PSI - Issue 70

Available online at www.sciencedirect.com

ScienceDirect

Procedia Structural Integrity 70 (2025) 453–460

Structural Integrity and Interactions of Materials in Civil Engineering Structures (SIIMCES-2025) Compressive Strength Prediction of FRP Confined Concrete Using Machine Learning Based Models N. Vignesh Kumar a, *, Ruhul Amin Mozumder a , Nirban Laskar a a Deparment of Civil Engineering, School of Engineering & Technology, Mizoram University, Tanhril, Aizawl-796004, India Abstract Fibre Reinforced Polymer (FRP) materials owing to its superior mechanical properties are widely used in improving the flexural and compressive strength capacity of structural members. Predicting confined concrete strength using FRP materials such as Carbon Fibre Reinforced Polymer (CFRP) and Glass Fibre Reinforced Polymer (GFRP) is of key importance for effective and economic wrapping of columns. Existing analytical FRP strength prediction models lack satisfactory performance for variety of reasons. In such cases, Machine Learning (ML) based data driven models may be leveraged for enhancing the predictive performance. To this end present study uses a large database comprising the experimental compressive strength of FRP-wrapped concrete cylinders for modelling the strength behavior of FRP confined concrete columns. A range of ML techniques incorporating both parametric models such as LASSO and STEPWISE regression algorithms, and nonparametric methods, such as Support Vector Machine (SVM) Regression, Gaussian Process Regression (GPR), and Regression Tree ensemble (RTE) have been developed. A comparative study of prediction performance of ML models indicated the superiority of nonparametric models over the parametric models, owing to their high flexibility. Among the non-parametric models, in particular GPR model outperformed the remaining models. For CFRP wrapping, GPR model showed R2 values of 0.917 (test) and 0.986 (train), while for GFRP wrapping it achieved 0.916 (test) and 0.997 (train). Sensitivity analysis (SA) resulted the thickness (t) of FRP material and unconfined concrete strength of concrete (fco) as the most influential variables in predicting the compressive strength of CFRP and GFRP wrapped concrete cylinder. © 2025 The Authors. Published by ELSEVIER B.V. This is an open access article under the CC BY-NC-ND license (https://creativecommons.org/licenses/by-nc-nd/4.0) Peer-review under the responsibility of International Conference on Structural Integrity Organizers

Keywords: FRP confinement; Machine Learning, CFRP; GFRP; Concrete strength prediction; Sensitivity analysis;

* Corresponding author. Tel.: +91-944-253-1198. E-mail address: mzut256@mzu.edu.in

2452-3216 © 2025 The Authors. Published by ELSEVIER B.V. This is an open access article under the CC BY-NC-ND license (https://creativecommons.org/licenses/by-nc-nd/4.0) Peer-review under the responsibility of International Conference on Structural Integrity Organizers 10.1016/j.prostr.2025.07.077