Finite Element Analysis And Nondestructive Testing Of The Ripe Navel Orange Damaged By Grasping

The vast area of southern Jiangxi has excellent soil quality,most of which belong to red soil.It is very suitable for the planting of high quality navel orange.The Gannan navel orange produced from the area is big and tasted sweet,juicy,appearance of the orange is orange red,bright and beautiful.It is the famous fruit products of Jiangxi.However,there are many hills and mountains in the south of Jiangxi Province,and the elevation of the terrain is high,so the navel orange planting on the slope makes it difficult to pick.In the past,the picking of navel orange mainly depended on artificial,but artificial picking was time-consuming.It is difficult to pick at night and in bad weather,and the fruit is easy to deteriorate because of improper picking,which affected the sale of navel orange in south Jiangxi.The dangerous coefficient of navel orange picking is increased due to topographic reasons,which is prone to safety accidents.Therefore,the work of picking navel orange urgently needs to realize intelligentized and mechanized operation.In this paper,the damage of navel orange caused by grasping force of manipulator picking navel orange is studied.First,the mechanical properties of navel oranges were analyzed.Then,the numerical value of the elastic modulus and other mechanical parameters of navel oranges as the initial conditions were used for the finite element simulation analysis of the navel orange.Finally,the results of the stress damage of the navel orange obtained by finite element simulation were tested by NIR.The qualitative and quantitative relationship between the stress of navel oranges and its injury is obtained.The basic data and models of relationship between stress and injury of navel orange are provided for the determination of its gripping force when that navel orange pic robot is developed for the future.The main research contents and results are as follows:(1)the peel and pulp of Gan Nan navel orange were studied.Three different compressible directions X、Y、Z were identified on the pulp of navel orange.using texture analyzer,compression of navel orange with three different compression distances,and the average value of compression force in each direction is calculated as the true value.Doing the navel orange peel experiment by stretching the pericarp,Texture tester stretch pericarp until peel breaks,the tensile force and absolute stretch length were measured when the pericarp edge was ruptured,and the instantaneous tensile force of breakage are used as the parameter to calculation.The modulus of elasticity of the flesh in navel orange is 0.33N/mm~2.The failure stress is 0.086 N/mm~2.The Poisson’s ratio is 0.48.The modulus of elasticity of the peel of the navel orange is 12.11N/mm~2.The failure stress is 1.29 N/mm~2.The Poisson’s ratio is 0.39.(2)the elastic modulus,failure stress and Poisson’s ratio obtained by Texture are used as parameters for finite element analysis of navel orange.The simulation of the mechanical process of picking navel orange is used by using finite element software.Set up the geometric model,define material properties,grid division,establish contact pair,Impose constraints and loads.When the applied load is less than 25N the navel orange is undamaged.When the load is 25N-50N the navel orange is slightly damaged.When the load is 50N-75N the navel orange is moderately damaged.When the applied load is greater than100N,the navel orange is severely damaged.It is deformed under pressure,and the damage is serious.(3)In order to verify the finite element analysis results,The simulation results were predicted by near infrared,and the damage spectrum of navel orange was modeled.In a diffuse near-infrared manner,Establish a qualitative discriminant model of damaged navel orange based on ELM algorithm,The results showed that the accuracy of ELM model in judging injured navel orange was 100%,There was no miscarriage of justice,establish Quantitative Detection Model of navel Orange damage on PLS algorithm,For four categories of undamaged navel oranges and mild,moderate and severe damaged navel oranges,the correlation coefficients of modeling and prediction are 0.91 and 0.85,the root mean square error is 0.240 and 0.315.Using damaged navel orange external samples to verify the accuracy of the model,The correct rate of ELM model recognition is 98.9%,Only 2 moderate injury navel oranges identified as severe injury navel orange.Using navel orange external samples to verify the accuracy of the PLS model,There was no miscarriage of justice,The classification accuracy is 100%.