Surface closest point with strange results


I was trying to populate a brep with points. For each face I want only those points lying on them separately in a list. To do so I used the “surface closest point” component to get the points that have 0 distance to a particular face. Strangely enough some points seem to be in 0 distance on several faces in the same time which is impossible. The equals zero component doesn’t work with my definition either.

Has someone an idea why this is happening?
Thanks in advance

Ps: I tried a different brep with other surfaces. Same problem, so its not the surfaces that are wrong
band geometry (22.2 KB)

The output of the Surface Closest Point component way at the start which measures the distances doesn’t return lists of distances for each point to all faces, but the distances of each face to all points.

However I think you want it that way. Doesn’t the first bit of the file already give you all the points in separate lists, one for each face?

band geometry (11.2 KB)

See, I can make a delaunay mesh of each point list and the meshes are definitely per face:

band geometry (10.5 KB)

Hi David.

yes. you are right, it gives me the distance for all points to a face but I am aware of that.
I wanted to separate for each face only those points that are lying on it. From those points and the normal direction to the surface its lying on I created planes that intersect the entire brep. At some points though, I get not only one plane but sereral. (and that happens mostly only on one face)
I narrowed it down to this distance thing: If there is only one point on a face this closest point component should NOT have several results being 0.

I hope this illustrats my problem a bit better for you.

Hi again.

When I replace the smaller than 0.01 with an “equals 0” mine/your difinition doesnt work anymore. I suppose thats because the “pop Geo” component creates points that aren’t exactly in 0 distance to the faces?

You were right (of course)
and I was wrong. I associated the distance result to the wrong
points. So the definition worked well there. The mistake was
somewhere else, can’t really put my finger on what it was but it
works now. I have corrected the definition for those who are
interested in in this.

It takes any BRep or surface-compound and intersects it with planes on randomly placed
points. You get a thread-like representation of the input geometry.
One could use it as a base to create something similar to this guys

Thank you guys for your
help and the great work you’re doing. I very appreciate it.
band geometry (18.8 KB)

1 Like