Research On Thermal Error Analysis And Modeling Of Machine Tool

The research into the thermal error of machine tools has been of practical value as the importance of machining accuracy has been recognized,which accounts for up to 80% of the total error sources.To improve the performance of machine tools,this paper,based on theoretical analysis,finite element simulation,and experiment,launches the research on the thermal error of machine tool in combination with the major special project "ADI series CNC lathe intelligent technology development and industrialization application" of Henan Provincial Department of Science and Technology.Firstly,a finite element thermal simulation model of the machine tool spindle system is proposed,which introduces cutting heat,improves heat transfer and displacement constraints.The spindle system is the key point in the study of the thermal characteristics of the machine tool.A three-dimensional model is established and simplified reasonably on the basis of the structure and mechanical indexes of the spindle and bearings,converted into a finite element model.The main heat transfer modes and the types of finite element thermal analysis of the machine tool are determined through the heat transfer theory and the structure of the machine tool.The main heat sources of the machine tool spindle system are the friction heat of the spindle bearings and the cutting heat during the processing of parts.In the light of the mathematical conditions of finite element thermal analysis,the specific calorific value is calculated,and the setting of the system convection heat transfer conditions is improved according to the diameter of the spindle,and the temperature field simulation environment is established.Based on the improvement of the displacement constraint conditions,the thermal structure coupling analysis is carried out,and the temperature change law of the spindle system of the machine tool and the thermal displacement of the machine tool under the steady temperature field are obtained.The axial and radial thermal displacement are 5.402?m and 9.137?m respectively.Then,a simple and effective experimental method for measuring the temperature and thermal error of the machine tool is carried out.Combined with the results of finite element thermal analysis of the spindle system of the machine tool,the predecessors' research on the arrangement of temperature measurement points and the structure of the machine tool parts,the specific locations of 16 temperature measurement points and 2 thermal displacement measurement points for the thermal characteristic experiment were established.By the experimental requirements,the types of temperature sensors and displacement sensors are selected and arranged scientifically,the parameters of the temperature data collector and displacement sensor controller are set,and the temperature and thermal displacement data of the machine tool are observed and obtained through the PC applications.Studying the change rules of temperature and thermal displacement,the time from turning on the machine to reaching the steady heat balance state is about the 7400 s.The axial and radial thermal displacements are 5.748?m?m and 10.009?m respectively under the steady-state temperature field.comparing the results of the finite element thermal analysis and experiment,it can be seen that the temperature error of the finite element thermal analysis of the spindle system is controlled within 2.49% when the machine tool is in the steady-state temperature field.The error between the simulation and the experimental results is 6.019% and 8.712% respectively.The accuracy of the axial thermal deformation of the machine tool is increased by 41.493%,which proves the rationality of the results of the finite element thermal analysis,that is to say,the rationality of the improved thermal boundary conditions of the temperature field of the spindle system according to the heat transfer mechanism of the machine tool.Finally,a thermal error modeling method based on random forest is proposed.According to the optimization strategy,the Pearson correlation coefficient and the fuzzy clustering method are used to determine the best temperature measurement points of the machine tool.In this paper,the existing thermal error modeling methods are compared and analyzed,and applying the random forest algorithm to thermal error modeling of machine tools is proposed.The thermal error prediction model of the machine tool is established by the random forest algorithm,taking the multiple linear regression method as a control group.The results show that the training modelbased on the random forest algorithm has better accuracy in the prediction of axial and radial thermal errors of the machine tool,compared with the traditional multiple linear regression,whose goodness of fit is 15.18% and 4.27% higher than the former.The research paves a foundation for the real-time compensation of thermal error of machine tools as well as provides a reference for the establishment of the best temperature measurement point of other machine tools.