Stability Analysis Of Regenerative Turning Chatter And Research On Preddiction Method

Chatter vibrations are present in almost all cutting operations,which can cause a series of problems such as dimensional accuracy reduce of the workpiece and the tool's damage.In this paper,the non-linear dynamics knowledge is applied to establish the model of regenerative turning chatter and analyse the stability of the chatter system.In addition,experimental method is applied to research the turning chatter signal.The main work of this paper is to predict the regenerative turning chatter,mainly including the following aspects:In chapter 1,the mechanism,model and feature extraction of cutting chatter are reviewed to summarize the current research status and development trend of cutting chatter prediction.Therefore,the research contents of this paper are given.In chapter 2,the dynamic model of regenerative turning chatter is constructed.Stability lobe diagram of regenerative turning chatter system is obtained and its stability is analyzed.It is found that increasing the equivalent damping ratio,natural frequency and equivalent stiffness of the cutting system can increase the stability of Stability lobe diagram,while increasing the cutting stiffness coefficient,the direction coefficient and the cutting overlap coefficient will reduce the stability of Stability lobe diagram.In chapter 3,a regenerative turning chatter experimental platform is set.The vibration acceleration is slected as the target signal for predicting chatter and then the experimental program of turning chatter is developed and the signal from stable turning to turning chatter is obtained.Finally,the chatter signal is analyzed in the time domain and the frequency domain.In chapter 4,two methods are proposed to predict the turning chatter.The former uses the Lempel-Ziv complexity algorithm to divide the complex interval of the different turning stages to carry out the chatter prediction.Latter,Elman neural network is applied to the prediction of turning chatter.It is found that the mean square error of the neural network can distinguish turning chatter.So,prediction function is constructed and its prediction accuracy is 92%.In chapter 5,Empirical Mode Decomposition is applied to extract feature of turning chatter.Then feature is analysed by Hilbert-Huang Transform(HHT)Result shows that the frequency band is narrowed and the energy is concentrated in the narrow frequency band when the chatter is generated in the time spectrum.So the turning chatter prediction is carried out.In chapter 6,the research work of this dissertation was summarized and prospected.