Simulation of Solid Meal Digestion in a Soft Gastric Robot using SOFA

The gastric simulator SoGut is a soft robotic stomach to imitate the gastric motility in the human stomach. The simulator contains seven ring-shaped chambers along the J-shape stomach to perform contractions by applying pneumatic actuation. Experiments conducted previously for gastric emptying focused on liquid meals instead of solid meals, which have higher stiffness and require higher break-down forces. In this paper, we evaluate the performance of a simulation model of SoGut when it grinds solid food contents that are beads (diameter 1.27 cm) containing agar gel and water at various mixing ratios. The simulation model is based on Simulation Open Framework Architecture (SOFA), with the advantages of finite element method modelling and real-time simulation. First, the relationships between pressure inputs and contraction ratios of the modelling stomach are identified. Next, a series of peristaltic contractions is successfully simulated. An average error at 0.03 of contraction ratio is found, except chamber #5 that has an error at 0.12. Later, beads with different stiffness are placed in the center of chambers and tested to investigate the amount of forces applied to these contents. The stiffness of beads depends on the mixing ratio of agar gel and water, and the motions of beads are constrained on the cross-section plane of the chambers in order to avoid food contents escaping from the contracting regions. The average forces from different chambers applied on the contents in the antrum part of the simulated stomach are less than 0.9 N, which corresponds to the findings in the human stomach reported by literature. Meanwhile, it is also found that the strength of forces generated from the chambers of SoGut depends on the stiffness of the food contents. Additionally, the results of the simulation indicate that when the initial diameters of chambers increase, the provided average forces reduce.

I. INTRODUCTION

Most common robots observed in daily life can be classified as rigid robots, containing components such as joints made of steel and aluminum alloys, wheels made of hard plastic and rigid electromagnetic parts. This type of robot provides precise positioning and reasonable payloads because of its rigid body and joints. However, these rigid systems are inadequate to handle complex unpredictable environments with limited degree of freedom, and have safe interaction with human.