Title: Deep Learning for Void Detection in Composite Oriented Strand Board

Authors: Wenyue Hu, Xiaoxing Wang, Christopher C. Bowland, Phu Nguyen, Carina Xiaochen Li, Steven Nutt, Bo Jin

DOI: 10.33599/nasampe/c.22.0132

Abstract: X-ray micro-computed tomography (micro-CT) offers the ability to assess and quantify microstructural characteristics in general, and the distribution of voids in particular, in composite laminates. Despite these capabilities, analysis of micro-CT data generally requires extensive training and human input regarding detection, number counting, morphology analysis, and predictions of mechanical performance. In recent years, advances in Deep Learning (DL) have been applied to challenging tasks of image processing, such as automatic detection and analysis of features in images, and rapid classifications of features with minimal human intervention/oversight. The application of DL simplifies the pipeline of tomography analysis by automating void detection and segregation, providing an accessible path to in-depth studies of porosity evolution. This study describes an automated void segmentation solution built within commercial software (MATLAB) and applied to cross-sectional scans of Composite Oriented Strand Board (COSB). By training labeled data produced by grayscale binary masking, three of the most representative neural networks are assessed, based on the respective performance and accuracy in void detection. A Fully Convolutional Network (FCN, a neural network) performed semantic segmentation at the pixel level. A modified FCN, SegNet, was created by making the encoder-decoder structure symmetrical. The third FCN, U-net (used in biomedical image segmentation), was thought to be the state-of-art segmentation solution. Compared with the manual-labeled dataset, FCN, which yields the most accurate results statistically and successfully incorporates boundary-aware segmentation, outperformed the other two networks. Furthermore, FCN could be combined with pre-processing binary masking to develop an autonomous annotation tool for void-content study. SegNet was intended for scene understanding and therefore falsely occupied a larger area than the labeled ground truth voids, while U-net exhibited limitations in continuous boundary depiction when irregularities and perturbations were engaged.

References: [1] R. A. Smith. Composite Defects and Their Detection. Materials Science and Engineering. Vol. III. Encyclopedia of Life Support Systems (EOLSS). 2009. [2] Mahoor Mehdikhani, Larissa Gorbatikh, Ignaas Verpoest and Stepan V Lomov. Voids in fiber-reinforced polymer composites: A review on their formation, characteristics, and effects on mechanical performance. Journal of Composite Materials, 2019, Vol. 53(12) 1579–1669. [3] João M. Machado, João Manuel R.S. Tavares, Pedro P. Camanho, and Nuno Correia. Automatic void content assessment of composite laminates using a machine-learning approach. Composite Structures, Volume 288, 2022, 115383, ISSN 0263-8223. [4] B. Wei, K. Hao, X. S. Tang, and Y. Ding, “A new method using the convolutional neural network with compressive sensing for fabric defect classification based on small sample sizes,” Textile Research Journal, vol. 89, no. 17, pp. 3539–3555, Sep. 2019. [5] Karayiannis, Y.A., Stojanovic, R., Mitropoulos, P., Koulamas, C., Stouraitis, T., Koubias, S., Papadopoulos, G.: Defect detection and classification on web textile fabric using multiresolution decomposition and neural networks. In: Electronics, Circuits and Systems, 1999. Proceedings of ICECS’99. The 6th IEEE International Conference on. Vol. 2, pp. 765–768. IEEE (1999) [6] Kopaczka, M., Saggiomo, M., Güttler, M., Kielholz, K., Merhof, D. (2019). Detection and Classification of Faulty Weft Threads Using Both Feature-Based and Deep Convolutional Machine Learning Methods. In: De Marsico, M., di Baja, G., Fred, A. (eds) Pattern Recognition Applications and Methods. ICPRAM 2018. Lecture Notes in Computer Science, vol 11351. Springer, Cham. [7] Daniel Sammons, William P. Winfree, Eric Burke, et al. Segmenting delaminations in carbon fiber reinforced polymer composite CT using convolutional neural networks. AIP Conference Proceedings 1706, 110014 (2016); https://doi.org/10.1063/1.4940585. [8] Bo Cheng Jin, Xiaochen Li, Atul Jain, Carlos González, Javier LLorca, Steven Nutt. Optimizing microstructures and mechanical properties of composite oriented strand board from reused prepreg. Composite Structures, Volume 174, 2017, Pages 389-398, ISSN 0263-8223. [9] Jonathan Long, Evan Shelhamer, Trevor Darrell. Fully convolutional networks for semantic segmentation. 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Year: 2015, Pages: 3431-3440. [10] V. Badrinarayanan, A. Kendall and R. Cipolla, ""SegNet: A Deep Convolutional Encoder-Decoder Architecture for Image Segmentation,"" in IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 39, no. 12, pp. 2481-2495, 1 Dec. 2017, DOI: 10.1109/TPAMI.2016.2644615. [11] Shenli Pei, Kaifeng Wang, Jingjing Li, Yang Li, Danielle Zeng, Xuming Su, Xianghui Xiao, Hui Yang. Mechanical properties predict injection molded short/long carbon fiber reinforced polymer composites using micro X-ray computed tomography. Composites Part A: Applied Science and Manufacturing, Volume 130, 2020, 105732, ISSN 1359-835X. [12] MathWorks MATLAB, online Help Center, the topic of “bfscore.” https://www.mathworks.com/help/images/ref/bfscore.html [13] Nikhil Mittal, Chidanand, Eswara Prasad. Carbon Fiber Reinforced Plastic (CFRP) Market by Raw Material (Polyacrylonitrile (PAN), Petroleum Pitch, and Others). Carbon Fiber Reinforced Plastic (CFRP) Market. Feb 2022. [14] Reed Kopp, Joshua Joseph, Xinchen Ni, Nicholas Roy, and Brian L. Wardle. Deep Learning Unlocks X-ray Microtomography Segmentation of Multiclass Microdamage in Heterogeneous Materials. Advanced Material, 2022, 2107817. [15] Reed Alan Kopp. X-ray Micro-Computed Tomography and Deep Learning Segmentation of Progressive Damage in Hierarchical Nanoengineered Carbon Fiber Composites. Massachusetts Institute of Technology. Department of Aeronautics and Astronautics. 2021. [16] Brian L. Wardle, Roberto Guzman de Villoria, Antonio Miravete, Southborough. Systems and methods for structural sensing. United States Patent, 2014. [17] Wardle B, et al., Fabrication and Characterization of Ultrahigh-Volume-Fraction Aligned Carbon Nanotube-Polymer Composites. Advanced Materials, vol. 20, pp. 2707-2714 (2008) [18] Dhimiter Bello, Brian L Wardle, Namiko Yamamoto, et al. Exposure to nanoscale particles and fibers during machining of hybrid advanced composites containing carbon nanotubes. Journal of Nanoparticle Research, Springer Netherlands, 11-1, 2009. [19] Maysam Shahedi, Anusha Devi T.T., James D. Dormer, Baowei Fei. A Study on U-Net Limitations in Object Localization and Image Segmentation. University of Texas at Dallas. SIIM 20 Virtual Meeting, 2020.

Conference: CAMX 2022

Publication Date: 2022/10/17

SKU: TP22-0000000132

Pages: 16

Price: $32.00