Boolean Difference problem

Hi guys,

I have been trying to create a fractal progression for regular tetrahedron based on Koch division. My problem is that when I get the new tetrahedrons I need to delete the previous tetrahedron's surface that covers the base of the new ones so that the resulting shape will have its structure visible both from outside and inside view. When I boolean difference manually in Rhino i get the result that I want. However when I do it in grasshopper, it works fine for the first new tetrahedron created but when I insert also the 3d, then the boolean difference just gives just the 3d tetrahedron. Any suggestions are very welcome.

Cheers,

Thanasis

Desired Brep

Works fine for 1st to new tetrahedrons

When I also put the 3d tetrahedron, it fails

Attachments:

kosch with regular tetrahedron boolean difference Question.gh, 47 KB

Replies to This Discussion

Permalink Reply by Marios Tsiliakos on October 7, 2012 at 1:26pm

Hi,

Booleans usually operate in strangely (while in your case this is more likely to happen because the two faces are co-planar thus gh or even rhino gets confused). My suggestion is to avoid using boolean operations in cases such as this. My approach would be either to rebuild the geometry from the points and lines you already got or to perform a region diff boolean or curve join (either works) on the coplanar faces to achieve the geometry you need then re-join the brep. I hope this is clear enough..

Btw your experiment desperately calls for recursion. look into Hoopsnake..

Best,

Permalink Reply by Thanasis Korras on October 7, 2012 at 1:56pm

Hi,

thanks very much for the reply. I think what you are describing is clear enough and I'll give it a shot as it seems it would be a more efficient way to get the final surface. At the moment I am close to a workaround for the the solid difference. It seems that if tou flip the normal of the surface that gives you the problem, then you get what you want. However I am completely unaware of the parameter that seperates the ones that need flipping and the ones that not. An it seems that as I continue the subdiviion (manually for the time being), one of the 9 new tetrahedrons doesn't work even when flipped, so I guess it may be a dead end road solving it through boolean difference.

I am aware that it calls for recursion, which has been the target from the start, but I am not yet familiar with hoopsnake, although I am trying getting there.Again, thanks a lot.

Best,

Thanasis

Permalink Reply by Marios Tsiliakos on October 7, 2012 at 2:03pm

No worries,

Yes it would definitely be a waste of time to try to "guess" which face needs flipping. I mean you got all the topology to describe the geometry so why not using it from scratch in a more clear way. This is the way to go from now on. Hoopsnake is simple enough, you have the basics from your def you just need to organise the recursion. Well good luck with your project.

regards,

RSS