Hi everyone, I’m new to Blender but I’ve been searching online for solutions to this issue and reading tutorials, so hopefully my post won’t come off as ignorant.
Anyway, to avoid confusion, I’ll just jump to explaining my end goal. I’m trying to model a simple puzzle for my friend’s kid that I can print out with my 3D printer. The puzzle consists of her first, middle, and last name, one on top of the other, with sections cut out of each layer to allow for the one above it to fit inside. The fully assembled puzzle would look something like this:
That’s “FIRST”, “MIDDLE”, and “LAST” in case it’s hard to read. To clarify, first would be subtracted from middle, and middle from last. First and last shouldn’t have any interaction.
So the first thing I tried was to simply create the text objects, convert them to meshes, then add a boolean difference modifier to the larger piece. Unfortunately, this apparently doesn’t work due to the way text gets converted to a mesh in Blender. I searched online and found someone who fixed a similar problem by first remeshing the text, then running the boolean modifier after that. This actually worked! Well, sort of. Remeshing kind of leaves me with a dirty text model (some lumpy, rounded edges that should be sharp, loss of fidelity, etc.) but no huge deal breakers there. However, now I’m left with another problem.
Remember how I said that I was creating this with the intention of it printing it out into a real life object? Well, the problem with that idea and dealing with something like the boolean difference modifier is that things don’t work 100% perfectly in real life like they do in the computer world, especially not when it comes to 3D printing. If the edges of the outer and inner objects match up mathematically perfectly, you’re probably going to need a sledgehammer to get them to fit together in real life.
So what I really need to do is make the inset pieces slightly smaller. Well, not smaller, but skinnier. If I were to just scale them down, any piece with inner space or multiple convex shapes would no longer fit (take the letter “O” as the most basic example). Of course, there’s a tool for this–shrink/fatten–however it doesn’t really work very well after the remesh modifier has been applied:
It might be hard to tell, but the corners of the text shapes are much less enthusiastic about shrinking down than the edges/faces, resulting in a pointy, slanted shape. Also, this is after only applying a very mild level of shrink (about -0.015, which is probably not enough). Anything more than that results in a shape that no longer resembles text.
Is there a simpler way to accomplish what I’m aiming for? The idea of “text shape subtracted from text shape” seems like such a conceptually simple idea, but the more I pursue it the more complicated it gets.