I've been thinking about this puzzle idea. I'm no puzzle builder and I don't really keep up with the latest creations. But I created some quick designs in MS Paint.



I don't if there are any rules builders follow to know where and whether a puzzle can be turned along certain lines. I also don't have knowledge of how internals should work. In the following images, I have highlighted what seems to be the logical turning lines for this puzzle idea:











Looks like if you were to turn two opposite "layers" 180 degrees, the shape would form a triangular prism. Lastly, it looks like maybe adding one more line, splitting the shape, would allow for more of these turns:



Really, I'm wondering if those "splitting lines" in the second image and the previous image are necessary. They may make the puzzle too easy.

If I didn't have work today, I'd stay home and mock this up in solidworks, just so we could see how it looks all shifted and stuff. D: If nobody does it by the time I get home, I'll do it.

I think the design on the red cube needs to be rotated by 90 degrees. This would make it "perfect" geometry and make the extra cuts you designed not needed. The only other issues I see are floating pieces and it getting too jumbled.

I gave this design a try. I did rotate the red face 90 degrees to make it more symmetrical. It looks like the cube would have to be pretty large for it to work. The piece that I'm having trouble with is the right triangle-shaped part at the edges. When the cube is at 65mm edge length, the stalk is a little under 1mm thick, which I predict would break pretty easily.

Yes!! Like this.

But the only non-jumbling move appears to be 360 degree rotation so in other words no non-jumbling moves. Can it even jumble? Its not obvious to me there is any place short of 360 degrees which will allow another rotation to take place. So can this even be scrambled?

Carl

Certainly would be disappointing if it couldn't scramble.

Adding back the 2x2x2 cuts (while not enabling 2x2x2 moves) would seem to allow 180° turns to create more moves:

This rotates around 2 points on each face. These pints are on the line that cuts each face in half. Becease of this it has two angles of rotation, one I'd say around 20 degrees and the other close to 180 degrees. The question is can it rotate from any of those two positions?

I'm not positive, but this concept appears to be a deeper-cut version of a Hexaminx. If that's the case, it might scramble in a manner similar to the Hexaminx. Does anyone know if the angles are indeed the same?

The angles are not the same as on a hexaminx. I tried experimenting to see what shape the cuts are based off of, and the shape is basically a deformed dodecahedron.

EDIT: Pictures added.

Just answered my own question...This shape would be the core of the cube, and it is not a regular dodecahedron. Each of it's 12 faces is a pentagon, but with only four equal sides, the fifth side being longer. Therefore, this slice pattern is not the equivalent of a deeper-cut hexaminx.

Whoops, you beat me to the answer...

After I finished drawing the original image and was trying to understand the turns, I decided to play around and draw the lines in a slightly different way. I later realized that it was exactly an Axis Cube. Are the two related in any way?

How is this related to an axis cube? An axis cube is a shape modification of a 3x3. This puzzle would have a completely different core and mechanism. What do you mean when you say you drew the lines "slightly differently"?

Just asking. I had hoped I made it clear in my first post that I don't have the puzzle building knowledge of many of the forum members. I simply had a puzzle idea that was appealing to me.

Instead of drawing the starting triangle on each face from two adjacent corners leading to a point at an edge (in other words the triangle has one vertex at the edge of the cube face and two vertices in corners of the face), I drew the lines from a corner leading to two adjacent edges of the cube face (the triangle has one vertex at a corner and two vertices at edges). Then of course the straight line is rotated along with them going from corner to corner now instead of edge to edge. I guess I could say I used some of my original design and rotated it 45 degrees.

If anyone wants to build and test this puzzle (if it even scrambles as the fear seems to be), please do. I of course am not able to and would never get around to learning how. I would love to see a working version.