钢铁钒钛2025，Vol.46Issue(5)：75-84,10.DOI:10.7513/j.issn.1004-7638.2025.05.008
人工神经网络驱动的P650无磁钻铤用钢高温流变行为研究
ANN-Driven modeling of high-temperature flow behavior in P650 for nonmagnetic drilling collars
王英虎 1程礼梅 2王建强 3王婀娜 2宋令玺 2盛振东2
作者信息
- 1. 成都先进金属材料产业技术研究院股份有限公司,四川成都 610000||北京科技大学国家材料服役安全科学中心,北京 100083
- 2. 成都先进金属材料产业技术研究院股份有限公司,四川成都 610000
- 3. 攀钢集团江油长城特殊钢有限公司,四川江油 621704
- 折叠
摘要
Abstract
High-temperature tensile tests on P650 high-nitrogen steel had been conducted under 1 000-1 150 ℃ and strain rates of 0.01-10 s-1,using a Gleeble-3500 thermomechanical simulator.Based on the obtained stress-strain data,a strain-compensated Arrhenius constitutive model and an artificial neural network(ANN)model were developed,with prediction accuracy evaluated by average absolute relative error,root mean square error,and correlation coefficient.Results demonstrated that the prediction by ANN model with a single hidden layer(17 neurons)achieved high-precision nonlinear mapping between input parameters(temperature,strain rate,strain)and flow stress.Besides,the ANN predic-tions exhibited good agreement with experimental data(r=0.996,EAARE=4.63％,ERMSE=6.721 MPa)com-pared to the Arrhenius model(r=0.975,EAARE=7.94％,ERMSE=16.032 MPa).This study reveals that artifi-cial neural networks can effectively capture constitutive relationship characteristics of complex thermal deformation behaviors,providing an improved strategy for establishing high-accuracy flow stress pre-diction models and optimizing material processing technologies.关键词
人工神经网络/高温流变行为/高氮奥氏体不锈钢/无磁钻铤/流变应力预测Key words
Artificial neural network(ANN)/high-temperature flow behavior/high-nitrogen austenitic stainless steel/nonmagnetic drilling collars/flow stress prediction分类
矿业与冶金引用本文复制引用
王英虎,程礼梅,王建强,王婀娜,宋令玺,盛振东..人工神经网络驱动的P650无磁钻铤用钢高温流变行为研究[J].钢铁钒钛,2025,46(5):75-84,10.
