Design of the optimum preform for near net shape manufacturing is a crucial step in upsetting process design. In this study, the same is arrived at using artificial neural networks (ANN) considering different unequal interfacial friction conditions between top and bottom die and billet interface. Back propagation neural networks are trained based on finite element analysis results considering four unequal interfacial frictional conditions and varying geometrical and processing parameters, to predict the optimum preform for commercial aluminum. Neural network predictions are verified for three new problems of commercial aluminum and observed that these are in close match with their simulation counterparts.
