# Help with Subtracting a Scanned 3D Mesh surface from a Solid

My current focus is on a new cleat design for athletic shoes. I meticulously aligned the output of a 3-D scanner, a mesh with a .35 mm spacing, with two uncomplicated solid models in Rhino.

Below are two images, the first shaded, the second wire frame.

What I usually do with the surfaces I model is split each one with the other, toss the parts I don’t need, join the result, and probably Join some naked edges to create a closed solid.

I tried that with the mesh and learned you can’t split a Mesh with a solid or a solid with a mesh. I then converted the Mesh to a SubD using the Mesh Edit Quad Remesh tool. I settled on a 1 mm edge length, converted to SubD, and Interpolated the SubD to get a pretty good surface that I can at least select in a Split command.

However, when I run the Split command, I end up with a line at the intersection point that is part of the SubD, and the rest of the SubD surface is untouched.

Any thoughts on how to do this are welcome!

Wireframe Image

Hi Brian - better to post the geometry or send to tech@mcneel.com with a link back here in your comments. What happens if you Explode the split object?

-Pascal

1 Like

Pascal, good point—here is the file. There are three elements: the mesh from the scanner and the two solids I’m trying to subtract from. The idea is that I’ll 3D print the two “Mesh Trimmed” solids and glue them onto the shoe the scan was taken from.

3DM Mesh & Solids.

1 Like

Hi Brian - the intersection of the subd and the pads is dodgy –

the pads have not very good edge tolerances (`What` command)
Geometry:
Valid polysurface.
closed solid polysurface with 3 surfaces.
Edge Tally:
1 seam edges
2 manifold edges
= 3 total edges
Edge Tolerances: 0.052 to 0.082 <<<<<<<<<<<<<<< THIS
median = 0.069 average = 0.068
Vertex Tolerances: 0.004 to 0.053
median = 0.028 average = 0.028

and the edges are mostly right along the intersection-

So it is unlikely to be a clean curve that you can trim with.

After a RebuildEdges on the surfaces in that polysurface, the edges are not even close:

What I think I’d do here is make the sides of the pads extend right through the scan, forget about the top surface. Normally I’d ExtendSrf but in this case the side surfaces are hooked in at the top so that does not do it

I moved some control points and extended .

Then splitting works - - eventually - it takes a while with that very dense SubD

-Pascal

1 Like

Pascal,
Thank you!
I understand what I did wrong; I patched the top surface of the “solid,” and that resulted in the crappy edge.
I’ll take the curve I built by placing points on the mesh, move that curve in via a 2 mm offset, and rebuild the outer surface. As for the long wait for the split, that’s what coffee is for!
I have 11 pairs of cleats to do this to—size USM 6 to 11—so you are a lifesaver!
Thank you!
Brian

1 Like

Pascal,

Thank you; this worked like a champ! After I add the cleat pattern, I will glue these blocks onto shoes for testing, so the quality of the fit matters! There are a ton of SubD blocks, but that’s how I get the good fit I need for contact cement.

Only 23 more shoes to go!

Brian

2 Likes

That’s awesome. I’m doing research on 3d scanners this afternoon, and found this at the top of my search