Research On Thermal Error Compensation Of Motorized Spindle And Fault Detection Method Of Important Parts

The motorized spindle technology of machining centers has gained increasing attention with the rapid development of the manufacturing industry.Both domestic and foreign scholars have conducted extensive research on the vibration and thermodynamic aspects of motorized spindles.As a crucial component of machining centers,motorized spindles are utilized not only in high-speed machining but also in precision machining.Nevertheless,the high integration and sealing of motorized spindles have led to thermal errors and failures of critical parts.To address the thermal deformation of the motorized spindle shaft and common faults in its crucial components in machining centers,we propose a real-time thermal error compensation calculation method.This method uses a thermal radiation sensor to collect temperature changes in the internal shaft of the motorized spindle and a fault diagnosis method for important parts.For thermal error compensation,we derive the calculation model of axial thermal deformation of the shaft using numerical analysis and basic thermodynamic knowledge through an internal thermal radiation temperature measurement node.We implement improved convolutional neural networks with multiple convolution kernel feature extraction for fault warning and detection of critical parts.Our approach utilizes Labview,a laboratory virtual instrument engineering platform,for data acquisition,processing,communication with the lower computer,calling deep learning models,and other functions.The proposed thermal error compensation method effectively improves the machining accuracy of the machining center,while fault detection enables timely fault warning and stop loss.Additionally,specific faults of critical parts can be identified for easy maintenance.The main contents of this paper are as follows :(1)To determine the internal working conditions of the motorized spindle,the mechanical structure and heat source inside are analyzed.Labview is utilized to select data acquisition,data storage,and external communication.Data acquisition is completed either by a data acquisition card or an improved SPI bus.Communication with PLC is achieved using the OPC module,and the encapsulated deep learning model is directly called by DLHUB.The collected and processed data are stored in the capital or database for various applications through program segments.(2)Using the thermal radiation full radiation temperature measurement method,the motorized spindle’s structure and temperature measurement nodes distribution were analyzed.Based on this information,a simplified model of the motorized spindle was created and a thermal deformation compensation calculation formula was derived.Labview software was utilized to design the control system and simulate analog signals.The simulation results confirmed the feasibility of the theory and control system.(3)The total radiation temperature measurement method can indirectly indicate the working condition of crucial components through temperature measurement nodes.By monitoring the temperature of these nodes,initial assessments of potential failures can be made.In the event of an anomaly,a convolutional neural network with adaptive fusion of multiple convolution kernel feature extraction is utilized to classify specific faults.The control system is designed using Labview software and simulated using analog signals.The simulation results confirm the feasibility of the control system.To verify the principle of fault detection based on full radiation temperature measurement,experiments are conducted.