NUMERICAL PREDICTION OF SURFACE ROUGHNESS IN TURNING OPERATION USING TAGUCHI METHOD AND ARTIFICIAL NEURAL NETWORK

Surface Roughness plays an important role in assessment of surfaces quality; however, minimizing of time and cost consumption in production processes is one of the very important target. In this research, selective cutting conditions and specimens are used in turning operations to produce machined surfaces. Mitutoyo SurfTest-SJ201® is used to obtain surface roughness parameter Ra for the resulted machined surfaces. The Taguchi Method with analysis of variance using Minitab 17® is used as numerical technique to predict Ra for non-machined specimens at suggested cutting conditions. An introduced artificial neural network ANN is modeled using MATLAB 2014a®, to predict Ra at the previous suggested cutting conditions. Comparison between results from turning operations and those obtained from both the Taguchi method and ANN prediction is introduced. Moreover, effect of cutting conditions on Ra is introduced throughout the topography of the machined surfaces.