基于磁记忆与极端随机树的20CrMo钢应力集中智能评估

陈玉凤; 马世博; 朴振华; 董含; 孙元玲; 穆振凯; 王奕博

doi:10.3969/j.issn.1674-6457.2026.09.011

PDF(2569 KB)

精密成形工程 ›› 2026, Vol. 18 ›› Issue (9) : 101-113. DOI: 10.3969/j.issn.1674-6457.2026.09.011

钢铁成形

基于磁记忆与极端随机树的20CrMo钢应力集中智能评估

陈玉凤, 马世博^*, 朴振华², 董含⁴, 孙元玲^1a, 穆振凯^1a,1b, 王奕博³

作者信息 +

Intelligent Assessment of Stress Concentration in 20CrMo Steel Based on Magnetic Memory and Extra Trees

CHEN Yufeng, MA Shibo^*, PIAO Zhenhua², DONG Han⁴, SUN Yuanling^1a, MU Zhenkai^1a,1b, WANG Yibo³

Author information +

文章历史 +

摘要

目的为提高金属结构应力集中程度评估的准确性和可靠性,探讨金属磁记忆检测技术在金属结构应力集中程度评估中的应用。方法通过COMSOL Multiphysics软件建立20CrMo钢试件的磁-力耦合仿真模型,制备不同缺陷类型（无损伤、孔洞缺陷、裂纹缺陷）的标准试件。引入应力集中系数K_t以量化应力集中程度,建立低、中、高三级应力集中评价体系。基于磁记忆法向信号提取6种特征量,进而利用极端随机树算法（Extra Trees,ET）对所选特征进行训练与建模,以预测应力集中程度的风险等级。结果仿真结果表明,不同缺陷类型对应的磁记忆法向信号特征差异显著,且信号强度与载荷大小及应力集中程度密切相关。ET算法在分类性能评估中表现优异,整体识别准确率达到91.43%。特征重要度分析结果显示,峰宽W与峭度k是区分构件应力集中程度较为关键的特征参数。实验结果与仿真结果一致,证实了磁-力耦合仿真的可靠性。结论金属磁记忆检测技术结合极端随机树算法可有效量化材料应力集中程度,为工程结构早期损伤诊断提供了理论与实验依据。

Abstract

In order to improve the accuracy and reliability of stress concentration in metal structures, the work aims to explore the application of metal magnetic memory testing technology in the assessment of stress concentration in metal structures. A magneto-mechanical coupling simulation model of 20CrMo steel specimens was established with COMSOL Multiphysics software, and standard specimens with different defect types (no damage, hole defect, crack defect) were prepared. The stress concentration factor K_t was introduced to quantify the degree of stress concentration, and the evaluation system of low, middle and high stress concentration was established. Six characteristic parameters were extracted from the normal component signals of magnetic memory and then the Extra-Trees (ET) algorithm was used to train and model the selected features for predicting the risk levels of stress concentration. Simulation results indicated that the normal component signals of magnetic memory corresponding to different defect types exhibited significant characteristic differences, and the signal intensity was closely related to the load magnitude and stress concentration level. The ET algorithm demonstrated excellent performance in classification evaluation, achieving an overall recognition accuracy of 91.43%. Feature importance analysis revealed that peak width W and kurtosis k were the key feature parameters for distinguishing the stress concentration levels in components. Experimental verification results were consistent with the simulation results, confirming the reliability of the magneto-mechanical coupling simulation. Metal magnetic memory testing technology combined with Extra-Trees algorithm can be used to effectively quantify and assess the degree of stress concentration in materials, providing theoretical and experimental basis for early damage diagnosis of engineering structures.

