Modelling and On-Line Monitoring of Machined Surface in Turning Operations

Machined surface profile and roughness are important parameters in evaluating the quality of a machining operation. They are resulted from the transformation of the complex tool-workpiece displacements involving the dynamics of the machine tool mechanical system, cutting process, and cutting motions. The focus of this study is the fundamental understanding of the surface profile formation during turning and development of regression and neural network (NN) models of surface roughness incorporating the effects of cutting parameters and tool-workpiece displacements. Also, a bifurcated optoelectrical transducer was developed for on-line monitoring of surface roughness based on the scattering of laser beams from machined surface. The feasibility of on-line monitoring was studied by comparing with actual roughness as well as the prediction results of the regression and NN models.