基于图神经网络的直接能量沉积工艺3D工件全局残余应力场预测

doi:10.3901/JME.260031

机械工程学报 ›› 2026, Vol. 62 ›› Issue (1): 421-435.doi: 10.3901/JME.260031

• 数字化设计与制造 • 上一篇

扫码分享

基于图神经网络的直接能量沉积工艺3D工件全局残余应力场预测

李玉梅^1,2, 陈健^1,2,3, 李亚冠^3,4, 陈韵之^3,5,6, 聂振国^3,5,6

1. 北京信息科技大学高动态导航技术北京市重点实验室北京 100192;
2. 北京信息科技大学现代测控技术教育部重点实验室北京 100192;
3. 清华大学机械工程系北京 100084;
4. 太原理工大学机械与运载工程学院太原 030024;
5. 清华大学高端装备界面科学与技术全国重点实验室北京 100084;
6. 清华大学精密超精密制造装备及控制北京市重点实验室北京 100084

收稿日期:2025-01-08 修回日期:2025-09-17 发布日期:2026-02-13
作者简介:李玉梅，女，1981年出生，博士，副研究员。主要研究方向为数值仿真方法及应用、医工交叉。E-mail：liyumei3680238@163.com
聂振国(通信作者)，男，1983年出生，博士，助理研究员，硕士生导师。主要研究方向为人工智能工程应用、增材制造、医工交叉等。E-mail：zhenguonie@tsinghua.edu.cn
基金资助:
北京市自然科学基金(L242044)和国家自然科学基金(52175237)资助项目。

3D Global Residual Stress Field Prediction of Direct Energy Deposition Workpiece Based on Graph Neural Networks

LI Yumei^1,2, CHEN Jian^1,2,3, LI Yaguan^3,4, CHEN Yunzhi^3,5,6, NIE Zhenguo^3,5,6

1. Beijing Key Laboratory of High Dynamic Navigation Technology, Beijing Information Science & Technology University, Beijing 100192;
2. Key Laboratory of Modern Measurement & Control Technology of Ministry of Education, Beijing Information Science & Technology University, Beijing 100192;
3. Department of Mechanical Engineering, Tsinghua University, Beijing 100084;
4. College of Mechanical and Vehicle Engineering, Taiyuan University of Technology, Taiyuan 030024;
5. State Key Laboratory of Tribology in Advanced Equipment, Tsinghua University, Beijing 100084;
6. Beijing Key Lab of Precision/Ultra-precision Manufacturing Equipments and Control, Tsinghua University, Beijing 100084

Received:2025-01-08 Revised:2025-09-17 Published:2026-02-13

摘要/Abstract

摘要： 增材制造过程产生的残余应力导致工件发生变形、开裂以及多种结构缺陷，在工业应用中严重制约了金属工件的形状控制与性能稳定性。提出了一种基于图神经网络的直接能量沉积工艺残余应力预测方法，该方法首先通过有限元计算将红外热像仪和结构光相机构建的3D工件表面温度场计算为3D工件全局温度场，然后利用图神经网络建立打印结束时瞬态温度场与冷却后残余应力场之间的映射关系，从而实现对工件冷却后3D全局残余应力场的快速预测。验证实验结果表明，所提出的方法能够在2 s预测不同形状和边界条件工件的3D全局残余应力场，比有限元计算速度提高大约7 200倍，平均相对误差为13.72%，满足实时性与准确性的双重需求。此外，通过对比实验得出，使用温度梯度场预测残余应力场比直接使用温度场预测更准确，整体精度提升28.61%。所提出的方法为AM过程中工艺参数动态调整提供了可行性数据支持。