导出引用

陈玉凤, 马世博, 朴振华, 董含, 孙元玲, 穆振凯, 王奕博. 基于磁记忆与极端随机树的20CrMo钢应力集中智能评估[J]. 精密成形工程. 2026, 18(9): 101-113 https://doi.org/10.3969/j.issn.1674-6457.2026.09.011

CHEN Yufeng, MA Shibo, PIAO Zhenhua, DONG Han, SUN Yuanling, MU Zhenkai, WANG Yibo. Intelligent Assessment of Stress Concentration in 20CrMo Steel Based on Magnetic Memory and Extra Trees[J]. Journal of Netshape Forming Engineering. 2026, 18(9): 101-113 https://doi.org/10.3969/j.issn.1674-6457.2026.09.011

中图分类号： TG115.28

参考文献

[1] 应华剑. 循环荷载下基于自发漏磁的钢筋应力状态监测试验研究[D]. 重庆: 重庆交通大学, 2024.
YING H J.Experimental Study on Stress State Monitoring of Steel Bars Based on Spontaneous Magnetic Flux Leakage Under Cyclic Loading[D]. Chongqing: Chongqing Jiaotong University, 2024.
[2] 曾慧敏, 高向东, 张南峰, 等. 焊接裂纹无损检测方法[J]. 热加工工艺, 2017, 46(11): 13-16.
ZENG H M, GAO X D, ZHANG N F, et al.Nondestructive Testing Methods of Welding Crack[J]. Hot Working Technology, 2017, 46(11): 13-16.
[3] LU G X, WANG Q, ATTARD B, et al.A Post-Treatment to Reduce Stress Concentration Sensitivity under Intermediate-Temperature Fatigue in GH4169: High-Energy Impact Composite Modification[J]. Transactions of Materials Research, 2025, 1(3): 100039.
[4] 杨杰, 李路可, 沈华刚. 新型焊接质量的无损检测技术[J]. 精密成形工程, 2017, 9(3): 111-114.
YANG J, LI L K, SHEN H G.New Nondestructive Testing Technology of Welding Quality[J]. Journal of Netshape Forming Engineering, 2017, 9(3): 111-114.
[5] YANG L J, ZHENG F Y, HUANG P, et al.Research on the Analytical Model of Improved Magnetic Flux Leakage Signal for the Local Stress Concentration Zone of Pipelines[J]. Sensors, 2022, 22(3): 1128.
[6] 郑子明. 应力集中导致钢板零件失效的问题分析[J]. 汽车工艺与材料, 2017(10): 32-35.
ZHENG Z M.Problem Analysis of Steel Plate Part Failure Caused by Stress Concentration[J]. Automobile Technology & Material, 2017(10): 32-35.
[7] 迟大钊, 齐聪成. 国内焊接缺陷声学无损检测研究综述[J]. 精密成形工程, 2018, 10(1): 74-81.
CHI D Z, QI C C.Research on NDT Based on Acoustic in Welding Defects Testing in China[J]. Journal of Netshape Forming Engineering, 2018, 10(1): 74-81.
[8] SHI P P, SU S Q, CHEN Z M.Overview of Researches on the Nondestructive Testing Method of Metal Magnetic Memory: Status and Challenges[J]. Journal of Nondestructive Evaluation, 2020, 39(2): 43.
[9] GONG D, ZHAO L W, HAN G.Design and Experimental Validation of MMM-Based Pipeline Stress Concentration Detection System[J]. Results in Engineering, 2025, 27: 106834.
[10] 王松雷, 沈恒山. 金属磁记忆法对起重机金属结构早期损伤检测的优化研究及工程应用[J]. 起重运输机械, 2025(16): 84-88.
WANG S L, SHEN H S.Optimization Research and Engineering Application of Metal Magnetic Memory Method in Early Damage Detection of Crane Metal Structure[J]. Hoisting and Conveying Machinery, 2025 (16): 84-88.
[11] 杨书宝, 王磊, 吴楠, 等. 金属磁记忆检测在某埋地钢制管道的应用[J]. 无损探伤, 2025, 49(4): 30-32.
YANG S B, WANG L, WU N, et al.Application of Metal Magnetic Memory Testing in a Buried Steel Pipeline[J]. Nondestructive Testing Technologying Technology, 2025, 49(4): 30-32.
