A computational strategy that can optimize the mechanical properties of jammed granular materials is reported online this week in Nature Materials. The approach makes possible the systematic exploration of the role that particle shape has in these materials' response to compression.
In a granular solid - such as rice or corn flakes - the shape of a particle (granule) determines how the granules pack and the network of contacts between them, and thus how the solid will respond to compression (arising from external pressure or the material's own weight). However, because shape is an inexhaustible parameter, a systematic exploration of the role of particle shape in the mechanical response of granular materials has so far been impractical. Marc Miskin and Heinrich Jaeger rephrased the problem in terms of artificial evolution. They demonstrate an algorithm that improves iteratively, through selection and mutation of shapes, the granule that leads to the best mechanical performance among an initial set of granules with varied shapes. The researchers show that after a few tens or hundreds of generations the evolutionary algorithm was able to find a granule for which its aggregates stiffen - rather than weaken - under compression, and verified this using three- dimensional printing.