journal article·2025-12Open access

A generative adversarial network enhanced ensemble learning-based prediction model for moment improvement effect of UHPC strengthened damaged RC beams

Woubishet Zewdu Taffese, Nima Khodadadi, Yanping Zhu, Seyedali Mirjalili, Antonio Nanni

Abstract

Reinforced Concrete (RC) structures are often compromised by cracks and corrosion, necessitating effective retrofitting strategies to restore their structural integrity. Ultra-High-Performance Concrete (UHPC) has emerged as a promising material for strengthening damaged RC beams, significantly enhancing their load-bearing capacity and durability. However, accurately predicting the flexural performance of UHPC-strengthened RC beams remains a challenge due to complex material interactions and limited datasets. This addresses this gap by developing a data-driven framework that combines generative data augmentation and explainable machine learning to predict the ultimate moment resistance ( M u ) of strengthened beams. A curated dataset of 160 experimental cases was expanded using a Tabular Generative Adversarial Network (TGAN), and six ensemble learning models were trained and evaluated. Among them, Categorical Boosting (CatBoost) demonstrated superior performance with an R² of 0.90 on the test set. SHapley Additive exPlanations (SHAP) were employed to explain model predictions, revealing that the UHPC reinforcement ratio, longitudinal reinforcement ratio of beam, beam width, and concrete compressive strength are the most influential factors. The proposed approach not only improves prediction accuracy and model robustness but also provides interpretable insights to support rational design decisions in structural retrofitting. • First use of GANs to predict moment improvement in UHPC-strengthened damaged RC beams. • CatBoost achieved high accuracy on both the training set ( R 2 =0.99) and test set ( R 2 =0.90). • The Shapely-based model explanations reveal features importance and sensitivity for engineering design.
Source: openalex_inverted_index · Quality: baseline

Research context

OpenAlex topics: Innovative concrete reinforcement materials · Concrete Corrosion and Durability · Structural Behavior of Reinforced Concrete

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

References (50)

