基于神经网络的复合材料含孔特征件应变场重构

doi:10.16183/j.cnki.jsjtu.2024.229

摘要/Abstract

摘要：

针对复合材料机械连接结构易损伤,且连接位置应变难监测的问题,提出一种基于神经网络的复合材料含孔特征件应变场重构方法.首先,根据复合材料机械连接结构设计不同孔径的特征样件,并确定多级载荷谱;其次,基于数值仿真构建数据集,对比支持向量机、极限学习机、随机森林与反向传播(BP)神经网络对应变场的重构精度,确定基于BP神经网络建立的重构模型精度最高,完成应变场重构预演;最后,对5 mm孔径样件进行多级加载试验,通过修正数值仿真结果获取拟测点,以实测点与拟测点应变信息为输入,完成力学响应最大位置的应变重构,重构平均误差为6.4%,验证了方法的可行性.

关键词: 复合材料, 应力集中, 神经网络, 应变重构

Abstract:

Composite mechanical connection structures are prone to damage, and strain monitoring at the connection location remains challenging. To address these issues, this paper proposes a neural network-based strain field reconstruction method for composite hole-containing feature parts. First, feature samples with different pore diameters are designed based on the composite mechanical joint structure, and a multilevel loading spectrum is determined. Then, a dataset is constructed via numerical simulation, and the strain field reconstruction accuracy of four methods, support vector machine, extreme learning machine, random forest, and back propagation (BP) neural network is compared. The results show the BP neural network-based model achieves the highest accuracy, enabling effective reconstruction of the strain field. Finally, multistage loading tests are conducted on the 5 mm hole diameter sample, and optimal measurement points are identified by correcting the numerical simulation results. Using strain data from both measured and optimized points as inputs, strain reconstruction is performed at the location of maximum mechanical response, yielding an average reconstruction error of 6.4%, which verifies the feasibility of the proposed method.

Key words: composite, stress concentration, neural network, strain reconstruction

中图分类号:

TB332

杨晨旭, 杨博文, 管振祥, 李桦, 邓书森, 霍军周. 基于神经网络的复合材料含孔特征件应变场重构[J]. 上海交通大学学报, 2026, 60(8): 1323-1335.

YANG Chenxu, YANG Bowen, GUAN Zhenxiang, LI Hua, DENG Shusen, HUO Junzhou. Reconstruction of Strain Field of Composite Porous Feature Based on Neural Network[J]. Journal of Shanghai Jiao Tong University, 2026, 60(8): 1323-1335.

