journal article·2026-06Open access

On the machine learning approach for computational homogenization of short fiber-reinforced composites with elastoplastic behavior

Ayşe Polat, Aiman Tariq, Babür Deliktaş, Murat Yazıcı, George Z Voyiadjis

Abstract

This study aims to develop machine learning (ML)-based homogenization models to efficiently predict the effective elastoplastic properties of short fiber-reinforced composites (SFRCs), reducing the reliance on computationally expensive micromechanical simulations. Sobol sampling is employed to generate training data by varying constitutive and microstructural parameters of representative volume element. Several ML models including artificial neural networks (ANN), support vector regression (SVR), random forest (RF), and extreme gradient boosting (XGB) are trained to predict homogenized stress-strain responses. A new approach is introduced that decomposes the stress-strain response into elastic and plastic components, allowing the ML models to learn these components separately and effectively. Additionally, the Taguchi method (L27 orthogonal array) is employed to minimize simulation runs and evaluate parameter sensitivity through ANOVA. The best-performing ML model is implemented in a finite element analysis (FEA) of an automotive component. Among all models, ANN demonstrated the highest accuracy in predicting the macroscopic elastoplastic response across a wide range of input parameters. Finally, the ANN-based elastoplastic material model is validated on a real-world macroscopic structure by implementing it into the finite element analysis of an automotive component. The results demonstrate that ML-based homogenization closely matches the traditional homogenization methods and also highlights its ability to effectively capture nonlinear material behavior.
Source: crossref · Quality: baseline

Research context

OpenAlex topics: Composite Material Mechanics · Topology Optimization in Engineering · Advanced Mathematical Modeling in Engineering

SciVX discovery tags are terminology-based. OpenAlex topics are source-provided classifications.

Affiliations

Bursa Uludag University

Louisiana State University

Funding

Türkiye Bilimsel ve Teknolojik Araştırma Kurumu

References (73)

