A methodology for the multi-resolution simulation of grinding wheel surface

A new analysis method, multi-resolution analysis which is based on wavelet transform is proposed in the study of a grinding wheel surface. The main advantages of the wavelet transform over the traditional signal processing techniques are its space-frequency localization and multi-scale view of the components of the signal. Using wavelet transformations, we can effectively eliminate the measurement noise in the wheel measurement process, and decompose the measured profile according to different scales. The wheel surface can be reconstructed from the multi-resolution profile using the combination of different scales according to practical needs, and particular to this paper, the characteristic wavelength of the wheel can be obtained from the multi-resolution study of the wheel. The characteristic wavelength associated with their characteristic grain can be used in the prediction of ground surface roughness. The wavelet transformation method can also be used in the data compression of the measurement data, which provides an efficient method for the modeling of the grinding wheel surfaces. The analysis and simulation results can be applied to the grinding process simulation and grinding process control. © 1997 Elsevier Science S.A