核电用高温合金Inconel 617B高温变形本构模型

High Temperature Deformation Constitutive Model of Superalloy Inconel 617B for Nuclear Power

传统的Arrhenius本构模型因对流变应力的预测精度较低而存在局限性。通过应变量补偿构建高温合金Inconel617B的唯象型本构模型,同时构建合金的BP人工神经网络(Back propagation-artificial neural network,BP-ANN)型本构模型。结果表明:合金的流变应力对变形温度、变形速率和应变量较为敏感。唯象型本构模型预测流变应力的相对误差δ为9.020%,相关系数R为0.9814;而BP-ANN型本构模型(最佳结构确定为3×20×1)预测结果的相对误差仅为1.527%,R达到0.9989,可更加准确地预测Inconel 617B的流变应力。说明BP-ANN构建本构模型可提高预测精度,更为合理地描述Inconel 617B高温合金的流变应力与变形参数间的本构关系。

The conventional Arrhenius constitutive model is limited by its low accuracy in predicting flow stress.The phenomenological Arrhenius constitutive model coupling true strain is constructed for superalloy Inconel 617 B.Furthermore,the BP-ANN(Back propagation-artificial neural network)constitutive model is constructed.The results show that the flow stress of the alloy is sensitive to the deformation temperature,deformation rate and true strain.The relative error(δ)of the flow stress predicted by the phenomenological model is 9.020%and the correlation coefficient(R)is 0.9814,while the relative error of the BP-ANN model(the optimum structure is determined as 3×20×1)is only 1.527%and R reaches 0.9989.The flow stress of Inconel 617 B is more accurately predicted by the BP-ANN model.Using BP-ANN to construct constitutive model can improve the prediction accuracy,which can describe the constitutive relationship between flow stress and deformation parameters of superalloy Inconel 617 B more rationally.