New ring primitive(s)

[Damien Roux]

In order to model 32004b I would need a 1-6ring with a 45 inner diameter and a 51.3 outer diameter.

I'm not sure what primitive(s) I would need to do it in a efficient way. Can someone help me please.

[Tim Gould]

Trial and error finds _an_ optimal solution...

45 -> 48 = 3 x rin15
48 -> 49.6 = 1.6 x rin30
49.6 -> 51.31 = 1.71 x rin29

although there may be others.

[Philippe Hurbain]

This tool written by Niels Karsdrop may help: http://www.philohome.com/isecalc/rings.htm. It doesn't manage fractionnal numbers, but
- you may scale everything by appropriate value
- you can generally round things to closest LDU.

[Steffen]

I also use that tool quite frequently.
I remember the old times without it, they were much more tedious.
Thanks go to Niels for it.

[Ronald Vallenduuk]

I know, I'm seven weeks late, but...
That's not a solution. In fact there isn't one, because we have no 1-6 primitives.

The real reason I'm replying here is because this got me thinking. Is it worth using three concentric ring primitives, especially for a partial ring? Is a slightly reduced file size worth tripling the number of quads used to draw the ring?

Creating a ring of any size with any number of segments is seconds work in Excel. That's how I created all the rim segments for 88517.

[Philippe Hurbain]

In this case the benefit of ring primitives is that these rings can be substituted by exact mathematical equivallent in hi-quality rendering (POV...). That said this substitution often doesn't work properly when joining to tri/quads in their neighbourhood...

[Damien Roux]

So what should I do.
In fact I would prefer to have primitive substitution available for high quality renders.

I've put a WIP 32004b in the tracker. I don't think the size is really big. Could you have a look.