Estimating the compressive strength of fly ash added concrete using artificial neural networks

Abstract

The aim of this study is to develop an artificial intelligence that predicts the compressive strength of fly ash substituted concretes using material mixing ratios. Within the scope of the study, 5 different fly ash mixed concrete samples were estimated. The strength values were estimated using artificial neural networks before the produced samples were subjected to the pressure test. In order to develop the artificial neural network, it is introduced as a dataset of 1030 different mixing ratios consisting of experimental results in the existing literature. In order to estimate the compressive strength, varying ratios of 8 different materials such as water, cement, fly ash entering the mixture are analyzed. As a result of the study, it has been observed that the predictions made using artificial neural networks are very close to the strength values obtained from the experiments