Hello everyone,
I am working on a workflow where I want to locally modify an area of an existing surface (for example a sphere). Conceptually, I have a target surface (eg sphere), and I would like to replace a local region of it with a custom deformed surface, while keeping the body watertight
Currently my idea is to do a kind of “ surface replacement”:
-
I project a reference surface on the target geometry.
-
I modify this reference surface into the surface I actually want to integrate (transformed surface).
-
I scale the region of the reference surface slightly to get some tolerance buffer.
-
I trim out the corresponding area from the target surface using the scaled surface.
-
I use lofting between the edges to connect the modified surface and the original body.
-
I try to rebuild it into a surface
This almost works, but in the end I always end up with very small gaps along the transition, most likely due to tolerance.
surface_replacing.gh (4.1 MB)
I think i could transform it into a mesh and then use ShrinkWrap to fix it, but thats to computationally.
Performance is important to me. I deliberately only want to manipulate a small region of the object, because the final models will become larger and more complex. I already tried more approaches such as:
-
converting everything to dense point grids and then deform the points
-
locally increasing mesh resolution (for deformation workflows like Mesh by Color)
Both worked conceptually but became too computationally expensive.
Do you guys have an ideas for:
-
a robust approach for replacing or locally deforming a surface region
-
ideally something that is not extremely computationally heavy
Any guidance, best practices, or example definitions would be highly appreciated.
Thanks!
