Tensile Strength Prediction of Glass/Epoxy Composite Structures Using Machine Learning and Response Surface Methodology

Shahsavari, Milad; Zamani, Mohammad Reza; Parsa, Hosein

doi:10.22075/macs.2026.37511.1839

Tensile Strength Prediction of Glass/Epoxy Composite Structures Using Machine Learning and Response Surface Methodology

Document Type : Research Article

Authors

Milad Shahsavari

Mohammad Reza Zamani

Hosein Parsa

Department of Mechanical Engineering, Malek Ashtar University of Technology, Tehran, Iran

10.22075/macs.2026.37511.1839

Abstract

Analytical methods for evaluating composite structures face complexities and limitations owing to the anisotropic nature of their material properties. Integrating numerical simulations with artificial intelligence (AI) techniques not only overcomes these limitations but also enhances design flexibility, diversity, speed, and accuracy. Tensile strength is a critical design parameter for such structures, and its precise determination significantly affects the design quality. This study calculates the tensile strength of glass/epoxy laminated composites using two approaches: Machine Learning (ML) and Response Surface Methodology (RSM). RSM is a widely used mathematical modeling technique to establish relationships between independent and dependent variables, while ML—a subset of AI—offers advanced predictive capabilities. Both methods require a Design of Experiments (DoE) and extensive datasets, which were optimized here using Latin Hypercube Sampling (LHS) to reduce sample tests while improving accuracy. Finite Element Analysis (FEA) was employed for dataset generation, validated through experimental tests. The developed models assessed the influence of design parameters (e.g., layer count and orientation) on tensile strength, demonstrating strong alignment with the dataset. This study accelerates the composite design process and enhances the understanding of mechanical behavior.

Keywords

Laminated Composite

Tensile Strength

Response Surface Methodology

Machine Learning

Stacking ensemble

Subjects

Composite Materials and Structures

[1] Vijayakumar, S. and Palanikumar, K., 2019. Mechanical property evaluation of hybrid reinforced epoxy composite. Materials Today: Proceedings, 16, pp. 430–438.

[2] Dalfi, H., Al-Obaidi, A., Tariq, A., Razzaq, A. and Rafiee, R., 2023. Optimization of the mechanical performance and damage failure characteristics of laminated composites based on fiber orientation. Frontiers of Structural and Civil Engineering, 17(9), pp. 1357–1369.

[3] Nagaraja, K.C., Rajanna, S., Prakash, G.S., Praveennath, G. and Alipour, M., 2020. Studying the effect of different carbon and glass fabric stacking sequence on mechanical properties of epoxy hybrid composite laminates. Composites Communications, 21, p. 100425.

[4] Kumar, N., Kumar, S., Tyagi, A., Kumar, D. and Shekhar, C., 2020. Finite element and micromechanical analysis of glass/epoxy laminated composite with different orientations. Materials Today: Proceedings, 28, pp. 1899–1903.

[5] Muralidhara, B., Kumaresh, S.P. and Suresha, B., 2020. The effect of fiber architecture on the mechanical properties of carbon/epoxy composites. Materials Today: Proceedings, 22, pp. 1755–1764.

[6] Ornaghi, H.L., Roberta, R.M.N., Francisco, M.M. and Sabu, T., 2021. Modeling of dynamic mechanical curves of kenaf/polyester composites using surface response methodology. Journal of Applied Polymer Science, 139(18), pp. 78–87.

[7] Saravanakumar, S., Sathiyamurthy, P. and Pathmanaban, P., 2024. Integrating machine learning and response surface methodology for analyzing anisotropic mechanical properties of biocomposites. Composite Interfaces, Taylor & Francis, pp. 101–140.

[8] Khanlou, H.M., Sadollah, A., Ang, B.C., Kim, J.H., Talebian, S. and Ghadimi, A., 2014. Prediction and optimization of electrospinning parameters for polymethyl methacrylate nanofiber fabrication using response surface methodology and artificial neural networks. Neural Computing and Applications, 25(3-4), pp. 767–777.

[9] Fnides, M., Amroune, S., Belaadi, A., Saada, K., Chai, B.X., Abdullah, M.M.S., Alshaikh, I.M.H., Ghernaout, D. and Al-Khawlani, A., 2024. Modeling and optimizing the alkaline treatment process to enhance the date palm fibers' tensile mechanical properties using RSM. Journal of Natural Fibers, 21(1), 2384663.

[10] Saha, M., Singh, H., Srivastava, V. and Singh, M.K., 2024. Analysis and optimization of delamination factor for microwave-cured pineapple leaf fiber polymer composite through ANOVA analysis. Applied Science and Engineering Progress, 17(4), Article 7531. https://doi.org/10.14416/j.asep.2024.08.010

[11] Singh, M.K., Verma, N. and Zafar, S., 2019. Optimization of process parameters of microwave processed PLLA/coir composites for enhanced mechanical behavior. Journal of Physics: Conference Series, 1240(1), p. 012038. doi.org/10.1088/1742-6596/1240/1/012038.

