Design Space Enumeration for Shape-Changing Structures Using a Bioinspired Heuristic

Abstract Shape-changing structures combine soft actuators with passive rigid members in a planar or spatial domain. These designs have traditionally been posed as multimaterial topology optimization problems with a predefined objective, which is computationally intensive for design space exploration. A thorough exploration of the feasible solution concepts greatly enhances the designer’s ability to select and refine solutions. In this article, we introduce an enumeration-based design framework for planar shape-changing structures, drawing inspiration from the biological arrangement of bipennate muscles. The concept of bipennates is generalized to a triangle whose edges consist of exactly two muscles (soft actuators) and one bone (passive members). We enumerate all possible configurations of a closed planar topology with its interior triangulated with the two-muscle, one-bone (M2B) units. A pseudo-rigid body model (PRBM) is used to analyze the deformations and identify feasible designs. Physics-based guidelines are used to further refine the selected design by either eliminating certain muscles or toggling a muscle to a bone. We demonstrate the efficacy of this framework in quickly exploring the space of all feasible configurations, including the solutions obtained from an evolutionary optimization framework.