Prediction of Compressive Strength and Design Parameters of C30/37, C35/45 and C40/50 Concrete Classes by ANN

Authors:
Publisher website
Google Scholar
Abstract
The quality of concrete used in the construction sector is increasing day by day with ready-mixed concrete production. The quality of concrete is directly related to its compressive strength and the related tests are labor-intensive and time-consuming. Therefore, different artificial intelligence-based models are used to predict the compressive strength of concrete. In this study, compressive strength and design parameters of concrete classes C30/37, C35/45 and C40/50 were predicted by ANN model. A total of 240 compressive strength results obtained from concretes produced in a ready-mixed concrete plant for the construction of columns, beams, decks and stairs. 70% of these data were used for training and remaining 30% of data were reserved for testing. The prediction accuracy of the ANN model was evaluated by R2, MAPE and RMSE statistical methods. According to results, the compressive strengths of concrete classes C30/37, C35/45 and C40/50 could be predicted with errors of -0.70%, 1.25% and 0.17% for 7 days and 0.99%, 0.03% and -0.69% for 28 days, respectively. Depending on the design parameters, it was found that prediction performance could be made with almost 100% accuracy for all concretes except high-performance superplasticizer admixture. As a result, it was concluded that ‘very good’ or ‘high accuracy’ predictions can be made with ANN models.

This publication has 18 references indexed in Scilit: