Optimizing CFRP Thickness and Fiber Orientation for Seismic Retrofitting Using LightGBM and Whale Optimization Algorithm

Salama, Ahmed; Hassan, Ahmad Hassan; Aljarrah, Njood

doi:10.22115/scce.2025.1980

Optimizing CFRP Thickness and Fiber Orientation for Seismic Retrofitting Using LightGBM and Whale Optimization Algorithm

Document Type : Regular Article

Authors

Ahmed Salama ¹^{, 2}

Ahmad Hassan Hassan ³

Njood Aljarrah ⁴

¹ Assistant Professor, Department of Civil Engineering, Faculty of Engineering, Ajloun National University (ANU), Ajloun 26810, Jordan

² Assistant Professor, Department of Civil Engineering, Al-Azhar University, Nasr City 11884, Cairo, Egypt

³ Assistant Professor, Department of Cybersecurity, Faculty of Information Technology, Middle East University (MEU), Amman, Jordan

⁴ Part-time Lecturer, Department of Computer Sciences, Faculty of Information Technology and Computer Sciences, Yarmouk University (YU), Irbid, Jordan

10.22115/scce.2025.1980

Abstract

This study aims to optimize Carbon Fiber Reinforced Polymer (CFRP) configurations—specifically, thickness and fiber orientation—for seismic retrofitting of reinforced concrete (RC) frames. To achieve this, the Whale Optimization Algorithm (WOA) is integrated with a Light Gradient Boosting Machine (LightGBM) model to enhance predictive accuracy and minimize seismic damage indicators such as inter-story drift and base shear. A synthetically generated dataset comprising over 10,000 samples was used to train and evaluate the model under varying structural and seismic conditions. Before optimization, the model achieved a Mean Absolute Error (MAE) of 0.005 and an R² score of 0.89 on the testing set. After WOA-based hyperparameter tuning, the MAE decreased to 0.004 and the R² improved to 0.92. The optimization also reduced inter-story drift from 0.035 to 0.029 and base shear from 2200 kN to 1950 kN. These results demonstrate the effectiveness of combining WOA and LightGBM in simultaneously optimizing CFRP parameters. The proposed framework offers a computationally efficient, data-driven approach to retrofitting design, with improved accuracy and generalizability compared to traditional methods. The model’s adaptability across various seismic intensities and structural profiles highlights its potential for broader application in performance-based seismic design.

Highlights

· A novel machine learning–metaheuristic framework was developed by integrating LightGBM with the Whale Optimization Algorithm (WOA) to optimize CFRP thickness and fiber orientation for seismic retrofitting of RC frames

· The LightGBM model, after WOA-based hyperparameter tuning, achieved high predictive accuracy with an R² score of 0.92 and a Mean Absolute Error (MAE) of 0.004

· Optimization led to measurable structural improvements, reducing inter-story drift from 0.035 to 0.029 and base shear force from 2200 kN to 1950 kN

· The approach is based on a large synthetically generated dataset comprising over 10,000 samples, allowing the model to generalize across varied structural geometries and seismic intensities

· Feature importance analysis revealed CFRP thickness and fiber orientation as the most critical factors, followed by seismic parameters like Peak Ground Acceleration and spectral acceleration

· Compared to traditional simulation-based methods, the proposed framework significantly reduced optimization time from 24 hours to 3 hours while maintaining high accuracy

· The study provides a scalable, data-driven tool for optimizing retrofitting strategies, supporting performance-based seismic design with practical and economic advantages

Keywords

Whale Optimization Algorithm (WOA)

Light GBM

Carbon Fiber Reinforced Polymer (CFRP)

Seismic Retrofitting

Optimizing

Machine Learning

Subjects

Machine Learning

[1] Aljabbri NAS, Karim AA, Majeed FH. Carbon Fiber-Reinforced Polymer Composites Integrated Beam–Column Joints with Improved Strength Performance against Seismic Events: Numerical Model Simulation. Eng 2024;5:1112–39. https://doi.org/10.3390/eng5020061.

