Hybrid SIO2/AL2O3 Filler Reinforcement in Glass Fiber/Epoxy Composites: Flexural and Tensile Behavior with Physics-Informed Neural Network Validation

Abstract

Glass fiber-reinforced polymer (GFRP) composites are widely used in structural applications, yet the inherent brittleness of the epoxy matrix limits their mechanical performance. While SiO2 and Al2O3 have individually been shown to improve polymer composite properties, the synergistic effect of their combined use as a hybrid filler on flexural behavior remains insufficiently explored. In this study, five composite formulations were fabricated with a fixed SiO2 content of 1 wt.% and Al2O3 contents of 0, 0.5, 1.0, 1.5, and 2.0 wt.%. Three-point bending tests were conducted at four support spans, and tensile tests were performed; the resulting flexural strength data were further modeled using a Physics-Informed Neural Network (PINN). Flexural strength decreased progressively with increasing support span across all sample groups. The sample containing 1 wt.% Al2O3 exhibited the highest flexural strength and modulus, corresponding to an increase of approximately 21% over the mono-filler reference at the 60 mm span. Tensile strength increased progressively with Al2O3 content up to 1.5 wt.%, corresponding to an improvement of approximately 21% over the mono-filler reference, while mechanical performance under both loading modes declined at 2.0 wt.% due to particle agglomeration. The PINN model achieved high predictive accuracy, with R2 values of 0.98 and 0.96 and mean absolute percentage errors of 3.81% and 5.37% on the training and test sets, respectively. These findings demonstrate that hybrid SiO2/Al2O3 reinforcement can meaningfully improve GFRP mechanical performance, with the optimum filler content depending on the dominant loading mode.

Keywords

• Physics-informed neural network
• Hybrid filler
• SiO2
• Al2O3
• Flexural strength
• Glass fiber-reinforced polymer

Ethical Statement

Ethics committee approval was not required for this study because of there was no study on animals or humans.

References

1. Abdi, A., Eslami-Farsani, R., & Khosravi, H. (2018). Evaluating the Mechanical Behavior of Basalt Fibers/Epoxy Composites Containing Surface-modified CaCO3 Nanoparticles [Article]. Fibers and Polymers, 19(3), 635-640.
2. Abebe, Y., Palani, S., & Sirahbizu, B. (2025). Effect of Al2O3 nanoparticle on mechanical properties of polyester/ glass-wool fiber reinforced polymer composites. Hybrid Advances, 10, 100472. /https://doi.org/10.1016/j.hybadv.2025.100472
3. Berladir, K., Antosz, K., Ivanov, V., & Mitaľová, Z. (2025). Machine Learning-Driven Prediction of Composite Materials Properties Based on Experimental Testing Data. Polymers, 17(5), 694.
4. Dinesh, C., & Yoganandh, J. (2026). Effect of Modified Silicon Carbide Nanofillers on the Mechanical Performance of Nettle Fiber Reinforced Polymer Composites. Polymer Composites, 47(7), 6536-6549. https://doi.org/10.1002/pc.70572
5. Ergün, R. K., Bayar, İ., & Köse, H. (2025). Data-Driven Prediction and Experimental Analysis of Wear in Micro/Nano SiO2-Reinforced Aramid Composites [Mikro/Nano SiO2 Takviyeli Aramid Kompozitlerde Aşınmanın Veriye Dayalı Tahmini ve Deneysel Analizi]. Journal of Materials and Mechatronics: A, 6(2), 425-441. https://doi.org/10.55546/jmm.1797773
6. Gowda, H. D. S., Raju, H., Kumar, V. G. P., Bharat, P. B., Suresha, B., & Bhat, S. K. (2026). A review of recent machine learning advances for predicting and optimizing mechanical and tribological properties of fiber-reinforced thermoset polymer nanocomposites. Discover Applied Sciences, 8(6), 636. https://doi.org/10.1007/s42452-026-08663-5
7. Guven, C., Kisa, M., Demircan, G., Ozen, M., & Kirar, E. (2024). Effect of seawater aging on mechanical, buckling, structural, and thermal properties of nano Al2O3 and TiO2-doped glass-epoxy nanocomposites [Article]. Polymer Composites, 45(8), 7376-7390. https://doi.org/10.1002/pc.28273
8. Khalid, S., Yazdani, M. H., Azad, M. M., Elahi, M. U., Raouf, I., & Kim, H. S. (2025). Advancements in Physics-Informed Neural Networks for Laminated Composites: A Comprehensive Review. Mathematics, 13(1), 17.

9. Kiran, M. D., Govindaraju, H. K., & Jayaraju, T. (2018). Evaluation of Mechanical Properties of Glass Fiber Reinforced Epoxy Polymer Composites with Alumina, Titanium dioxide and Silicon Carbide Fillers. Materials Today: Proceedings, 5(10, Part 3), 22355-22361. https://doi.org/10.1016/j.matpr.2018.06.602
10. Kose, H., Bayar, I., & Koray Ergün, R. (2025). Optimization of Hybrid Micro/Nano-SiO2 Particle Ratios for Enhancing Mechanical Performance of Aramid Composites: Optimization and Validation Using Taguchi and ANOVA Analyses. Advanced Engineering Materials, 27(23), e202501977.
11. Li, J., Zhang, L., & Qiao, J. (2026). Static and Dynamic Compressive Properties of Nano-Al2O3-Reinforced Epoxy Matrix Composites. Polymers, 18(10), 1228.
12. Malashin, I., Tynchenko, V., Gantimurov, A., Nelyub, V., & Borodulin, A. (2025). Physics-Informed Neural Networks in Polymers: A Review. Polymers, 17(8), 1108.
13. Mani, M. (2025). Silica nanoparticle-enhanced mechanical properties and energy absorption behavior of hybrid fiber-reinforced polymer composites for structural applications. Next Materials, 9, 101213.
14. Navas-Pinto, W., Díaz-Leime, P., Barrionuevo, G. O., Luna-Jaén, J., Sánchez-Sánchez, X., Navas-Cárdenas, C., & Cree, D. E. (2026). Machine Learning Approach for Mechanical Property Prediction of a Bio-Epoxy and Glass Fiber Composite Reinforced with Titanium Dioxide Nanoparticles [Article]. Journal of Composites Science, 10(3), 123.
15. Shettar, M., Bhat, A., Katagi, N. N., & Gowrishankar, M. C. (2025). Experimental Investigation on Mechanical Properties of Glass Fiber–Nanoclay–Epoxy Composites Under Water-Soaking: A Comparative Study Using RSM and ANN. Journal of Composites Science, 9(4), 195.
16. Shubham, S. K., Purohit, R., Meravi, M., Bajpai, G., & Bhagyalaxmi, M. (2025). Development and testing of polymer matrix hybrid composites reinforced with nano SiO2. AIP Conference Proceedings,
17. Singh, T., Gangil, B., Ranakoti, L., & Joshi, A. (2021). Effect of silica nanoparticles on physical, mechanical, and wear properties of natural fiber reinforced polymer composites [Article]. Polymer Composites, 42(5), 2396-2407.
18. Wang, W., & Thai, H.-T. (2025). A physics-informed neural network framework for laminated composite plates under bending. Thin-Walled Structures, 210, 113014. https://doi.org/10.1016/j.tws.2025.113014