Topological Self-Organisation: Using A Particle-Spring System Simulation To Generate Structural Space-Filling Lattices


The problem being addressed relates to the filling of a certain volume with a structural space frame network lattice consisting of a given number of nodes. A method is proposed that comprises a generative algorithm including a physical dynamic simulation of particle-spring system. The algorithm is able to arrange nodes in space and establish connections among them through local rules of self-organisation, thus producing space frame topologies.

A method for tackling the problem was proposed that involved a generative algorithm including a physical dynamic simulation of a particle-spring system. The aim of the thesis was set to determine the appropriateness of the method. Preliminary manual testing of the algorithm showed that it was able to robustly generate structures with observable good properties. Certain assumptions were formulated based on these observations that led to the more precise definition of the system’s physical parameters and algorithm steps. A formal experiment was conducted that involved filling a cubic envelope with a population of structures generated by the algorithm. Some quantitative features of the generated space frames were documented, reflecting the system’s performance. The geometrical, topological and structural aspects of the generated structures were analysed using statistical methods and compared to three engineered structural topologies. The evaluation of the results confirmed that the algorithm is capable of producing lattice structures with advantageous properties.

The outcome of this investigation is supportive to the claim that architectural design can benefit from incorporating physical dynamic simulation in parametric and algorithmic approaches of certain design problems. In this study, by modelling the elements of design as carriers of behavioural information that physically interact with each other, a region of possibilities in space frame topological layouts has been mapped, containing solutions that would have less likely been conceptualised and discovered in a conventional, top-down design approach. Even though the problem that this this addressed relates more to sculptural or artistic formal pursuits, the method can very likely be proven to be applicable to other architectural considerations through appropriate adaptations.

Title: Topological Self-Organisation: Using A Particle-Spring System Simulation To Generate Structural Space-Filling Lattices

Author: Sean Hanna
Author: A Kanellos

Publication: Proceedings of the 26th eCAADe Education and Research in Computer Aided Architectural Design in Europe 2008

Year: 2008

Tags: Sean Hanna A Kanellos Computer Aided Architectural Design dynamic simulation eCAADe generative algorithm particle string system structural network lattice topology volumetrics