Evaluation of Compressive Strength of Fibre Reinforced Concrete using Randomizable Filtered Classifier

Read full text Summarise with VIRAT
Authors: AA Ganiyu
Cited by: 0
References: 35

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

Abstract. Fibre reinforced concrete (FRC) is a matrix material that adds discrete fibres to mixed concrete with the aim of increasing its mechanical properties. Conventional techniques for the evaluation of compressive strength of concrete entail time-consuming, laborious laboratory tests. Machine Learning (ML) models have the capabilities to utilise data on mix proportions to train a model that can accurately estimate strength outcomes for new mixes. This research employs a codeless ML tool to predict the compressive strengths of FRC using 173 datapoints with 7 input features of binder, water, fine aggregate, coarse aggregate, volume fraction of steel, aspect ratio and age. An 80:20 data split was adopted for training and testing sets, while pre-processing of data, visualisation, hyper-parameter optimisation, and feature selection were all done in the model development stage. Randomizable Filtered Classifier (RFC) algorithm yielded the optimal model, with a coefficient of determination, R2 of 0.9886 and root-mean-square-error of 14.4529 as values of the performance measures. This implies that the model’s predictions are closely aligned with the true values, underscores the strong correlation between the input features and the compressive strength, and asserts the model’s validity. The study leverages data-driven soft computing methods to provide timely, cost-effective, accurate and reliable processes for the design of compressive strength of fibre-reinforced concrete for practical usage.

Fields of Study & Subjects

Innovative concrete reinforcement materials Structural Behavior of Reinforced Concrete Concrete and Cement Materials Research

Publication details

Journal / source
Materials research proceedings
Volume
63
Pages
47–55
Publication date
1 January 2026
Language
EN