基于机器学习的黏钢构件黏接层缺陷识别方法研究

姚浩; 夏桂然; 刘泽佳; 周立成

doi:10.21656/1000-0887.440365

基于机器学习的黏钢构件黏接层缺陷识别方法研究

doi: 10.21656/1000-0887.440365

1.
广州交通投资集团有限公司，广州 510330
2.
华南理工大学土木与交通学院，广州 510641

基金项目:

广东省自然科学基金 2023A1515012942

详细信息

作者简介:
姚浩(1991—)，男，硕士(E-mail: yyaohao@foxmail.com)

通讯作者:
周立成(1987—)，男，副教授，博士(通讯作者. E-mail: ctlczhou@scut.edu.cn)

(我刊青年编委周立成来稿)
中图分类号: O3
计量
- 文章访问数: 58
- HTML全文浏览量: 21
- PDF下载量: 19
- 被引次数: 0
出版历程
- 收稿日期: 2023-12-25
- 修回日期: 2024-01-24
- 刊出日期: 2024-04-01

A Defect Identification Method for Bonding Layers of Adhesive Steel Members Based on Machine Learning

1.
Guangzhou Communications Investment Group Co., Ltd., Guangzhou 510330, P.R.China
2.
School of Civil Engineering & Transportation, South China University of Technology, Guangzhou 510641, P.R.China

(Contributed by ZHOU Licheng, M.AMM Youth Editorial Board)

摘要

摘要: 对黏钢加固结构黏接层缺陷对超声检测信号的影响进行了深入研究，并提出了一种基于机器学习的黏接层缺陷识别的新型方法.首先，该文基于直接接触式的脉冲回波反射法对黏钢构件进行有限元模拟，并阐述了超声波在黏钢构件中的传播规律；其次，通过分析局部段超声回波信号及相关信号特征，讨论了不同缺陷变量对超声回波信号的影响规律；最后，建立了黏钢构件超声时程响应数据集，并对比了不同机器学习模型对缺陷大小、位置的分类识别性能，形成了黏钢构件黏接层缺陷识别方法.结果表明，局部段超声回波信号及其特征随着缺陷大小、位置的改变呈规律性变化，能够对缺陷信息进行初步区分.同时，该文提出的基于RF模型的黏钢构件黏接层缺陷识别方法能够有效识别黏钢构件黏接层缺陷，具有较广阔的工程应用前景.
- 超声检测 /
- 机器学习 /
- 黏钢构件 /
- 黏接层缺陷
Abstract: The effects of bonding layer defects on ultrasonic detection signals of bonded steel reinforced structures were deeply studied and a new method for the bonding layer defect identification based on machine learning was proposed. Firstly, based on the direct contact pulse-echo reflection method, the finite element simulation of the viscous steel member was carried out, and the propagation law of ultrasonic waves in the viscous steel member was expounded. Secondly, the characteristics of local ultrasonic echo signals and related signals were analyzed, and the effects of different defect variables on ultrasonic echo signals were discussed. Finally, the ultrasonic time-history response data set of the adhesive steel member was established, and the classification and recognition performances of different machine learning models for the size and location of defects were compared, and the defect identification method for the adhesive layer of the bonded steel member was built. The results show that, the local ultrasonic echo signal and its characteristics change regularly with the defect size and location, which can help preliminarily distinguish the defect information. Meanwhile, the proposed RF model-based defect identification method can effectively identify the defects of the adhesive layer in the bonded steel member, and has a broad engineering application prospect.
- ultrasonic testing /
- machine learning /
- bonded steel component /
- bonding layer defect
(Contributed by ZHOU Licheng, M.AMM Youth Editorial Board)
注释:

1) (我刊青年编委周立成来稿)

HTML全文

图 1 黏钢构件中超声波传播示意

Figure 1. Ultrasonic propagation in the bonded steel member

下载: 全尺寸图片幻灯片

图 2 黏钢构件有限元模型

Figure 2. The finite element model for the adhesive steel member

下载: 全尺寸图片幻灯片

图 3 超声激励信号

Figure 3. The ultrasonic excitation signal

下载: 全尺寸图片幻灯片

图 4 黏钢构件整体概况及测点分布情况

Figure 4. The general layout and measuring points distribution of the steel members

下载: 全尺寸图片幻灯片

图 5 黏接层超声检测与有限元模拟结果对比结果

注为了解释图中的颜色，读者可以参考本文的电子网页版本，后同.

