Prediction of Characteristic Strength of Sustainable Concrete Containing Mineral Additives by Artificial Neural Networks.

Authors

Mohamed Elgamal, Master of Science Researcher., Structural Engineering Department., El-Mansoura University., Mansoura., Egypt.Follow
Ahmed Tahwia, Professor of Structural Engineering Department., Faculty of Engineering., El-Mansoura University., Mansoura., Egypt.Follow
Ashraf Heniegal, Professor., Civil Engineering Department., Suez University., Suez., Egypt.Follow

Corresponding Author

Elgamal, Mohamed

Subject Area

Civil and Environmental Engineering

Article Type

Original Study

Abstract

It this study, the application of artificial neural networks (ANN) for estimating the characteristic strength of sustainable concrete that contains various amounts of fly ash, silica fume, slag and steel fiber have been investigated. Using ANN model, it is possible to establish a linear and nonlinear correlation between known input data like concrete ingredients and a certain output like characteristic strength, because ANN is an excellent tool to determine concrete properties. For the training of ANN models, an experimental data base (1410 concrete mixtures from earlier published papers) has been utilized. Then experimental tests were performed on some mixes of concrete to validate the model. The ANN model parameter statistics R2 is 0.888, 0.93, 0.9 for training, validation and test steps and indicate that ANN model makes effective prediction for characteristic strength of sustainable concrete. The application of ANN in predicting characteristic strength is considered to the quality assurance of manufacturing of concrete.

Keywords

artificial neural network; Sustainable Concrete; characteristic strength; Pozzolanic materials

Recommended Citation

Elgamal, Mohamed; Tahwia, Ahmed; and Heniegal, Ashraf (2021) "Prediction of Characteristic Strength of Sustainable Concrete Containing Mineral Additives by Artificial Neural Networks.," Mansoura Engineering Journal: Vol. 46 : Iss. 1 , Article 21.
Available at: https://doi.org/10.21608/bfemu.2021.160427