Prediction of machinability of AA7075 using artificial neural networks

Abstract

Abstract In this study, machinability behavior of AA7075 aged at different time lengths was examined experimentally and by using artificial neural network prediction model. The hardness values were measured after the heat treatment processes. Homogenized reference samples and aged samples were machined by turning processes. On the one hand, the wear occurring on the cutting edge during machining, and the cutting forces depending on cutting speed and surface roughness were investigated. Surface roughness values for each reference material and aged sample were measured using processing parameters. Acquired surface roughness values formed a surface roughness prediction model by using artificial neural networks. The results showed that the surface roughness of the samples decreases while the cutting speed of the lathe increases. In the prediction model formed by using surface roughness acquired after the machinability tests, cutting force, cutting speed and aging process were used as input parameters. Surface roughness as a result of machinability tests were used as output parameters of the proposed prediction model. High coefficient of determination, R2 rate, obtained in the formed prediction model showed that the model is successful in the prediction of surface roughness.