基于条件生成对抗网络的高强钢热变形微观组织预测Microstructure Prediction of High-Strength Steel during Hot Deformation Based on Conditional Generative Adversarial Network
张焱城,肖贵乾,孙朝远,周杰,权国政
摘要(Abstract):
针对大型锻件热变形微观组织难以由传统模型直接预测的问题,提出一种以变形温度、应变速率和变形量为连续条件的二值晶界条件生成对抗网络预测方法。基于Gleeble热压缩试验建立包含8个温度、5个应变速率和6个变形量的240幅显微组织图像数据库;经图像变换得到720幅训练样本,并分别设置30幅验证图像和30幅独立测试图像。模型以96维随机噪声和3维工艺条件为输入,通过卷积生成器输出96×128晶界概率图,条件判别器同时学习晶界真实性与工艺条件一致性。结果表明,在840℃的16组工况下,预测平均晶粒尺寸的相对误差为-5.16%~8.34%,绝对相对误差均小于10%,其中13组不超过5%。预测组织在晶界密度、晶粒数量、平均尺度及其随应变速率和变形量的变化趋势方面与试验结果总体一致。结果表明,该方法能够建立热变形参数与晶界图像之间的映射,为高强钢热变形微观组织的快速图像化预测提供一种可行方法。
关键词(KeyWords): 高强钢;热变形;微观组织预测;条件生成对抗网络;晶界图像生成
基金项目(Foundation): 国家科技重大专项(2025ZD1605900/高性能锻造工艺智能生成模型与软件研发及验证);; 德阳市科技计划项目(2025CCZYYS007/梯度异质结构长寿命锻模关键技术及应用)
作者(Author): 张焱城,肖贵乾,孙朝远,周杰,权国政
DOI: 10.14147/j.cnki.51-1396/tg.2026.05.008
参考文献(References):
- [1]Ran Xianzhe,Liu Dong,Li An,et al. Microstructure characterization and mechanical behavior of laser additive manufactured ultrahigh-strength Aer Met100steel[J]. Materials Science and Engineering A,2016,663:69-77.
- [2]Ji Guoliang,Li Fuguo,Li Qinghua,et al. Research on the dynamic recrystallization kinetics of Aer Met100steel[J]. Materials Science and Engineering A,2010,527(9):2350-2355.
- [3]王鑫,董洪波,邹忠波,等.Aer Met100钢的热变形显微组织演变及动态再结晶行为[J].特种铸造及有色合金,2016,36(2):121-125.
- [4]Sun Chaoyuan,Qin Yi,Liu Yang,et al. Research on the hot deformation process of A100 steel based on high-temperature rheological behavior and microstructure[J]. Materials,2024,17(5):991.
- [5]Yuan Zhanwei,Li Fuguo,Qiao Huijuan J,et al.Constitutive flow behavior and hot workability of AerMet100 at elevated temperatures[J]. Journal of Materials Engineering and Performance,2014,23:1981-1999.
- [6]Zhao Zhanglong,Min Xiaonan,Xu Wenxin,et al.Dynamic recrystallization models of Aer Met100 ultrahigh-strength steel during thermo-mechanical processing[J]. Rare Metal Materials and Engineering,2020,49(10):3285-3293.
- [7]Sellars C M,Whiteman J A. Recrystallization and grain growth in hot rolling[J]. Metal Science,1979,13(3-4):187-194.
- [8]Jonas J J,Quelennec X,Jiang L,et al. The Avrami kinetics of dynamic recrystallization[J]. Acta Materialia,2009,57(9):2748-2756.
- [9]Zhu Huajia,Chen Fei,Zhang Haiming,et al. Review on modeling and simulation of microstructure evolution during dynamic recrystallization using cellular automaton method[J]. Science China Technological Sciences,2020,63:357-396.
- [10]Jin Y,Bozzolo N,Rollett A D,et al. 2D finite element modeling of misorientation dependent anisotropic grain growth in polycrystalline materials:level set versus multi-phase-field method[J]. Computational Materials Science,2015,104:108-123.
- [11]Kingma D P,Welling M. Auto-encoding variational Bayes[C]//International Conference on Learning Representations. Banff,2014.
- [12]Bostanabad R,Zhang Y C,Li X,et al. Computational microstructure characterization and reconstruction:review of the state-of-the-art techniques[J].Progress in Materials Science,2018,95:1-41.
- [13]Cang R,Xu Y,Chen S,et al. Microstructure representation and reconstruction of heterogeneous materials via deep belief network for computational material design[J]. Journal of Mechanical Design,2017,139(7):071404.
- [14]Lyu X R,Ren X D. Variational autoencoder guided conditional diffusion generative model for material microstructure reconstruction and inverse design[J].Materials Today Communications, 2025, 48:113087.
- [15]Ho J,Jain A,Abbeel P. Denoising diffusion probabilistic models[C]//Advances in Neural Information Processing Systems 33. 2020:6840-6851.
- [16]Düreth C,Seibert P,Rücker D,et al. Conditional diffusion-based microstructure reconstruction[J].Materials Today Communications, 2023, 35:105608.
- [17]Goodfellow I,Pouget-Abadie J,Mirza M,et al.Generative adversarial nets[C]//Advances in Neural Information Processing Systems 27. Montreal:Curran Associates,2014:2672-2680.
- [18]Mirza M,Osindero S. Conditional generative adversarial nets[EB/OL].(2014-11-06).
- [19]Mao X D,Li Q,Xie H R,et al. Least squares generative adversarial networks[C]//Proceedings of the IEEE International Conference on Computer Vision.Venice:IEEE,2017:2794-2802.
- [20]Isola P,Zhu J Y,Zhou T H,et al. Image-to-image translation with conditional adversarial networks[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. Honolulu:IEEE,2017:1125-1134.
- [21]Lee J W,Goo N H,Park W B,et al. Virtual microstructure design for steels using generative adversarial networks[J]. Engineering Reports, 2021, 3(1):e12274.
- [22]Wang Z F,Chen H W,Zhang J Y,et al. Prediction of multi-stage recrystallization behavior of Aer Met100high-strength steel based on deep learning[J]. Journal of Materials Research and Technology,2025,36:2501-2517.