In the realm of architecture, I frequently come across situations where clients request panelizations or where the project naturally leans in that direction. In the initial phases of involvement, my role does not revolve around panelizing for construction; rather, I focus on executing panelization for concept design and design intents.
Even though I am curious on the tools and workflows to create panelizations for manufacturing, involving planarization, and piece optimization (to not have millions of different pieces), etc. I do not need to go there to make a render look nice. Yet, I still need some control over panel size and direction of the pattern.
I have seen some previous discussions on panelization as well as posted my own discussions.
Despite the lengthy discussions on the subject, I have never understood the clear workflow to achieve a panelization on any 3D model/surface.
I am aware of Kangaroo, Panelling Tools, etc (if others please recommend).
Still, none of these help me actually panelize any 3D model I make, with some sort of control and that does not mean making it entirely manual by drawing each panel as a 3D face.
I often work with SubDs, so any SubD model could work as an example of geometry I need to panelize. In this case I am working with rectangular panels, but it could be anything honestly, hexagonal, triangular, etc.
Hi Thomas, my goal is basically design intent. Meaning, I need to be able to panelize certain models with x shape and approximately controlling panel size.
Basically those are my two parameters. Polygon side count and size.
For instance, rectangular panels of 2x4 meters.
I was never able to do this except using roughly using Quadremesh. I am not an expert on Panelling tools but it always gave me a headache when trying to panelize subD models which actually need to be converted to surfaces?
Hello. Geometry for architecture is not like a rhino or a bunny in your image. As far as I can tell, you want to know the panelization workflow for architectural design, right? I’m asking this because general quad-meshing of sub-D surfaces is a completely different topic, as the central topic is how to identify/control singularity points.
For architecture, there are usually floors or columns that are hiding behind the shape, making the problem far simpler and more manageable. Because the shape must be supported by the columns or beams supporting the floors or any other types of structures, one of the two directions of the grid is very likely restrained to a given set of curves. So, in many cases, one set of directions is given, and only you need to do is to determine the other set of directions that intersects with the first set of directions.
That said, the process generally starts with slicing the geometry into ribbons and dividing the ribbons into panels.
Studying patterns without assuming any internal structure and a panelization considering where are supports are two different things. The former is actually a realm of computer graphics, in my opinion.
I believe there should be a moment somewhere in the project when you need to start thinking about the structure, and that’s the moment you need to throw away the scripts you have made in the conceptual phase. However, thinking about structure in the conceptual phase may slow down the process, especially when you want to test multiple ideas. Good luck!