Estimation of slump value and Bingham parameters of fresh concrete mixture composition with artificial neural network modelling

High performance is the most important expectation from concrete which is commonly used in today’s construction technology. To form a high performance concrete “HPC”, two fundamental properties are required. These properties are optimization of the materials used to form the concrete and the workability of fresh concrete during shaping. Many scientists have used rheological properties in conjunction with Bingham model to determine the workability of fresh concrete. Bingham model is represented by two parameters: yield stress and plastic viscosity. Even though, many models are developed to explain rheological properties, there is no acceptable easy to use method. In this study, artificial neural network “ANN” is used to determine the rheological properties of fresh concrete. Ferraris and de Larrard’s experimental slump, yield stress and viscosity data from different composed concretes is used in this study. Slump, yield stress and viscosity are estimated with respect to mixture design parameters. Obtained results from this study indicates that ANN is a utilizable method to determine the rheological proporties (Bingham model) of fresh concrete.