If the purpose of this project was to illustrate how straightforward and incredible the 3d printing-verse is these days, it absolutely succeeded. My favorite thing about this project was how easy it was to make something really really fun.
I decided which object I wanted to print the day we did the showcase in class. I really enjoyed the portable gyroscope fidget toy, and so my decision was based on my selfish desire to print an extra one and keep it for myself. I started with essentially the same design as one of the showcased pieces, except that it had seven rings instead of the five that the already printed design had. The “impossible” part of this piece is that it’s held together by a small protrusion on the outside of each ring that fits into a divot on the inside of the ring bigger than it. There are two of these “joints” on the opposite sides of each ring, which allows the ring to pivot on a single axis. It would be impossible to fit these together if they were crafted separately, and it would be extremely difficult to carve the entire thing out of a single piece.
However, this design did not scale particularly well in 3dPrinterOS because those protrusions and divots were scaled as well. I tried scaling both uniformly and then only scaling in the x- and y- planes, but this reduced the size of the joints too much to make a secure connection, and they either slipped out or broke each time. If I were to accomplish this myself, I would probably just choose to design it myself to fit the given parameters, while making the joints as large as possible. Here’s all the broken pieces plus a sketch of how the jointing system works.
The ring-based gyroscope failed, but I still wanted to make a similar item, so I chose a different design that would scale better. Instead of being held together by two points, each ring stays in place by fitting into the curvature of the next layer all the way around the circle. Because it’s the angle of curvature that matters, not the distance of any part, it scaled much better. I did most of my designs as large as possible, but I did one tiny one just for kicks. I did learn the limit of the scaling because the innermost layer was too small to print well, but it still functions well otherwise. This design is also cool because it comes with a little fitting for a ring, so whoever gets it could use it as a keychain if they wanted to.
I’m really happy with how the FDM prints turned out, but the design definitely did not work out so well with SLA. The gap between the layers that allows each ring to rotate seems to be too small, and I believe extra resin seeped into these gaps and fused the different layers together. I tried to chip away at the layers within an Exacto-knife, but it was impossible to follow the internal curvature in each joint. Even if I was able to fully separate each layer, it would only be able to rotate in the same plane as the next bigger ring. It couldn’t turn on the axis that goes through the diameter, which is the exact feature that makes this part so fun to play with. Explaining spatial things well in words is always a little confusing (at least when I’m the reader), so I sketched the problem vs. how it’s supposed to be, and a picture of the failed part. It’s not the clearest, but if you look closely you can see the shiny part that runs across all layers.
To compensate for the failed SLA print, I did a quick mini version of the FDM print so I could have 5 parts to turn in. Here’s the final result of a successful FDM print!
Cost Analysis:
The FDM prints were $0.37 each, and the SLA prints are estimated to be a dollar each (at a rate of $50 per kilogram). I spent approximately two hours on this project if you don’t count the hours spent waiting for my prints to finish. All together, including the failed prints, this project comes in at $19.60.
