Li Chuan, Chen Shuang, Du Shiyu, Yu Juhong, Zhang Yiming
School of Mechanical and Electrical Engineering, Jiangxi University of Science and Technology, Ganzhou 341000, China.
Zhejiang Key Laboratory of Data-Driven High-Safety Energy Materials and Applications, Ningbo Key Laboratory of Special Energy Materialss and Chemistry, Ningbo Institute of Materials Technology and Engineering, Chinese Academy of Sciences, Ningbo 315201, China.
Materials (Basel). 2025 Jun 25;18(13):3007. doi: 10.3390/ma18133007.
Numerical simulation is a vital tool in the development of FeCrAl alloy cladding tubes, with its reliability closely tied to the predictive accuracy of the thermal deformation constitutive model used. In this study, hot compression tests on 0Cr23Al5 alloy were conducted using a Gleeble-3800 thermal compression testing machine (Dynamic Systems Inc., located in Albany, NY, USA), across a temperature range of 850-1050 °C and a strain rate range of 0.1-10 s. Based on the data obtained, both the Arrhenius constitutive model and the artificial neural network (ANN) model were developed. The ANN model demonstrated significantly superior predictive accuracy, with an average absolute relative error (AARE) of only 0.70% and a root mean square error (RMSE) of 1.99 MPa, compared to the Arrhenius model (AARE of 4.30% and RMSE of 14.47 MPa). Further validation via the VUHARD user subroutine in ABAQUS revealed that the ANN model has good applicability and reliability in numerical simulations, with its predicted flow stress showing high consistency with the experimental data. The ANN model developed in this study can effectively predict the rheological stress of FeCrAl alloys during hot deformation. It provides methodological support for high-fidelity constitutive modeling of the flow stress of FeCrAl alloys and offers a reliable constitutive model for simulating the thermomechanical load response behavior of FeCrAl alloys.
数值模拟是FeCrAl合金包覆管开发中的重要工具，其可靠性与所使用的热变形本构模型的预测精度密切相关。在本研究中，使用Gleeble - 3800热压缩试验机（位于美国纽约奥尔巴尼的Dynamic Systems Inc.）对0Cr23Al5合金进行了热压缩试验，试验温度范围为850 - 1050°C，应变速率范围为0.1 - 10 s⁻¹。基于所获得的数据，开发了阿伦尼乌斯本构模型和人工神经网络（ANN）模型。与阿伦尼乌斯模型（平均绝对相对误差（AARE）为4.30%，均方根误差（RMSE）为14.47 MPa）相比，ANN模型显示出显著更高的预测精度，其平均绝对相对误差仅为0.70%，均方根误差为1.99 MPa。通过ABAQUS中的VUHARD用户子程序进行的进一步验证表明，ANN模型在数值模拟中具有良好的适用性和可靠性，其预测的流动应力与实验数据显示出高度一致性。本研究中开发的ANN模型能够有效地预测FeCrAl合金在热变形过程中的流变应力。它为FeCrAl合金流动应力的高保真本构建模提供了方法支持，并为模拟FeCrAl合金的热机械载荷响应行为提供了可靠的本构模型。