[2] Kang S, Chai C, Hong S. Evaluation of Carbon Fiber Grid Reinforced Concrete Panel for Disaster Response and Improved Seismic Performance. Appl Sci 2021;11:5223. https://doi.org/10.3390/app11115223.

[3] Medeghini F, Guhathakurta J, Tiberti G, Simon S, Plizzari GA, Mark P. Steered fiber orientation: correlating orientation and residual tensile strength parameters of SFRC. Mater Struct 2022;55:251. https://doi.org/10.1617/s11527-022-02082-9.

[4] Li W, Chen F, Cen K, Wang Z, Li L, Sun L. Experimental and simulation study on seismic performance of steel‐ PVA hybrid fiber cementitious composites‐encased CFST. Struct Concr 2024;25:2141–63. https://doi.org/10.1002/suco.202300714.

[5] Wang Z, Wang J, Liu T, Zhang F. Modeling seismic performance of high-strength steel–ultra-high-performance concrete piers with modified Kent–Park model using fiber elements. Adv Mech Eng 2016;8. https://doi.org/10.1177/1687814016633411.

[6] Golias E, Zapris AG, Kytinou VK, Osman M, Koumtzis M, Siapera D, et al. Application of X-Shaped CFRP Ropes for Structural Upgrading of Reinforced Concrete Beam–Column Joints under Cyclic Loading–Experimental Study. Fibers 2021;9:42. https://doi.org/10.3390/fib9070042.

[7] El Yassari S, EL Ghoulbzouri A, El Janous S. Seismic Fragility of FRC Columns using Incremental Dynamic Analysis and eXtended Finite Element Method. Int J Eng 2024;37:268–82. https://doi.org/10.5829/IJE.2024.37.02B.05.

[8] Xia Y, Shen C, Xia Y, Yang H, Wang Y, Tian J. Seismic performance analysis and control method of composite coupled shear wall with steel plate-fiber reinforced concrete coupling beams. Mater Res Express 2024;11:075501. https://doi.org/10.1088/2053-1591/ad592e.

[9] Lindsey NJ, Rademacher H, Ajo‐Franklin JB. On the Broadband Instrument Response of Fiber‐Optic DAS Arrays. J Geophys Res Solid Earth 2020;125. https://doi.org/10.1029/2019JB018145.

[10] Zhou Y, Liu X, Zhang X, Guo X. Investigation on Seismic Performance of Reinforced Concrete Frame Retrofitted by Carbon Fiber-Reinforced Polymer. Buildings 2024;14:1604. https://doi.org/10.3390/buildings14061604.

[11] Lakshmipriya N. Retrofitting of Beam-Column Joints in RC Buildings using Jacketing Techniques along with Cross Bars- Review Paper. Int J Res Appl Sci Eng Technol 2018;6:1408–10. https://doi.org/10.22214/ijraset.2018.3217.

[12] Jafari S, Mahini SS. Enhancement of the Fragility Capacity of RC Frames Using FRPs with Different Configurations at Joints. Polymers (Basel) 2023;15:618. https://doi.org/10.3390/polym15030618.

[13] Tatar J, Sattar S, Goodwin D, Milev S, Ahmed S, Dukes J, et al. Performance of externally bonded fiber-reinforced polymer retrofits in the 2018 Cook Inlet Earthquake in Anchorage, Alaska. Earthq Spectra 2021;37:2342–71. https://doi.org/10.1177/87552930211028609.

[14] Laseima SY, Mutalib AA, Osman SA, Hamid NH. Seismic Behavior of Exterior RC Beam-Column Joints Retrofitted using CFRP Sheets. Lat Am J Solids Struct 2020;17. https://doi.org/10.1590/1679-78255910.

[15] Hejazi F, Azarm R, Firoozi AA. Efficiency of Flange-Bonded CFRP Sheets in Relocation of Plastic Hinge in RC Beam–Column Joints. Appl Sci 2023;13:11870. https://doi.org/10.3390/app132111870.

[16] Afshar A, Nouri G, Ghazvineh S, Hosseini Lavassani SH. Machine-Learning Applications in Structural Response Prediction: A Review. Pract Period Struct Des Constr 2024;29. https://doi.org/10.1061/PPSCFX.SCENG-1292.

