Enhancing Structural Evaluation: Machine Learning Approaches for Inadequate Reinforced Concrete Frames
No Thumbnail Available
Date
2024
Journal Title
Journal ISSN
Volume Title
Abstract
This paper provides comprehensive analyses of the performance of inadequate reinforced concrete frames from different aspects. First, a one-story, single-span, 1/3 scaled frame is constructed. The ultimate lateral load-bearing capacity of the sample is experimentally determined. Second, the behavior of the sample is obtained by finite element analysis. In the analyses, axial load, rebar diameter, and concrete strength values are taken as variables. Load–displacement behaviors determined from the tests and finite element analyses are compared. Third, machine learning models are developed to estimate the ultimate load-bearing capacity of the frames. Random Forest, ElasticNet, RANSAC, Decision Tree, K-Nearest Neighbors, and Gaussian Naive Bayes are employed to assess the load-bearing capacities of the frames. Coefficient of determination, Root Mean Squared Error, Mean Absolute Error, and Mean Absolute Percentage Error, which are widely recognized performance indicators, are also employed to assess the effectiveness of machine learning methods. The findings reveal that the Random Forest method is the most precise and effective in both regression and classification analyses. It has the highest Coefficient of determination of 87% in predicting the load-bearing capacity of the frame and the highest accuracy of 100% in classifying the frames based on their load-bearing capacity. This is the first attempt to employ machine learning approaches to assess the load-bearing capacity of inadequate reinforced concrete frames. The proposed model can provide a better understanding of inadequate frames and has advantages over other analysis methods with respect to simplicity in application, flexibility, reducing the time and cost associated with the process. © The Author(s), under exclusive licence to Shiraz University 2024.