Ti微合金化集装箱用钢流变应力模型  

Flow stress model of Ti microalloyed container steel

作　　者：姜珊 马鑫 周晓光[1] 刘振宇[1] JIANG Shan;MA Xin;ZHOU Xiaoguang;LIU Zhenyu(State Key Laboratory of Rolling and Automation,Northeasten University,Shenyang 110819,China)

机构地区：[1]东北大学轧制技术及连轧自动化国家重点实验室,辽宁沈阳110819

出　　处：《轧钢》2022年第6期153-158,共6页Steel Rolling

基　　金：国家重点研发计划项目(2017YFB0304104)。

摘　　要：为精准预测轧制力并合理制定轧制工艺,采用MMS-300热模拟试验机对Ti微合金化集装箱用钢进行单道次压缩实验,研究其高温变形行为,得到了实验钢的应力-应变曲线,建立了实验钢基于物理冶金学、神经网络以及两者相结合的流变应力模型。结果表明:基于物理冶金学的流变应力模型有较好的外延性但是精度偏低;基于神经网络的流变应力模型精度极高但是外延扩展能力较差;而基于物理冶金学和神经网络相结合的流变应力模型,在符合物理冶金学规律的同时具有较高的精度和外延性,可以更好地预测实验钢的流变应力。To accurately predict the rolling force and reasonably formulate the rolling process, the high temperature deformation behavior of Ti microalloyed container steel was studied by single pass compression test on MMS-300 thermal simulation test machine. The stress-strain curves of the experimental steel were obtained. The flow stress models of experimental steel based on physical metallurgy, neural network and their combination were established. The results show that the flow stress model based on physical metallurgy has good extrapolation ability, but the prediction accuracy is relatively low. The flow stress model based on neural network has high accuracy but poor generalization ability. The flow stress model based on physical metallurgy and neural network not only comply with the law of physical metallurgy, but also has high accuracy and extrapolation ability, which can better predict the flow stress of experimental steel.

关 键 词：集装箱用钢 流变应力 物理冶金学 模型 神经网络 

分 类 号：TG335[金属学及工艺—金属压力加工] TG142.1[一般工业技术—材料科学与工程]