Reconfigurable materials through chirped design

Evolution in time-varying environments naturally leads to adaptable biological systems that can switch functionalities easily. Here we exploit this principle to design materials that can switch between two distinct functionalities with minimal changes in design parameters. We consider high-dimensional materials design problems where any particular functionality can be effectively realized by numerous equivalent choices of design parameters. We run native design algorithms like gradient descent or CMA-ES for these problems; however, we switch the target functionality between two distinct targets at regular intervals, even if the design is incomplete. By systematically increasing the frequency of such switches in a `chirp’, we naturally uncover pairs of parameter sets that exhibit the two target functionalities but are minimally different from each other. We apply these ideas to design mechanical networks that can switch deformation modes with minimal bond changes and to design 3d shapes that can switch between specified packing fractions with minimal shape changes.