Evaluation of Recycled Steel Fiber Effect on Concrete Performance Using Artificial Intelligence Technique

Abstract

Reusing waste materials is critical for sustainability and preventing adverse impacts on human life and the environment. Waste vehicle tires have become a big problem due to high consumption. It is possible to separate waste tires into different materials through technological means. Recycled steel fiber is a material obtained from these tires, and various studies have been conducted on its use in concrete. In addition to the geometric properties, such as the length and diameter, the percentage of steel fiber also affects the strength of concrete. In this study, the effect of recycled steel fiber on concrete's compressive and flexural strength values was estimated using artificial intelligence functions with high statistical significance. The relationship between the strength results and the recycled steel fiber properties was determined using literature data. The model's accuracy was demonstrated by comparing the obtained compressive and flexural strengths with the laboratory results. Thanks to the model with a high correlation coefficient created as a result of the study, the effect of recycled steel fiber on concrete performance as an alternative to laborious laboratory tests can be predicted with artificial intelligence-supported functions. With the proposed neural network method, R2 values of 0.83 for compressive strength measurements and 0.96 for flexural strength measurements were obtained. Based on the findings, it is concluded that the recycled steel fiber-reinforced concrete parameters can be well represented by artificial neural networks, and the presented model can be used as a good alternative to laboratory studies for further research.