Identifying different features and phases within the biaxial warp-knitted composites enables accurate characterization of the internal changes of the composites under complex loading conditions and different processing parameters. However, obtaining 3D reconstructions that can be accurately segmented and quantitatively analyzed is still a challenge, primarily due to the low attenuation of materials, especially the low contrast between the weft yarn and warp yarns. Here, we use deep learning approaches to identify the boundary of the section of the tow on the 2D image, extract individual warp or weft tows automatically, and this is followed by computational modeling of composite materials. The results show that the performance of ResUNet++ model is better than that of other models for the biaxial warp-knitted composite specimen segmentation tasks. The special improved dataset augmentation algorithm is also a positive effective way to enhance network performance.
Heide-JørgensenSTeixeira de FreitasSBudzikMK. On the fracture behaviour of CFRP bonded joints under mode I loading: effect of supporting carrier and interface contamination. Compos Sci Technol2018; 160: 97–110. DOI: 10.1016/j.compscitech.2018.03.024.
2.
LiuXZhangDQiuH, et al.On-axis fatigue behaviors and failure characterization of 3D5D braided composites with yarn-reduction using X-ray computed tomography. Compos Sci Technol2021; 203. DOI: 10.1016/j.compscitech.2020.108585.
3.
GuYZhangDZhangZ, et al.Torsion damage mechanisms analysis of two-dimensional braided composite tubes with digital image correction and X-ray micro-computed tomography. Compos Struct2021; 256. DOI: 10.1016/j.compstruct.2020.113020.
4.
DongKPengXZhangJ, et al.Temperature-dependent thermal expansion behaviors of carbon fiber/epoxy plain woven composites: experimental and numerical studies. Compos Struct2017; 176: 329–341. DOI: 10.1016/j.compstruct.2017.05.036.
5.
ZhaoYSongLLiJ, et al.Multi-scale finite element analyses of thermal conductivities of three dimensional woven composites. Appl Compos Mater2017; 24: 1525-1542. DOI: 10.1007/s10443-017-9601-0.
6.
SevenoisRDBGarozDVerbovenE, et al.Multiscale approach for identification of transverse isotropic carbon fibre properties and prediction of woven elastic properties using ultrasonic identification. Compos Sci Technol2018; 168: 160–169. DOI: 10.1016/j.compscitech.2018.09.016.
7.
FuTXuJHuiZ. Analysis of induction heating temperature field of plain weave CFRP based on finite element meso model. Appl Compos Mater2021; 28: 149–163. DOI: 10.1007/s10443-020-09852-0.
8.
LomovSVGusakovAVHuysmansG, et al.Textile geometry preprocessor for meso-mechanical models of woven composites. Compos Sci Technol2000; 60: 2083–2095. DOI: 10.1016/S0266-3538(00)00121-4.
9.
VerpoestILomovSV. Virtual textile composites software WiseTex: integration with micro-mechanical, permeability and structural analysis. Compos Sci Technol2005; 65: 2563–2574. DOI: 10.1016/j.compscitech.2005.05.031.
10.
LinHZengXSherburnM, et al.Automated geometric modelling of textile structures. Textil Res J2012; 82: 1689–1702. DOI: 10.1177/0040517511418562.
11.
PotluriPRawalARivaldiM, et al.Geometrical modelling and control of a triaxial braiding machine for producing 3D preforms. Compos Appl Sci Manuf2003; 34: 481–492. DOI: 10.1016/S1359-835X(03)00061-7.
WangY-bLiuZ-gLiuN, et al.A new geometric modelling approach for 3D braided tubular composites base on Free Form Deformation. Compos Struct2016; 136: 75–85. DOI: 10.1016/j.compstruct.2015.09.036.
14.
SinchukYKibleurPAeltermanJ, et al.Variational and deep learning segmentation of very-low-contrast X-ray computed tomography images of carbon/epoxy woven composites. Materials2020; 13: 936. DOI: 10.3390/ma13040936.
