Thermal Characteristic Analysis And Temperature Prediction Of Highspeed Motorized Spindle

As the core component of NC machine tools,the performance of the motorized spindle determines the development of high-speed machine tools.The built-in motor and bearings of the motorized spindle are unavoidable to generate heat.The resulting temperature rise will cause thermal deformation of the spindle,which will affect the operating state and machining accuracy of spindle.Therefore,in order to ensure the processing accuracy of the motorized spindle at high speed,it is necessary to conduct a detailed analysis of the thermal characteristics of the motorized spindle and accurately predict the temperature rise.In this paper,the experimental motorized spindle 170SD24Q15 is taken as the research object,a thermodynamic model of the motorized spindle considering the surface roughness of stator and rotor is proposed,and the influence of roughness on it is studied.The simulation model of the motorized spindle is established by using the finite element method.The thermal experiment of the motorized spindle is designed.Due to the strong non-linearity and complexity of the internal temperature change of the motorized spindle,different artificial intelligence algorithms are put forward to predict the internal temperature of the motorized spindle.The main research contents of this article include:1)A thermodynamic model of motorized spindle considering the surface roughness of stator and rotor is proposed.During the high-speed operation of the motorized spindle,the surface roughness of the stator and rotor has different effects on the heat transfer coefficient and air friction loss of the stator and rotor.The higher th e speed is,the more obvious the effect is.It can be seen from the simulation results:when the speed is constant,the air friction loss increases as the surface roughness of the fixed rotor increases;when the surface roughness of the fixed rotor is cons tant,the air friction loss increases as the speed increases;With the increase of surface roughness,the coefficient shows a trend of decreasing first and then increasing.2)A finite element thermal analysis model of the motorized spindle is established.The finite element model of the motorized spindle is built in ANSYS Workbench.The relevant material properties are set,the heat generation rate of the two major heat sources,the traditional heat transfer coefficient,and the heat transfer coefficient of the stator and rotor under different roughness are loaded on the finite element model.The distribution laws of the steady-state temperature field of the motorized spindle under two kinds of considering the surface roughness are obtained.When the surface roughness of the rotor is 1 ?m,its temperature is closer to the experimental value.The smaller the surface roughness of the fixed rotor,the smaller the air friction loss.Under the effect of the fixed rotor gap heat transfer coefficient,more heat can be taken away.3)An internal temperature prediction model of motorized spindle based on BP neural network is established.On the basis of the design of thermal experiments to obtain the temperature data of the motorized spindle,the BP neural network is used to study the temperature rise of the motorized spindle with complicated changes by its powerful nonlinear problem processing ability.Based on traditional method of selecting hidden layer neuron nodes,the value of (6 is expanded to get the best number of hidden layer neuron nodes,and a better topology structure is constructed.A BP neural network prediction model of motorized spindle internal temperature is established,and the temperature of the outer ring of front bearing,rear bearing block and stator end of the motorized spindle is predicted.The prediction results show that the BP neural network can predict the internal temperature of the motorized spindle.4)An internal temperature prediction method for motorized spindle based on improved BP neural network is proposed.Particle Swarm Optimization(PSO)has fast convergence speed and strong global search ability.Combining the advantages of PSO with BP neural network,a PSO-BP model is established,which realizes the prediction of the internal temperature of the motorized spindle.In order to improve the prediction accuracy,a mutation operator is introduced into the PSO algorithm to optimize the PSO-BP model,and an improved PSO-BP model for predicting the internal temperature of the motorized spindle is established.The results show that the improved PSO-BP prediction model has high robustness,smaller errors of the predicted and expected values of the temperature measurement points of the motorized spindle,and higher prediction accuracy than the PSO-BP and BP model.