Thursday, December 16, 2010


Chris K. Palmer modeled these novel Sphericons based on a mathematical discussion by Colin J. Roberts and Paul J. Roberts. His models included 4mm holes to accommodate widely-available 3.2mm magnets (like these) plus glue to cement them into place.

I scaled to model to 75% so the holes would be 3mm, slightly smaller than MagCraft N5NO566 disc magnets (3.2mm x 1.6mm). The resilient PVC material had sufficient 'flex' to permit the magnets to be forced into them.

Once the magnets were installed, I briefly soaked the parts in PVC welding solvent.

The solvent caused the PVC material to return to its square-ish shape around the magnets, securing them neatly in the holes.

The magnets allow the pieces to be paired to form round, solid objects. The pairing on the left can roll in a straight line, whereas the pair on the right can roll in a circular path.

But the same pieces can be paired so the curves look quite different. These pieces can be rolled along irregular paths.

Here's how the sphericon at left would roll. It follows a straight path with an occasional backstep.

And here's it's counterpart, at right in the previous picture. Its path is a bit more crooked.

1 comment:

  1. Scott, awesome blog! I just ordered my sd300 today. As a complete newby to g-code, I was wondering if you had school for it or if you're just self taught? I would like Solido classes or something to help get me started. I will be building firearm add-on parts and there is no code available for it, it will all need to be built from scratch. I will use them as mold interiors for foundry molds, mostly casting aluminum. Are you getting some of your code from a site somewhere? Thanks and as soon as I get my machine in and going, i'll send you some pics! Thanks! Terry