This is an unedited manuscript accepted for publication and provided as an Article in Press for early access at the author’s request. The article will undergo copyediting, typesetting, and galley proof review before final publication. Please be aware that errors may be identified during production that could affect the content. All legal disclaimers of the journal apply.
Dibya Tripathi,
Kheelraj Pandey,
Akhilesh Kumar Chauhan,
Ashwani Sharma,
- Research Scholar, Department of Mechanical Engineering, Amity University, Lucknow, Campus, Uttar Pradesh, India
- Assistant Professor, Department of Mechanical Engineering, Amity University, Lucknow, Campus, Uttar Pradesh, India
- Associate Professor, Kamla Nehru Institute of Technology (KNIT), Sultanpur, Uttar Pradesh, India
- Assistant Professor, Department of Mechanical Engineering, Amity University, Lucknow, Campus, Uttar Pradesh, India
Abstract
Defect detection in fibre-reinforced composite materials is important for maintaining product quality and ensuring structural reliability. However, manual inspection can be time-consuming and may depend on the experience of the inspector. This study presents a hybrid Convolutional Neural Network–Grey Wolf Optimization (CNN–GWO) approach for automated detection and classification of defects in fibre-reinforced composite materials. The proposed approach combines CNN-based feature learning with GWO-based hyperparameter optimization to improve the selection of the network configuration. The image dataset was divided into training, validation, and testing sets in the ratio of 70:15:15, with augmentation applied to the training data. GWO was used to minimize validation loss and identify an effective CNN configuration. The optimized model consisted of 128 filters in the first convolutional layer with a kernel size of 5, 192 filters in the second convolutional layer, and 256 dense units. The model achieved a reported training accuracy of 99.23%, validation accuracy of 99.82%, precision of 99.80%, recall of 99.51%, F1-score of 99.77%, and AUC of 0.999. The validation loss was approximately 0.02, compared with 0.1119 for the Keras-Tuner-based model. These results show that GWO can effectively support CNN hyperparameter selection and improve the reported classification performance. The proposed framework provides a promising basis for automated composite-defect inspection and its potential future application in industrial inspection systems.
Keywords: Fibre-Reinforced Composites; Defect Detection; Convolutional Neural Network (CNN); Grey Wolf Optimization (GWO); Hyperparameter Optimization; Automated Inspection
References
[1] Medeghini F, Tiberti G, Guhathakurta J, Simon S, Plizzari GA, Mark P. Fiber orientation and orientation factors in steel fiber‐reinforced concrete beams with hybrid fibers: a critical review. Structural Concrete. 2025 Feb;26(1):481-500.
[2] Fraile DH, Faccin E, Minelli F, Plizzari G, Muttoni A. Fibre orientation in SFRC slabs and consequences for punching shear and flexural resistance. Engineering Structures. 2024 Mar 1;302:117364.
[3] Mujalli MA, Dirar S, Mushtaha E, Hussien A, Maksoud A. Evaluation of the tensile characteristics and bond behaviour of steel fibre-reinforced concrete: An overview. Fibers. 2022 Dec 2;10(12):104.
[4] Deng Y, Zhang Z, Shi C, Wu Z, Zhang C. Steel fiber–matrix interfacial bond in ultra-high performance concrete: A review. Engineering. 2023 Mar 1;22:215-32.
[5] Girshick R, Donahue J, Darrell T, Malik J. Rich feature hierarchies for accurate object detection and semantic segmentation. InProceedings of the IEEE conference on computer vision and pattern recognition 2014 (pp. 580-587).
[6] Girshick R. Fast r-cnn. InProceedings of the IEEE international conference on computer vision 2015 (pp. 1440-1448).
[7] Ren S, He K, Girshick R, Sun J. Faster r-cnn: Towards real-time object detection with region proposal networks. Advances in neural information processing systems. 2015;28.
[8] He K, Gkioxari G, Dollár P, Girshick R. Mask r-cnn. InProceedings of the IEEE international conference on computer vision 2017 (pp. 2961-2969).
[9] BELDAR P. Optimized convolutional neural networks for defect classification in fiber-reinforced composites. Sigma. 2026 Jun;44(3):2207-18.
[10] Lu H, Cantero-Chinchilla S, Yang X, Gryllias K, Chronopoulos D. Deep learning uncertainty quantification for ultrasonic damage identification in composite structures. Composite Structures. 2024 Jun 15;338:118087.
[11] Huang H, Liu Y, Zhan J, Jia B, Dong Z, Zhao M, Liu Y, Li B. Recent progress of AI in CFRP: From design and manufacturing to structural health monitoring. Thin-Walled Structures. 2026 Apr 3:114936.
