基于应变补偿和BP神经网络的BT25钛合金本构关系研究

Constitutive relationship research on BT25 titanium alloy based on strain compensation and BP neural network

利用Gleeble-3500热模拟试验机对BT25钛合金在变形温度为950~1100℃,应变速率为0.001~1 s^-1和最大压下率为60%条件下进行热压缩实验,研究了BT25钛合金的热变形行为,并分别建立了α+β两相区和β单相区的基于应变补偿的Arrhenius本构模型,同时构建了基于BP神经网络的本构关系模型。结果表明,合金对变形温度和应变速率较为敏感,其流变应力随变形温度的降低和应变速率的增加而增大。在α+β两相区,基于应变补偿的本构模型的相关系数R为0.981;在β单相区,其相关系数R为0.984;基于BP神经网络的本构模型相关系数R达到0.996,表明BP神经网络模型具有更高的精度,能更好地预测BT25钛合金的高温流变应力。

The hot compression tests were carried out to study the hot deformation behavior of BT25 titanium alloy by Gleeble-3500 thermal simulator at deformation temperature of 950-1100℃,strain rate of 0.001-1 s-1 and maximum reduction of 60%.The Arrhenius constitutive models based on strain compensation ofα+βtwo-phase region andβsingle-phase region were established,and the constitutive relation model based on BP neural network was constructed.The results show that the alloy is sensitive to the deformation temperature and strain rate,and the flow stress of the alloy increases with the decrease of deformation temperature and the increase of strain rate.Inα+βtwo-phase region,the correlation coefficient R of the constitutive model based on strain compensation is 0.981 and inβsingle-phase region,the correlation coefficient R is 0.984.The correlation coefficient R of the constitutive model based on BP neural network is 0.996,which indicates that BP neural network model has higher accuracy and can predict the high temperature rheological stress of BT25 titanium alloy better.