I came across Eva Friedrich’s work last year and thought it was really thought provoking material. I’ve always had a fascination with Voronoi diagrams (I shamelessly attached a picture of a piece I will be printing with the DShape soon – made with Voronoi diagrams.) Here’s more about what they are, where you see them and how they’re being used to create computational iteration of structures.

Voronoi Diagrams

A Voronoi diagram is a pattern of space and structure. These patterns are found everywhere: fur, cities, plant leaves, soap bubbles and more. They are regular in their irregularity, a type of motif that would be difficult to reproduce. How does one make these? To start, a Voronoi ‘cell’ encompasses a single point. Within a cloud of points, lines are draw between them, keeping equal distance from each other. If your lines are, let’s say, blood capillaries, this type of patterning is useful if you are trying to cover an area efficiently. To get a better idea, here’s a video of a Hyperbolic Voronoi diagram and how it changes under certain conditions.

What Friedrich developed was a program to evolve a 20-cell table to adapt to pressure. She would bounce her table between a Processing-based program and a structural analysis program called Oasys GSA. Then she would apply ‘pressure’ the table surface, coaxing her table to try different shapes using Voronoi-filtering, then re-testing it, searching for an optimal structure with greater physical fitness. After a few hundred iterations, her table managed to minimize deformation by 40%. Further iterations had no effect. In a few cases, 60% optimization was reached with little geometric deformation. This evolutionary process managed to balance tension and compression to increase the structural strength, without sacrificing much geometric shape or resources at hand.

To be honest, I think my cup would slid right off this table. But let’s not ignore the bigger picture. With the advent of large-scale 3D printing, this little known paper raises a salient point about the power of computational architecture. Enrico Dini’s DShape has been able to make sizable structures and buildings using recycled stone material, giving an architect total freedom of design. Friedrich’s work explains that it would be completely possible to iterate a building’s structure under load-testing and optimize its internal geometry. We see this where humans have adapted to their conditions. So I wonder to myself why we haven’t picked up on this unique form of structural optimisation, because it seems that the new architecture is right here – in our blood and in our bones.