Abstract
The inverse design of soft actuators refers to a design methodology that derives the actuator's geometric and material parameters based on the desired target shape. This approach offers an efficient solution for developing soft actuators with complex geometries, enabling the precise matching of intricate deformation targets. Furthermore, it facilitates personalized customization and design optimization for complex application scenarios. In this article, we present a novel inverse design method for three-dimensional (3D) variable curvature pneumatic soft actuators and incorporate buckling deformation to improve shape matching. By analyzing variable cross-sectional actuators, we demonstrate that cross-sectional curvature is determined solely by its shape, enabling specific curvatures to be achieved by adjusting the cross-sectional geometry. To process input 3D point coordinates and calculate curve curvatures, we introduce cubic spline interpolation and the three-point circle method. Additionally, a minimum discretization method is proposed to identify the helical axis for helical curves. Finally, based on the actuator's deformation model, we develop an inverse design program to achieve the desired actuator shape. Finite element simulations and experimental validation confirm the effectiveness of this approach, demonstrating that the fabricated actuators can precisely match complex 3D variable curvature shapes. The proposed approach has broad potential applications in situations that require specific, repetitive use, such as specific object grasping and customized medical rehabilitation devices.