关键词: 增材制造, 残余应力, 图神经网络, 3D全局残余应力场, 温度梯度场

Abstract: Residual stress induced by additive manufacturing is associated with deformation, cracking, and various structural defects in the workpieces, thereby limiting geometry control and performance stability in industrial applications. A residual stress prediction method based on direct energy deposition is proposed, employing a graph neural network framework. The approach begins with the computation of the 3D surface temperature field of the workpieces, captured using an infrared thermal imager, which is subsequently transformed into the global temperature field through finite element analysis. A mapping is then established between the transient temperature field at the end of the printing process and the residual stress field after cooling by means of a graph neural network. This enables the rapid prediction of the 3D global residual stress field following the cooling stage. Experimental results indicate that the 3D residual stress field of workpieces with varying geometries and boundary conditions can be predicted within approximately 2 seconds, about 7 200 times faster than traditional finite element simulations, while maintaining an average relative error of 13.72%, satisfying both real-time and accuracy requirements. Furthermore, comparative analyses reveal that the use of the temperature gradient field for prediction yields higher accuracy than the direct temperature field, achieving an overall improvement of 28.61%. This method offers a practical data-driven solution for the dynamic adjustment of process parameters in additive manufacturing.

Key words: additive manufacturing, residual stress, graph neural networks, 3D global residual stress field, temperature gradient field

中图分类号:

TP183

李玉梅, 陈健, 李亚冠, 陈韵之, 聂振国. 基于图神经网络的直接能量沉积工艺3D工件全局残余应力场预测[J]. 机械工程学报, 2026, 62(1): 421-435.

LI Yumei, CHEN Jian, LI Yaguan, CHEN Yunzhi, NIE Zhenguo. 3D Global Residual Stress Field Prediction of Direct Energy Deposition Workpiece Based on Graph Neural Networks[J]. Journal of Mechanical Engineering, 2026, 62(1): 421-435.

