when subdividing a trimmed surface it still uses its untrimmed size

I have a trimmed surface. i am using the divide domain and isotrim command on to subdivide however the subdivided surface is giving me a subdivision of the original untrimmed surface. I even tried to surface rebuild and am still getting same results... whats the reason, and how do i solve? any help is appreciated thanks forum !

Replies to This Discussion

Permalink Reply by peter fotiadis on June 15, 2015 at 12:04am

The reason is that the subdivide etc is working on a "surface" (either on a per se basis or using the underlying data) whilst you have a brep ("surface" + trimming info) actually on hand.

I have several C# that do a variety of "similar" jobs. I'll post the simplest possible here soon.

Permalink Reply by Ethan on June 15, 2015 at 3:38am

Thanks Peter!!!

Permalink Reply by peter fotiadis on June 15, 2015 at 7:21am

Well ... the good news are that this is very easy (works in any collection of Breps either trimmed or untrimmed or mixed or whatever). The bad ones are that R is INCREDIBLY slow with regard the "filtering" method (the only one working in any condition, mind).

I'll upload the thingy as soon as I'll add some divisions with zero more (tradition above all) and some chaotic options more (ditto).

Permalink Reply by peter fotiadis on June 15, 2015 at 1:32pm

This is the simplest possible (accepts mixed [trimmed/untrimmed] brep Lists etc etc).

Nerve braking slow (kinda like a Harley Davidson) due to just one method: the bool diff between the underlying surface and the brep (that way we get in a list the "culling" breps ... and the rest are history).

Note: If the "culling" brep is turned into a closed polysurface via the available option (offset BrepFace both sides etc etc) ... the whole thingy is 3 times slower: Mama Mia! Avoid at any cost.

Notice the by-pass (ignore trim info - if present) option.

Load Rhino file first (for the demo cases).

best

Attachments:

Trimmed_surf_subdivide_14A_V2.gh, 313 KB
Trimmed_surf_subdivide_14A_V2.3dm, 248 KB

Permalink Reply by Ethan on June 15, 2015 at 10:13pm

wow! had no idea this simple request would result in this... thanks!!!!!!!! will study.

Permalink Reply by peter fotiadis on June 16, 2015 at 12:00am

It's not a big deal at all (I've just removed various things instead of adding new code)... it's just ONE line (sTB = underlying surface as brep, b = OEM brep, tol = tolerance, k [not used] = Karma) ...

List<Brep> cullingBList = Brep.CreateBooleanDifference(sTB, b, tol).ToList();

... that performs the bool diff ONCE that turns me crazy (s l o w).

Permalink Reply by peter fotiadis on June 16, 2015 at 11:18pm

Spend 5 minutes more on that one: by separating hare and tortoise ... well .. at least you can achieve "some" interactive capability with regard U/V values.

Spot the Elapsed values in these 2 C# (Only offset curve on surface is maybe slower ... )

best

Attachments: