Conway polyhedron operators in python for Blender and Sverchok
Switch branches/tags
Nothing to show
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.

Conway Operators in python

Conway Polyhedra are formed by applying various operators to a seed polyhedron such as one of the platonic solids.

from Wikipedia

conway operators

This repo includes python code to implement a subset of these operators.

The code is designed to be used with Blender and Sverchok scripted nodes but the only dependency is Blender's mathutils library. This means the code can be run outside Blender using a standalone version of the mathutils module.

Usage Notes

Using the module in Blender with Sverchok's Scripted Node Lite.

  • Install Sverchok in Blender.
  • Download the zip file from github
  • Open as a text block in Blender.
  • Open and as text blocks in Blender. These contain code for each Sverchok Scripted Node Lite.
  • In a Node Editor view create a new Node Tree and add two Scripted Node Lite nodes.
  • Use the notebook icon on the node to select on the left node and on the right node. Click the plug icon on each node to load the code.
  • Wire up the nodes along with a Viewer Draw node as shown below.

conway nodes

Wire up multiple copies of in a row to produce more complex shapes.


Two of the operators kis and chamfer can take parameters such as the height of the kis pyramid or the height and thickness of the chamfer. There is a separate Scripted Node Lite given for these two operators with sliders for the parameters.

Some operators, particularly gyro, propellor and whirl and chamfer give polyhedra that are not particularly smooth or convex, the faces may not be flat or symmetric.

conway cgC

The canonical form of a convex polyhedra has all faces planar and all edges tangential to the unit sphere. The centre of gravity of the tangential points is also at the centre of the same unit sphere.

The module contiains functions that attempt to shift the points of a polyhedron to satisfy these conditions. This is a iterative process and can take several hundred steps to converge.

To try this in Sverchok, add the and files as text blocks in your Blender file and add as a Scripted Node Lite. The node has two parameters iterations and scale_factor. At each iteration the vertices are moved a scale_factor fraction of the calculated distance. Setting this parameter too high may cause the shape to become unstable. Increasing the iterations will increase the calculation time.


The canonicalization can also be applied after each operator. In the example below just enough iterations have been applied to form a pleasing shape. The proper canonical form of this polyhedra should be the same whether the canonicalization is performed once or twice.


These Conway operators can be applied to any manifold (ie. a closed solid) mesh not just the platonic solids. They currently don't work on planar grids unless one applies a solidify node to the grid first.


Other Sverchok nodes of course can be used interspersed with the Conway operators for other effects.

I've only implemented a subset of the operators defined on the Wikipedia page. Many of the operators are equivalent to a combination of other operators as shown in the chart

Conversion chart

The operator order is given as the left to right node order. Note that this is the opposite to the order given in the Conway notation.

Operator Description Implementation
kis poke face node
dual faces become vertices, vertices become faces node
ambo full vertex bevel node
chamfer hexagons replace edges node
gyro faces divided into pentagons node
whirl insets a smaller rotated copy of the face node
propellor insets a rotated copy of the face node
zip dual of kis kis dual
expand edge bevel ambo ambo
bevel vertex bevel applied twice ambo dual kis dual
snub dual of gyro gyro dual
join dual of ambo ambo dual
needle dual of truncate dual kis
ortho single subdivide ambo ambo dual
meta poke face and subdivide edges ambo dual kis
truncate half vertex bevel dual kis dual

See my Look Think Make blog for more info.