| The CNC machine tools are the main machine tool of the equipment manufacturing industry,which technical quality is the symbol of a country's comprehensive national strength.With the progress of industrial technology,the functions and performance of domestic CNC machine tools are continuously improved.However,in terms of reliability,a large gap with other international technologies remains.The motorized spindle is the key functional component of the CNC machine tool,and its main function is to provide cutting torque for the latter.However,in the process of long-term service,due to the wear of parts and other reasons,the service performance of motorized spindles gradually declines,seriously affecting the machining accuracy of CNC machine tools.Therefore,research on the degradation analysis and life prediction of motorized spindles is of great significance in achieving preventive maintenance and reliability improvement of CNC machine tools.As an important support of motorized spindle,the performance degradation of the bearings caused by wear and contact fatigue is the main reason for the degradation of motorized spindle accuracy and reliability.Accurately mastering the performance degradation law of motorized spindle bearing is an important basis for improving its reliability level.However,in the actual machining process,the motorized spindle is subjected to the combined action of multiple directions,time-varying static,and dynamic loads.Furthermore,due to the complex assembly relationship of parts in the motorized spindle,there is a strong coupling relationship between the loads,making it difficult to establish the mapping relationship between the motorized spindle load and the degradation of bearing performance.In addition,as a long-life and high reliability product,the motorized spindle of an CNC machine tool cannot obtain enough degradation fault samples in a limited time,which is difficult to accurately analyze the performance degradation process of motorized spindle bearing by using conventional data-driven method.To solve the above problems,the physical model with less sample demand can be used to establish the relationship between motorized spindle and bearing from the perspective of mechanism.In this paper,the research method of degradation analysis and life prediction of motorized spindle bearing based on dynamics is proposed.The purpose is to construct the force transfer law between structures with dynamic model,and analyze the degradation and life of motorized spindle by establishing the physical model of bearing performance degradation.The main work and innovative achievements of this paper are as follows:(1)The nonlinear dynamic model of motorized spindle is established.The structure of motorized spindle was simplified based on analyzing the structure and function of motorized spindles,and the nonlinear dynamic model of motorized spindle is established that is used to establish the relationship between dynamic load of motorized spindle and the bearing load.Among them,the discrete dynamic model of spindle multi node is established by using the finite element model of Timoshenko beam element model.Based on the analysis of the non-uniform contact pressure of the joint surface under the action of radial force,the dynamic model of the nonlinear contact stiffness of the joint surface of motorized spindles is proposed and verified using the finite element method.Based on Hertz contact theory,the dynamic model of bearing multi-degree of freedom nonlinear support stiffness is established.The dynamic model of motorized spindle is verified by modal test and loading test.The nonlinear dynamic model of motorized spindle is verified by modal test and loading test.(2)A method to evaluate the bearing life of motorized spindles based on dynamics is proposed.The dynamic cutting force under actual working conditions is predicted by the milling dynamic model,which is used as the input force of the motorized spindle dynamic model and the time history of bearing dynamic load is obtained.The rain flow counting method and the Miner linear fatigue accumulation method are used to convert the bearing dynamic load into equivalent load,facilitating the calculation and evaluation of the theoretical life of each bearing and bearing group of motorized spindle,and the relationship between the dynamic load of motorized spindle and the equivalent load of bearing is established.The effects of bearing preload,cutting force,broach force,and rotating speed on the life of each bearing and bearing group of motorized spindle are analyzed.(3)This thesis proposes a bearing damage state evaluation method based on dynamic interactive multi-model particle filter.In particular,a dynamic model consistent with the degradation mechanism of each stage of bearing damage is established by analyzing the performance degradation process of bearing damage,taking roughness,crack growth coefficient,crack length,and crack depth as parameters.Among them,a dynamic model with crack length parameter growth and constant crack depth parameter is proposed to explain the decreasing trend of vibration amplitude caused by smooth crack edge during bearing crack propagation,thus solving the problem of unclear mechanism during bearing degradation.Then,an interactive multi model particle filter method based on dynamics is also proposed.Taking the root mean square of bearing vibration as the observation vector,the state of dynamic model parameters of bearing damage degradation is evaluated,and the real-time update of dynamic model parameters is realized.Moreover,taking the crack length in the evaluation results as the threshold,the remaining useful life of bearing damage is predicted via exponential fitting.The feasibility and accuracy of the method are investigated based on the bearing test data set of Xi'an Jiaotong University.(4)In order to solve the problem of variable working condition of motorized spindle,the accelerated degradation tests of the motorized spindles are carried out under constant stress loading and the step-down stress loading,whcih consistency of the degradation mechanism of the motorized spindle test scheme by bearing contact load limit.Next,the degradation test data of motorized spindles is denoised by combining nonlocal mean filtering and variational mode decomposition method,after which the frequency domain analysis method is used to diagnose the damage of motorized spindle bearing.Furthermore,the degradation mechanism of motorized spindle precision degradation according to the degradation data,and the degradation stages are divided.(5)A method to evaluate bearing performance degradation under variable load conditions of motorized spindles based on dynamics is proposed,and the remaining useful life of the bearing and the accuracy degradation of the motorized spindle are predicted.Based on Archard wear theory,the parameter state vector of wear degradation process of motorized spindle bearing is established.Taking the motorized spindle degradation data as the observation vector,the parameter states of bearing wear degradation and damage degradation is estimated by the interactive multi-model particle filter method based on dynamics,which analyze the performance degradation of motorized spindle bearing by the characterization data.Taking the bearing wear depth as the life threshold,the remaining life of motorized spindle bearing is predicted.By predicting the wear depth of each bearing,the degradation trend of the mean radial displacement of the motorized spindle is predicted,and the accuracy degradation of the motorized spindle is analyzed with the mean radial displacement at the end of the test as the accuracy index.This thesis propose a method for the degradation analysis and life prediction of motorized spindle bearing based on dynamics,which mainly studies the bearing life evaluation method of motorized spindle in the design stage and the performance degradation analysis and life prediction method of motorized spindle bearing in the service stage,so as to provide technical support for the reliability design and reliability guarantee of motorized spindle. |
PDF Full Text Request