Dodecahedron consisting of 20 fused C60s (still incomplete)

I am working on a beaded dodecahedral structure which is going to consist of 20 fused C60. The construction strategy is similar to certain bead models made by Mr. Kazunori Horibe. I called those models as "Carbon poti donut （波提甜甜圈）".
In the carbon poti donut, every C60 is connected to two neighbored C60 through two particular 5-fold axes. It is easy to see which two 5-fold axes I am talking about. Here I am trying to connect every C60 to three other C60 such that all 20 C60 locate on the vertices of a dodecahedron. But the bond angles generated by the trivalent C60 do not match with the angle of a dodecahedron. So the whole structure distorts quite significantly.

This model is very similar to a dodecahedron consisting of 20 dodecahedra (Clathrate cluster). I tried to make this model last month. But because each dodecahedron in this model is highly strained, I gave it up.