参考文献

[1] GU Dongdong, SHI Xinyu,POPRAWE R,et al.Material-structure-performance integrated laser-metal additive manufacturing[J].Science,2021,372(6545):eabg1487.
[2] Cooke S,AHMADI K,WILLERTH S,et al.Metal additive manufacturing:Technology,metallurgy and modelling[J].Journal of Manufacturing Processes,2020,57:978-1003.
[3] TUNCER N and BOSE A.Solid-state metal additive manufacturing:A review[J].Jom,2020,72(9):3090-3111.
[4] SHGAEV M V,HE M,LEVY Y,et al.Laser-induced thermal processes:heat transfer,generation of stresses,melting and solidification,vaporization,and phase explosion[M]//Handbook of Laser Micro-and Nano-Engineering.Cham:Springer International Publishing,2021:83-163.
[5] 颜江涛,郑雪鹏,石张平.金属增材制造检测技术与质量控制研究进展[J].无损检测,2024,46(9):90-100.YAN Jiangtao,ZHENG Xuepeng,SHI Zhangping.Research progress of metal additive manufacturing inspection technology and quality control[J].Non-Destructive Testing,2024,46(9):90-100.
[6] BIAN P,SHAO Xiaodong,DU Jingli.Finite element analysis of thermal stress and thermal deformation in typical part during SLM[J].Applied Sciences,2019,9(11):2231.
[7] YOU M,LONG W,ZHANG G,et al.Relieving residual stress and enhanced mechanical properties by novel sandwich composites filler metal[J].Intermetallics,2025,181:108751.
[8] SOORI M,AREZOO B.A review in machining-induced residual stress[J].Journal of New Technology and Materials,2022,12(1):64-83.
[9] HUANG Xiaoming,SUN Jie,LI J.Effect of initial residual stress and machining-induced residual stress on the deformation of aluminium alloy plate[J].Strojniški vestnik-Journal of Mechanical Engineering,2015,61(2):131-137.
[10] RENDLER N J,VIGNESS I.Hole-drilling strain-gage method of measuring residual stresses:Authors indicate that this method permits the magnitudes and principal directions of residual stresses at the hole location to be determined[J].Experimental Mechanics,1966,6:577-586.
[11] TOTEMEIER T C,WRIGHT J K.Residual stress determination in thermally sprayed coatings-a comparison of curvature models and X-ray techniques[J].Surface and Coatings Technology,2006,200(12-13):3955-3962.
[12] BALASINGH C,SINGH A K.Residual stresses and their measurements by X-ray diffraction methods[J].Metals materials and Processes,2000,12(2/3):269-280.
[13] WEBSTER G A,WIMPRY R C.Non-destructive measurement of residual stress by neutron diffraction[J].Journal of Materials Processing Technology,2001,117(3):395-399.
[14] NORONHA P J,CHAPMAN J R,WERT J J.Residual stress measurement and analysis using ultrasonic techniques[J].Journal of Testing and Evaluation,1973,1(3):209-214.
[15] 伍建文,潘学荣,芦丽莉.基于XRD的Inconel 690和321不锈钢异种金属焊接残余应力测试与分析[J].无损检测,2018,40(10):49-52.WU Jianwen,PAN Xuerong,LU Lili.Measurement and analysis of welding residual stress of Inconel 690 and 321 stainless steel dissimilar metals based on XRD[J].Non-Destructive Testing,2018,40(10):49-52.
[16] SCHOINOCHORITIS B,CHANTZIS D,SALONITIS K.Simulation of metallic powder bed additive manufacturing processes with the finite element method:A critical review[J].Proceedings of the Institution of Mechanical Engineers,Part B:Journal of Engineering Manufacture,2017,231(1):96-117.
[17] 董为勇,倪嘉,吴可凡.玻璃与金属焊接残余应力有限元模型模拟计算分析[J].建材世界,2016,37(5):35-39.DONG Weiyong,NI Jia,WU Kefan.Finite element model simulation and analysis of residual stress in welding of glass and metal[J].Building Materials World,2016,37(5):35-39.
[18] HE T,LIU L,MAKEEV A.Uncertainty analysis in composite material properties characterization using digital image correlation and finite element model updating[J].Composite Structures,2018,184:337-351.
[19] WOLFER A J,AIRES J,WHEELER K,et al.Fast solution strategy for transient heat conduction for arbitrary scan paths in additive manufacturing[J].Additive Manufacturing,2019,30:100898.
[20] NIE Zhengguo,JIANG H,KARA L B.Stress field prediction in cantilevered structures using convolution neural networks[J].Journal of Computing and Information Science in Engineering,2020,20(1):011002.
[21] YANG Z,YU C H,BUEHLER M J.Deep learning model to predict complex stress and strain fields in hierarchical composites[J].Science Advances,2021,7(15):eabd7416.
[22] KHADILKAR A,WANG J,RAI R.Deep learning-based stress prediction for bottom-up SLA 3D printing process[J].The International Journal of Advanced Manufacturing Technology,2019,102:2555-2569.
[23] MAURIZI M,GAO C,BERTO F.Predicting stress,strain and deformation fields in materials and structures with graph neural networks[J].Scientific Reports,2022,12(1):21834.
[24] 马泽超,刘小明,夏汗青.基于图神经网络的路面病害态势预测方法[J].浙江大学学报,2024,58(12):2596-2608.MA Zechao,LIU Xiaoming,XIA Hanqing.Pavement distress situation prediction method based on graph neural network[J].Journal of Zhejiang University,2024,58(12):2596-2608.
[25] SUN B,ZHAO D,SHI X,et al.Modeling global spatial-temporal graph attention network for traffic prediction[J].IEEE Access,2021,9:8581-8594.
[26] HAJI-SHEIKH A,BECK J V,AGONAFER D.Steady-state heat conduction in multi-layer bodies[J].International Journal of Heat and Mass Transfer,2003,46(13):2363-2379.
[27] KANJANAKIJKASEM W.A finite element method for prediction of unknown boundary conditions in two-dimensional steady-state heat conduction problems[J].International Journal of Heat and Mass Transfer,2015,88:891-901.
[28] 孙聪磊,蔡来强,王旭东.基于无网格伽辽金法的板坯传热及力学行为[J].钢铁,2022,57(10):91-100.SUN Conglei,CAI Laiqiang,WANG Xudong.Heat transfer and mechanical behavior of slabs based on meshless Galerkin method[J].Iron & Steel,2022,57(10):91-100.
[29] VELICKOVIC P,CUCUEULL G,CASANOVA A,et al.Graph attention networks[J].Stat,2017,1050(20):10-48550.
[30] MAREAU C.Constitutive equations for thermo-elasto- plastic metallic materials undergoing large temperature variations[J].Mechanics of Materials,2023,181:104637.
[31] FANG Z C,WU Z L,HUANG C G,et al.Review on residual stress in selective laser melting additive manufacturing of alloy parts[J].Optics & Laser Technology,2020,129:106283.
[32] LI X,SIAHPOUR S,LEE J,et al.Deep learning-based intelligent process monitoring of directed energy deposition in additive manufacturing with thermal images[J].Procedia Manufacturing,2020,48:643-649.
[33] YANG S,CLARE A T,JIN X.Understanding residual stress in functionally graded directed energy deposition[J].Additive Manufacturing,2024,96:104581.

