Prediction of Cutting Forces during Hard Turning of AISI 52100 Steel Using Regression Analysis and Artificial Neural Networks

The measuring of cutting forces in hrad turning is an important factor. The cutting force of a certain material represents the machinibility of such materail, it also reflects how quickly a tool can wear and also the vibrations that can occur during machining. So it is clear that for condition monitoring aswell as for process optimization, the measurement of cutting forces becomes fundamental. This Paper deals with a prediction of Cutting forces in a Hard turning process of AISI 52100 steel with the help of regression models and artificial neural networks. The objective of this dissertation is to develop artificial neural networks and multiple regression models to predict the cutting forces developed during hard turning of A.I.S.I 52100 steel using PCBN cutting tools. In this research, a hard turning model was developed to evaluate the effect of machining parameters and nose radius of cutting tool on the cutting force. The model was authenticated by comparing the cutting forces with the experimental results. In this ANN was used which is a network of artificial neurons (sometimes called perceptron) inspired by the biological neural networks that constitute human brain. The artificial neuron that constitutes the neural network is a simplified mathematical representation of the real neuron found in human brains. The experiments are planned based on taguchi design and measured cutting forces were compared with the predicted forces in order to validate the feasibility of the proposed design. The percentage contribution of each process parameter had been analyzed using Analysis of Variance (ANOVA).