AerMet100钢热压缩过程流变应力的神经网络预报  

Flow Stress Numerical Prediction of Steel AerMet100 during Hot Compression Deformation Process Based on BP Artificial Neural Network Principle

作　　者：汪向荣[1] 张雁祥[1] 闫牧夫[1] 

机构地区：[1]哈尔滨工业大学材料科学与工程学院,哈尔滨黑龙江150001

出　　处：《热处理技术与装备》2008年第4期10-13,共4页Heat Treatment Technology and Equipment

摘　　要：本文研究了变形温度、应变速率和应变量对固溶和轧制态的AerMet100钢热压缩变形过程流变应力的影响规律。结果表明:变形温度、应变速率和应变量对流变应力有显著影响,预处理方式对流变应力的影响不大。基于人工神经网络原理,建立了流变应力与变形温度、应变速率和应变量的数学模型,由此预报的AerMet100钢热压缩过程流变应力与实测结果相吻合。This paper researches the effects of deformation temperature, strain rate and strata to me flow stress of hot compression deformation for solution - treating and rolling steel AerMet100. The results show that the deformation temperature, strain rate and strain influence significantly on the flow stress, the effect of pretreatment manner to the flow stress is not evident. The mathematical model for predicting the flow stress as a function of deformation temperature, strain rate and strain has been established based on BP artificial neural network principle. The simulated flow stresses of steel AerMet100 agree with the experimental ones under different hot compression deformation.

关 键 词：AERMET100钢 热压缩变形 流变应力 BP神经网络 预报 

分 类 号：TG115.53[金属学及工艺—物理冶金]