surface solid difference problem

Hey Everybody,

I would like to build a soild difference between surfaces and the" project onto a plane" component Brep/surface but it tells me that the round surface is invalid.

My aim is to print out an boolean if the green surface is completely cut away by the round surfaces.

Is this possible somehow?

Any help is much appreciated!

Replies to This Discussion

Permalink Reply by peter fotiadis on July 19, 2016 at 4:12am

Well ... I guess that the attached is what you want ... BUT ... it's slower than any Harley Davidson.

Moral: you tell me.

Attachments:

Brep_ops_6_V1.gh, 133 KB

Permalink Reply by Armin Seltz on July 19, 2016 at 8:51am

I think its much easier to just do this with mesh outline and curves. Then just test a bunch of points on the rectangular surface (something like a 100) and see if they are all inside the "shadow" of the spheres. Easy peasy.

Attachments:

inside_projection.gh, 18 KB

Permalink Reply by Armin Seltz on July 20, 2016 at 1:51pm

That's true, but i would use more than just the corner points, because they might all lie inside, but some other parts of the rectangle might not. I think its still pretty fast. But nice way to reduce it even more ;)

Permalink Reply by Max Rind on July 21, 2016 at 6:29am

Many thanks! I got it done by using a mixture of your answers.

The last small problem I struggle now is how to trim all surfaces by the spheres alltogether.

So that every sphere cuts away the part of the surfaces that are inside the spheres.

I basically just want the surfaces parts outside the spheres and when a surface is completely covered by a sphere I need to print out an empty data tree or "True"for it.

Attachments:

spheretrimssurfaces.3dm, 119 KB
spheretrimssurfaces.gh, 6 KB

Permalink Reply by mark zirinsky on July 21, 2016 at 6:37am

check if centroid is inside, if yes and no ineresections, ....

Permalink Reply by Max Rind on July 22, 2016 at 12:08am

So you would use point in brep and polygon center and combine that with true and false ?

Permalink Reply by Hyungsoo Kim on July 22, 2016 at 1:14am

Hi. Max. Why not use the nice code Marcus made for you?

Attachments:

spheretrimssurfaces_re.gh, 15 KB

Permalink Reply by Max Rind on July 22, 2016 at 3:41am

I am sorry I try hard, but every here and then I struggle with some little parts of my model, that makes it inaccurate then.

I found a cutting problem in your definition while studying it. When the spheres are bigger than 3 from the radius. It doesn't cut one surface.

By using point of curve it picks the wrong boolean at 0;0;1 I found out.

Isn't it the savest way to use evaulate curve like Armin did and have a look that all points are inside from the projection curves and sort by dispatch.

But then I have a data tree probelm because every projection curve is picked seperately as you can see in the second picture.