Research On Wear Condition Monitoring Of Milling Cutter Under Variable Parameters Using Spindle Current

Impeller is one of the key components in many fields such as petrochemicals,nuclearpower,and aerospace,and its processing quality is directly related to the performance of related equipment.Milling cutter wear is an unavoidable tool degradation phenomenon caused by mechanical,thermal,chemical,and abrasive particles acting together on the cutting edge,which seriously affects the machining quality of the impeller and the production cost.Accurately and quickly identifying the state of wear of the cutter has important significance for improving the machining quality of the impeller,saving the production cost and ensuring the performance of the relevant equipment.The method of monitoring wear condition of milling cutter using spindle current is mainly researched in this paper,and the main contents are as follows:(1)In view of the disadvantages of the force signal monitoring of the wear state of the milling cutter,the feasibility of monitoring the spindle current instead of the milling force is demonstrated from the theoretical and experimental perspectives,respectively.Theoretically,according to the spindle dynamics model and the milling processing characteristics,the relationship between the RMS of the spindle current and the tangential milling force is established to demonstrate the feasibility of the scheme.In the experimental perspective,the similarity of the average trend vector of force signal and current signal with tool wear is measured by the cosine similarity.The results show that they have a significant linear correlation and the feasibility of the spindle current instead of the milling force to monitor the wear state is proved.(2)Due to the shortcomings of the cutting force coefficient as a monitoring index,a cutting current model was established based on the spindle drive model and the cutting force model,and the wear state of the milling cutter was identified by using the cutting current coefficient and the degree of deviation of the cutting current.The experimental results show that there is a significant linear correlation between the cutting force coefficient and the cutting current coefficient.The cutting current coefficient can be used to identify the wear state instead of the cutting force coefficient.Meanwhile,for the lack of directly using the cutting coefficient,the deviation degree of the cutting current is used instead.The experimental result shows that this method can effectively monitor the wear state of the milling cutter.(3)To address the limitation of the model parameters-based cutter wear monitoring method,using order spectrum and convolutional neural network(CNN)for wear condition monitoring is proposed to more fully consider the machining factor effect of tool wear monitoring.The experimental results show that the method can identify the milling cutter wear state with an accuracy of up to 99.12% and is independent of the machining parameters.At the same time,the superiority of this method in the recognition accuracy is proved by compared with other classification algorithms.(4)Based on Lab VIEW and Python,the real-time monitoring system of milling cutter wear condition is developed.Current sensors and NI 9234 acquisition cards are used to collect and store the current signals.The milling cutter condition monitoring methods based on model parameters and convolutional neural network are integrated into the system.Finally,the validity and practicability of the system are verified based on the experimental data.