Prediction of ultimate load capacity of demountable shear stud connectors using machine learning techniques

Abstract This study investigates the use of machine learning (ML) models to predict the ultimate load capacity of demountable shear connectors in steel–concrete composite structures. A dataset of 239 experimental and numerical records was assembled, incorporating critical features such as bolt diame...

Full description

Saved in:
Bibliographic Details
Main Authors: Ahmed I. Saleh, Nabil S. Mahmoud, Fikry A. Salem, Mohamed Ghannam
Format: Article
Language:English
Published: Nature Portfolio 2025-08-01
Series:Scientific Reports
Subjects:
Online Access:https://doi.org/10.1038/s41598-025-15711-4
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849226465009205248
author Ahmed I. Saleh
Nabil S. Mahmoud
Fikry A. Salem
Mohamed Ghannam
author_facet Ahmed I. Saleh
Nabil S. Mahmoud
Fikry A. Salem
Mohamed Ghannam
author_sort Ahmed I. Saleh
collection DOAJ
description Abstract This study investigates the use of machine learning (ML) models to predict the ultimate load capacity of demountable shear connectors in steel–concrete composite structures. A dataset of 239 experimental and numerical records was assembled, incorporating critical features such as bolt diameter, bolt yield and ultimate strengths, concrete and grout compressive strengths, and multiple interfacial friction coefficients. Eight supervised ML algorithms were evaluated: Linear Regression, Ridge, Lasso, K-Nearest Neighbors (KNN), Support Vector Regression (SVR), Decision Tree, Random Forest, and XGBoost. Model performance was assessed using R², Mean Absolute Error (MAE), and Mean Squared Error (MSE). Among these, XGBoost and Random Forest delivered the best predictive accuracy, with R² values of 0.9477 and 0.9255, respectively, outperforming other methods across all evaluation metrics. SHAP analysis was employed to explain model behavior and identify the most influential features. The analysis revealed that bolt diameter, steel ultimate strength, and bolt-to-concrete friction were the most impactful predictors of shear capacity. These insights were supported by feature importance heatmaps generated from the top-performing models. Despite the robust performance of advanced algorithms, the study acknowledges challenges such as data imbalance and minimal influence of certain parameters, which may affect model generalizability.
format Article
id doaj-art-eccf8f83de37487f8f6ab3e378337476
institution Kabale University
issn 2045-2322
language English
publishDate 2025-08-01
publisher Nature Portfolio
record_format Article
series Scientific Reports
spelling doaj-art-eccf8f83de37487f8f6ab3e3783374762025-08-24T11:20:14ZengNature PortfolioScientific Reports2045-23222025-08-0115111810.1038/s41598-025-15711-4Prediction of ultimate load capacity of demountable shear stud connectors using machine learning techniquesAhmed I. Saleh0Nabil S. Mahmoud1Fikry A. Salem2Mohamed Ghannam3Civil Engineering Department, Faculty of Engineering, Delta University for Science and TechnologyStructural Engineering Department, Faculty of Engineering, Mansoura UniversityStructural Engineering Department, Faculty of Engineering, Mansoura UniversityStructural Engineering Department, Faculty of Engineering, Mansoura UniversityAbstract This study investigates the use of machine learning (ML) models to predict the ultimate load capacity of demountable shear connectors in steel–concrete composite structures. A dataset of 239 experimental and numerical records was assembled, incorporating critical features such as bolt diameter, bolt yield and ultimate strengths, concrete and grout compressive strengths, and multiple interfacial friction coefficients. Eight supervised ML algorithms were evaluated: Linear Regression, Ridge, Lasso, K-Nearest Neighbors (KNN), Support Vector Regression (SVR), Decision Tree, Random Forest, and XGBoost. Model performance was assessed using R², Mean Absolute Error (MAE), and Mean Squared Error (MSE). Among these, XGBoost and Random Forest delivered the best predictive accuracy, with R² values of 0.9477 and 0.9255, respectively, outperforming other methods across all evaluation metrics. SHAP analysis was employed to explain model behavior and identify the most influential features. The analysis revealed that bolt diameter, steel ultimate strength, and bolt-to-concrete friction were the most impactful predictors of shear capacity. These insights were supported by feature importance heatmaps generated from the top-performing models. Despite the robust performance of advanced algorithms, the study acknowledges challenges such as data imbalance and minimal influence of certain parameters, which may affect model generalizability.https://doi.org/10.1038/s41598-025-15711-4Demountable shear stud connectorsUltimate shear capacity predictionLinear regressionRidgeLassoK-Nearest neighbors (KNN)
spellingShingle Ahmed I. Saleh
Nabil S. Mahmoud
Fikry A. Salem
Mohamed Ghannam
Prediction of ultimate load capacity of demountable shear stud connectors using machine learning techniques
Scientific Reports
Demountable shear stud connectors
Ultimate shear capacity prediction
Linear regression
Ridge
Lasso
K-Nearest neighbors (KNN)
title Prediction of ultimate load capacity of demountable shear stud connectors using machine learning techniques
title_full Prediction of ultimate load capacity of demountable shear stud connectors using machine learning techniques
title_fullStr Prediction of ultimate load capacity of demountable shear stud connectors using machine learning techniques
title_full_unstemmed Prediction of ultimate load capacity of demountable shear stud connectors using machine learning techniques
title_short Prediction of ultimate load capacity of demountable shear stud connectors using machine learning techniques
title_sort prediction of ultimate load capacity of demountable shear stud connectors using machine learning techniques
topic Demountable shear stud connectors
Ultimate shear capacity prediction
Linear regression
Ridge
Lasso
K-Nearest neighbors (KNN)
url https://doi.org/10.1038/s41598-025-15711-4
work_keys_str_mv AT ahmedisaleh predictionofultimateloadcapacityofdemountableshearstudconnectorsusingmachinelearningtechniques
AT nabilsmahmoud predictionofultimateloadcapacityofdemountableshearstudconnectorsusingmachinelearningtechniques
AT fikryasalem predictionofultimateloadcapacityofdemountableshearstudconnectorsusingmachinelearningtechniques
AT mohamedghannam predictionofultimateloadcapacityofdemountableshearstudconnectorsusingmachinelearningtechniques