|
| Predicting tensile behavior of additively manufactured copper alloys by convolutional neural network |
| Received:July 22, 2024 Revised:August 13, 2024 |
| View Full Text View/Add Comment Download reader |
| DOI：10.7511/jslx20240722002 |
| KeyWord:additive manufacturing convolutional neural networks crystal plasticity copper alloys |
| Author | Institution |
| 肖庆晖 |
南京航空航天大学 航空航天结构力学及控制全国重点实验室, 南京 |
| 张仁嘉 |
北京宇航系统工程研究所, 北京 |
| 刘士杰 |
北京航天动力研究所 低温液体推进技术实验室, 北京 |
| 胡文轩 |
南京航空航天大学 航空航天结构力学及控制全国重点实验室, 南京 |
| 吕晨晞 |
南京航空航天大学 航空航天结构力学及控制全国重点实验室, 南京 |
| 朱思瑛 |
南京航空航天大学 航空航天结构力学及控制全国重点实验室, 南京 |
| 易敏 |
南京航空航天大学 航空航天结构力学及控制全国重点实验室, 南京 |
|
| Hits: 50 |
| Download times: 52 |
| Abstract: |
| Deep learning has gained significant attention due to its remarkable advantages in handling complex data and tasks,and has been successfully applied in material property prediction.Here,a mathematical framework combining the convolutional neural network (CNN) model with the crystal plasticity finite element method (CPFEM) that considers the strengthening contributions from solid solution,dislocation and grain boundary is proposed to predict the uniaxial tensile mechanical behavior of additively manufactured CuCrZr copper alloy by using its crystallographic texture polar figure,microstructure figure and grain size.The crystal plasticity model parameters are calibrated by using experimental results to verify the model’s accuracy and predictive ability.Subsequently,a series of CPFEM simulations is conducted for different representative volume elements using the calibrated crystal plasticity model.These simulation results are used to train,validat,and test the CNN model.The results show that the CNN model significantly reduces the computation time while guaranteeing the prediction accuracy,demonstrating its promising application in mechanical property prediction. |
|
|
|