Estimating the effects of heat treatment on aluminum alloy with artificial neural networks

Abstract

In this study, after T6 heat treatment was applied to two different kinds of aluminum alloy (AA) 6061 formed with different compositions, their mechanical properties and microstructures were observed and then the mechanical properties were estimated by using artificial neural networks (ANNs). First, AA 6061 aluminum alloys were cast. After extrusion, profiles were taken into solution at 530 degrees C for 2 h. Then, aluminum was aged at different temperatures and different times, and elongation, yield and tensile strengths and hardness were obtained. After T6 heat treatment was applied to alloys, it was observed whether the mechanical properties had changed over time. With ANNs, the results of the long and costly process can be achieved in a shorter time and at a lower cost. In the second stage, the yield strength, tensile strength, hardness and elongation of the material were estimated with different ANN models. The model with the lowest error among the different ANN models was chosen and used in the study. The mechanical properties of AA 6061 obtained by experiments were used in ANN training, and ANN estimation results were compared with experimental results. The developed ANN model estimated the mechanical properties of AA 6061 by 90%.