Show source-provided references
1. 10.1016/j.compositesa.2023.10792910.1016/j.compositesa.2023.107929 ↗
2. 10.1016/j.ijsolstr.2021.11114110.1016/j.ijsolstr.2021.111141 ↗
3. 10.1177/002199831557431410.1177/0021998315574314 ↗
4. 10.1007/s00707-010-0338-110.1007/s00707-010-0338-1 ↗
5. 10.47852/bonviewAAES3202172310.47852/bonviewAAES32021723 ↗
6. 10.1016/j.compstruct.2023.11772610.1016/j.compstruct.2023.117726 ↗
7. 10.1061/(ASCE)0733-9399(2001)127:7(636)10.1061/(ASCE)0733-9399(2001)127:7(636) ↗
8. 10.1016/0001-6160(73)90064-310.1016/0001-6160(73)90064-3 ↗
9. 10.1016/0167-6636(87)90005-610.1016/0167-6636(87)90005-6 ↗
10. 10.1016/S0020-7683(03)00013-110.1016/S0020-7683(03)00013-1 ↗
11. 10.1016/j.compstruct.2010.09.00210.1016/j.compstruct.2010.09.002 ↗
12. 10.1016/j.compositesa.2015.02.01710.1016/j.compositesa.2015.02.017 ↗
13. 10.1016/j.mechmat.2023.10474710.1016/j.mechmat.2023.104747 ↗
14. 10.1016/0045-7825(90)90148-F10.1016/0045-7825(90)90148-F ↗
15. 10.1016/S0045-7825(03)00348-710.1016/S0045-7825(03)00348-7 ↗
16. 10.1016/j.ijmecsci.2011.06.00210.1016/j.ijmecsci.2011.06.002 ↗
17. 10.1177/073168441877944110.1177/0731684418779441 ↗
18. 10.1016/S0020-7683(03)00346-910.1016/S0020-7683(03)00346-9 ↗
19. 10.1016/j.jcp.2006.09.01910.1016/j.jcp.2006.09.019 ↗
20. 10.1016/j.addma.2022.10259710.1016/j.addma.2022.102597 ↗
21. 10.1016/j.cma.2016.02.00110.1016/j.cma.2016.02.001 ↗
22. Tariq A, Abualshar B, Deliktas B, et al. ANN-based evaluation system for erosion resistant highway shoulder rocks. Int J Geoeng 2024; 15(1): 17.
23. 10.1177/0309324724127839110.1177/03093247241278391 ↗
24. Polat A, Tariq A, Okay F, et al. Investigation of the critical buckling load of a column with linearly varying moment of inertia using analytical, numerical, and hybrid machine learning approaches. J Strain Anal Eng Des. 2025; 03093247251337987.
25. 10.1007/s11340-024-01136-z10.1007/s11340-024-01136-z ↗
26. 10.1016/j.compositesb.2021.10873610.1016/j.compositesb.2021.108736 ↗
27. Tariq A, Polat A, Deliktaş B. Boosting machine learning algorithms for predicting the macroscopic material behavior of continuous fiber reinforced composite. J Reinforc Plast Compos. 2024; 07316844241292694.
28. 10.1016/j.ijsolstr.2024.11309310.1016/j.ijsolstr.2024.113093 ↗
29. 10.1016/j.compstruct.2014.11.05210.1016/j.compstruct.2014.11.052 ↗
30. 10.1016/j.ijsolstr.2023.11245210.1016/j.ijsolstr.2023.112452 ↗
31. 10.1016/j.ijsolstr.2023.11233410.1016/j.ijsolstr.2023.112334 ↗
32. 10.1002/nme.495310.1002/nme.4953 ↗
33. 10.1016/j.mechmat.2023.10459710.1016/j.mechmat.2023.104597 ↗
34. 10.1016/j.cma.2025.11774310.1016/j.cma.2025.117743 ↗
35. 10.1088/1361-651X/aced5a10.1088/1361-651X/aced5a ↗
36. 10.3390/polym1519396210.3390/polym15193962 ↗
37. Tariq A, Uzun B, Akpinar M, et al. Size dependent dynamics of a bi-directional functionally graded nanobeam via machine learning methods. Adv Nano Res 2025; 18(1): 33–52.
38. 10.1007/s10570-022-04631-510.1007/s10570-022-04631-5 ↗
39. 10.1080/15376494.2020.186407110.1080/15376494.2020.1864071 ↗
40. 10.1016/j.engappai.2025.11158910.1016/j.engappai.2025.111589 ↗
41. 10.1007/s00289-022-04636-x10.1007/s00289-022-04636-x ↗
42. Karna SK, Sahai R. An overview on Taguchi method. Int J Eng Math Sci 2012; 1(1): 1–7.
43. Gopalsamy BM, Mondal B, Ghosh S. Taguchi method and ANOVA: an approach for process parameters optimization of hard machining while machining hardened steel. J Sci Ind Res 2009; 68: 686–695.
44. 10.1016/j.compscitech.2024.11096110.1016/j.compscitech.2024.110961 ↗
45. 10.1016/j.compstruct.2024.11817510.1016/j.compstruct.2024.118175 ↗
46. 10.3390/polym1211248110.3390/polym12112481 ↗
47. 10.3390/mi202012910.3390/mi2020129 ↗
48. 10.3390/su13201117210.3390/su132011172 ↗
49. 10.1002/mats.20170010410.1002/mats.201700104 ↗
50. 10.1016/j.mechmat.2021.10405910.1016/j.mechmat.2021.104059 ↗
51. 10.1016/j.ijplas.2010.06.00410.1016/j.ijplas.2010.06.004 ↗
52. 10.1016/j.euromechsol.2024.10548510.1016/j.euromechsol.2024.105485 ↗
53. 10.1016/j.mtcomm.2019.10055210.1016/j.mtcomm.2019.100552 ↗
54. 10.1122/1.54994510.1122/1.549945 ↗
55. 10.1016/j.compstruct.2017.03.02110.1016/j.compstruct.2017.03.021 ↗
56. 10.1016/j.promfg.2018.03.04010.1016/j.promfg.2018.03.040 ↗
57. 10.1016/j.matpr.2020.12.79210.1016/j.matpr.2020.12.792 ↗
58. 10.1002/nme.721110.1002/nme.7211 ↗
59. 10.1063/1.495335510.1063/1.4953355 ↗
60. 10.1016/j.polymer.2015.11.04910.1016/j.polymer.2015.11.049 ↗
61. 10.1016/j.ijplas.2013.02.00910.1016/j.ijplas.2013.02.009 ↗
62. 10.1016/0041-5553(67)90144-910.1016/0041-5553(67)90144-9 ↗
63. 10.1016/j.mbs.2021.10859310.1016/j.mbs.2021.108593 ↗
64. Kucherenko S Albrecht D Saltelli A. Exploring multi-dimensional spaces: a comparison of Latin hypercube and quasi Monte Carlo sampling techniques. arXiv preprint arXiv:1505.02350. 2015.
65. 10.1098/rsta.2015.020210.1098/rsta.2015.0202 ↗
66. 10.1007/s00419-025-02776-w10.1007/s00419-025-02776-w ↗
67. Chen T Guestrin C. Xgboost: a scalable tree boosting system. In: Proceedings of the 22nd ACM SIGKDD international conference on knowledge discovery and data mining San Francisco CA USA 13–17 August 2016 pp. 785–794.10.1145/2939672.2939785 ↗
68. 10.1038/nature1453910.1038/nature14539 ↗
69. 10.1016/j.compscitech.2024.11095110.1016/j.compscitech.2024.110951 ↗
70. Ross PJ. Taguchi techniques for quality engineering: loss function orthogonal experiments parameter and tolerance design. New York: McGraw-Hill 1996.
71. 10.1007/978-3-319-95342-7_710.1007/978-3-319-95342-7_7 ↗
72. Montgomery DC. Design and analysis of experiments. 7th ed. Hoboken, NJ: Wiley, 2008.
73. 10.1016/j.rser.2022.11291710.1016/j.rser.2022.112917 ↗