Analysis and validation for mechanical damage of apple by gripper in harvesting robot based on finite element method

Abstract: In order to reduce the impact and crush damage of apple during the process of robot apple picking, the variation of stress on and inside the apple skin in the contact process of apple with different type's robot end-effectors was investigated mainly using the finite element analysis software ANSYS. Considering the mechanical characteristics difference of apple tissues, in this paper, apple component was simplified into skin, cortex and core, and each part contained the same material with same parameters such as elastic modulus, Poisson's ratio, stress intensity and density attributes, and then apple tissue was simplified to linear elastic material. Through the method combination of calculation and experiments, the mechanical parameters of apple skin, cortex and core, such as elastic modulus, breaking stress and poisson ratio, were obtained. And then the solid mechanics model constituted by skin, cortex and core of a single apple was established. On the basis above, taking the arithmetic mean of the elastic modulus for 3 parts of apple, the finite element model of apple was set up by ANSYS. Considering the stability of apple grasping and adopting the method to get the apple using cut knife to cut off the stem for picking robot, the load force by robot finger was set as 5，20，35 and 50 N respectively to simulate the contract process of apple with plane finger and arc-shaped finger. Based on the simulation above the Von Mises stress nephograms of apple skin, cortex and core were also obtained under different load force by different finger types. The Von Mises stress nephograms showed that when load force was 5，20，35 and 50 N respectively, the stress in skin caused by plane finger was 0.0258，0.0898，0.1559 and 0.3647 MPa, respectively, the stress in cortex was 0.0184, 0.0654, 0.1347 and 0.3245 MPa, respectively, the stress in core was 0.0017, 0.0058, 0.0136 and 0.0498 MPa, respectively. The stress caused by arc-shaped finger was relatively small. As for the deformation quantity, when the loading force on apple was 5, 20, 35 and 50 N respectively, the deformation quantity caused by arc-shaped finger was 0.0289, 0.104, 0.181, 0.564 mm respectively, and the deformation quantity caused by plane finger was 6.7%, 12.1%, 12.4% and 14.5% larger than that caused by arc-shaped finger respectively. The results showed that the stress of apple skin was the largest and stress of cortex takes the second place both for plane finger and arc-shaped finger. But the cortex was easier to get damaged as the result of its relatively small breaking stress. Meanwhile, under the precondition that the load force was same, compared to plane finger, the deformation and the stress of each part caused by arc-shaped finger was relatively small. That was to say, the probability to cause internal mechanical damage of the apple by arc-shaped finger was relatively small. The experiment for apple grasping damage of robot gripper validated the reliability of simulated results with relative error smaller than 10% between simulated and measured values. The results are helpful to make fast and accurate prediction along with evaluation of apple damage, and also provide valuble information for optimization of the injury-reduce devices for apple mechanization and automation harvesting robot as well as processing equipments.