This needs to render as a bubble. The inner sphere normals need to point toward the center and away from the normals of the outer shell.
ttt.3dm (28.3 KB)
Select the two spheres and run the command NonManifoldMerge with the DeleteInput=Yes option. Select the new object and run the command CreateRegions. Select the resulting inner Surface (not the Polysurface) and delete. You now have a hollow sphere.
HTH
Jeremy
p.s. This used not to be possible in Rhino, which was designed as a surface modeller rather than a solid modeller. But by popular demand (and possibly the advent of 3d printing) âŚ
It groups the surfaces but does not flip the normals of the inner surface. This is still not correct.
For rendering, just turn it into a mesh and then you can flip the normals on the inner one, the fact theyâre not the same âobjectâ makes no difference for rendering.
Yeah, for rendering the raytracer only cares about the direction of the normals it encounters. I could split the sphere in half and flip itâs normals that way, but not practical for more complex shapes. Iâll take note of your meshes suggestion.
I was hoping there was a way to supress Rhinoâs desire to think of it as a solid. It should let me flip any surface closed or not.
Well in more complex shapes are you not going to at some point wind up with a connection between the inside and outside?
I do notice the SurfaceThickness feature lets you âshellâ a closed solid like this.
Yes, the shell thickness would work for uniform thicknesses. There are cases where the glass envelope has uneven geometry.
Imagine a decorative glass egg, that has something inside the capsule.
Seriously, you are trolling me now. I just showed a screengrap that it doesnât.
Sorry you donât appreciate my taking time to offer you some help. As a final offering, here is a video showing how the normals are flipped if you correctly follow the steps I set out:
You are right. I did not notice this step: Rhinoceros Help
It works as you say and I retract my earlier frustration.