The Plan

I’m using trueSpace3 for Win95/NT. I modeled a solid 12-facet dodecahedron a long time ago (took forever to figure that one out!). Then today I used a "smooth quad divide" function which turned it into quadrangles for me. It's kind of an accident how it turned out, but since the whole dome is made from equal kite-shapes (where short to long ratio is 1:1.5) it should be relatively easy to construct, and uses less sticks then a regular geodesic dome. (though perhaps not quite as strong?). I still haven't figured out all the angles, once I do I'll post the info. I won't know if they're right for sure until I come up with a way to build a real model.

Let's say I made a small 10 x 10 x 5 dome with the quadrilateral design out of wood or metal. Then I stapled or used binding wire poked through the canvas to tie pieces of it tightly to the framework. Could you then mix up a bunch of concrete in some buckets and dump it over the top until you get a thick layer, possibly doing this for a few days to build layer after layer? Or would rain, time, or other things prevent this? Is concrete too expensive? Or would simply creating a thin outer canvas layer that is waterproof, like a tent, serve better? A dome tent might not blow over like regular tents can.

I think I know a trick to making a dome framework very easily. The dome-builder would only have to know how many segments to cut and what the lengths are. The builder would cut pipes of some sort to the proper lengths, and then run cable or rope through them to build small sections, which can then be roped together to form the finished dome. Obviously a roped-together dome might be a tad flimsy, so then you would have to wrap binding wire very tightly around each vertex, going over and under the pipes until it's strong. Then you could wire fabric to the framework to cover it like a tent and/or dump concrete on top. Or whatever. The best part is that the builder will not have to measure any angles, or any of that crap. Perhaps only a picture of the finished dome or a real good description would suffice. Does this sound reasonable to anyone? I'm going to try out some small models of various geometries.

Offered by Joe.