Predicting Performance of Self-Compacting Concrete Mixtures Using Artificial Neural Networks

Abstract

This paper shows that artificial neural networks (ANN) can be used to effectively predict the performance of self-compacting concrete (SCC) mixtures. Inspired by the internal operation of the human brain, the ANN method has learning, self-organizing, and auto-improving capabilities. Thus, it can capture complex interactions among input/output variables in a system without any prior knowledge of the nature of these interactions, and without having to explicitly assume a model form. Indeed, such a model form is generated by the data points themselves. The database assembled, the architecture of the network selected, and the training process of the ANN model used are described. Initial tests show that the ANN method can accurately predict the slump flow, filling capacity, segregation, and compressive strength test results of SCC mixtures. A model for the acceptance or rejection of SCC mixtures based on knowledge of their mixture proportions is proposed and may be used after sufficient development of a more comprehensive database on an industrial scale for the proportioning of SCC with tailor-made properties.