基于图神经网络的直接能量沉积工艺3D工件全局残余应力场预测

3D Global Residual Stress Field Prediction of Direct Energy Deposition Workpiece Based on Graph Neural Networks

PDF

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 15

编辑推荐

Metrics

本文评价

[1]	乔春光, 王中华, 杨廷贵, 王荣东, 姜潮. 一种(B₄C+Gd)/316L中子屏蔽复合材料设计制备[J]. 机械工程学报, 2025, 61(9): 112-122.
[2]	申昕昕, 钱明芳, 张学习, 孙良博, 耿林. 不同烧结温度对粘结剂喷射增材制造Ni-Mn-Ga合金的马氏体相变和磁性能影响研究[J]. 机械工程学报, 2025, 61(9): 157-167.
[3]	杜心伟, 魏艳红, 沈泳华, 赵文勇, 刘仁培. 电弧增材制造装备系统与应用的发展现状[J]. 机械工程学报, 2025, 61(6): 92-102.
[4]	付玲, 易斌, 尹莉, 薛丁琪, 任会礼, 王江超. 随焊感应加热尺寸和热源数对焊接失稳变形的影响[J]. 机械工程学报, 2025, 61(6): 112-120.
[5]	宋和川, 周晓敏, 张清东, 熊青霞, 张勃洋. 磁处理对铁磁材料残余应力调控规律及机理研究[J]. 机械工程学报, 2025, 61(6): 174-183.
[6]	汪汉明, 董庆兵, 陈壮, 赵波, 史修江. 行星轮系竞争失效的寿命预测及试验研究[J]. 机械工程学报, 2025, 61(3): 284-298.
[7]	高壮, 刘雨欣, 朱明亮, 轩福贞. 点阵结构设计制造与疲劳性能研究进展[J]. 机械工程学报, 2025, 61(3): 347-375.
[8]	蔡成恒, 邝伟峰, 李振华, 石学智. 车载运输振动环境对电弧增材制造碳钢形貌、组织及性能的影响[J]. 机械工程学报, 2025, 61(3): 411-421.
[9]	闫五柱, 王耀伟, 李振, 陈西锋, 王星, 张志远. 基于芯棒挤压力的孔的冷挤压最优干涉量确定新方法[J]. 机械工程学报, 2025, 61(24): 129-138.
[10]	郑洋, 赵岑芽, 熊瑞泽, 牛伟, 程芳, 刘伟, 臧立彬. 空间增材制造技术研究现状与展望[J]. 机械工程学报, 2025, 61(23): 217-239.
[11]	何天英, 余圣甫, 周杰, 肖宇. 多电弧增材制造高强钢组织转变规律研究[J]. 机械工程学报, 2025, 61(23): 259-269.
[12]	秦国华, 王纪国, 林锋, 吴竹溪. 多工序加工过程的飞机结构件变形预测与控制方法[J]. 机械工程学报, 2025, 61(23): 361-372.
[13]	卢书林, 王建峰, 占小红, 段宇航, 程立宏, 黄舒文, 赵耀邦. 预成形和热处理工序对2219铝合金旋压薄壁结构应力分布的影响[J]. 机械工程学报, 2025, 61(22): 57-68.
[14]	赵文勇, 魏艳红, 胥国祥, 胡庆贤, 王耀伟. 基于一体化模型的铝合金MIG增材单道形貌、传热及流动行为研究[J]. 机械工程学报, 2025, 61(22): 97-108.
[15]	李信贤, 祝弘滨, 李会军, 何宾, 钟华, 黄一鸣, 崔雷. 工艺参数对铝合金搅拌摩擦增材力学性能影响机理研究[J]. 机械工程学报, 2025, 61(20): 62-71.