Adaptive energy dissipation in disordered viscoelastic rod networks

Disordered networks of viscoelastic rods, with their large number of tuning parameters, offer the potential to be optimized for a wide range of mechanical properties. A central question of active exploration is how materials can self-tune their parameters and thereby adapt to unknown environments without a centralized computing system. Here, we focus on their energy dissipation, which is governed by resonance modes and thus highly sensitive to the frequency, network topology, and directions of excitations. Using a spectral method for linear truss networks with a continuum description of rods, we examine how truss network materials can learn to maximize dissipation and perform other tasks such as cloak specific regions under given frequencies and excitation locations. We explore a broad range of material cost constraints, and study how the networks evolve towards topologically distinct configurations. These results have implications for biomimetic devices and sound-absorbing materials.