Figure 5. Comparison of ultrasonic detection and finite element simulation results of the adhesive layer

下载: 全尺寸图片幻灯片

图 6 超声检测回波信号

Figure 6. Ultrasonic detection echo signals

下载: 全尺寸图片幻灯片

图 7 超声检测声场云图

Figure 7. Sound fields of the ultrasonic detection

下载: 全尺寸图片幻灯片

图 8 黏钢构件缺陷示意图

Figure 8. Diagram of defects in the bonded steel member

下载: 全尺寸图片幻灯片

图 9 不同缺陷尺寸对超声信号的影响

Figure 9. Influences of different defect sizes on ultrasonic signals

下载: 全尺寸图片幻灯片

图 10 不同缺陷位置对超声信号的影响

Figure 10. Influences of different defect positions on ultrasonic signals

下载: 全尺寸图片幻灯片

图 11 无噪工况下RF、DT、KNN模型缺陷大小、位置识别准确率

Figure 11. Defect size and location recognition accuracies of RF, DT, KNN models under the noiseless condition

下载: 全尺寸图片幻灯片

图 12 不同噪声条件下RF、DT、KNN模型缺陷大小、位置识别准确率

Figure 12. Defect size and location recognition accuracies of RF, DT, KNN models under different SNR conditions

下载: 全尺寸图片幻灯片

表 1 材料参数

Table 1. Material parameters

simulation material	density ρ/(kg/m³)	elasticity modulus E/MPa	Poisson’s ratio υ
steel	7.85×10³	2.00×10⁵	0.28
epoxy structural adhesive	1.20×10³	2.02×10³	0.48
concrete	2.50×10³	2.50×10⁴	0.20

下载: 导出CSV

表 2 超声检测数据集工况统计

Table 2. Working condition statistics of the ultrasonic test data set

thickness of adhesive h_a/mm	number of changes in the defect location	number of changes in defect diameter	number of working conditions corresponding to various thicknesses
1	1	29	29
2	11	29	319
3	21	29	609
4	31	29	899

下载: 导出CSV

参考文献(21)

