3D Printing

#&%!*@% boolean unions; apparently mesh interior finding is a nontrivial problem

My excitement coming into this project was slightly tempered by my lack of any meaningful experience with 3D modeling software. I'd used Blender to do animations before, but its UI was esoteric and inscrutable and I originally didn't want to rely on it for modeling. After trying out both SketchUp and FreeCAD, however, I realized that the type of things I wanted to make - figurines as opposed to parts, models versus mechanics - were not what those programs were designed to create. So after some wailing and gnashing of teeth, I went back to Blender.

I decided to start with something simplistic and work my way up to the "can't be built subtractively" goal, and someone at the IDC hobby shop appeared to be a big Tuzki bunny fan. This was my first 3D model in Blender.

I was given a couple of definitions of "can't be built subtractively". One was that the object had to have interior detail, like a hypercube or a Klein bottle. Another was a bit more lax and only specified that a miller moving on 3 axes (l/r, forward/backward, up/down) would not be able to produce the object; hence, an object with detail on its "underside" would be fair game. In order to add more interior/difficult-to-reach detail, I gave Tuzki some ancient Han Chinese clothes.

Unfortunately, when I exported the file as an .stl I forgot to perform a boolean union on all of the separate meshes I'd used to construct Tuzki, in order to combine them into a single closed surface. Therefore, the following was all that was printed of Chinese Warlord Tuzki the first time around.

By the time I figured out how to do Boolean unions properly, I was running out of time to print another Tuzki iteration on the Form1 in the IDC labs, so I decided to make a second, simpler model that could be printed directly on the Makerbot. Since the makerbot was less robust with regards to parts, I built this model closer to the ground and with less protruding parts, but with some interior/bottom structure that would fit the assignment.


Thankfully, the makerbot worked as intended for this model. (I've also asked Charles G. to reprint the Tuzki model - hopefully I'll be able to update this page later with it too)

The last thing I did was attempt to scan an object into a model using 123D Catch (since both TAs strongly recommended we stay away from the lab scanner). I took about 40 pictures of two action figures I had lying around and uploaded them; however, I ran into the problem of having too cluttered a background and not being close enough to the actual object I wanted to model, so both models ended up being lower resolution than I wanted.


Also this is what happens when you have a lot of background in your pictures!

Index