Wear And Stress Analysis Of Cycloid Outer Rotor Shaping Die Based On Neural Network

Cycloid rotor is the key component of oil pump in automobile engine lubrication system.Most of them are mass-produced by powder metallurgy method,which has the advantages of high efficiency,high precision and high performance.Powder metallurgy is a process technology that uses metal powder(or a mixture of metal powder and nonmetal powder)as raw material to prepare various metal materials and products by pressing,sintering and post-processing.The post-shaping process based on forming and sintering is a key step to ensure product quality.During the post-shaping process,the plastic deformation resistance is generated on the surface of the sintered metal material,and the mold will produce a large stress concentration,resulting in mold wear(about70 %),deformation and fatigue fracture failure.Therefore,optimizing the shaping process parameters is of great significance to reduce the wear and stress of the mold and improve the service life of the mold.In order to consider the wear and stress of the mold at the same time,this paper takes the powder metallurgy cycloid outer rotor shaping mold as an example.Based on the Archard wear model,the finite element simulation method is used to analyze the influence of relevant process parameters on the wear and stress of the mold during the cycloid rotor shaping process.At the same time,the BP neural network prediction model is established with the orthogonal experimental simulation data as the sample,and the relevant parameters affecting the wear and stress of the mold are taken as variables.Combined with the multi-objective particle swarm optimization algorithm,the process parameters are optimized to improve the service life of the mold.Firstly,the forming technology of iron-based cycloid rotor is studied,including material preparation,pressing,sintering and shaping.The shaping method and related mathematical models of the cycloid outer rotor are analyzed(elastic-plastic model and wear model),which provides theoretical support for the subsequent numerical simulation of the extrusion shaping process of the cycloid outer rotor.Secondly,the three-dimensional software UG was used to complete the modeling and assembly of the sintered parts and the shaping mold of the cycloid outer rotor.After setting the relevant simulation parameters in the DEFORM-3D finite element software,the simulation calculation of the extrusion shaping process of the cycloid outer rotor was completed.Under the premise of ensuring the shaping effect,the wear and stress of the main mold were analyzed.Based on Archard ’s wear model and die structure design in actual production,four influencing parameters related to the wear and stress of the female die were determined : die pressing speed,friction coefficient,initial hardness and fillet radius of the female die.The influence of four process parameters on the wear and stress of the female die was analyzed by single factor variable method.Based on the established orthogonal experimental simulation results,the range method is used to determine the influence weight of each process parameter on the wear and stress of the female mold.The order of influence on the wear of the female mold is determined as follows : initial hardness of the mold > fillet radius of the female mold > friction coefficient > pressing speed of the mold.The order of influence on the stress of the female mold is : friction coefficient > fillet radius of the female mold > pressing speed of the mold > initial hardness of the mold.Finally,based on the orthogonal test simulation data,the BP neural network is used to construct the prediction model of the wear and stress of the die,and the process parameters are optimized by combining with the multi-objective particle swarm optimization algorithm.The results show that the average relative errors between the predicted values and the simulated values of the mold wear depth and stress of the neural network are only 4.12% and 4.99%,and the maximum relative errors are-5.93%and 7.98%,which are less than 10%,indicating that the BP neural network model has a good prediction effect.The optimal parameter combination is obtained : the mold pressing speed is 1.95mm/s,the friction coefficient is 0.122,the initial hardness of the mold is 59.63 HRC,and the radius of the die fillet is 4.85 mm.At this time,the wear depth of the mold is 1.71310-5mm,and the mold stress is 2036 MPa.The relative error between the numerical simulation results and the numerical simulation results is only2.60% and 4.41%,both less than 5%,indicating that the multi-objective particle swarm optimization algorithm has high accuracy of parameter optimization.Through experimental verification,it is found that the error between the actual wear amount of the mold and the optimization result of the particle swarm optimization algorithm is only 2.95%,indicating that the optimized parameters can effectively reduce the wear of the mold and improve the stress concentration of the mold and improve the service life of the mold.