An empirical model of soft bellows actuator.

Abstract

Soft robotics has emerged as a highly promising field, particularly for handling interactions in unstructured environments such as food factories and agricultural warehouses. This potential is largely attributed to the inherent flexibility and compliance of soft robots. A critical aspect in the development of these robots lies in the selection and utilization of appropriate soft actuators and materials. Nevertheless, the modeling of soft robots presents considerable challenges owing to their intricate properties and continuum nature. In this article, we focus on the design and modeling of a three dimensional (3D) printed soft bellows actuator. The primary objective is to assess its efficacy in creating suitable soft grippers for handling various practical products. We propose an empirical model to predict the output forces of the soft bellows actuator. This model comprehensively integrates parameters such as bellows geometry and material properties, thereby providing valuable insights for the actuator's design and control. To ascertain the precision of our model, we conducted a series of finite element simulations considering different designed parameters of the bellows, and performed experimental validations using 3D printed bellows actuators. The empirical model demonstrated high accuracy in predicting the output forces of the bellows actuator, with average absolute and relative errors of N and , respectively. As an application, a robotic gripper with two parallel bellows actuators was developed, and its grasping force was validated using the empirical model. Building on this, a robotic gripper incorporating three bellows actuators was designed and fabricated based on the empirical model, and pick-and-place experiments were effectively conducted for handling a range of products.