Workpiece surface roughness prediction in grinding process for different disc dressing conditions

Abstract

The surface roughness of workpiece in grinding process is influenced and determined by the disc dressing conditions due to effects of dressing process on the wheel surface topography. In this way, prediction of the surface roughness helps to optimize the disc dressing conditions to improve surface roughness. The objective of this study is to design of a feed forward back propagation neural network (FFBP-NN) for estimation of surface roughness in grinding process using the data generated based on experimental observations when the wheel is dressed using a rotary diamond disc dresser. The input parameters of model are dressing speed ratio, dressing depth and dresser cross-feed rate and output parameter is surface roughness. In the experiment procedure the grinding conditions are constant and only the dressing conditions are varied. The comparison of the predicted values and the experimental data indicates that the predictive model has an acceptable performance to estimation of surface roughness.