Prediction of forging force and barreling behavior in isothermal hot forging of AlCuMgPb aluminum alloy using artificial neural network

Journal of Advanced Materials and Processing

مقاله 5، دوره 6، شماره 1، خرداد 2018، صفحه 29-45   

نوع مقاله: Research Paper

نویسندگان

hamidreza Rezaei Ashtiani^{* 1}؛ p shahsavari²

¹Mechanical Engineering dep., arak university of technology, iran

²mechanical eng. dep., arak university of technology, Arak, Iran

چکیده

In the present investigation, an artificial neural network (ANN) model is developed to predict the isothermal hot forging behavior of AlCuMgPb aluminum alloy. The inputs of the ANN are deformation temperature, frictional factor, ram velocity and displacement whereas the forging force, barreling parameter and final shape are considered as the output variable. The developed feed-forward back-propagation ANN model is trained with Leven berg–Marquardt learning algorithm. Since the finite element (FE) simulation of the process is a time-consuming procedure, the ANN has been designed and the outputs of the FE simulation of the hot forging are used for training the network and then, the network is employed for prediction of the behavior of the output parameters during the isothermal forging process. Experimental data is compared with the FE predictions to verify the model accuracy. The performance of the ANN model is evaluated using a wide variety of standard statistical indices. Results show that the ANN model can efficiently and accurately predict isothermal hot forging behavior of AlCuMgPb alloy. Finally the extrapolation ability and noise sensitivity of the ANN model are also investigated. It is found that the extrapolation ability is very high in the proximity of the training domain, and the noise tolerance ability very robust.

کلیدواژه‌ها

AlCuMgPb aluminum alloy؛ isothermal hot forging؛ artificial neural network؛ finite element simulation