15.
LiuXZhangDSunJ, et al.Refine reconstruction and verification of meso-scale modeling of three-dimensional five-directional braided composites from X-ray computed tomography data. Compos Struct2020; 245: 112347. DOI: 10.1016/j.compstruct.2020.112347.
16.
AiSSongWChenY. Stress field and damage evolution in C/SiC woven composites: image-based finite element analysis and in situ X-ray computed tomography tests. J Eur Ceram Soc2021; 41: 2323–2334. DOI: 10.1016/j.jeurceramsoc.2020.12.026.
17.
JiangYChenX. Geometric and algebraic algorithms for modelling yarn in woven fabrics. J Textil Inst2005; 96: 237–245. DOI: 10.1533/joti.2005.0005.
18.
BalokasGKriegesmannBCzichonS, et al.Stochastic modeling techniques for textile yarn distortion and waviness with 1D random fields. Compos Appl Sci Manuf2019; 127: 105639. DOI: 10.1016/j.compositesa.2019.105639.
19.
NilakantanGVirtual microstructure generation using thermal growth: case study of a plain-weave Kevlar fabric. Finite Elem Anal Des2018; 147: 21–33. DOI: 10.1016/j.finel.2018.05.002.
20.
Wehrkamp-RichterTDe CarvalhoNVPinhoST. A meso-scale simulation framework for predicting the mechanical response of triaxial braided composites. Compos Appl Sci Manuf2018; 107: 489–506. DOI: 10.1016/j.compositesa.2018.01.028.
21.
SononBMassartTJ. A level-set based representative volume element generator and XFEM simulations for textile and 3D-reinforced composites. Materials2013; 6: 5568–5592. DOI: 10.3390/ma6125568 (accessed on 28 November 2013).
22.
PierreuxGVan HemelrijckDMassartTJ. Automated generation of 3D orthogonal woven composites RVEs including yarn cross-section variations. Compos Sci Technol2019; 176: 90–102. DOI: 10.1016/j.compscitech.2019.03.023.
23.
UllahHHarlandARLucasT, et al.Finite-element modelling of bending of CFRP laminates: multiple delaminations. Comput Mater Sci2012; 52: 147–156. DOI: 10.1016/j.commatsci.2011.02.005.
24.
JacquesSDe BaereIVan PaepegemW. Application of periodic boundary conditions on multiple part finite element meshes for the meso-scale homogenization of textile fabric composites. Compos Sci Technol2014; 92: 41–54. DOI: 10.1016/j.compscitech.2013.11.023.
25.
ZengXBrownLPEndruweitA, et al.Geometrical modelling of 3D woven reinforcements for polymer composites: prediction of fabric permeability and composite mechanical properties. Compos Appl Sci Manuf2014; 56: 150–160. DOI: 10.1016/j.compositesa.2013.10.004.
26.
LiuXChenF. Defects characterization in CFRP using X-ray computed tomography. Polym Polym Compos2016; 24: 149–154. DOI: 10.1177/0967391116024002.
27.
SenckSScheererMRevolV, et al.Microcrack characterization in loaded CFRP laminates using quantitative two- and three-dimensional X-ray dark-field imaging. Compos Appl Sci Manuf2018; 115: 206–214. DOI: 10.1016/j.compositesa.2018.09.023.
28.
RoseMNivertySSchmidtB, et al.X-ray computer tomography (XCT) of fatigue damage in laser-machined versus milled carbon fiber reinforced polymer matrix composites. Eng Fract Mech2021; 252: 107820. DOI: 10.1016/j.engfracmech.2021.107820.
29.
SketFRodríguez-HortaláMMolina-AldareguíaJM, et al.In situtomographic investigation of damage development in ±45° carbon fibre reinforced laminates. Mater Sci Technol2014; 31: 587–593. DOI: 10.1179/1743284714Y.0000000561.
30.
