Soft Modeling Approach in Predicting Surface Roughness, Temperature, Cutting Forces in Hard Turning Process Using Artificial Neural Network: An Empirical Study

Hard Turning means turning of steel having hardness more than 50 HRC (Rockwell Hardness C). It is used for metal removal from hardened steels or difficult to machine steels. In this research hard turning of EN-31 steel (tool steel) 48 HRC was done with Carbon Nano Tubes based coated insert and two tools. Taguchi L27 orthogonal array was used for design of experiments. The input parameters taken in this research were cutting speed, feed, depth of cut, type of coating and cutting conditions. The output responses were surface roughness, temperature and cutting forces. 5-5-1 Feed forward artificial neural network was used in simulation of actual cutting conditions and prediction of responses before actual machining was done. The simulation results of ANN were in unison to those predicted by actual experimental procedure. It was concluded that relative error in values of surface finish, temperature and cutting forces as predicted with ANN versus those achieved during experimental procedure were 1.04%, 2.889% and 1.802% respectively, which were very close to actual values.