Continuing the study on inter-connected structures, an aggregation is modified using the part transform functionality in WASP and Galapagos with the goal to find a closed cell in a given sequence of parts.
The genetic solver tests transformations of a single part type, with the goal to minimize the distance between the open ends of the part – sequence. Here, the only transformation parameters are rotations of branches. The simulation shows that after iterative testing, a closed cell is found.
In aggregations with this transformed part, closed cells emerge occasionally, sometimes even in configurations that are different than the cell initially created in the simulations

It can be observed, that cells are usually formed only at the origin and don’t occur again in the periphery. This might indicate, that aggregations with this setup are mainly producing structures with open ends (L-system patterns) with cells emerging only locally.
Variations with different rule grammar, but identical part:




Sometimes, identical gaps between parts appear: an additional bridging part could be used to interconnect, as in the previous study.
Larger cells
based on the same simulation logic, larger cells are formed.




Cells occur locally. However, no larger cell emerges as “designed” in the transform simulation above.
Overall, the study shows, that when parts are designed to form larger cells (in the initial simulation), they are less likely to occur naturally in stochastic aggregations with that transformed part.
Continue with next study…