Analytical Stiffness Modeling and Experimental Validation for a Pneumatic Artificial Muscle

Abstract

Since the introduction of assistive technologies for enhancing human mobility, there has been a high demand for compact, lightweight, powerful and energy efficient actuator. The Pneumatic Artificial Muscle (PAM) is a distinctive compliant pneumatic actuator, which has properties similar to the biological skeletal muscle, making it a great candidate for applications in human mobility assistive devices. Whereas the PAMs can be used actively or passively, until now, it has been mostly used in active applications to power various mechanisms, such as robotic arms, most notably the Shadow Robot Company Dextrous hand. For those applications, static and dynamic models of PAM have been developed by researchers to fairly accurately predict the muscle-force carrying capabilities and muscle contraction distance behavior, respectively. However, limited passive models have been developed, with results varying from acceptable to poor in terms of accuracy. Recognizing the significance of characterizing the PAM passive behavior, especially for legged locomotion application, this paper proposes a PAM stiffness model that is based on Newtonian mechanics and considers geometric, mechanical and material properties of the muscle. The proposed stiffness model is experimentally validated for a wide range of operating conditions.