Show source-provided references
1. Flexural performance of V-shaped RC folded plates: the role of plate thickness and fiber hybridization10.1016/j.conbuildmat.2023.132815 ↗
2. Data-driven moment-carrying capacity prediction of hybrid beams consisting of UHPC-NSC using machine learning-based models
3. Mechanical properties of ultra-high-performance fiber-reinforced concrete: a review10.1016/j.cemconcomp.2016.08.001 ↗
4. Flexural strengthening of reinforced concrete beams or slabs using ultra-high performance concrete (UHPC): a state of the art review10.1016/j.engstruct.2019.110035 ↗
5. A review on ultra high performance concrete: part I. Raw materials and mixture design10.1016/j.conbuildmat.2015.10.088 ↗
6. International perspective on UHPC in bridge engineering10.1061/(ASCE)BE.1943-5592.0001630 ↗
7. Effects of nano-silica and micro-steel fiber on the engineering properties of ultra-high performance concrete
8. Ensemble-learning model based ultimate moment prediction of reinforced concrete members strengthened by UHPC10.1016/j.engstruct.2024.117705 ↗
9. Ultra-high performance concrete: from fundamental to applications
10. Seismic performance of high-performance fiber-reinforced cement-based composite structural members: a review
11. Fiber-reinforced polymer (FRP) in concrete: A comprehensive survey10.1016/j.conbuildmat.2024.136634 ↗
12. Prediction of shear capacity of UHPC-concrete composite members based on existing codes10.3846/jcem.2018.6484 ↗
13. Flexural behavior of RC beams retrofitted with ultra-high strength concrete10.1016/j.conbuildmat.2018.04.174 ↗
14. Flexural behaviors and capacity prediction on damaged reinforcement concrete (RC) bridge deck strengthened by ultra-high performance concrete (UHPC) layer10.1016/j.conbuildmat.2019.04.229 ↗
15. Effect of overlay thickness, fiber volume, and shrinkage mitigation on flexural behavior of thin bonded ultra-high-performance concrete overlay slab10.1016/j.cemconcomp.2022.104752 ↗
16. Numerical modeling for damaged reinforced concrete slab strengthened by ultra-high performance concrete (UHPC) layer10.1016/j.engstruct.2019.110031 ↗
17. Flexural capacity prediction of composite RC members strengthened with UHPC based on existing design models
18. Flexural behavior of reinforced concrete beams repaired with ultra-high performance fiber reinforced concrete (UHPFRC)10.1016/j.compstruct.2016.09.010 ↗
19. Interpretable machine-learning models for maximum displacements of RC beams under impact loading predictions10.1016/j.engstruct.2023.115723 ↗
20. Explainable machine learning model for predicting punching shear strength of FRC flat slabs10.1016/j.engstruct.2023.117276 ↗
21. Explainable machine learning models for predicting the axial compression capacity of concrete filled steel tubular columns10.1016/j.conbuildmat.2022.129227 ↗
22. A novel approach to explain the black-box nature of machine learning in compressive strength predictions of concrete using shapley additive explanations (SHAP)
23. Prediction of shear strength in UHPC beams using machine learning-based models and SHAP interpretation10.1016/j.conbuildmat.2023.133752 ↗
24. Peak and ultimate Stress-Strain model of confined Ultra-High-Performance concrete using machine learning model with conditional tabular generative adversarial network10.1016/j.asoc.2024.111353 ↗
25. Data-driven prediction and optimization of axial compressive strength for FRP-reinforced CFST columns using synthetic data augmentation10.1016/j.engstruct.2023.117225 ↗
26. Predicting ultra-high-performance concrete compressive strength using tabular generative adversarial networks10.3390/ma13214757 ↗
27. Generative adversarial networks10.1145/3422622 ↗
28. Generative adversarial network: an overview of theory and applications
29. Synthesizing tabular data using generative adversarial networks
30. Experimental and numerical study of the performance of ultra high performance fiber reinforced concrete for the flexural strengthening of full scale reinforced concrete members10.1016/j.conbuildmat.2018.07.123 ↗
31. Flexural strengthening of RC beams using UHPFRC laminates: bonding techniques and rebar addition10.1016/j.conbuildmat.2017.08.056 ↗
32. Strengthening and repair of RC beams with fiber reinforced concrete10.1016/j.cemconcomp.2010.07.001 ↗
33. Flexural behaviour of damaged RC beams strengthened with ultra high performance concrete
34. Structural behavior of ultra-high performance fiber reinforced concrete-normal strength concrete or high strength concrete composite members10.1016/j.conbuildmat.2015.05.030 ↗
35. Repair of pre-cracked RC beams using several cementitious materials10.9734/JSRR/2013/5238 ↗
36. Retrofitting of reinforced concrete beams with CARDIFRC10.1061/(ASCE)1090-0268(2003)7:3(174) ↗
37. Flexural behavior of reinforced concrete beams strengthened with ultra-high performance fiber reinforced concrete10.1016/j.conbuildmat.2016.12.094 ↗
38. Strengthening of reinforced concrete beams using ultra high performance fibre reinforced concrete (UHPFRC)10.1016/j.engstruct.2015.10.042 ↗
39. Experimental study on flexural behavior of damaged reinforced concrete (RC) beam strengthened by toughness-improved ultra-high performance concrete (UHPC) layer10.1016/j.compositesb.2020.107834 ↗
40. Experimental investigation on the behaviour of reinforced concrete slabs strengthened with ultra-high performance concrete10.1016/j.conbuildmat.2017.08.077 ↗
41. Experimental and analytical study on the effect of different repairing and strengthening strategies on flexural performance of corroded RC beams
42. Flexural strengthening of RC beams with prefabricated ultra high performance fibre reinforced concrete laminates10.1016/j.engstruct.2017.08.048 ↗
43. Theoretical modelling and acoustic emission monitoring of RC beams strengthened with UHPC10.1016/j.conbuildmat.2017.10.063 ↗
44. Acoustic emission and flexural behaviour of RC beams strengthened with UHPC overlay10.1016/j.conbuildmat.2016.07.033 ↗
45. Experimental validation of damaged reinforced concrete beam strengthened by pretensioned prestressed ultra-high-performance concrete layer10.1016/j.engstruct.2022.114251 ↗
46. Developments in the use of ultra high performance fiber reinforced concrete as strengthening material10.1016/j.engstruct.2021.111914 ↗
47. Behaviour of RC beams strengthened in flexure with hybrid CFRP-reinforced UHPC overlays10.1016/j.engstruct.2022.114356 ↗
48. Flexural performance of reinforced concrete (RC) beam strengthened by UHPC layer
49. Flexural performance of pre-damaged RC beams strengthened with different configurations of UHPFRC layer–experimental and analytical investigation
50. Ahmed, Sayed, et al. Experimental and numerical investigation of flexural behavior of RC beams retrofitted with reinforced UHPFRC layer in tension surface. Structures. Vol. 49. Elsevier, 20.10.1016/j.istruc.2023.01.113 ↗