An efficient short-time Fourier transform algorithm for grinding wheel condition monitoring through acoustic emission

Indirect methods to monitor the surface integrity of grinding wheels by acoustic emission (AE) have been proposed, aiming to ensure their optimal performance. However, the time-frequency analysis of the content of these signals has not been addressed in the literature. AE signal analysis performed only in the frequency domain makes it impossible to locate faults on the grinding wheel surface during the dressing operation and examine the behavior of the frequencies contained in these signals over time. In this regard, the time-frequency analysis of AE signals during dressing through STFT (short-time Fourier transform) can contribute toward the proposal of new monitoring methodologies, thus reflecting the optimization of the grinding process. This paper proposes an algorithm based on the Kaiser window to adjust the STFT parameters to ensure an appropriate balance between time-frequency resolutions. Besides, this algorithm is used to investigate the characteristic frequencies in the aluminum oxide grinding wheel in dressing operation. The results indicate that the spectral content of the AE signals during dressing follows a uniform behavior, but their amplitude changes depending on the characteristics of topography and sharpness of the grinding wheel cutting edges.