Designing a fiber-reinforced actuator to perform a specific task can be a challenging problem and often involves a lot of trial and error. To speed up the process and find the optimal actuator design for the task at hand, we can use mathematical models which relate actuator geometry and materials to actuator output motion.
For example, we may wish to design an actuator which upon pressurization replicates the motion of the thumb or index finger. To achieve this, we put (1) the kinematics of the desired motion and (2) analytical models of actuator motions into an optimization algorithm. The algorithm then outputs a recipe for designing an actuator which will achieve the required motion.
More details on our design strategy can be found in the following paper:
F. Connolly, C.J. Walsh and K. Bertoldi, Automatic Design of Fiber-Reinforced Soft Actuators for Trajectory Matching, PNAS 2016, doi:10.1073/pnas.1615140114
The Matlab scripts which implement the design strategy can be downloaded below. The main file is run_optimization.m, and the other files are functions called by the main script.