Explainable machine learning model and reliability analysis for flexural capacity prediction of RC beams strengthened in flexure with FRCM

← Back to papers

Physical Sciences › Engineering › Building and Construction

★

☆

SHARE

Overview

Paper Summary

Conflicts of Interest

Identified Weaknesses

Rating Explanation

Good to know

Topic Hierarchy

File Information

Paper Summary

Paperzilla title

Boosting Beams: An AI Predicts How to Make Concrete Stronger with FRCM

This study uses machine learning, specifically the xgBoost model, to predict the flexural capacity of reinforced concrete beams strengthened with fabric-reinforced cementitious matrix (FRCM). The xgBoost model outperformed existing analytical models and a SHAP analysis identified the area of tensile steel reinforcement, width of the beam, FRCM area, and effective beam depth as the most influential factors.

Possible Conflicts of Interest

The authors acknowledge funding from the Qatar National Research Fund and the Natural Sciences and Engineering Research Council of Canada. However, no specific conflicts related to FRCM manufacturers or commercial interests are identified.

Identified Weaknesses

Limited generalizability due to lab-based dataset

The dataset used to train and validate the models comes from experiments conducted under lab conditions. This raises questions about the model's generalizability to real-world scenarios, where unpredictable factors might influence the actual response differently.

Exclusion of FRCM mortar properties

Despite claiming to use a wide range of FRCM systems, the study does not consider the properties of FRCM mortar. This omission affects the models' accuracy since the mortar characteristics directly impact the overall strength.

Oversimplification of input parameters

The simplification of input variables by representing the combined effect of fabric layers, plate width, and thickness as a single area parameter neglects the potential influence of individual parameter variations on FRCM behavior. This oversimplification could compromise model fidelity.

Lack of consideration for the impact of number of fabric layers

The authors admit that the number of fabric layers can influence FRCM behavior, yet the model does not consider this factor. This impacts the models' predictive accuracy when dealing with varying FRCM configurations.

Limited transparency despite SHAP analysis

Although the SHAP approach offers insight into the model's decision-making, it does not provide a fully transparent representation of the complex interactions between the variables. The simplified interpretation offered by SHAP may not completely capture the nuances of the real-world phenomenon.

Rating Explanation

This paper presents a novel approach to predicting flexural capacity in FRCM-strengthened RC beams using machine learning. The use of xgBoost and SHAP analysis is commendable. While the reliance on a lab-generated dataset and the simplification of some input parameters are limitations, the study's innovative methodology and thorough comparative analysis warrant a strong rating. The identified funding sources do not raise significant conflict-of-interest concerns.

Good to know

This is our free standard analysis. Paperzilla Pro fact-checks every citation, researches author backgrounds and funding sources, and uses advanced AI reasoning for more thorough insights.

Explore Pro →

Topic Hierarchy

Domain:

Physical Sciences

Field:

Engineering

Subfield:

Building and Construction

File Information

Original Title:

Explainable machine learning model and reliability analysis for flexural capacity prediction of RC beams strengthened in flexure with FRCM

File Name: