Research On Milling Tool Breakage Monitoring Based On Acoustic Emission

Real-time tool condition monitoring is crucial to realize automaticmanufacturing. Effective tool breakage monitoring plays a significant andpractical role in enhancing the productive efficiency, reducing theproduction cost and improving the product quality. In order to realize thisgoal, carry out researches in this paper as follow:Based on reviewing the development and current situation ofmethods and technologies in sensing, signal processing and staterecognition for tool failure monitoring, A method of acousticemission-base milling tool breakage is presented.Firstly, through a LabVIEW-based experimental platform, anorthogonal cutting test is carried out and normal and breakage AE signalsunder different cutting conditions are acquired. Secondly, the AE signalsare analysed in time domain, frequency domain and wavelettransformation. Based on wavelet transformation, three kinds of featuresare abstracted, including energy, standard deviation and kurtosis. Themethod is demonstrated through comparing three kinds of signals-normal,random and breakage. And then the features are optimized by analysingthe sensitivity to tool condition and cutting parameters. Thirdly, the feature difference of AE signal during cut-in and cut-out process arestudied in this paper, and its cause is also discussed. At last, the theory ofsupport vector machine and its application in milling tool breakagemonitoring is deliberated. According to the test, the method is provedsuccessful. And also the optimised combination features of energy,standard and kurtosis are proved more successful than single features intool breakage monitoring.