Chatter detection using vibrational and acoustic emissions in turning process

Abstract

Chatter vibration is a longstanding problem for machining industry. It lowers both surface finish quality and life of cutting tool. In this study, real time chatter vibration detection system is designed and tested experimentally. A traditional lathe, equipped with two accelerometers placed on cutting tool in two orthogonal coordinates and a microphone, is used for machining a steel rod in face turning operation. During cutting process analog data of accelerometers and microphone is acquired by data acquisition card that is connected to a computer. A Matlab graphical user interface that has the capability of acquiring and processing accelerometer and microphone data simultaneously is generated. Processing data is done by means of calculating coherences between accelerometer vs. accelerometer data and accelerometer vs. microphone data and comparison of coherences to threshold values that are tuned experimentally. Coherence function determines if two data signal come from the same source or not. Coherence functions value of data pair reaches unity at the chatter frequency. Three chatter conditions are described as stable, slight and severe chatter conditions. If both coherences exceed threshold values, severe chatter occurs. If one of the coherences exceeds and other doesn't than slight chatter occurs. If both of the coherences don't exceed threshold values then the system is stable. Depending on the chatter condition Matlab graphical user interface alerts the operator by indicator that changes color for different conditions. Graphical user interface also shows Fast Fourier Transform of the microphone data. Real time detection allows operator to change lathe speed or cutting depth in case of chatter occurs during the cutting process. By preventing chatter vibration, surface quality of the work-piece and life time of the machine tool is increased.