[12] Cheng J, Tan W, Yuan Y, Zhao Z, Cheng Y. Research on defect detection method for composite materials based on deep learning networks. Applied Sciences. 2024 May 14;14(10):4161.
[13] Han J, Wang Z, Liu W, Song B, Wang C, Liu X. Data-centric deep learning for defect detection in fiber composite materials: A review of algorithms, challenges, and emerging trends. Knowledge-Based Systems. 2026 Aug 4:116784.
[14] Peng S, Addepalli S, Farsi M. Machine learning in thermography non-destructive testing: a systematic review. Applied Sciences. 2025 Sep 1;15(17):9624.
[15] Vasilache MM, Tayong RB, Velisavljevic V. Comprehensive review on the use of machine learning techniques applied to the ultrasound data for the characterisation of porosity across carbon fibre reinforced polymer layers. Applied Composite Materials. 2025 Aug;32(4):1315-39.
[16] Yazdani MH, Azad MM, Khalid S, Kim HS. A hybrid deep transfer learning framework for delamination identification in composite laminates. Sensors. 2025 Jan 30;25(3):826.
[17] Azad MM, Shah AU, Prabhakar MN, Kim HS. Deep learning-based microscopic damage assessment of fiber-reinforced polymer composites. Materials. 2024 Oct 29;17(21):5265.
[18] Silenzi A, Castorani V, Tomassini S, Falcionelli N, Contardo P, Bonci A, Dragoni AF, Sernani P. Quality control of carbon look components via surface defect classification with deep neural networks. Sensors. 2023 Sep 1;23(17):7607.
[19] Daghigh V, Daghigh H, Lacy Jr TE, Naraghi M. Review of machine learning applications for defect detection in composite materials. Machine Learning with Applications. 2024 Dec 1;18:100600.
[20] Azad MM, Kumar P, Kim HS. Delamination detection in CFRP laminates using deep transfer learning with limited experimental data. Journal of Materials Research and Technology. 2024 Mar 1;29:3024-35.
[21] Liu C, Chen Y, Xu X. Structural digital Twin for damage detection of CFRP composites using meta transfer Learning-based approach. Expert Systems with Applications. 2025 Feb 1;261:125458.
[22] Deng K, Liu H, Cao J, Yang L, Du W, Xu Y, Zhao Y. Attention mechanism enhanced spatiotemporal-based deep learning approach for classifying barely visible impact damages in CFRP materials. Composite Structures. 2024 Jun 1;337:118030.
[23] Liu J, Long X, Jiang C, Liao W. Multi-feature vision transformer for automatic defect detection and quantification in composites using thermography. NDT & E International. 2024 Apr 1;143:103033.
[24] Li Z, Li Y, Lu J, Zhu H, Zheng Y, Xue J, Dong K, Yang Z, Luo K. Baseline-free assisted lamb wave-based damage detection in CFRP composites using graph convolutional networks and Transformer models. Measurement. 2025 Jan 1;242:116159.
[25] Wen R, Tao C, Ji H, Qiu J. Classification, localization and quantization of eddy current detection defects in cfrp based on edc-yolo. Sensors. 2024 Oct 21;24(20):6753.
[26] Wang G, Zhang Z, Yin W, Zhou G, He C, Chen H, Xie Q. Real-Time defect detection in CFRP with enhanced YOLO via differencing and attention for line laser thermography. Infrared Physics & Technology. 2026 May 25:106673.
[27] Liu W, Liu J, Zhu J, Han K, Zhuo L, Liu Z. CM-YOLO: an improved YOLOv8 for defect detection and classification of CFRP composites using infrared thermography. Nondestructive Testing and Evaluation. 2026 Aug 19:1-21.
[28] Fan L, Li Z, Dong F, Gao M, Sun Y, Tang Z, Du L, Wei C, Yue H, Wang F, Liu J. Infrared image defect detection for carbon fiber reinforced polymer composites via infrared preprocessing and improved YOLOv8. Infrared Physics & Technology. 2026 Sep 1:106886.
[29] Kandiyil AP, Atkinson GA, Damghani M. Localisation of defects on carbon fibre surfaces using deep learning. NDT & E International. 2025 Aug 7:103498.
[30] Liang C, Hu Q, Gao X, Wu J, Mei H, Qi F, Cheng L, Zhang L. Accurate segmentation and quantitative evaluation of Cf/SiC fiber fracture defects using an enhanced deep learning method. Materials Characterization. 2025 Feb 1;220:114712.
