Design of resilient structures by randomization and bistability

TL;DR

This study explores how randomization and bistability in lattice structures can enhance impact resilience by increasing energy dissipation and controlling damage spread, offering new design strategies for protective structures.

Contribution

It introduces a novel approach combining randomization, topology, and bistability to improve impact resilience in elastic-brittle lattices, with experimental validation.

Findings

Random removal of 8% links maximizes energy dissipation.

Randomized bistable lattices outperform traditional designs in energy absorption.

Nonperiodic topologies and bistability reduce damage connectivity.

Abstract

This paper examines various ways of improving the impact resilience of protective structures. Such structures' purpose is to dissipate an impact's energy while avoiding cracking and failure. We have tested the reaction of plane elastic-brittle lattices to an impulse. Four topologies are compared: periodic triangular, square, and hexagonal topologies, and aperiodic Penrose topology. Then, structures with random variations of the links' stiffness, node positions, and random holes are compared. Combinations of these random factors are also considered, as well as the resilience of bistable elastic-brittle lattices with sacrificial links. Several parameters are introduced to measure the structural resilience of the compared designs: (i) the amount of dissipated impact energy, (ii) the size of broken clusters of links, and (iii) the spread of damage. The results suggest new routes for rationally designing protective structures using nonperiodic topology, bistability, and structural randomness. In particular, we find that some quantities of interest can be maximized by tuning the randomized design appropriately -- for example, randomly removing 8\% of links maximizes energy dissipation. We also find that randomization of bistable lattices can offer superior energy dissipation while reducing the connectivity between broken clusters of links.