Improved Transfer Learning Method for Eddy Current Testing and Identification of CFRP Defects

doi:10.3901/JME.2025.06.053

Abstract

Abstract: In recent years, automated detection and identification of defects in carbon fiber reinforced polymer (CFRP) composites has emerged as a hot topic in the field of nondestructive testing. However, insufficient defect data of CFRP components might result in overfitting and low defect recognition accuracy. Therefore, an eddy current testing approach for CFRP defects using improved migration learning is proposed. Firstly, the methods of eddy current C-scan imaging and complex plane signal feature extraction are used to obtain CFRP defect samples and implement data enhancement, which solves the problem of insufficient samples. The MobileNet V2 network with K-means clustering is then used to pick out source domain images with similar features in the hot rolled strip surface defect data set for pre-training, to complete the extraction of similar features in the target domain and reduce the negative migration. Finally, the convolutional attention module is integrated into the feature extraction network to reduce the influence of background feature information in the feature map. The network weights trained on the source domain are transferred to the improved Faster R-CNN object detection model through model transfer, to establish a CFRP defect detection model. Through comparison tests, the method efficiently overcomes the problem of less defect data for CFRP components while maintaining high accuracy and robustness. The constructed CFRP defect detection model achieves high-precision recognition of three types of defects, namely, crack, delamination, and wrinkle. The mean average precision reaches 94.62%, which is 29.31% and 2.79% higher than that of the traditional training method and the original network migration learning detection accuracy, respectively. The recognition accuracy of wrinkle defects in particular is significantly improved, which achieves a better result and meets the requirements for CFRP defect detection.

Key words: eddy current testing, CFRP, transfer learning, K-means clustering, convolutional neural network

CLC Number:

CHENG Jun, ZHU Yulong, WANG Buyun, LIANG Yi, XU Dezhang, LIU Mengmeng, LIU Youyu. Improved Transfer Learning Method for Eddy Current Testing and Identification of CFRP Defects[J]. Journal of Mechanical Engineering, 2025, 61(6): 53-65.

