Prediction of Concrete Compressive Strength & Slump using Artificial Neural Networks (ANN)

Concrete is the most used building material in the world, due to its high compressive strength and durability. Those properties are measured and assessed in fresh and hardened states of concrete, with standard methods which are time and cost consuming. In the present study, the compressive strength and slump of concrete has been predicted using Artificial Neural Network (ANN), which is constructed using different input parameters involving concrete mix design (i.e. coarse & fine aggregates properties, cement content, water/cement (W/C) ratio, admixtures type and dosage …etc.).The predicted strength was compared with the experimentally obtained actual compressive strength and slump data collected in many years for different materials and mix designs in the Sudan. An ANN model has been developed by using MATLAB neural network toolbox. A good co-relationships with regression values of 0.915 and 0.931 for strength and slump respectively have been obtained between the predicted and experimental values. It is concluded that the ANN method can gain acceptable predictions for compressive strength and slump.