机器手采摘苹果抓取损伤机理有限元分析及验证

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

为了减少机器人自动采摘过程中夹持器抓取苹果时的碰撞、挤压损伤,对抓取过程中苹果与不同指面类型夹持器手指接触时果实内部的应力变化进行研究。通过压缩试验后计算得到了苹果果皮、果肉和果核3个不同部分的力学参数,建立了单个苹果的3层实体力学模型。将苹果果皮、果肉和果核3部分弹性模量分别取多次试验的平均值,通过ANSYS软件建立了苹果所对应的有限元模型,模拟苹果与平面和弧面手指的接触过程,进而得到苹果果皮、果肉和果核3部分的节点Von Mises应力云图。结果显示,加载过程中,果皮处的应力最大,果肉处的应力次之,但由于果肉破坏应力较小,果肉最易受到损伤;同时,当加载力相同时,弧面手指比平面手指对苹果各部分的作用应力要小,苹果的形变也较小,当加载力分别为5、20、35、50 N时,平面手指对苹果所造成的形变量分别比弧面手指大6.7%、12.1%、12.4%、14.5%,因此,弧面手指对果皮内部造成机械损伤的概率相对较小。最后,利用自行研制的采摘机器人2弧面手指夹持器苹果实物抓取损伤试验验证了所研究方法的有效性。该研究结果可以实现苹果损伤的较准确预测和评估,并为采摘机器人末端夹持器减损结构优化设计提供了一定参考依据。

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.