The hot deformation behavior of Mn-Ni-Mo system nuclear power steel was conducted at 950-1250℃ and strain rates of 0.01-10 s-1. Combining Arrhenius hyperbolic sine constitutive equation, the response mechanism of activation energyQ, structural factorA, materials constantsn andα on the strain were obtained through multiple linear regression analysis, and the Arrhenius model of flow stress with variable parameters was established. At the same time, the artificial neural network (ANN) model was developed based on the true stress-strain curves, where the inputs parameters were deformation temperature (T), strain rate ((ε)), strain ((ε)), and flow stress (σ) was the output parameter. The results show that the ANN model is more accurate in predicting the flow stress, and the average relative error is 2.53%. Based on the dynamic material model (DMM), the processing map of the studied alloy at strain 0.9 is constructed to recognize optimum hot deformation process parameters: deformation temperate ranges of 950-1250℃ and strain rates of 0.06-0.3 s-1 with a peak power dissipate efficiency (η) of about 0.54, deformation temperate ranges of 1100- 1250℃and strain rates of 0.3-1 s-1 with a peak power dissipate efficiency (η) of about 0.44.%在变形温度为950~1250℃、变形速率为0.01~10 s-1的条件下对Mn-Ni-Mo系核电用钢进行高温热压缩实验.结合Arrhenius双曲正弦本构方程,通过多元线性回归分析获得热激活能Q、结构因子A及材料常数n和α对应变的响应规律,从而建立流变应力与应变量、温度和应变速率之间的变参数Arrhenius本构模型.同时,基于真应力-应变曲线,建立输入参数为温度(T)、变形速率((ε))、应变((ε))和输出参数为流变应力(σ)的神经网络预测模型(ANN).研究结果表明:神经网络模型(ANN)的预测精度更高,其预测流变应力的平均相对误差为1.31%.根据动态材料模型理论(DMM),构建并分析合金在应变为0.9时的热加工图,确定了最佳热变形工艺参数,即当变形温度为950~1250℃,应变速率为0.06~0.3 s-1时,峰值功率耗散系数(η)约为0.54;当变形温度为1100~1250℃,应变速率为0.3~1 s-1时,峰值功率耗散系数(η)约为0.44.
展开▼
