2 x 4 with Power Pickup'
> > (Step 8 in: http://library.brickshelf.com/scans/0000/0727/0727-03.html)
> > This part contains some curved surfaces that cannot be done with primitives
> > (very odd angles, like 41.112 degrees, 36.082 degrees and 21.786 degrees, so I
> > have to create these using triangles, quads and conditional lines.
> > My question now is: how accurate do I have to create these surfaces, ie. how
> > many angles?
>
> Generally, I create curved surfaces by using the appropriate primitives. Once I've
> got the primitive sized to create the surface, I inline it and clip out anything that
> falls outside of the desired surface. So I'd use the primitive to determine the
> number of angles.
>
> However, the number of segments to use probably depends more upon the radius of
> curvature, rather than angles in degrees. Up to about radius 30LDU, I'd definitely
> stick with the regular primitives, which have 16 segments per circle, or 1 segment =
> 22.5 degrees. Above radius 60LDU, I'd definitely use the 48-segment primitives (1
> segment = 7.5 degrees). Between 30LDU and 60LDU radius, it's a judgement call. IMO.
I forgot to mention the radii, these are between 50 and 65 LDU.
I first tried 11.25 degrees, this didn´t look round enough. Later I changed this
to 3.75 degrees. Now I changed this back to 7.5 degrees, which looks good enough
in L3Lab and MLCad.
Improvement for PovRay can be achieved with a L3PParts library, like discussed
here: http://news.lugnet.com/cad/dev/?n=9469
Niels
(Curious about the status of L3PParts)
|
|
Message is in Reply To:
 | | Re: Accuracy of curved surfaces
|
| (...) Generally, I create curved surfaces by using the appropriate primitives. Once I've got the primitive sized to create the surface, I inline it and clip out anything that falls outside of the desired surface. So I'd use the primitive to (...) (20 years ago, 7-Apr-04, to lugnet.cad)
|
3 Messages in This Thread:
 
- Entire Thread on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
|
|
|
|
