mesh union trouble

Hello,

I am trying to use mesh union for a set of extrusions. It works when I use breps and solid union, but this is part of a larger file, and with breps it is taking too much computation. Some times it is computed correctly, but some other times it is as in the file, with some face errors.

Are there any settings I could use to make the mesh union to work?

I tried simple mesh in the attached example too, but it doesn´t work either.

I am not sure if this a bug or not, so if this post shouldn´t be here let me know and I will move to general discussions.

Thanks,

Javier Zaratiegui

Attachments:

meshunionproblem.gh, 5 KB

Replies to This Discussion

Permalink Reply by David Stasiuk on February 3, 2014 at 3:45am

Yes, the mesh boolean doesn't work very well...I find that avoiding it altogether is the best bet, when possible. Are you always going to be extruding elements the same amount, and along the same plane? If so, you may find it easier to try to construct them as a unified mesh to begin with (using weaverbird here as well):

Permalink Reply by javier zaratiegui on February 3, 2014 at 3:52am

That sounds good. One of my cases is in the same plane, so I will give this now a try.

One question...the planar surface element output, is it connected to something?

Permalink Reply by David Stasiuk on February 3, 2014 at 4:22am

oh, whoops...at first I was passing it into the mesh brep component, but realized you don't need it...you can just pass your region-union curves into it.

Permalink Reply by javier zaratiegui on February 3, 2014 at 4:23am

I just did...but I needed this to compute the volume after the union and compare with the volume of all of them, so I can have the intersected one. I was doing with regions but when the elements group together in a "circular pattern", there might be empty regions in the middle, so I would need to do a script to check if these regions are inside to substract the value...but I am not very good with scripting. So i was using solid union, and solid intersection...but it was taking too long. i thought meshes would help.I think a script to check containment when there are more than 2 regions...and if there is containment then substract the area. In the picture is simple resulting in two regions, but I am running a multi-objective optimization, so the algorithm will try many solutions, and there will be cases when the rectangles don´t intersect, and I would need to check this containment thing, as I said.