[12] QIU G T, CAI Y Q, LI Z J.Multiscale Investigation of Magnetic Field Distortion on Surface of Ferromagnetic Materials Caused by Stress Concentration for Metal Magnetic Memory Method[J]. Computational Materials Science, 2022, 209: 111353.
[13] HAN G, HUANG H H.A Dual-Dipole Model for Stress Concentration Evaluation Based on Magnetic Scalar Potential Analysis[J]. NDT & E International, 2021, 118: 102394.
[14] BAO S, LOU H, FU M, et al.Application of Metal Magnetic Memory Testing for Quantitative Evaluation of Stress Concentration in Ferromagnetic Steels[J]. Materials Evaluation, 2020, 78(12): 1301-1309.
[15] 陈善功, 王光存, 郭艺. 液压缸表面应力金属磁记忆检测试验研究[J]. 装备制造技术, 2020(11): 18-20.
CHEN S G, WANG G C, GUO Y.Experimental Research on Hydraulic Cylinder Surface Stress Test Based on Metal Magnetic Memory Methods[J]. Equipment Manufacturing Technology, 2020(11): 18-20.
[16] YAO K, WANG Z D, DENG B, et al.Experimental Research on Metal Magnetic Memory Method[J]. Experimental Mechanics, 2012, 52(3): 305-314.
[17] 姚凯. 基于金属磁记忆法的铁磁材料早期损伤检测与评价的实验研究[D]. 北京: 北京交通大学, 2014.
YAO K.Experimental Research on Testing and Evaluation of Early Damage of Ferromagnetic Materials Based on Metal Magnetic Memory Method[D]. Beijing: Beijing Jiaotong University, 2014.
[18] LIU B, HE Y Y, ZHANG H, et al.Study on Characteristics of Magnetic Memory Testing Signal Based on the Stress Concentration Field[J]. IET Science, Measurement & Technology, 2017, 11(1): 2-8.
[19] LI Y, WILSON J, TIAN G Y.Experiment and Simulation Study of 3D Magnetic Field Sensing for Magnetic Flux Leakage Defect Characterisation[J]. NDT & E International, 2007, 40(2): 179-184.
[20] HUANG H H, JIANG S L, YANG C, et al.Stress Concentration Impact on the Magnetic Memory Signal of Ferromagnetic Structural Steel[J]. Nondestructive Testing and Evaluation, 2014, 29(4): 377-390.
[21] 武莉波. 基于金属磁记忆方法的应力和缺陷检测的定量化研究[D]. 北京: 北京交通大学, 2020.
WU L B.Quantitative Study on Stress and Defect Detection Based on Metal Magnetic Memory Method [D]. Beijing: Beijing Jiaotong University, 2020.
[22] ZHANG Y L, HU D W, CHEN J, et al.Research on Non-Destructive Testing of Stress in Ferromagnetic Components Based on Metal Magnetic Memory and the Barkhausen Effect[J]. NDT & E International, 2023, 138: 102881.
[23] JILES D C.Theory of the Magnetomechanical Effect[J]. Journal of Physics D: Applied Physics, 1995, 28(8): 1537-1546.
[24] 杨晓惠, 陈龙, 何长春, 等. 利用能量守恒定律的力磁耦合模型研究[J]. 机械科学与技术, 2025, 44(2): 352-360.
YANG X H, CHEN L, HE C C, et al.Research on Magneto-Mechanical Coupling Model Using Law of Energy Conservation[J]. Mechanical Science and Technology for Aerospace Engineering, 2025, 44(2): 352-360.
[25] YANG X H, HE C C, PU H F, et al.An Extended Magnetic-Stress Coupling Model of Ferromagnetic Materials Based on Energy Conservation Law and Its Application in Metal Magnetic Memory Technique[J]. Journal of Magnetism and Magnetic Materials, 2022, 544: 168653.