[17] Alshboul O, Almasabha G, Shehadeh A, Al-Shboul K. A comparative study of LightGBM, XGBoost, and GEP models in shear strength management of SFRC-SBWS. Structures 2024;61:106009. https://doi.org/10.1016/j.istruc.2024.106009.

[18] Kostinakis K, Morfidis K, Demertzis K, Iliadis L. Classification of buildings’ potential for seismic damage using a machine learning model with auto hyperparameter tuning. Eng Struct 2023;290:116359. https://doi.org/10.1016/j.engstruct.2023.116359.

[19] Zhang X, Chen J, Wu Y, Tang L, Ling X. Predicting the Maximum Seismic Response of the Soil-Pile-Superstructure System Using Random Forests. J Earthq Eng 2022;26:8120–41. https://doi.org/10.1080/13632469.2021.1988766.

[20] Tao H, Ali ZH, Mukhtar F, Al Zand AW, Marhoon HA, Goliatt L, et al. Coupled extreme gradient boosting algorithm with artificial intelligence models for predicting compressive strength of fiber reinforced polymer- confined concrete. Eng Appl Artif Intell 2024;134:108674. https://doi.org/10.1016/j.engappai.2024.108674.

[21] Cui Y, Fang J, Li Y, Liu H. Assessing effectiveness of a dual-barrier system for mitigating granular flow hazards through DEM-DNN framework. Eng Geol 2022;306:106742. https://doi.org/10.1016/j.enggeo.2022.106742.

[22] Wang C, Zou X, Sneed LH, Zhang F, Zheng K, Xu H, et al. Shear strength prediction of FRP-strengthened concrete beams using interpretable machine learning. Constr Build Mater 2023;407:133553. https://doi.org/10.1016/j.conbuildmat.2023.133553.

[23] Ma L, Zhou C, Lee D, Zhang J. Prediction of axial compressive capacity of CFRP-confined concrete-filled steel tubular short columns based on XGBoost algorithm. Eng Struct 2022;260:114239. https://doi.org/10.1016/j.engstruct.2022.114239.

[24] Nguyen H-H, Truong V-H. Machine Learning-based prediction of seismic lateral deflection of steel trusses using nonlinear time-history analysis. Structures 2024;69:107369. https://doi.org/10.1016/j.istruc.2024.107369.

[25] Yahiaoui A, Dorbani S, Yahiaoui L. Machine learning techniques to predict the fundamental period of infilled reinforced concrete frame buildings. Structures 2023;54:918–27. https://doi.org/10.1016/j.istruc.2023.05.052.

[26] Nguyen H, Vu T, Vo TP, Thai H-T. Efficient machine learning models for prediction of concrete strengths. Constr Build Mater 2021;266:120950. https://doi.org/10.1016/j.conbuildmat.2020.120950.

[27] Guo J, Yun S, Meng Y, He N, Ye D, Zhao Z, et al. Prediction of heating and cooling loads based on light gradient boosting machine algorithms. Build Environ 2023;236:110252. https://doi.org/10.1016/j.buildenv.2023.110252.

[28] Makomere RS, Koech L, Rutto HL, Kiambi S. Precision forecasting of spray-dry desulfurization using Gaussian noise data augmentation and k-fold cross-validation optimized neural computing. J Environ Sci Heal Part A 2024;59:1–14. https://doi.org/10.1080/10934529.2024.2317670.

[29] Mirjalili S. Handbook of Whale Optimization Algorithm: Variants, Hybrids, Improvements, and Applications. Elsevier; 2023.

[30] Kushwah R, Kaushik M, Chugh K. A modified whale optimization algorithm to overcome delayed convergence in artificial neural networks. Soft Comput 2021;25:10275–86. https://doi.org/10.1007/s00500-021-05983-z.

[31] Li L. LightGBM integration with modified data balancing and whale optimization algorithm for rock mass classification. Sci Rep 2024;14:23028. https://doi.org/10.1038/s41598-024-73742-9.