StamopoulosAGTserpesKIPantelakisSG. Multiscale finite element prediction of shear and flexural properties of porous CFRP laminates utilizing X-ray CT data. Theor Appl Fract Mech2018; 97: 303–313. DOI: 10.1016/j.tafmec.2017.04.020.
31.
PathakAKYokozekiTImadaM, et al.In-situ observation of tensile failure mode in cross-ply CFRP laminates using Talbot-Lau interferometry. Compos Struct2020; 253: 112758. DOI: 10.1016/j.compstruct.2020.112758.
32.
StraumitIHahnCWintersteinE, et al.Computation of permeability of a non-crimp carbon textile reinforcement based on X-ray computed tomography images. Compos Appl Sci Manuf2016; 81: 289–295. DOI: 10.1016/j.compositesa.2015.11.025.
33.
LiuYStraumitIVasiukovD, et al.Prediction of linear and non-linear behavior of 3D woven composite using mesoscopic voxel models reconstructed from X-ray micro-tomography. Compos Struct2017; 179: 568–579. DOI: 10.1016/j.compstruct.2017.07.066.
34.
WijayaWAliMAUmerR, et al.An automatic methodology to CT-scans of 2D woven textile fabrics to structured finite element and voxel meshes. Compos Appl Sci Manuf2019; 125: 105561. DOI: 10.1016/j.compositesa.2019.105561.
35.
NaouarNVasiukovDParkC, et al.Meso-FE modelling of textile composites and X-ray tomography. J Mater Sci2020; 55: 16969–16989. DOI: 10.1007/s10853-020-05225-x.
36.
WojakJAngeliniEDBlochI. Joint variational segmentation of CT-PET data for tumoral lesions. In: 2010 IEEE international symposium on biomedical imaging: from nano to macro. IEEE, 2010; 217–220.
37.
VittiA. The Mumford–Shah variational model for image segmentation: an overview of the theory, implementation and use. ISPRS J Photogrammetry Remote Sens2012; 69: 50–64. DOI: 10.1016/j.isprsjprs.2012.02.005.
38.
NaouarNVidal-SalleESchneiderJ, et al.3D composite reinforcement meso FE analyses based on X-ray computed tomography. Compos Struct2015; 132: 1094–1104. DOI: 10.1016/j.compstruct.2015.07.005.
39.
MumfordDShahJ. Optimal approximations by piecewise smooth functions and associated variational problems’. Commun Pure Appl Math1989; 42: 577–685. DOI: 10.1002/CPA.3160420503.
40.
HofbauerHJalilianEUhlA. Exploiting superior CNN-based iris segmentation for better recognition accuracy. Pattern Recogn Lett2019; 120: 17–23. DOI: 10.1016/j.patrec.2018.12.021.
41.
ReinhardtMJacobASadeghnejadS, et al.Benchmarking conventional and machine learning segmentation techniques for digital rock physics analysis of fractured rocks. Environ Earth Sci2022; 81: 1–20. DOI: 10.1007/s12665-021-10133-7.
42.
ShiQLiuYZhangS, et al.Channel estimation for WiFi prototype systems with super-resolution image recovery. In: ICC 2019-2019 IEEE international conference on communications (ICC). IEEE, 2019; 1–6.
43.
NguyenLDLinDLinZ, et al.Deep CNNs for microscopic image classification by exploiting transfer learning and feature concatenation. In: 2018 IEEE international symposium on circuits and systems (ISCAS). IEEE, 2018; 1–5.
44.
KimHJungW-KParkY-C, et al.Broken stitch detection method for sewing operation using CNN feature map and image-processing techniques. Expert Syst Appl2022; 188: 116014. DOI: 10.1016/j.eswa.2021.116014.
45.
HeKZhangXRenS, et al.Deep residual learning for image recognition. In: Proceedings of the IEEE conference on computer vision and pattern recognition, 2016; 770–778.
46.
RonnebergerOFischerPBroxT. U-net: convolutional networks for biomedical image segmentation. In: International Conference on Medical image computing and computer-assisted intervention. Springer, 2015; 234–241.
