Dr. Varda F. Hagh, University of Chicago
Traditionally, the properties of bulk materials such as elastic moduli or plasticity have been understood from the characteristic scales and symmetries of underlying ordered structures, e.g., atomic crystals or colloidal lattices. However, disordered materials, such as glasses or granular media, have great untapped potential: they can exist in a multitude of metastable states that are distinguished by their microstructure. Recent work has shown that while the vast majority of these states have similar (boring!) bulk properties, some rare cases have spectacular behavior, e.g. greatly enhanced stability to plastic rearrangements or allosteric actuation. The challenge, of course, is to direct the material to such targeted useful states that would never be discovered by chance alone. Analyzing the mechanics of networks and packings as prototypical disordered systems, we meet this challenge by introducing material training protocols that result in systematic evolution toward those desired states. One very successful and broadly applicable strategy we suggest is the transient introduction of extra degrees of freedom (such as particle radii in packings). In the augmented state space, extraordinary states can be found much more straightforwardly, after which the additional degrees of freedom are removed again. Choosing different transient degrees of freedom leads to different behaviors, such as the mechanical ultra-stability of the trained system. Whether employed for Soft Matter, biomaterials, or glassy solids, the framework presented here provides a systematic approach (into which data-driven and AI-based tools can be easily incorporated) to create novel materials and meta-materials by recognizing disorder as a crucial opportunity for versatile design of function.