All Issue

2024 Vol.37, Issue 4

Research Paper

Deep Learning-based Fracture Mode Determination in Composite Laminates

복합 적층판의 딥러닝 기반 파괴 모드 결정

Muhammad Muzammil Azad1
,
Atta Ur Rehman Shah2
,
M.N. Prabhakar3
,
Heung Soo Kim4†
무하마드 무자밀 아자드1
,
아타 우르 레만 샤2
,
M.N. 프라브하카르3
,
김 흥수4†
1Ph.D. Student, Department of Mechanical Engineering, Dongguk University, Seoul, 04620, Korea
2Associate Professor, Department of Mechanical Engineering, COMSATS University, Wah Campus, 47040, Pakistan
3Research Professor, Department of Mechanical Engineering, Changwon National University, Changwon, 51140, Korea
4Professor, Department of Mechanical, Robotics and Energy Engineering, Dongguk University, Seoul, 04620, Korea
1동국대학교 기계공학과 박사과정
2COMSATS 대학교 기계공학과 부교수
3창원대학교 기계공학과 연구교수
4동국대학교 기계로봇에너지공학과 교수
†Corresponding Author
31 August 2024. pp. 225-232
PDF XML Citation
Abstract

This study focuses on the determination of the fracture mode in composite laminates using deep learning. With the increase in the use of laminated composites in numerous engineering applications, the insurance of their integrity and performance is of paramount importance. However, owing to the complex nature of these materials, the identification of fracture modes is often a tedious and time-consuming task that requires critical domain knowledge. Therefore, to alleviate these issues, this study aims to utilize modern artificial intelligence technology to automate the fractographic analysis of laminated composites. To accomplish this goal, scanning electron microscopy (SEM) images of fractured tensile test specimens are obtained from laminated composites to showcase various fracture modes. These SEM images are then categorized based on numerous fracture modes, including fiber breakage, fiber pull-out, mix-mode fracture, matrix brittle fracture, and matrix ductile fracture. Next, the collective data for all classes are divided into train, test, and validation datasets. Two state-of-the-art, deep learning-based pre-trained models, namely, DenseNet and GoogleNet, are trained to learn the discriminative features for each fracture mode. The DenseNet models shows training and testing accuracies of 94.01% and 75.49%, respectively, whereas those of the GoogleNet model are 84.55% and 54.48%, respectively. The trained deep learning models are then validated on unseen validation datasets. This validation demonstrates that the DenseNet model, owing to its deeper architecture, can extract high-quality features, resulting in 84.44% validation accuracy. This value is 36.84% higher than that of the GoogleNet model. Hence, these results affirm that the DenseNet model is effective in performing fractographic analyses of laminated composites by predicting fracture modes with high precision.

Keywords
deep learning
fracture mode
composite laminates
transfer learning
DenseNet
GoogleNet

본 논문에서는 딥러닝을 활용하여 복합재 적층판의 파괴 모드를 결정하는 방법을 제안하였다. 수많은 엔지니어링 응용 분야에서 적층 복합재의 사용이 증가함에 따라 무결성과 성능을 보장하는 것이 중요해졌다. 그러나 재료의 이방성으로 인해 복잡하게 나타나는 파괴모드를 식별하는 것은 도메인 지식이 필요하고, 시간이 많이 드는 작업이다. 따라서 이러한 문제를 해결하기 위해 본 연구에서는 인공 지능(AI) 기술을 활용하여 적층 복합재의 파괴 모드 분석을 자동화하는 것을 목표로 하였다. 이 목표를 달성하기 위해 적층된 복합재에서 파손된 인장 시험편의 주사 전자 현미경(SEM) 이미지를 얻어 다양한 파괴 모드를 확보하였다. 이러한 SEM 이미지는 섬유 파손, 섬유 풀아웃, 혼합 모드 파괴, 매트릭스 취성 파손 및 매트릭스 연성 파손과 같은 다양한 파손 모드를 기준으로 분류하였다. 다음으로 모든 클래스의 집합 데이터를 학습, 테스트, 검증 데이터 세트로 구분하였다. 두 가지 딥 러닝 기반 사전 훈련 모델인 DenseNet과 GoogleNet을 이용해 각 파괴 모드에 대한 차별적 특징을 학습하도록 훈련하였다. DenseNet 및 GoogleNet 모델은 각각 (94.01% 및 75.49%) 및 (84.55% 및 54.48%)의 훈련 및 테스트 정확도를 보여주었다. 그런 다음 훈련된 딥 러닝 모델은 검증 데이터 세트를 활용해 검증하였다. 더 깊은 아키텍처로 인해 DenseNet 모델이 고품질 특징을 추출하여 84.44% 검증 정확도(GoogleNet 모델보다 36.84% 더 높음)를 얻을 수 있음을 확인하였다. 이는 DenseNet 모델이 높은 정밀도로 파괴 모드를 예측함으로써 적층 복합재의 파손 분석을 수행하는 데 효과적이라는 것을 알 수 있다.

