Beryl crystal structure (a.k.a. FCS nightmare)

This probably caused some headaches at shapeways, although it came out very nice in the end. I have not yet mustered the courage to inquire about the number of reprints necessary to achieve this, but I doubt that it was straightforward even though I made the connecting spheres bigger than I would have liked...

The model shows a section of the crystal structure of the mineral beryl (named aquamarine or emerald as a gemstone) in the polyhedral representation commonly used in solid state chemistry where ball-and-stick models would quickly become too confusing. (The red tetrahedra each contain a silicon atom in the center, while there is an aluminum atom inside the blue octahedra and beryllium in the green polyhedra. All vertices are occupied by oxygen atoms - the red spheres.)

I have left off the oxygens on the outside of the model as they would have added more weight to the more or less free-standing polyhedra and would have had a high probability of breaking off or crumbling away themselves - hopefully recent developments like the ProJet4500 will make printing similar models easier in the near future.

This is geometrically mesmerizing on the eyes. Very cool!

Thank you for the feedback. Obviously I had not truely appreciated the "evil" sandcastle rule and the somewhat unexpected fragility of "raw" FCS prints before I tried this. And rest assured that I will not try to offer it for sale here, or otherwise try to bankrupt shapeways through repeat orders of it.
Now if there was a multicolor material that did not require this diligent post-processing that seems diametrical to high-throughput printing...