Time-varying process model for intelligent prediction of strain and temperature in micro-turning process

Time-varying process models for micro-machining processes are important as they aid in control of machining parameters. In this research, a state-space-based process model for the temperature and strain generated near the cutting edge of the tool tip is identified using system identification approach. Fiber Bragg grating sensors were placed rigidly near the cutting edge of the tool tip in a micro-turning setup. Subsequently, micro-turning operations were carried out on aluminum and mild steel. The computer numerically controlled program was such that the machining parameters (feed velocity, depth of cut and RPM) change with machining time. The time-varying machining parameters act as inputs to the model, and the dynamic values of strain and temperature serve as model output. A state-space model was generated using the experimental data. Subsequently, a Kalman filter was used to intelligently predict the values of strain and temperature at the cutting edge of tool tip in advance using the model parameters identified by state-space modeling. Experimental results confirm that the time-varying model and the Kalman filter proposed in this research are effective in predicting the strain and temperature in advance with high accuracy. The maximum error in prediction of temperature was 0.4 degrees C, whereas for strain prediction, the maximum error was 0.3 mu is an element of.