키워드
딥 러닝
파괴 모드
적층 복합재
전이 학습
DenseNet
GoogleNet
References
1

Ahmad, H., Shah, A.U.R., Afaq, S.K., Azad, M.M., Arif, S., Siddiqi, M.U.R., Xie, L. (2024) Development and Characterization of Kevlar and Glass Fibers Reinforced Epoxy/vinyl Ester Hybrid Resin Composites, Polym. Compos., 45(9), pp.8133~8146.

10.1002/pc.28329
2

Avilés-Cruz, C., Aguilar-Sanchez, M., Vargas-Arista, B., Garfias-Garcia, E. (2024) A New Machine Learning-Based Evaluation of Ductile Fracture, Eng. Fract. Mech., 302, p.110072.

10.1016/j.engfracmech.2024.110072
3

Azad, M.M., Ejaz, M., Shah, A.U.R., Afaq, S.K., Song, J.-I. (2022) A Bio-Based approach to Simultaneously Improve Flame Retardancy, Thermal Stability and Mechanical Properties of Nano-Silica Filled Jute/Thermoplastic Starch Composite, Mater. Chem. & Phys., 289, p.126485.

10.1016/j.matchemphys.2022.126485
4

Azad, M.M., Kim, H.S. (2024) Hybrid Deep Convolutional Networks for the Autonomous Damage Diagnosis of Laminated Composite Structures, Compos. Struct., 329, p.117792.

10.1016/j.compstruct.2023.117792
5

Azad, M.M., Kim, S., Cheon, Y.B., Kim, H.S. (2023) Intelligent Structural Health Monitoring of Composite Structures using Machine Learning, Deep Learning, and Transfer Learning: A Review, Adv. Compos. Mater.,33(2) pp.162~188.

10.1080/09243046.2023.2215474
6

Azad, M.M., Kumar, P., Kim, H.S. (2024) Delamination Detection in CFRP Laminates using Deep Transfer Learning with Limited Experimental Data, J. Mater. Res. & Technol., 29, pp.3024~3035.

10.1016/j.jmrt.2024.02.067
7

Bastidas-Rodriguez, M.X., Prieto-Ortiz, F.A., Espejo, E. (2016) Fractographic Classification in Metallic Materials by using Computer Vision, Eng. Fail. Anal., 59, pp.237~252.

10.1016/j.engfailanal.2015.10.008
8

Chen, X. (2020) Fractographic Analysis of Sandwich Panels in a Composite Wind Turbine Blade using Optical Microscopy and X-ray Computed Tomography, Eng. Fail. Anal., 111, p.104475.

10.1016/j.engfailanal.2020.104475
9

Ejaz, M., Azad, M.M., Shah, A.U.R., Afaq, S.K., Song, J.I. (2022) Synergistic Effect of Aluminum Trihydrate and Zirconium Hydroxide Nanoparticles on Mechanical Properties, Flammability, and Thermal Degradation of Polyester/Jute Fiber Composite, Cellul., 29(3), pp.1775~1790.

10.1007/s10570-022-04417-9
10

Ejaz, M., Azad, M.M., Shah, A.U.R., Afaq, S.K., Song, J.I. (2020) Mechanical and Biodegradable Properties of Jute/Flax Reinforced PLA Composites, Fibers & Polym., 21(11), pp.2635~2641.

10.1007/s12221-020-1370-y
11

Huang, G., Liu, Z., Maaten, L. van der, Weinberger, K.Q. (2017) Densely Connected Convolutional Networks, 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pp.4700~4708.

10.1109/CVPR.2017.243PMC5598342
12

Khalid, S., Kim, H.S. (2019) Progressive Damage Modeling of Inter and Intra Laminar Damages in Open Hole Tensile Composite Laminates, J. Comput. Struct. Eng. Inst. Korea, 32(4), pp.233~240.

10.7734/COSEIK.2019.32.4.233
13

Khalid, S., Kim, H.S. (2022) Recent Studies on Stress Function-Based Approaches for the Free Edge Stress Analysis of Smart Composite Laminates: A Brief Review, Multiscale Sci. & Eng., 4(3), pp.73~78.

10.1007/s42493-022-00079-8
14

Khan, A., Azad, M.M., Sohail, M., Kim, H.S. (2023) A Review of Physics-based Models in Prognostics and Health Management of Laminated Composite Structures, Int. J. Precis. Eng. & Manuf. - Green Technol., 10(6), pp.1615~1635.

10.1007/s40684-023-00509-4
15

Khan, A., Kim, H.S. (2022) A Brief Overview of Delamination Localization in Laminated Composites, Multiscale Sci. & Eng., 4(3), pp.102~110.