[12] Theobald, O., 2017. Machine Learning for Absolute Beginners: A Plain English Introduction (2nd ed.). United States Press, pp. 287–296.

[13] Flach, P., 2012. Machine Learning: The Art and Science of Algorithms That Make Sense of Data. Cambridge University Press.

[14] Okafor, C., Iweriolor, S., Ani, O.I., Shahnawaz, A., Mehfuz, S., Onyekachukwu, G., Emmanuel, O., Emeka, S., Echezona, I. and Peter, O., 2023. Advances in machine learning-aided design of reinforced polymer composite and hybrid material systems. Hybrid Advances, 2, p. 100026.

[15] Tao, F., Liu, X., Du, H. and Yu, W., 2021. Learning composite constitutive laws via coupling Abaqus and deep neural network. Composite Structures, 272, p. 114137.

[16] Huang, J.S., Liew, J.X., Ademiloye, A.S. and Liew, K.M., 2021. Artificial intelligence in materials modeling and design. Archives of Computational Methods in Engineering, 28(5), pp. 3399–3413.

[17] Sharma, H., Arora, G., Kumar, R., Debnath, S. and Siengchin, S., 2025. Machine learning-based study of hardness in polypropylene/carbon nanotube and low-density polyethylene/carbon nanotube composites. Discover Materials, 5(1), Article 176. doi.org/10.1007/s43939-025-00176-z.

[18] Olsson, A.M.J. and Sandberg, G.E., 2002. Latin hypercube sampling for stochastic finite element analysis. Journal of Engineering Mechanics, 128(1), pp. 121–125.

[19] Singh, V.P., Jain, S.K. and Tiwari, A., 2007. Risk and Reliability Analysis. American Society of Civil Engineers.

[20] Kwon, H.H. and Moon, Y.I., 2006. Improvement of overtopping risk evaluations using probabilistic concepts for existing dams. Stochastic Environmental Research and Risk Assessment, 20(4), pp. 223–237.

[21] Chung, H., Senior, M., Wong, K.P., Lee, H.W. and Zhang, J.H., 2009. Probabilistic load flow evaluation with hybrid Latin hypercube sampling and Cholesky decomposition. IEEE Transactions on Power Systems, 24(2), pp. 120–126.

[22] Choi, Y., Song, D., Yoon, S. and Koo, J., 2021. Comparison of factorial and Latin hypercube sampling designs for meta-models of building heating and cooling loads. Energies, 14(2), 512.

[23] Wang, X., Liu, N., Zhau, J., Li, Z., Meng, Q., Guan, Z. and Du, S., 2023. Highly efficient simulation of composites by determining failure initiation and fracture angle with artificial neural networks. Composite Structures, 307, p. 116644.

[24] Qian, Z., Zhang, N., Liu, Y. and Cheng, W., 2019. Prediction of mechanical properties of carbon fiber based on cross-scale FEM and machine learning. Composite Structures, 212, pp. 199–206.

[25] Fontes, A. and Shadmehri, F., 2023. Data-driven failure prediction of fiber-reinforced polymer composite materials. Engineering Applications of Artificial Intelligence, 120, p. 105834.

[26] ASTM D3039, 2017. Standard test method for tensile properties of polymer matrix composite materials. Annual Book of ASTM Standards.

[27] Kaw, A.K., 2006. Mechanics of Composite Materials. Taylor & Francis.

[28] Shin, H., Ko, K.J., Park, W.J., Han, D.H., Yeom, I. and Lee, K.S., 2024. Machine learning models for the noninvasive diagnosis of bladder outlet obstruction and detrusor underactivity in men with lower urinary tract symptoms. International Neurourology Journal, 28(Suppl 2), pp. S74–S81.

[29] Kanaparthi, V., 2023. Credit risk prediction using ensemble machine learning algorithms. In: 2023 International Conference on Inventive Computation Technologies (ICICT), pp. 41–47.

[30] Chagahi, M., Mohammadi, D.M., Moshiri, B. and Piran, J., 2024. Cardiovascular disease detection using a novel stack-based ensemble classifier with aggregation layer, DOWA operator, and feature. Computers in Biology and Medicine, 173, 108345.

[31] Gay, D., Hoa, S.V. and Tsai, S.W., 2003. Composite Materials: Design and Applications. CRC Press.

[32] Daniel, I.M. and Ishai, O., 2006. Engineering Mechanics of Composite Materials. Oxford University Press.