My Kitchen Cupboard Doors | Week 3
This week I have started using blender’s geometry nodes. I made this:

It is the first step in recreating my kitchen’s cupboard doors in a fully non-destructive way. As you can see below, the door comes in three distinct sizes. So it is a good fit for a non-destructive workflow. We only need to model it once, and then we can just adjust the size parameters.
Obviously, even all three doors would still be quicker to model destructively. However, my goal is not to build a “procedural furniture generator” or anything like that. My goal is to extract common modeling steps and/or design elements into reusable, composable components.


As it stands right now, I just have some curve. In the next step, I will add a mitered inset. This will look something like this:


The importance of topology
Sometimes modelers view geonodes as a handy tool to quickly iterate on visual ideas, but not as anything you would want to make a final mesh with. The reason being (presumably) that a naive node setup will lead to a bad topology. This is completely true, as many nodes don’t expose much control over the result’s topology.
Here you can see an image of the topology that is spit out by Curve to Mesh. Next to it, a quick and dirty sketch of what the desired topology might look like.


A quick note on time tracking
I haven’t really been tracking my time very well this week. Part of the reason is that I don’t know what should count as “track-worthy” time. I watched quite a few geonodes tutorials on YouTube which doesn’t really feel like productive work time. So yeah, I’ll have to think about it.