10.1007/s42493-022-00085-w
16

Lee, S.-Y., Huynh, T.-C., Park, J.-H., Kim, J.-T. (2019) Bolt-Loosening Detection using Vision-Based Deep Learning Algorithm and Image Processing Method, J. Comput. Struct. Eng. Inst. Korea, 32(4), pp.265~272.

10.7734/COSEIK.2019.32.4.265
17

Li, Z., Shah, A.R., Prabhakar, M.N., Song, J. (2017) Effect of Inorganic Fillers and Ammonium Polyphosphate on the Flammability, Thermal Stability, and Mechanical Properties of Abaca-Fabric/Vinyl Ester Composites, Fibers & Polym., 18(3), pp.555~562.

10.1007/s12221-017-6859-7
18

Mohammadi, R., Assaad, M., Imran, A., Fotouhi, M. (2024) Fractographic Analysis of Damage Mechanisms Dominated by Delamination in Composite Laminates: A Comprehensive Review, Polym. Testing, 134, p.108441.

10.1016/j.polymertesting.2024.108441
19

Prabhakar, M.N., Rehman Shah, A. ur, Song, J.-I. (2017) Improved Flame-Retardant and Tensile Properties of Thermoplastic Starch/Flax Fabric Green Composites, Carbohydr. Polym., 168, pp.201~211.

10.1016/j.carbpol.2017.03.03628457441
20

Prabhakar, M.N., Venakat Chalapathi, K., Atta Ur Rehman, S., Song, J. (2021) Effect of a Synthesized Chitosan Flame Retardant on the Flammability, Thermal Properties, and Mechanical Properties of Vinyl Ester/Bamboo Nonwoven Fiber Composites, Cellul., 28(18), pp.11625~11643.

10.1007/s10570-021-04252-4
21

Rangappa, S.M., Puttegowda, M., Parameswaranpillai, J., Siengchin, S., Gorbatyuk, S. (2022) Advances in Bio-Based Fiber: Moving Towards a Green Society, Woodhead Publishing, p.834.

22

Shah, A.U.R., Prabhakar, M.N., Saleem, M., Song, J.I. (2017) Development of Biowaste Encapsulated Polypropylene Composites: Thermal, Optical, Dielectric, Flame Retardant, Mechanical, and Morphological Properties, Polym. Compos., 38(2), pp.236~243.

10.1002/pc.23580
23

Shah, A.U.R., Prabhakar, M.N., Wang, H., Song, J.I. (2018) The Influence of Particle Size and Surface Treatment of Filler on the Properties of Oyster Shell Powder Filled Polypropylene Composites, Polym. Compos., 39(7), pp.2420~2430.

10.1002/pc.24225
24

Szegedy, C., Liu, W., Jia, Y., Sermanet, P., Reed, S., Anguelov, D., Erhan, D., Vanhoucke, V., Rabinovich, A. (2015) Going Deeper with Convolutions, 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), IEEE, Boston, MA, USA, pp.1~9.

10.1109/CVPR.2015.7298594
25

Tang, K., Zhang, P., Zhao, Y., Zhong, Z. (2024) Deep Learning-Based Semantic Segmentation for Morphological Fractography, Eng. Fract. Mech., 303, p.110149.

10.1016/j.engfracmech.2024.110149
26

Tsopanidis, S., Osovski, S. (2021) Unsupervised Machine Learning in Fractography: Evaluation and Interpretation, Mater. Charact., 182, p.111551.

10.1016/j.matchar.2021.111551
27

Zulfiqar, A., Shah, A.U.R., Khalil, M.S., Azad, M.M., Zulfiqar, Y., Naseem, M.S., Song, J.-I. (2024) Enhancing Properties of Jute/Starch Bio-Composite Material through Incorporation of Magnesium Carbonate Hydroxide Pentahydrate: A Sustainable Approach, Mater. Chem. & Phys., 314.

10.1016/j.matchemphys.2023.128690
Information
  • Publisher :Computational Structural Engineering Institute of Korea
  • Publisher(Ko) :한국전산구조공학회
  • Journal Title :Journal of the Computational Structural Engineering Institute of Korea
  • Journal Title(Ko) :한국전산구조공학회 논문집
  • Volume : 37
  • No :4
  • Pages :225-232
  • Received Date : 2024-06-04
  • Revised Date : 2024-06-24
  • Accepted Date : 2024-06-24
  • DOI :https://doi.org/10.7734/COSEIK.2024.37.4.225
Journal Informaiton Journal of the Computational Structural Engineering Institute of Korea Journal of the Computational Structural Engineering Institute of Korea
Online Submission
  • NRF
  • KOFST
  • crossref crossmark
  • crossref cited-by
  • crosscheck
  • orcid
  • open access
  • ccl
Journal Informaiton Journal Informaiton - close