图/表 26

图1

表1

图2

表2

图3

图4

图5

图6

表3

图7

图8

图9

图10

表4

图11

图12

图13

图14

图15

表5

图16

表6

图17

图18

表7

图19

参考文献 27

[1]	边文凤, 金平, 杜善义. 含损伤复合材料的湿热弹性响应[J]. 复合材料学报, 2012, 29(1): 196-206.
	BIAN Wenfeng, JIN Ping, DU Shanyi. Hygrothermal elastic response of composites with damage[J]. Acta Materiae Compositae Sinica, 2012, 29(1): 196-206.
[2]	杨成鹏, 冯炜森, 林江嵘, 等. 复合材料层压板疲劳性能评估的模型和方法[J]. 航空制造技术, 2023, 66(15): 40-59.
	YANG Chengpeng, FENG Weisen, LIN Jiangrong, et al. Models and methodologies for fatigue property prediction of composite laminates[J]. Aeronautical Manufacturing Technology, 2023, 66(15): 40-59.
[3]	刘嘉鸣, 全东, 赵国群. 航空复合材料连接成形技术研究进展[J]. 机械工程学报, 2023, 59(20): 119-142. doi: 10.3901/JME.2023.20.119
	LIU Jiaming, QUAN Dong, ZHAO Guoqun. Progress in joint forming technology of aviation composites[J]. Journal of Mechanical Engineering, 2023, 59(20): 119-142. doi: 10.3901/JME.2023.20.119
[4]	王奕首, 卿新林. 复合材料连接结构健康监测技术研究进展[J]. 复合材料学报, 2016, 33(1): 1-16.
	WANG Yishou, QING Xinlin. Progress on study of structural health monitoring technology for composite joints[J]. Acta Materiae Compositae Sinica, 2016, 33(1): 1-16.
[5]	王彬文, 肖迎春, 白生宝, 等. 飞机结构健康监测与管理技术研究进展和展望[J]. 航空制造技术, 2022, 65(3): 30-41.
	WANG Binwen, XIAO Yingchun, BAI Shengbao, et al. Research progress and prospect of aircraft structural health monitoring and management technology[J]. Aeronautical Manufacturing Technology, 2022, 65(3): 30-41.
[6]	陈国材, 汪雪良, 杨华伟, 等. 基于iFEM和虚实结合的加筋板应变场构建研究[J]. 船舶力学, 2024, 28(3): 409-421.
	CHEN Guocai, WANG Xueliang, YANG Huawei, et al. Strain field construction method of stiffened plate based on iFEM and virtual-real combination method[J]. Journal of Ship Mechanics, 2024, 28(3): 409-421.
[7]	LI M Y, DIRIK Y, OTERKUS E, et al. Shape sensing of NREL 5 MW offshore wind turbine blade using iFEM methodology[J]. Ocean Engineering, 2023, 273: 114036. doi: 10.1016/j.oceaneng.2023.114036 URL
[8]	YU M, GUO J J, LEE K M. A modal expansion method for displacement and strain field reconstruction of a thin-wall component during machining[J]. IEEE/ASME Transactions on Mechatronics, 2018, 23(3): 1028-1037. doi: 10.1109/TMECH.3516 URL
[9]	陈玉刚, 毕雨晗, 刘越, 等. 基于多模态传递比的旋转叶片动应变场重构方法[J]. 中南大学学报(自然科学版), 2023, 54(10): 3915-3925.
	CHEN Yugang, BI Yuhan, LIU Yue, et al. Dynamic strain field reconstruction method of rotating blade based on multi-mode transfer ratio[J]. Journal of Central South University (Science and Technology), 2023, 54(10): 3915-3925.
[10]	赵飞飞, 曹开拓, 保宏, 等. Timoshenko梁的变形场重构及传感器位置优化[J]. 机械工程学报, 2020, 56(20): 1-11.
	ZHAO Feifei, CAO Kaituo, BAO Hong, et al. Deformation field reconstruction of Timoshenko beam and optimization of sensor placement[J]. Journal of Mechanical Engineering, 2020, 56(20): 1-11. doi: 10.3901/JME.2020.20.001
[11]	DING G P, YUE S Y, ZHANG S C, et al. Strain-deformation reconstruction of CFRP laminates based on Ko displacement theory[J]. Nondestructive Testing and Evaluation, 2021, 36(2): 145-157. doi: 10.1080/10589759.2019.1707200 URL
[12]	DING G, JIANG S, ZHANG S, et al. Strain-deformation reconstruction of carbon fiber composite laminates based on BP neural network[J]. Materials Research, 2019, 22(5): e20190393. doi: 10.1590/1980-5373-mr-2019-0393 URL
[13]	LI J, YAN J, ZHU J, et al. K-BP neural network-based strain field inversion and load identification for CFRP[J]. Measurement, 2022(187): 187.
[14]	COOPER S, DIMAIO D. Static load estimation using artificial neural network: Application on a wing rib[J]. Advances in Engineering Software, 2018, 125: 113-125. doi: 10.1016/j.advengsoft.2018.01.007 URL
[15]	GENG X, LU X, JIANG M, et al. Research on FBG-based CFRP structural damage identification using BP neural network[J]. Photonic Sensors, 2018, 8: 168-175. doi: 10.1007/s13320-018-0466-0
[16]	YANG Z, BUEHLER M. Fill in the blank: Transferrable deep learning approaches to recover missing physical field information[J]. Advanced Materials, 2023, 35(23): 2301449. doi: 10.1002/adma.v35.23 URL
[17]	霍军周, 葛利涵, 张占葛, 等. 基于BP神经网络的TBM刀盘关键位置应变重构[J]. 东北大学学报(自然科学版), 2023, 44(10): 1455-1463.
	HUO Junzhou, GE Lihan, ZHANG Zhange, et al. Strain reconstruction of TBM cutterhead at key positions based on BP neural network[J]. Journal of Northeastern University (Natural Science), 2023, 44(10): 1455-1463.
[18]	田阔, 孙志勇, 李增聪. 面向结构静力试验监测的高精度数字孪生方法[J]. 航空学报, 2024, 45(7): 429134.
	TIAN Kuo, SUN Zhiyong, LI Zengcong. High-precision digital twin method for structural static test monitoring[J]. Acta Aeronautica et Astronautica Sinica, 2024, 45(7): 429134.
[19]	张翊东. T800碳纤维复合材料及结构的疲劳性能研究[D]. 哈尔滨: 哈尔滨工业大学, 2021.
	ZHANG Yidong. Investigation on fatigue performance of T800 carbon fiber composite material and structure[D]. Harbin: Harbin Institute of Technology, 2021.
[20]	NAJAFI A, GARG M, ABDI F. Failure analysis of composite bolted joints in tension[C]// 50th AIAA/ASME/ASCE/AHS/ASC Structures, Structural Dynamics, and Materials Conference. California, USA: Palm Springs, 2009: 2505.
[21]	王建旗, 曹增强. 大直径高干涉螺栓安装应用研究[J]. 航空制造技术, 2012, 412: 87-89.
	WANG Jianqi, CAO Zengqiang. Study on application of large diameter and high-interference bolt installation[J]. Aeronautical Manufacturing Technology, 2012, 412: 87-89.
[22]	中国飞机强度研究所, 中国航空工业集团公司北京航空材料研究院. 聚合物基复合材料层合板开孔/受载孔性能试验方法.第3部分: 开孔拉伸强度试验方法:GB/T 30968.3—2014[S]. 北京: 中国标准出版社, 2015.
	China Aircraft Strength Research Institute, Beijing Institute of Aeronautical Materials, Aviation Industry Corporation of China. Test method for open/loaded hole performance of polymer based composite laminates. Part 3: Test method for open hole tensile strength: GB/T 30968.3—2014[S]. Beijing: China Standard Press, 2015.
[23]	FUNK S, AIROUD B, NACKENHORST U. Globally supported surrogate model based on support vector regression for nonlinear structural engineering applications[J]. Archive of Applied Mechanics, 2023, 93(2): 825-839. doi: 10.1007/s00419-022-02301-3
[24]	谭增强, 牛拥军, 李元昊, 等. 基于麻雀算法和深度极限学习机的NO_x预测研究[J]. 中国电机工程学报, 2022, 42(Sup.1): 187-192.
	TAN Zengqiang, NIU Yongjun, LI Yuanhao, et al. Research on NO_x prediction of thermal power plant based on SSA-DELM[J]. Proceedings of the CSEE, 2022, 42(Sup.1): 187-192.
[25]	常琦, 杨维希, 赵恒, 等. 基于多传感器的裂纹扩展监测研究[J]. 航空学报, 2020, 41 (2): 223336. doi: 10.7527/S1000-6893.2019.23336
	CHANG Qi, YANG Weixi, ZHAO Heng, et al. A multi-sensor based crack propagation monitoring research[J]. Acta Aeronautica et Astronautica Sinica, 2020, 41(2): 223336. doi: 10.7527/S1000-6893.2019.23336
[26]	LUO W, HU T, YE Y, et al. A hybrid predictive maintenance approach for CNC machine tool driven by Digital Twin[J]. Robotics and Computer-Integrated Manufacturing, 2020, 65: 101974. doi: 10.1016/j.rcim.2020.101974 URL
[27]	PAPA U, RUSSO S, LAMBOGLIA A, et al. Health structure monitoring for the design of an innovative UAS fixed wing through inverse finite element method (iFEM)[J]. Aerospace Science and Technology, 2017, 69: 439-448. doi: 10.1016/j.ast.2017.07.005 URL

