Research On Temperature Measuring Point Optimization And Thermal Error Modeling Of High Speed Motorized Spindle

High-speed motorized spindle system is the main component and the largest heat source of high-end precision CNC machine tools and its thermal deformation has a significant adverse impact on machining accuracy,which is the focus of the research on thermal problems of high-end precision CNC machine tools.Different from the traditional spindle system,the high-speed motorized spindle produces a lot of heat due to copper loss,iron loss and mechanical loss when the current flows into the built-in motor.At the same time,due to the friction between the rolling body and the bearing inner and outer ring,a large amount of heat is generated near the bearing contact point.In addition,because of its complex structure,poor cooling conditions and complex internal coupling,the thermal problem of high-speed motorized spindle is more serious.Therefore,it is urgent to study the thermal error of the motorized spindle system of precision CNC machine tools.In this paper,the finite element analysis method is used to analyze the thermal performance of the high-speed motorized spindle,including the establishment of the three-dimensional geometric model of the motorized spindle,and the calculation of the thermal load and convective heat transfer coefficient of the motorized spindle system based on the heat transfer theory.The engineering parameters and meshing methods of the motorized spindle system are determined,and finally the finite element model of the motorized spindle is established,and the overall temperature field distribution of the motorized spindle system is obtained.It provides a theoretical basis for thermal error detection of motorized spindle.Based on the thermal performance analysis of motorized spindle,an experimental platform for detecting temperature and thermal error of high-speed motorized spindle is designed and built.The temperature data are collected by multi-channel dynamic data acquisition instrument and dynamic signal acquisition and analysis system,and the axial thermal drift data are collected by laser displacement sensor and supporting software.Thus a complete dynamic detection system of electric spindle thermal error is established.It provides data basis for temperature measurement point optimization and thermal error modeling.According to the temperature and thermal drift data obtained by the motorized spindle temperature and thermal error detection system,the fuzzy C-means clustering(FCM)algorithm is used to cluster the temperature measuring points,so as to reduce the multiple correlations of different temperature measuring points.The grey comprehensive correlation degree in the grey correlation analysis(GRA)is used to rank the correlation degree between temperature and axial thermal drift,so as to obtain the optimal temperature measuring point.In view of the above temperature measurement point optimization method,the adaptive neural fuzzy inference system(ANFIS)is used to model the thermal error of the motorized spindle,and compared with the traditional artificial neural network model,the results show that the ANFIS model has higher prediction accuracy and robustness.