[31] Fan W. Composite material surface microscopic defect detection and classification combining diffusion models and zero-shot learning. Scientific Reports. 2025 Dec 19;16(1):388..
[32] Al-Shamasneh AA, Mahmoodzadeh A, Karim FK, Saidani T, Alghamdi A, Alnahas J, Sulaiman M. Application of machine learning techniques to predict the compressive strength of steel fiber reinforced concrete. Scientific Reports. 2025 Aug 21;15(1):30674.
[33] Khan A, Ahmad MS, Siddiqui AA. Machine learning methods for prediction of compressive strength of steel fibre reinforced concrete at elevated temperatures: a comparative analysis. Discover Civil Engineering. 2026 Feb 2;3(1):30.
[34] Beldar P, Chandramore R, Jagtap S, Bendale D. Enhanced convolutional neural networks for defect detection in fiber-reinforced composites: a hyperparameter optimization approach. Journal of Engineering and Applied Science. 2025 Dec;72(1):125.
[35] Mirjalili S, Mirjalili SM, Lewis A. Grey wolf optimizer. Advances in engineering software. 2014 Mar 1;69:46-61.
[36] Ding Y, Zhou SX, Yuan HQ, Pan Y, Dong JL, Wang ZP, Yang TL, She AM. Crack identification method of steel fiber reinforced concrete based on deep learning: a comparative study and shared crack database. Advances in Materials Science and Engineering. 2021;2021(1):9934250.
[37] Hampe M, “Micro-CT dataset of a steel fibre reinforced concrete sample,” DOI 10.5281/zenodo.10785684.
[38] Jung C, Redenbach C, Schladitz K. VoroCrack3d: An annotated semi-synthetic 3d image data set of cracked concrete. Data in Brief. 2024 Jun 1;54:110474.
[39] Souza JP, Silva AC, Congro M, Roehl D, Paiva AC, Pereira S, Cunha A. Automated Crack Detection in Micro-CT Scanning for Fiber-Reinforced Concrete Using Super-Resolution and Deep Learning. Electronics. 2025 Oct 28;14(21):4208.
[40] Wang P, Yun X, Hong S, Fang G, Dong B, Zhang Y, Gao Y. Identification and characteristic analysis of internal corrosion-expansion cracks in reinforced concrete based on deep learning. Journal of Building Engineering. 2026 Sep 5:117287..
[41] Mathiyazhagan GR, Paramasivam SK. Effect on AI-driven Ant Lion Optimization framework using fibre reinforced concrete with dual-stage building crack for structural applications. Matéria (Rio de Janeiro). 2026 Feb 16;31:e20250705.
[42] Huang M, Xu Z, Cai C, Hu C, Wei X, Yin W, He X. Multi-structure transfer damage detection for composite materials based on singular value decomposition and 2D-CNN network. InStructures 2025 Apr 1 (Vol. 74, p. 108656). Elsevier.
[43] Jiang L, Xiao S, Wu J, Liao Z, Zhu G, Zhang B. Ct and image post-processing for fiber composites: defect analysis, deep learning, digital volume correlation, and FE simulation–a review. Materials & Design. 2026 Feb 22:115698.
[44] Kojima Y, Hirayama K, Endo K, Harada Y, Muramatsu M. Transfer-learning-aided defect prediction in simply shaped CFRP specimens based on stress distribution obtained from finite element analysis and infrared stress measurement. Composites Part B: Engineering. 2025 Feb 15;291:111958.
[45] Buzpinar MA. Intelligent defect detection in composite materials: the role of deep learning and machine learning fusion. Nondestructive Testing and Evaluation. 2026 Aug 9:1-38.
[46] Abid H, Hassan D, Nauman S. Comparative study on the implementation of Custom Convolutional Neural Networks (CNNs), MobileNet V3 and ResNet-50 Architecture for Defect Detection in Carbon Fiber Reinforced Plastics (CFRPs). Next Research. 2026 Aug 1:102260.
[47] Li M, Leonetti D, Zappalá D, Snijder HB. Infrared-based damage detection in thick composites via transfer learning on simulated and experimental data. Composites Part B: Engineering. 2025 Oct 20:113112.
[48] Kawkabi KW, Nziko ASNT, Ngangura PM, Long X. Morphology-aware CNN with statistical descriptor regression and conditional image synthesis for predicting crack patterns in steel fiber-reinforced concrete. Comput Mater Contin. 2026. doi:10.32604/cmc.2026.086951.
Journal of Polymer & Composites
| Volume | 14 | |
| 05 | ||
| Received | 07/10/2026 | |
| Accepted | 15/10/2026 | |
| Published | 17/10/2026 | |
| Publication Time | 10 Days |