Research On Wear Condition Monitoring Of Multi-sensor Fusion Vibration Drilling

CFRP/aluminum alloy is widely used in the fields of aerospace,automobile industry,ship and so on because of its high specific strength,friction resistance and fatigue resistance.However,due to the different hardness and strength of the two materials,the traditional drilling method is easy to produce burr,tear,delamination and other defects,which is difficult to meet the high-precision requirements of aerospace and other special fields.Axial vibration drilling is a new type of pulse special machining method,which has the advantages of improving the cutting performance,reducing the cutting force and improving the surface quality of the hole.In the process of hole making,the tool wear state directly affects the production efficiency,processing cost and the quality of hinge hole.Drilling is carried out in a semiclosed state.It is impossible to directly observe and monitor the tool wear state in the process of hole making.It is mainly through the experience of operators to judge the tool wear degree and make a tool change plan.However,this method has poor reliability and is difficult to be used in automation equipment.Therefore,the study of tool wear monitoring technology is conducive to the development of automation and intelligence of machine tools.In this paper,the axial vibration drilling device is used as the test platform.Firstly,a multi-sensor fusion bit wear monitoring system is set up.With the drilling force signal,acoustic emission signal and vibration signal as monitoring signals,the tool wear test of CFRP / aluminum alloy workpiece drilling process is carried out by using carbide tools.Secondly,the wavelet threshold method is used to denoise the drilling force signals,acoustic emission signals and vibration signals to reduce the impact of noise.Then,the characteristics of the denoised signals are analyzed.Based on the time-domain analysis method,the mean value,square difference and root mean square value of the signals are calculated.Based on the frequency-domain analysis method,the spectrum structure and power distribution with frequency are obtained.In the time-frequency domain,the signals are analyzed and processed based on the wavelet decomposition method,and the energy ratio coefficients of each frequency band are extracted.The characteristic values of the bit in different wear state are compared and analyzed,and the characteristic values which have strong correlation with the bit wear state are selected as the eigenvectors.Finally,the application effect of BP neural network and support vector machine in vibration drilling bit wear state monitoring is compared and analyzed.The results show that both methods can effectively identify the bit wear state,but SVM has fast convergence speed and high recognition accuracy,which is more suitable for the monitoring of bit wear state in vibration drilling.In this paper,the wear monitoring system of vibration drilling bit based on multi-sensor fusion technology is of great significance for prolonging the service life of bit and improving the surface quality of machining hole.