I think I got sufficiently mad during the last week to call myself an artist. I upscaled to a degree that my calculations were quite off the target, and a lot of re-adjustment was required. Three structures with 95 cm struts emerged. The first one, attached to its current destination in a stormy night before the rain hit, still survives the wild weather. The second structure didn't stay for 24 hours in the wild, the latest one still needs some painting and still blocks some of my lounge.
While the tetrahedron exposes some efficiency in the use of materials (roughly a 1:1 ratio of string and tendon), the six-strut icosahedron seems wasteful. The initial string/tendon ratio was about 2.7:1, and as I was running low on nylon strings, I wanted more bang for my bucks. I can't really prove in any fashion that in order to create a tension triangle you can equally to from the corner along its sides or to its center, and the central joints introduced another additional node in the network of strings, but in practise it works just splendidly.
The six-strut icosahedron is a strange tensegrity, connecting the 12 corners of the icosa in a neatly way highly symmetrical through its center. In a way, the minimal 3 strut tensegrity can be understood as minimal octahedron. I better test this experimentally :) Being able to reduce to string/ratio to 1.5:1 made the idea of larger icosas much more viable, and I can't stop experimenting with the result.
I hit a sweet spot with the tendon lengths. Any three struts will balance (all 20 faces), drop and squeezing tests showed a lot of robustness, and the wigglyness can be hypnotising. I won't test it to breaking strength without camera, but I love the options offered by this design in a larger scale. It's simple to suspend small scale objects in the center, some sort of generic 'picture frame' to showcase more complex models.
I wonder whether I can use a set of centrally joined tendons to build from the scratch.