Robotic Slicing of Fruits and Vegetables: Modeling the Effects of Fracture Toughness and Knife Geometry

Abstract

Slicing is an important skill for a robot to learn as it is more efficient and results in less deformation in comparison with cutting by pressing. Cutting experiments with foods have indicated that the ease of slicing is caused by a decrease in fracture toughness. In this paper, we formally characterize this decrease based on the work needed to maintain the critical strain for fracture. Forces generating fracture and deformation and overcoming friction are predicted using the finite element method (FEM) and based on fracture mechanics. Extending our previous work [1] on cutting by pressing with a straight knife edge, we model general slicing and knife geometry (i.e., a curved edge). Experiments over potatoes and eggplants have demonstrated the accurate modeling of the overall cutting force during slicing, which could be leveraged for control of cutting by the robot to demonstrate human-level skills in the near future.