A platform for research: civil engineering, architecture and urbanism
A platform for research: civil engineering, architecture and urbanism
A Machine Learning Based Model to Assess Flexural Strength of Corroded Reinforced Concrete Beams
https://orcid.org/http://orcid.org/0009-0004-0626-408X
https://orcid.org/http://orcid.org/0000-0002-1159-1892
https://orcid.org/http://orcid.org/0000-0001-8580-9236
Corrosion in members is a significant durability problem in reinforced concrete; it reduces load-carrying capacity. This study collected corroded reinforced concrete beam specimens tested under flexural loads from the published literature. The whole database comprises 177 corroded beam specimens. A few parameters of the corroded beams, such as width, beam depth, compressive strength of concrete, yield strength of steel reinforcement, percentage weight loss etc., were collected from the literature. Two different machine-learning-based model was trained to predict the residual flexural strength of the corroded beam. K-Nearest Neighbor (KNN) and Support Vector Machine (SVM), were used to train the models for predictions. Comparative analysis of the models was done using six statistical indices R2, Mean Absolute Percentage Error (MAPE), Mean Absolute Error (MAE), Root Mean Square Error (RMSE), a-20 index, and Nash Sutcliffe, to propose the best of the two model for prediction. The results from the SVM model show an R2-value of 0.989 and that of the KNN model show 0.809. The proposed ML models are reliable, accurate, fast, and cost-effective. This model can be utilized as a structural health-monitoring tool to detect the early damages in the RC beams.
A Machine Learning Based Model to Assess Flexural Strength of Corroded Reinforced Concrete Beams
https://orcid.org/http://orcid.org/0009-0004-0626-408X
https://orcid.org/http://orcid.org/0000-0002-1159-1892
https://orcid.org/http://orcid.org/0000-0001-8580-9236
Corrosion in members is a significant durability problem in reinforced concrete; it reduces load-carrying capacity. This study collected corroded reinforced concrete beam specimens tested under flexural loads from the published literature. The whole database comprises 177 corroded beam specimens. A few parameters of the corroded beams, such as width, beam depth, compressive strength of concrete, yield strength of steel reinforcement, percentage weight loss etc., were collected from the literature. Two different machine-learning-based model was trained to predict the residual flexural strength of the corroded beam. K-Nearest Neighbor (KNN) and Support Vector Machine (SVM), were used to train the models for predictions. Comparative analysis of the models was done using six statistical indices R2, Mean Absolute Percentage Error (MAPE), Mean Absolute Error (MAE), Root Mean Square Error (RMSE), a-20 index, and Nash Sutcliffe, to propose the best of the two model for prediction. The results from the SVM model show an R2-value of 0.989 and that of the KNN model show 0.809. The proposed ML models are reliable, accurate, fast, and cost-effective. This model can be utilized as a structural health-monitoring tool to detect the early damages in the RC beams.
A Machine Learning Based Model to Assess Flexural Strength of Corroded Reinforced Concrete Beams
https://orcid.org/http://orcid.org/0009-0004-0626-408X
https://orcid.org/http://orcid.org/0000-0002-1159-1892
https://orcid.org/http://orcid.org/0000-0001-8580-9236
Corrosion in members is a significant durability problem in reinforced concrete; it reduces load-carrying capacity. This study collected corroded reinforced concrete beam specimens tested under flexural loads from the published literature. The whole database comprises 177 corroded beam specimens. A few parameters of the corroded beams, such as width, beam depth, compressive strength of concrete, yield strength of steel reinforcement, percentage weight loss etc., were collected from the literature. Two different machine-learning-based model was trained to predict the residual flexural strength of the corroded beam. K-Nearest Neighbor (KNN) and Support Vector Machine (SVM), were used to train the models for predictions. Comparative analysis of the models was done using six statistical indices R2, Mean Absolute Percentage Error (MAPE), Mean Absolute Error (MAE), Root Mean Square Error (RMSE), a-20 index, and Nash Sutcliffe, to propose the best of the two model for prediction. The results from the SVM model show an R2-value of 0.989 and that of the KNN model show 0.809. The proposed ML models are reliable, accurate, fast, and cost-effective. This model can be utilized as a structural health-monitoring tool to detect the early damages in the RC beams.
A Machine Learning Based Model to Assess Flexural Strength of Corroded Reinforced Concrete Beams
Lecture Notes in Civil Engineering
Nehdi, Moncef (editor) / Hung, Mo Kim (editor) / Venkataramana, Katta (editor) / Antony, Jiji (editor) / Kavitha, P. E. (editor) / Beena B R (editor) / Sharma, Arjun (author) / Sharma, Somain (author) / Kumar, Kuldeep (author)
International Conference on Structural Engineering and Construction Management ; 2023 ; Angamaly, India
2023-11-03
14 pages
Article/Chapter (Book)
Electronic Resource
English
Flexural strength of corroded reinforced concrete beams
| Online Contents | 2010
Residual flexural strength of corroded reinforced concrete beams
| Online Contents | 2016
Residual flexural strength of corroded reinforced concrete beams
| Elsevier | 2016
RESIDUAL FLEXURAL STRENGTH OF CORRODED REINFORCED CONCRETE BEAMS
| TIBKAT | 2020