47.
CaoKZhangX. An improved res-unet model for tree species classification using airborne high-resolution images. Rem Sens2020; 12: 1128. DOI: 10.3390/rs12071128.
48.
ZhouZSiddiqueeMMRTajbakhshN, et al.UNet++: a nested U-net architecture for medical image segmentation. Deep Learn Med Image Anal Multimodal Learn Clin Decis Support2018; 11045: 3–11. DOI: 10.1007/978-3-030-00889-5_1 (accessed on 01 September 2018).
49.
DuGCaoXLiangJ, et al.Medical image segmentation based on u-net: a review. J Imag Sci Technol2020; 64: 1-12. DOI: 10.2352/J.ImagingSci.Technol.2020.64.2.020508.
50.
ChaudharyVButtarPKSachanMK. Satellite imagery analysis for road segmentation using U-Net architecture. J Supercomput2022: 1–16. DOI: 10.1007/s11227-022-04379-6.
51.
LinDLiYPrasadS, et al.CAM-Guided multi-path decoding U-net with triplet feature regularization for defect detection and segmentation. Knowl Base Syst2021; 228: 107272. DOI: 10.1016/j.knosys.2021.107272.
52.
BertoldoJPDecencièreERyckelynckD, et al.A modular U-Net for automated segmentation of X-ray tomography images in composite materials. Frontiers in Materials2021; 8: 761229. DOI: 10.3389/fmats.2021.761229.
53.
PanZGuBSunB, et al.Progressive failure of 3-D textile composites under impact loadings. Compos Struct2017; 168: 710–724. DOI: 10.1016/j.compstruct.2017.02.067.
54.
KernbachJMStaartjesVE. Foundations of machine learning-based clinical prediction modeling: Part II—generalization and overfitting. Machine Learning in Clinical Neuroscience2022: 15–21. DOI: 10.1007/978-3-030-85292-4_3.
55.
BhaduriAGuptaAGraham-BradyL. Stress field prediction in fiber-reinforced composite materials using a deep learning approach. Compos B Eng2022; 238: 109879. DOI: 10.1016/j.compositesb.2022.109879.
56.
JungK-CChangS-H. Advanced deep learning model-based impact characterization method for composite laminates. Compos Sci Technol2021; 207: 108713. DOI: 10.1016/j.compscitech.2021.108713.
57.
SchwendickeFElhennawyKParisS, et al.Deep learning for caries lesion detection in near-infrared light transillumination images: a pilot study. J Dent2020; 92: 103260. DOI: 10.1016/j.jdent.2019.103260.
58.
WurmMStarkTZhuXX, et al.Semantic segmentation of slums in satellite images using transfer learning on fully convolutional neural networks. ISPRS J Photogrammetry Remote Sens2019; 150: 59–69. DOI: 10.1016/j.isprsjprs.2019.02.006.
59.
KiyassehDSwistonAChenR, et al.Segmentation of left atrial MR images via self-supervised semi-supervised meta-learning. Cham: Springer International Publishing, 2021; 13–24.
60.
JhaDSmedsrudPHJohansenD, et al.A comprehensive study on colorectal polyp segmentation with ResUNet++, conditional random field and test-time augmentation. IEEE J Biomed Health Inform2021; 25: 2029–2040. DOI: 10.1109/JBHI.2021.3049304 (accessed on 01 June 2021).
61.
TolkachevASirazitdinovIKholiavchenkoM, et al.Deep learning for diagnosis and segmentation of pneumothorax: the results on the kaggle competition and validation against radiologists. IEEE J Biomed Health Inform2021; 25: 1660–1672. DOI: 10.1109/JBHI.2020.3023476 (accessed on 22 September 2020).
62.
El AsnaouiKChawkiY. Using X-ray images and deep learning for automated detection of coronavirus disease. Journal of Biomolecular Structure Dynamics2021; 39: 3615–3626. DOI: 10.1080/07391102.2020.1767212.