References

[1] Rajak D K，Wagh P H，Linul E. Manufacturing technologies of carbon/glass fiber-reinforced polymer composites and their properties：A review [J]. Polymers，2021，13(21)：3721.
[2] 陈吉平，李岩，刘卫平，等. 连续纤维增强热塑性树脂基复合材料自动铺放原位成型技术的航空发展现状[J]. 复合材料学报，2019，36(4)：784-794. CHEN Jiping，LI Yan，LIU Weiping，et al. Aerospace development status of continuous fiber-reinforced thermoplastic resin matrix composite material automatic placement and in-situ molding[J]. Acta Materiae Compositae Sinica，2019，36(4)：784-794.
[3] Sun Z，Shan Z，Huang H，et al. Multi-scale modeling and finite element analyses of thermal conductivity of 3D C/SiC composites fabricating by flexible-oriented woven process[J]. Chinese Journal of Mechanical Engineering，2024，37：60.
[4] 张磊，许帅康，陈洁，等. 列车车体轻量化设计研究进展[J]. 机械工程学报，2023，59(24)：177-196. ZHANG Lei，XU Shuaikang，CHEN Jie，et al. Research progress in lightweight design of train body[J]. Journal of Mechanical Engineering，2023，59(24)：177-196.
[5] 刘明良，唐颀，田小永，等. 连续纤维增强加筋圆柱壳回转3D打印工艺及其轴压性能研究[J]. 机械工程学报，2024，60(15)：283-290. LIU Mingliang，TANG Qi，TIAN Xiaoyong，et al. Study on the 3D printing process and axial compression performance of reinforced cylindrical shell[J]. Journal of Mechanical Engineering，2024，60(15)：283-290.
[6] 苏飞，欧阳晨恺，李枫，等. 碳纤维增强树脂复合材料齿槽加工中分层缺陷的形成机制[J]. 复合材料学报，2021，38(12)：4042-4051. SU Fei，OUYANG Chenkai，LI Feng，et al. Formation mechanism of the slotting delamination of carbon fiber reinforced plastics[J]. Acta Materiae Compositae Sinica，2021，38(12)：4043-4052.
[7] 程军，游勇，汪步云，等. 基于空频去混叠的CFRP材料纤维纹路涡流成像[J]. 机械工程学报，2022，58(4)：14-21. CHENG Jun，YOU Yong，WANG Buyun，et al. Eddy current imaging of fiber textures in cfrp composite based on spatial frequency domain de-aliasing method [J]. Journal of Mechanical Engineering，2022，58(4)：14-21.
[8] 程旭，宋晨，史金钢，等. 基于深度学习的通用目标检测研究综述[J]. 电子学报，2021，49(7)：1428-1438. CHENG Xu，SONG Chen，SHI Jingang，et al. A survey of generic object detection methods based on deep learning[J]. Acta Electronica Sinica，2021，49(7)：1428-1438.
[9] Bang H T，Park S，Jeon H. Defect identification in composite materials via thermography and deep learning techniques[J]. Composite Structures，2020，246：112405.
[10] 路浩. 基于机器视觉的碳纤维预浸料表面缺陷检测方法的研究[D]. 济南：山东大学，2020. LU Hao. Surface defect detection method of carbon fiber prepreg based on machine vision[D]. Jinan：Shandong University，2020.
[11] 周建民，陈超，涂文兵，等. 红外热波技术、有限元与SVM相结合的复合材料分层缺陷检测方法[J]. 仪器仪表学报，2020，41(3)：29-38. ZHOU Jianmin，CHEN Chao，TU Wenbing，et al. Composite layer defect detection method based on infrared heat wave technology，finite element and SVM[J]. Chinese Journal of Scientific Instrument，2020，41(3)：31-40.
[12] Saeed N，King N，Said Z，et al. Automatic defects detection in CFRP thermograms，using convolutional neural networks and transfer learning[J]. Infrared Physics & Technology，2019，102：103048.
[13] Cheng J，Zhu Y，Wang B，et al. Noncontact visualization of multiscale defects in CFRP composites using eddy current testing with T-R probe[J]. NDT and E International，2024，145：103138.
[14] 程军，徐帅，汪步云，等. 一种自调零平切型电涡流探头及其在碳纤维复合材料细观成像中的应用[J]. 机械工程学报，2021，57(4)：27-35. CHENG Jun，XU Shuai，WANG Buyun，et al. Application of self-nulling pancake-tangent eddy current probe for mesoscopic imaging of CFRP composites[J]. Journal of Mechanical Engineering，2021，57(4)：27-35.
[15] 胡嘉沛，李震，黄河清，等. 采用改进YOLOv4-Tiny模型的柑橘木虱识别[J]. 农业工程学报，2021，37(17)：197-203. HU Jiapei，LI Zhen，HUANG Heqing，et al. Citrus psyllid detection based on improved YOLOv4-Tiny model[J]. Transactions of the Chinese Society of Agricultural Engineering，2021，37(17)：197-203.
[16] Sinaga K P，Yang M S. Unsupervised K-means clustering algorithm[J]. IEEE Access，2020，8：80716- 80727.
[17] 何倩，李双富，黄焕，等. 一种海量数据快速聚类算法[J]. 北京邮电大学学报，2020，43(3)：118-124. HE Qian，LI Shuangfu，HUANG Huan，et al. A fast clustering algorithm for massive data [J]. Journal of Beijing University of Posts and Telecommunications，2020，43(3)：118-124.
[18] Cao C，Wang B，Zhang W，et al. An improved faster r-cnn for small object detection[J]. IEEE Access，2019，7：106838-106846.
[19] 王忠塬. 基于改进Faster RCNN的小目标检测技术研究[D]. 长春：长春理工大学，2021. WANG Zhongyuan. Research on small target detection technology based on improved faster RCNN[D]. Changchun：Changchun University of Science and Technology，2021.
[20] Woo S，Park J，Lee J Y，et al. CBAM：Convolutional block attention module[C]// Springer. 15th European Conference on Computer Vision (ECCV2018)，2018，Sep. 8-14，Munich，Germany. Springer Nature Switzerland AG，2018，LNCS 11211：3-19.
[21] Roh M C，Lee J Y. Refining faster-RCNN for accurate object detection[C]// IEEE. 15th IAPR International Conference on Machine Vision Applications (MVA) Nagoya University，2017，May 8-12，Nagoya，Japan. NewYork：IEEE，2017：514-517.
[22] Alwyn M，Jimson M，Mahesh G，et al. An improved transfer learning approach for intrusion detection [J]. Procedia Computer Science，2017，115：251-257．