[32] Khan MN, Sinha AK. Whale optimization algorithm for scheduling and sequencing. Handb. Whale Optim. Algorithm, Elsevier; 2024, p. 487–94. https://doi.org/10.1016/B978-0-32-395365-8.00041-5.

[33] Zara A, Belaidi I, Khatir S, Oulad Brahim A, Boutchicha D, Abdel Wahab M. Damage detection in GFRP composite structures by improved artificial neural network using new optimization techniques. Compos Struct 2023;305:116475. https://doi.org/10.1016/j.compstruct.2022.116475.

[34] Shen Y, Wu S, Cheng H, Zhang H, Wang J, Yang Z, et al. Uncertainty analysis method of slope safety factor based on quantile-based ensemble learning. Bull Eng Geol Environ 2023;82:87. https://doi.org/10.1007/s10064-023-03091-w.

[35] Zhao J, Li D, Jiang J, Luo P. Uniaxial Compressive Strength Prediction for Rock Material in Deep Mine Using Boosting-Based Machine Learning Methods and Optimization Algorithms. Comput Model Eng Sci 2024;140:275–304. https://doi.org/10.32604/cmes.2024.046960.

[36] Kaveh A, Dadras Eslamlou A. Metaheuristic Optimization Algorithms in Civil Engineering: New Applications. vol. 900. Cham: Springer International Publishing; 2020. https://doi.org/10.1007/978-3-030-45473-9.

[37] Gautam GD, Shrivastava Y. Advancements in metaheuristic optimization techniques for laser beam cutting of FRP composites: A review. Proc Inst Mech Eng Part E J Process Mech Eng 2024. https://doi.org/10.1177/09544089241278097.

[38] Habibi SA, Hemmati A, Naderpour H. An Innovative Approach to Estimate Chloride Diffusion Coefficient in Submerged Concrete Structures Using Soft Computing. J Rehabil Civ Eng 2023;11:88–106. https://doi.org/10.22075/jrce.2022.28128.1697.

[39] Al Yamani WH, Bisharah M, Alumany HH, Al Mohammadin NA. Machine learning in seismic structural design: an exploration of ANN and tabu-search optimization. Asian J Civ Eng 2024;25:2367–77. https://doi.org/10.1007/s42107-023-00913-w.

[40] Al Yamani WH, Ghunimat DM, Bisharah MM. Modeling and predicting the sensitivity of high-performance concrete compressive strength using machine learning methods. Asian J Civ Eng 2023;24:1943–55. https://doi.org/10.1007/s42107-023-00614-4.

[41] Al-Rawashdeh M, Al Nawaiseh M, Yousef I, Bisharah M, Alkhadrawi S, Al-Bdour H. Predicting building damage grade by earthquake: a Bayesian Optimization-based comparative study of machine learning algorithms. Asian J Civ Eng 2024;25:253–64. https://doi.org/10.1007/s42107-023-00771-6.

[42] Chou J-S, Ngo N-T, Pham A-D. Shear Strength Prediction in Reinforced Concrete Deep Beams Using Nature-Inspired Metaheuristic Support Vector Regression. J Comput Civ Eng 2016;30. https://doi.org/10.1061/(ASCE)CP.1943-5487.0000466.

[43] Chamanzari S, Dorostkar A, Yosefi A, Nasseri H. Investigation of Different Artificial Intelligence Algorithms in Predicting the Bending Strength of Reinforced Concrete Beams n.d.

[44] Chen L, Fakharian P, Rezazadeh Eidgahee D, Haji M, Mohammad Alizadeh Arab A, Nouri Y. Axial compressive strength predictive models for recycled aggregate concrete filled circular steel tube columns using ANN, GEP, and MLR. J Build Eng 2023;77:107439. https://doi.org/10.1016/j.jobe.2023.107439.

[45] Ezami N, Özyüksel Çiftçioğlu A, Mirrashid M, Naderpour H. Advancing Shear Capacity Estimation in Rectangular RC Beams: A Cutting-Edge Artificial Intelligence Approach for Assessing the Contribution of FRP. Sustainability 2023;15:16126. https://doi.org/10.3390/su152216126.