Computational framework for bioinspired, mechanically programmable soft robots actuated by negative pressure

Designing soft, pneumatically driven robots has attracted increasing interest for their capability to mimic or assist functions of biological systems due to their inherent mechanical compliance. Soft machines powered by negative pressure can obviate disadvantages seen in positive pressure-driven actuators, such as too high stresses upon actuation, resulting in device failure. The actuation of vacuum channels embedded in elastomeric matrices presents potential for creating biomimetic controlled complex motion, but so far, analytical tools that efficiently predict their deformation upon vacuum are lacking. This limits the development of negative pressure-driven soft machines capable of meeting desired performances. Therefore, in this work, we introduce a 2D mathematical framework that precisely captures the deformation behavior of arrays of lens-shaped voids and the surrounding elastomeric matrix and define design rules that guarantee a stable operation. The accuracy of the framework is validated using finite element analysis and physical prototypes. Finally, we apply the framework to build a soft walking robot capable of cheetah-like locomotion, which we characterize in terms of walking speed as a function of actuation vacuum and frequency. The developed model unlocks the potential of negative pressure-driven soft machines and opens up the possibility of automated processes for mechanically programmed biologically-inspired designs.