Investigation of acoustic emission activities under different grinding conditions using short time Fourier transform

Grinding is the final production process used when sufficient measurement precision and surface quality cannot be achieved by machining methods such as grinding, turning, milling, reaming, etc. This thesis presents the use of short time Fourier transform (STFT) in the analysis of acoustic emission activities detected under different operating conditions (i.e. cutting speed and infeed rate). Experimental data are obtained from a fully automatic plunge-type cylindrical Computer Numerical Control (CNC) grinding machine and conditions of the grinding process in three different cases (i.e. normal grinding, grinding with a higher cutting speed and greater infeed) are monitored using acoustic emission (AE). Detected AE signal is evaluated primarily using time and frequency domain analysis, and the symptoms of different working conditions are determined. Spectrogram shows how the energy content of AE activities is distributed over the time-frequency plane. It has been found that the AE characteristics up to 300kHz provide the greatest contribution to signal energy for all grinding conditions and can be ideal source of data for monitoring process parameters. Moreover, it has been found that any increase in the cutting speed or radial infeed causes an increase in the instantaneous energy obtained from the STFT analysis.