Development of Predictive Model for Surface Roughness Using Artificial Neural Networks

Abstract

AbstractThe need for quality products has been a constant driving force for manufacturing industries. The surface properties are the determinant of the product quality. Surface roughness prediction is now an area of interest in the machining industry. Feed rate, speed of cutting and depth of cut are some of the parameters that influence the prediction of surface roughness. The combined effect of all the three parameters influences the surface roughness to a much significant extent. Data driven prediction is the way ahead. In this study, an artificial neural network is developed fusing the speed of cutting, feed rate and depth of cut. The ANN model is trained using the experimental data already present in the research papers for the prediction as well as optimisation of parameters in CNC lathe for the least possible value of surface roughness of mild steel using statistical techniques and regression models. Further, the ANN model is validated on the basis of two other unseen sets of experimental data on mild steel. From the validation, it has been found that prediction of surface roughness by the ANN has higher accuracy as compared to other existing methods.KeywordsMild steelSurface roughnessArtificial neural networks