Prediction of the hot deformation constitutive behavior of superalloy IN-718 using an artificial neural network
In order to optimize the hot working process of superalloy IN-718, an artificial neural network (ANN) model is developed to predict the hot deformation constitutive behaviour of superalloy IN-718 using the published experimental data. The inputs of the neural network are strain, strain rate and temperature, whereas flow stress is the output. A three layer back-propagation (BP) neural network with a single hidden layer consisting of 18 neurons is established, and the resilient learning algorithm is applied to train the artificial neural network. The predicted stress-strain curves agree well with experimental results, which indicate the developed ANN model can accurately and efficiently predict the hot deformation constitutive behavior of superalloy IN-718 in the wide range of deformation temperature and strain rate. In addition, the relative importance of process variables for flow stress predictions of superalloy IN-718 is investigated based on Garsons algorithm. The results from both ANN model and experiments show that the deformation temperature and strain rate have the pronounced effects on the flow stress of superalloy IN-718, while the strain is not significant.
Guoliang Ji Fuguo Li
School of Materials Science and Engineering, Henan Polytechnic University, Jiaozuo 454000, China School of Materials Science and Engineering, Northwestern Polytechnical University, Xi’an 710072, Ch
国际会议
桂林
英文
1-8
2010-11-16（万方平台首次上网日期，不代表论文的发表时间）