[26] 黄炳炎. 焊接裂纹磁记忆效应力磁耦合的数值模拟研究[D]. 天津: 天津大学, 2007.
HUANG B Y.Numerical simulation study of coupling between stress and magnetism for magnetic memory effect of the welding crack[D]. Tianjin: Tianjin University, 2007.
[27] 周先军, 彭煜轩, 刘瑾, 等. 磁记忆PE全包覆抽油杆SVM应力预测研究[J]. 兵器材料科学与工程, 2025, 48(1): 108-113.
ZHOU X J, PENG Y X, LIU J, et al.Magnetic Memory PE Full-Encapsulation Sucker Rod SVM Stress Prediction Research[J]. Ordnance Material Science and Engineering, 2025, 48(1): 108-113.
[28] 王贵生, 李炜, 杨勇, 等. 基于磁记忆信号特征的管道缺陷分类识别和分级识别方法[J]. 腐蚀与防护, 2022, 43(11): 68-73.
WANG G S, LI W, YANG Y, et al.Identification Methods for Pipeline Defect Classification and Grading Based on Magnetic Memory Signal Characteristics[J]. Corrosion and Protection, 2022, 43(11): 68-73.
[29] 邢海燕, 葛桦, 秦萍, 等. 基于遗传神经网络的焊缝缺陷等级磁记忆定量化研究[J]. 材料科学与工艺, 2015, 23(2): 33-38.
XING H Y, GE H, QIN P, et al.MMM Quantifying of Welded Joint Defect Levels Based on BP Neural Network Optimized by Genetic Algorithm[J]. Materials Science and Technology, 2015, 23(2): 33-38.
[30] 冷建成, 赵雷, 张新, 等. 基于磁记忆在线监测的再制造毛坯疲劳寿命预测方法[J]. 材料导报, 2025, 39(2): 204-209.
LENG J C, ZHAO L, ZHANG X, et al.Fatigue Life Prediction Method of Remanufacturing Blank Based on Magnetic Memory Online Monitoring[J]. Materials Reports, 2025, 39(2): 204-209.
[31] 徐滨士, 董丽红. 再制造质量控制中的金属磁记忆检测技术[M]. 北京: 国防工业出版社, 2015.
XU B S, DONG L H.Metal Magnetic Memory Testing Technology in Remanufacturing Quality Control[M]. Beijing: National Defense Industry Press, 2015.
[32] PILKEY W D, PILKEY D F.Peterson’s Stress Concentration Factors[M]. New York, USA: John Wiley & Sons, 2007.
[33] PETERSON R E.Stress Concentration Factors[M]. New York: Wiley-Interscience, 1974.
[34] DUBOV A A.Principle Features of Metal Magnetic Memory Method and Inspection Tools as Compared to Known Magnetic NDT Methods[C]// Proceedings of the 7th European Conference on Non-Destructive Testing. Copenhagen, Denmark, 1997, 2: 920-927.
[35] ROSKOSZ M, GAWRILENKO P.Analysis of Changes in Residual Magnetic Field in Loaded Notched Samples[J]. NDT & E International, 2008, 41(7): 570-576.
[36] WANG Z D, YAO K, DENG B, et al.Quantitative Study of Metal Magnetic Memory Signal versus Local Stress Concentration[J]. NDT & E International, 2010, 43(6): 513-518.
[37] 任吉林, 林俊明, 宋凯, 等. 金属磁记忆检测技术[M]. 北京: 中国电力出版社, 2000.
REN J L, LIN J M, SONG K, et al.Metal Magnetic Memory Detection Technology[M]. Beijing: China Electric Power Press, 2000.
[38] KONG W Y, HU J, HUANG W, et al.The Prediction of Adhesive Failure of Asphalt Mixture Based on Extremely Randomized Trees[J]. Science China Technological Sciences, 2025, 68(9): 1920201.
[39] LIU B, WANG F C, WU Z H, et al.Research on Magnetic Memory Inspection Signal Characteristics of Multi-Parameter Coupling Pipeline Welds[J]. NDT & E International, 2024, 143: 103019.