Research On Temperature Field Prediction And Thermal Error Modeling Of Variable Pressure Preload Motorized Spindle

High-speed motorized spindle has integrated structure,compact structure and high processing efficiency make it used in a variety of machining environments.With the continuous processing time,the heat inside the motorized spindle is difficult to discharge and cause accumulation,resulting in thermal elongation at the shaft end,thus reducing the machining accuracy and efficiency.The proper adjustment of preload is the benign pursuit of different machining environment,the internal temperature field and shaft end thermal elongation will also change.For variable pressure preload motorized spindle,the real-time change of temperature field after changing preload becomes the focus of machining,and the shaft end thermal displacement is also used as the basis of machining compensation.Therefore,the research on temperature field prediction and thermal error modeling of variable pressure preload motorized spindle is carried out.Firstly,the components of high-speed motorized spindle are introduced.For each part of the structural design,including the selection of spindle motor,spindle bearing layout and type selection,spindle cooling strategy,and spindle tool change structure of the working principle of analysis.By discussing the relationship between bearing preload and stiffness,the composition and working principle of variable pressure preload module are described.Secondly,the temperature and thermal displacement acquisition platform of the variable pressure preload motorized spindle was built by using the variable pressure preload motorized spindle as the carrier.Under the experimental conditions of the same speed and different preload,the data of temperature change and shaft end thermal elongation of key components of motorized spindle were collected to provide experimental reference for the temperature field prediction and thermal error modeling of motorized spindle.Thirdly,on the basis of heat transfer,the heat generation and heat transfer modes of variable pressure preloaded motorized spindle were analyzed,and the heat transfer coefficient between heat generation and internal parts was calculated.According to the temperature field simulated by finite element method and the experimental data of the same working conditions,the heat transfer coefficient was optimized by gradient descent method,and the temperature field of the variable pressure preload motorized spindle was predicted.By comparing the temperature of key components before and after optimization,it is found that the temperature field prediction method based on gradient descent method to optimize the heat transfer coefficient has better prediction accuracy and real-time performance.Finally,on the basis of experimental data,the attention mechanism module is optimized to improve the time convolution network,the motorized spindle based variable pressure preload thermal error prediction model,and with the time convolution network were compared with both short-term and long-term memory of artificial neural network,the results show that based on attention mechanism to improve the time convolution model of higher prediction accuracy and robustness.