ANCF-based dynamic modeling of variable curvature soft pneumatic actuators with experimental verifications

Abstract

Accurate and computationally efficient models of soft pneumatic actuators are crucial for utilizing their compliance in various fields. However, existing research primarily relies on the piecewise constant curvature assumption or the quasi-static assumption, only valid in limited situations. In this paper, we present a dynamic model based on absolute nodal coordinate formulation (ANCF) that simultaneously accounts for variable curvature deformation and dynamic properties. To this end, deformed configurations of soft pneumatic actuators are firstly discretized into ANCF-based beam elements. Based on this parameterization method, the dynamic model is derived by the principle of virtual work. After identifying model parameters, Newmark algorithm is utilized to solve the dynamic model in real-time, averagely consuming 6.76 s of a 10 s simulation. The derived dynamic model is experimental verified using a soft pneumatic actuator. The experimental results demonstrate that the maximum simulation errors of the tip remain below 2.5% of the actuator’s length when the actuator is subjected to various pressure and tip loads. In addition, the overshoot behavior and period of vibration in the oscillations are also predicted by the dynamic model. Moreover, the dynamic model exhibits an average 46.53% reduction in simulation error compared with the static ANCF-based model. Overall, this work paves the way to a deeper insight to dynamic motion analysis of soft pneumatic actuators.