[1]	阿林香. 桥梁钢T梁梁底贴钢板施工质量控制[J]. 中国高新科技, 2021(2): 40-41. https://www.cnki.com.cn/Article/CJFDTOTAL-GXKE202102016.htm A Linxiang. Construction quality control of steel plate attached to the bottom of bridge steel T-beam[J]. China High and New Technology, 2021(2): 40-41. (in Chinese) https://www.cnki.com.cn/Article/CJFDTOTAL-GXKE202102016.htm
[2]	缪飞, 王庆曌. 钢筋混凝土结构粘钢加固质量检测[J]. 中华民居, 2014(5): 160. https://www.cnki.com.cn/Article/CJFDTOTAL-ZHMJ201405140.htm MIAO Fei, WANG Qingzhao. Quality inspection of reinforced concrete structure bonded to steel reinforcement[J]. China Homes, 2014(5): 160. (in Chinese) https://www.cnki.com.cn/Article/CJFDTOTAL-ZHMJ201405140.htm
[3]	佟阳. 粘贴钢板补强钢筋混凝土梁抗剪性能试验研究[J]. 公路交通科技(应用技术版), 2011, 7(5): 4-5. https://www.cnki.com.cn/Article/CJFDTOTAL-GLJJ201105003.htm TONG Yang. Experimental study on shear behavior of reinforced concrete beams reinforced by pasted steel plates[J]. Journal of Highway and Transportation Research and Development, 2011, 7(5): 4-5. (in Chinese) https://www.cnki.com.cn/Article/CJFDTOTAL-GLJJ201105003.htm
[4]	林学春. 钢筋混凝土桥梁粘钢加固试验研究[J]. 中外公路, 2013, 33(1): 167-172. https://www.cnki.com.cn/Article/CJFDTOTAL-GWGL201301043.htm LIN Xuechun. Experimental study on reinforced concrete bridge reinforced with steel[J]. Journal of China & Foreign Highway, 2013, 33(1): 167-172. (in Chinese) https://www.cnki.com.cn/Article/CJFDTOTAL-GWGL201301043.htm
[5]	刘茂钊, 杨博, 杨英武. 基于涡流热激励的粘钢加固混凝土结构粘结层缺陷热像识别试验研究[J]. 中国测试, 2023, 49(5): 52-59. https://www.cnki.com.cn/Article/CJFDTOTAL-SYCS202305008.htm LIU Maozhao, YANG Bo, YANG Yingwu. Experimental study on thermal image identification of bonded layer defects in reinforced concrete structures based on eddy current thermal excitation[J]. China Measurement & Test, 2023, 49(5): 52-59. (in Chinese) https://www.cnki.com.cn/Article/CJFDTOTAL-SYCS202305008.htm
[6]	杨英武, 张欣, 杨小青, 等. 红外热像法识别混凝土结构粘钢加固缺陷的试验研究[J]. 低温建筑技术, 2018, 40(5): 21-27. https://www.cnki.com.cn/Article/CJFDTOTAL-DRAW201805010.htm YANG Yingwu, ZHANG Xin, YANG Xiaoqing, et al. Experimental study on identification of reinforced defects of concrete structures by using infrared thermography[J]. Low Temperature Architecture Technology, 2018, 40(5): 21-27. (in Chinese) https://www.cnki.com.cn/Article/CJFDTOTAL-DRAW201805010.htm
[7]	YAN D, NEILD S A, DRINKWATER B W. Modelling and measurement of the nonlinear behaviour of kissing bonds in adhesive joints[J]. NDT & E International, 2012, 47: 18-25.
[8]	ADAMS R D, DRINKWATER B W. Nondestructive testing of adhesively-bonded joints[J]. NDT & E International, 1997, 30(2): 93-98.
[9]	TITOV S A, MAEV R G, BOGACHENKOV A N. Pulse-echo NDT of adhesively bonded joints in automotive assemblies[J]. Ultrasonics, 2008, 48(6/7): 537-546.
[10]	孙朝明, 汤光平, 李建文. 脉冲反射法检测粘接缺陷的有限元模拟[J]. 无损检测, 2014. 36(7): 6-10. https://www.cnki.com.cn/Article/CJFDTOTAL-WSJC201407002.htm SUN Chaoming, TANG Guangping, LI Jianwen. Finite element simulation of detection of bonding defects by pulse reflection method[J]. Nondestructive Testing, 2014, 36(7): 6-10. (in Chinese) https://www.cnki.com.cn/Article/CJFDTOTAL-WSJC201407002.htm
[11]	SUN G, ZHAO L, DONG M, et al. Non-contact characterization of debonding in lead-alloy steel bonding structure with laser ultrasound[J]. Optik, 2018, 164: 734-744. doi: 10.1016/j.ijleo.2018.03.075
[12]	陈军, 乔丹, 崔哲, 等. 黏接结构弱黏接缺陷的非线性超声评价[J]. 无损检测, 2019, 41(9): 60-64. https://www.cnki.com.cn/Article/CJFDTOTAL-WSJC201909016.htm LI Jun, QIAO Dan, CUI Zhe, et al. Nonlinear ultrasonic evaluation of weak bonding defects in bonding structure[J]. Nondestructive Testing, 2019, 41(9): 60-64. (in Chinese) https://www.cnki.com.cn/Article/CJFDTOTAL-WSJC201909016.htm
[13]	郝威, 李明, 徐莹, 等. 复合材料蜂窝夹芯缺陷超声检测模拟研究[J]. 机械科学与技术, 2023, 42(8): 1362-1365. https://www.cnki.com.cn/Article/CJFDTOTAL-JXKX202308024.htm HAO Wei, LI Ming, XU Ying, et al. Simulation study on ultrasonic detection of defects in honeycomb sandwich of composite materials[J]. Mechanical Science and Technology for Aerospace Engineering, 2022, 42(8): 1362-1365. (in Chinese) https://www.cnki.com.cn/Article/CJFDTOTAL-JXKX202308024.htm
[14]	李伟, 李建增, 周海林, 等. 多层复合材料超声检测的数值模拟[J]. 系统仿真技术, 2012, 8(1): 32-36. https://www.cnki.com.cn/Article/CJFDTOTAL-XTFJ201201009.htm LI Wei, LI Jianzeng, ZHOU Hailin, et al. Numerical simulation of ultrasonic testing of multilayer composites[J]. System Simulation Technology, 2012, 8(1): 32-36. (in Chinese) https://www.cnki.com.cn/Article/CJFDTOTAL-XTFJ201201009.htm
[15]	WOJTCZAK E, RUCKA M. Wave frequency effects on damage imaging in adhesive joints using lamb waves and RMS[J]. Materials, 2019, 12(11): 1842. doi: 10.3390/ma12111842
[16]	WANG H, FAN Z, CHEN X, et al. Automated classification of pipeline defects from ultrasonic phased array total focusing method imaging[J]. Energies, 2022, 15(21): 8272. doi: 10.3390/en15218272
[17]	LIU Q, JIANG A, FANG D, et al. Intelligent recognition of defects in vermicular graphite cast iron engine cylinder head by ultrasonic testing[J]. Journal of Physics: Conference Series, 2021, 1894(1): 12034. doi: 10.1088/1742-6596/1894/1/012034
[18]	LV G, GUO S, CHEN D, et al. Laser ultrasonics and machine learning for automatic defect detection in metallic components[J]. NDT & E International, 2023, 133: 102752.
[19]	SAMBATH S, NAGARAJ P, SELVAKUMAR N. Automatic defect classification in ultrasonic NDT using artificial intelligence[J]. Journal of Nondestructive Evaluation, 2011, 30(1): 20-28. doi: 10.1007/s10921-010-0086-0
[20]	徐猛, 李宇涛, 徐彦霖, 等. 粘接层厚度对粘接质量超声检测的影响分析[J]. 兵器材料科学与工程, 2008, 31(3): 62-65. https://www.cnki.com.cn/Article/CJFDTOTAL-BCKG200803019.htm XU Meng, LI Yutao, XU Yanlin, et al. Analysis of influence of bonding layer thickness on ultrasonic detection of bonding quality[J]. Ordnance Material Science and Engineering, 2008, 31(3): 62-65. (in Chinese) https://www.cnki.com.cn/Article/CJFDTOTAL-BCKG200803019.htm
[21]	马晓磊. 基于COMSOL仿真的材料缺陷超声检测模式识别[D]. 南昌: 南昌航空大学, 2019. MA Xiaolei. Pattern recognition of ultrasonic testing of material defects based on COMSOL simulation[D]. Nanchang: Nanchang Hangkong University, 2019. (in Chinese)