参数	取值
E₁/MPa	98 700
E₂/MPa	11 500
G₁₂/MPa	6 845
G₁₃/MPa	6 845
G₂₃/MPa	4 066
ν₁₂	0.3

W/mm	h/mm	L/mm	D/mm	F/kN
25	2	250	5	1~16
25	2	250	7.5	1~14
25	2	250	10	1~12

D/mm	训练集样本数	验证集样本数	测试集样本数	总样本数
5	27 878	6 970	4 896	39 744
7.5	28 851	7 213	5 152	41 216
10	25 958	6 490	4 992	37 440

D/mm	训练集平均M	验证集平均M	测试集平均M	训练集平均R²	验证集平均R²	测试集平均R²
5	1.310×10^-4	1.342×10^-4	7.657×10^-4	0.997 5	0.997 4	0.948 2
7.5	1.306×10^-4	1.343×10^-4	7.426×10^-4	0.997 2	0.997 0	0.950 4
10	1.547×10^-4	1.599×10^-4	7.562×10^-4	0.995 8	0.995 5	0.948 8

测点	a₀	b₀
1-1	0.642 1	3.687×10^-7
1-2	0.658 2	-2.374×10^-6
1-3	0.598 1	7.671×10^-6
1-4	0.554 6	1.137×10^-6