Prediction of surface roughness in hard turning under high pressure coolant using Artificial Neural Network

摘要: In this study, an artificial neural network (ANN) based predictive model of average surface roughness in turning hardened EN 24T steel has been presented. The prediction was performed by using Neural Network Tool Box 7 MATLAB R2015a for different levels cutting speed, feed rate, material hardness and conditions. To be specific the dry high pressure coolant (HPC) jet environments were explored as experimental runs determined full factorial design experiment. Afterward 3-n-1, 3-n-2 4-n-1 ANN architectures trained utilizing Levenberg–Marquardt (LM), Bayesian regularization (BR) scaled conjugate gradient (SCG) algorithms, evaluated on lowest root mean square error (RMSE). 3-10-1 3-4-2 models, BR, revealed RMSE. A good fit models established regression coefficients higher than 0.997. At last, behavior respect speed-feed-hardness HPC conditions analyzed. reduced efficient cooling lubrication whereas induced due to restraining force against tool imposed force.