After setting on the idea for quite some time, I have finally finished my first 3D Printed Puzzle. I call it the Hex Crystal
due to its crystalline shape. Mechanically, it is the 3-Layer counterpart of the Rubik's Cheese and Rubik's UFO. The shape is something I have dubbed a Golden Elongated Hexagonal Dipyramid, golden referring to the fact that its faces are golden triangles and golden rectangles.
Before I get to the pictures, I would like talk about the history of my design:
I first hinted at this puzzle in May of 2010 when I posted the following in Hints of Secret Ideas:
Jeffery Mewtamer wrote:
12 Golden Triangles and 6 Golden Rectangles.
3-Layer version of a mass-produced puzzle.
At that time, I just had the idea: The shape and how it would turn.
In May 2011, I followed up with this more detailed hint:
Jeffery Mewtamer wrote:
12 Golden Triangles
6 Golden Rectangles
3-D printed shell
~$240 in printing costs without hollows.
Design complete, just need to hollow to make it prototyping affordable.
This post was based on a V1 design that I never had printed. It was much larger(short edge length of 4cm) and used the core from a Rubik's UFO. Even hallowed, it was still too expensive for my liking to print, and the UFO becoming harder to come by cheaply only added to the price. So I decided to scrap the design and make the V2 design have its own core.
I finally printed V2 in November of 2011, but could not produce a working prototype as the grooves in the center layer were too narrow for the feet of the outer layers to fit. I was however, able to make a shapemodded Cheese from the throwaways that I have dubbed the Hex
Nut. I made a V2.1 revision to fix the fitting between center and outer layers, and I also made the core looser fitting for easier assembly/disassembly. The prototype displayed below uses the V2.1 Core and Center Layer with the V2 Outer Layers, which I did not modify.
Anyways, here are the pictures of the Hex Crystal
File comment: Hex Crystal view 1
Hex Crystal.png [ 181.79 KiB | Viewed 1873 times ]
File comment: Hex Crystal view 2
Hex Crystal 2.png [ 184.45 KiB | Viewed 1873 times ]
File comment: Hex Crystal turning on a deep cut.
Big Turn.png [ 267.23 KiB | Viewed 1873 times ]
File comment: Hex Crystal turning on its shallow cuts.
Small Turn.png [ 220.05 KiB | Viewed 1873 times ]
File comment: Hex Crystal scrambled
Hex Crystal Scambled.png [ 180.93 KiB | Viewed 1873 times ]
And of the Hex
File comment: Hex Nut WSF
Hex Nut.png [ 928.98 KiB | Viewed 1873 times ]
File comment: Hex Nut Colored
Hex Nut Colored.png [ 120.93 KiB | Viewed 1873 times ]
As you can see, I chose to go with colored plastic for finishing my prototypes. I chose this as I felt custom stickers would be cost prohibitive for such a small puzzle and the fact that classic cheeses use the same kind of finish. I chose to use Sharpie instead of dye as it seem the cheaper and easier option. For some reason, even though the parts were coloered at the same time, the Hex
Nut remains vibrant while the Hex Crystal
has gotten dull and dingy.
Turning wise: The Hex
Nut, using the tighter fitting V2 core turns smooth as smoothly, on par with some of the mass produced puzzles in my collection. The Hex
Nut can even cut corners. On the Hex Crystal
, the tips turn quite well after breaking in, though they still catch from time to time. The turns along the puzzle's length tend to go easily if everything is lined-up, but catch if the tips are out of alignment.
I would like offer this puzzle on Shapeways, but I want to get public opinion to ensure my design is unique enough before I make the model public. Also, I feel like the core needs at least one more revision: The V2 core, which is designed to fit snuggly with the stems on the puzzle's centers makes the puzzle difficult to assembly and disassemble, but makes for a very stable puzzle once assembled(as the Hex
Nut shows). The V2.1 core which has a small gap(~1 mm) between itself and the center stems makes for easy assembly, but results in a very loose puzzle as the following image shows:
File comment: That gap is entirely due to looseness, and stayed open long enough for me to take the photo.
Gap.png [ 46.68 KiB | Viewed 1873 times ]
Assembling the Hex Crystal
with the V2 core proved impossible, and using one of half of each core produced tighter build, but more catches in Cheese turns. I am thinking a V2.2 core that is halfway between the previous revisions might be the best.
For those interested, here are the measurements:
Long: 32.36mm or ~20phi mm
In-radius: 17.32mm or ~10sqrt(3)mm
Out-radius: 20 mm
Total Length: ~83.24mm or ~2sqrt((20phi)^2 - 20^2) + 20phi
Weight: Unknown as I lack an appropriate scale, but the above should be enough to calculate total volume and the pieces are solid and the puzzle has negligible cavities.
Also: no hardware required for assembly.
File comment: Center layer and core
Parts1.png [ 1002.5 KiB | Viewed 1873 times ]
File comment: Outer layers
Parts2.png [ 1.83 MiB | Viewed 1873 times ]