Computational design synthesis of multi-state, 4D printed, architected active composites

Abstract 4D printing, the process of additively manufacturing active materials, enables the assembly-free creation of stimuli-responsive systems that would usually require sensors, actuators, structural components and control systems. Computational design methods have been proposed to help designers navigate the complex relations between material distribution and structural behavior to design shape morphing structures. However, they either neglect mechanical loads or demonstrate only small shape changes. This intersection is of particular relevance since large shape changes are often essential to change the function of a design, e.g. to change the shape so that a structure can form a hook that can be mechanically loaded. In this paper, a computational design method is proposed that can design structures that undergo two targeted shape changes, one in response to environmental actuation and a subsequent one in response to mechanical loading. The foundation for this approach is a library of parametrized unit cell (UC) architectures that have adjustable mechanical properties and stimulus responses that strongly differ from those of the constituent materials. A custom genetic algorithm is used to optimize a design composed of UCs such that the resulting structure shows the desired behavior. The method is applied to three case studies, demonstrating that design solutions are found that not only deform in response to environmental actuation, but also show a targeted shape change in response to subsequent mechanical loading. Overall, the method makes the technological potential of 4D printing more available to designers by enabling a means for computational, systematic exploration of the vast design space.