Statistical analyses of hooked-end steel fibers in concrete

Concrete, owing to its brittle nature is weak in tension and therefore, a significant amount of published research has focused on the addition of fibers to improve the flexural performance of concrete. The addition of fibers in a concrete matrix improves its overall performance in terms of strength, ductility, and durability by the mechanism of restraining crack width. Steel fibers are found to be the most efficient in improving the flexural strength of concrete. The design mix for Steel Fiber-Reinforced Concrete (SFRC) is chosen mostly based on the trial mixes. To aid the mix design of SFRC, this study attempts to provide predictive models using machine learning techniques such as non-linear regression on a database of 146 samples, collated from literature. 80% of the randomly chosen samples are used for training the model and the remaining 20% are used to test it. The Root Mean Squared Error (RMSE), Mean Absolute Error (MAE), the statistical significance of the model coefficients, and the coefficient of determination are adopted as criteria to check the performance of the model on both the training and test data. Interaction terms such as fiber factor and number of fibers within a volume were inferred from literature to improve the efficacy of the predictive modeling. Two non-linear models for the prediction of compressive and flexural strength of hooked-end SFRC are proposed and validated. Using the proposed models, the variations in compressive and flexural strength with the key parameters were plotted and discussed. With a decrease in the water-to-binder ratio, there was a significant improvement in compressive strength from 55 MPa to 85 MPa at an aspect ratio of 60. A significant improvement of about 200% to 300% in flexural strength was noted with an increase in fiber factor from 0 to 150. Copyright (c) 2022 Elsevier Ltd. All rights reserved. Selection and peer-review under responsibility of the scientific committee of the F-EIR Conference 2021 on Environment Concerns and its Remediation: Materials Science