PREDICTION OF SURFACE ROUGHNESS OF Al/SiC COMPOSITE MATERIAL WITH ARTIFICIAL NEURAL NETWORKS

In this study, surface roughness of al/sic composite material depending on the cutting parameters were predicted with high accuracy using approach of artifical neural network. surface roughness values obtained as experimentally result of machining with ticn+tin coated cementide carbide cutting element of al/sic composite material are trained in nine different ann models with feed forward back propogation. the numbers of neuron in network structure of ann models are 3-5-6-1, 3-6-4-1, 3-6-6-1, 3-4-3-5-1, 3-4-5-3-1, 3-6-2-3-1, 3- 7-1, 3-8-1 ve 3-9-1. the values obtained from the ann training and testing were evaluated by applying the statistical analyses that are widely used in ann models. in the face of diffuculty of experimental studies and complexity of the analitical expression, as with many studies, this study also showed that ann is a usable method for predicting the surface roughness value depending on cutting parameters.