I’m trying to create a sphere surface to run a shape optimization in another software.
Basically, the idea is that I want to create a sphere and alter the position of control points along it’s surface to deform it and then generate the geometry to be exported.
I tried multiple ways of doing this and have encountered the following problem in one approach: I can’t close the surface! The loft function only works with curves.. but at the poles I just have a point. Does anyone have a suggestion on how to fix this? Maybe is there a more straightfoward way of doing it?
Do you need it to be a surface? Your approach and goal seems more suited for a mesh.
Simply create a sphere mesh, and deconstruct it; you get the vertices and the faces’ topology, with which you can reconstruct the shape after modifying the position of the vertices.
I managed to do what I wanted with other scheme… Let me show you. I want to deform the sphere surface through the control points. But what I’m not satisfied is that I have a higher concentration of control points near the poles.. with the file that I sent you first I have a even distribution of points across the sections, if I manage to create closed surface with that file I can repeat the procedure to create this maleable sphere.
You can create a spherical mesh with the Mesh Sphere component. However, this mesh will have the longitudinal / latitudinal divisioning which you are trying to avoid.
You can create a custom mesh from points using Convex Hull here, since the object is a sphere. In reality, you are not going to get a more equidistant divisioning than in an “Icosphere” which you can get from the Polyhedra plug-in.
I feel like an attractor point system might be better. You will have to change the Gene Pool parameter count for every change in the divisioning and moving a single “control point” of the surface might be too localized of a change.
