Hello everyone,

I'd like to present to you a puzzle that I completed a while back and mentioned here. It doesn't work as well as I would've liked, but I'll present it anyway and hope that future designers can take inspiration from it.



It is essentially a Square-2 with added functionality on the middle layer, or a fully unbandaged Square-1.







The mechanism I used is a shells buildup of three puzzles: Rubik's Cheese, a Hexagonal Prism, and the Square-2. With some internal bandaging on one of the cheese slices I was able to achieve a mechanism that doesn't misalign (at least, it's not supposed to misalign). Due to the sharp 30° angles of the Square-2 slices, the mechanism can jam and wiggle around internally a bit. For that reason, it's often difficult to turn the three out of the six slices of the puzzle that correspond with the Hexagonal Prism shell. Also, I had some problems with tolerance in the design. The inner Rubik's Cheese slice is way too loose (mainly from the nylon breaking in), and the outermost layer is way too tight.



I'm not sure whether or not I'll make a second prototype. The puzzle was a bit expensive, and I'm not sure if I can risk having another one that barely works. I already know of another designer that's making something similar to this. I hope everyone can learn from my mistakes and use a more efficient mechanism.

VIDEO

Chandler

Wow. This looks quite complicated and cool. I don't like the normal Square-1 puzzle but this one is great!

That is a very impressive design. Will you be posting a video?

Carl

Nice puzzle, too bad it didn't work too well. I second the request for a video!

Maybe I'm just weird and alone with my opinion, but I consider a square-2 to be fully unbandaged, merely having stored cuts.

Unblievable! Great job, CJ!

Thanks, everyone.

I'll try to make a video soon.
Edit: The video has been uploaded here: http://www.youtube.com/watch?v=Tay4LCiBQIQ

I didn't consider that. I guess it can be viewed either way, but I usually associate stored cuts with jumbling puzzles.

Chandler

Very cool no matter how it turns.

Nice. Though I may need to watch that again. It didn't appear that you turned the puzzle along any of the horizontal cuts (refering to their orientation in the first picture of this thread). Could you add a few turns of those cuts to your video?

Carl

Oh, I forgot to mention that. This is one of the reasons I don't think this mechanism will ever lead to a working puzzle. The 12 slices wobble around so much that they are almost impossible to line up. The top and bottom layers are pretty tight to begin with, so turns on those layers are pretty much impossible to do without pushing each piece across the track carefully by hand. If you really wanted to see it, I suppose I could add it to the video. For now, here is a picture:

Ooo I really like this puzzle, even though it doesnt turn that great, it is still an impressive proof of concept.

Isn't this very much like a 12-sectioned, 3-layered Masterball? The Masterball mechanism is notoriously finicky. It's a tough puzzle to get right, with so many cuts going through the center.

This is pretty good, though! Any chance of a second iteration?

If the outer layers are troublesome... any chance of a 1-layer version (unbandaged Puck)?

I like this puzzle a lot if it were only in a non fragile state. I was thinking that you could probably use a 3 layer puck puzzle for this but then the question would be would be is how to hide the black center turning that moves the puzzle pieces? Am I the only one that doesn't like the black center piece that moves that piece showing on other puzzles? Don't puck puzzles come with ball bearing and springs?

Well, the "black center piece" is essentially the entire mechanism of the puck puzzle. Pucks typically use one single turning slice that can be rotated to move any plane. The same goes for the original Square-1, except it has a middle layer that does the same job as the puck's center piece. The concept behind my puzzle is to avoid the use of a movable center piece by using a 6-cut (12 piece) prism, rather than a single cut as the mechanism. Until an efficient way of building that 12 piece prism is found, it'll be pretty difficult to construct a non-fragile version.


It is very much like a 12-sectioned, 3-layered Masterball. I agree, I wanted to avoid the Masterball mech at all costs, but now I'm beginning to believe it's the only mechanism that could hold the puzzle together tight enough to allow for the addition of layers.

I can't really see a second iteration in the near future. I'd have to design it first. And so far, I haven't done even that. I'd really like to build a better one, though I have a feeling someone will beat me to it, which I wouldn't mind.


That's what I wanted to build originally, but I realized it'd have to be either a really large cuboid, or cubic. That's why I made it three layers; the extra layers didn't really take up much space in the cube.

Well, I hope you'll do a cylindrical one someday...