论文部分内容阅读
在变形温度为1 173~1 323 K,应变速率0.01~5.00 s-1的条件下,采用Gleeble-1500热模拟实验机对45Cr4NiMoV钢进行等温压缩试验,获得了其高温流变行为曲线。以所得热压缩实验数据为基础,建立了BP人工神经网络模型,结果表明模型预测值与实验值吻合良好,预测值和实验值的相关系数为0.999 7,平均误差为0.04%,即该模型具有较高的预测精度,且模型预测值可以追踪热压缩在宽泛变形条件下的高温变形行为,包括加工硬化阶段和应变软化阶段,故ANN模型可以描述45Cr4NiMoV钢在热变形过程中的高度非线性的流变行为。
Under the conditions of deformation temperature of 1 173 ~ 1 323 K and strain rate of 0.01 ~ 5.00 s-1, the isothermal compression test of 45Cr4NiMoV steel was carried out by Gleeble-1500 thermal simulation machine, and its high temperature rheological behavior curve was obtained. Based on the experimental data obtained, the artificial neural network model of BP was established. The results show that the model predictive value is in good agreement with the experimental data. The correlation coefficient between predicted and experimental values is 0.999 7 with an average error of 0.04% High predictive accuracy, and the model predictive value can trace the hot deformation behavior of hot compression under a wide range of deformation conditions, including the work-hardening stage and the strain softening stage. Therefore, the ANN model can describe the high nonlinearity of the 45Cr4NiMoV steel during hot deformation Rheological behavior.