A New Estimation Approach for Fly Ash Incorporated High Strength Concrete Using Artificial Neural Network

The fly ash obtained from the extraction of exhaust gases from coal and silt furnaces, is non-plastic and fine, which is a different combination based on natural coal fuel. Fly ash is one of the waste materials in thermal power plants. The use of waste materials such as fly ash in the concrete industry offers an alternative and valuable solution to create an environmentally friendly environment. However, experimental work is time-consuming and expensive, and the use of soft computing techniques can speed up the process of predicting concrete's resistance properties. In this study, artificial neural network (ANN) was used to predict the compressive strength of fly ash-based high-performance concrete. A number of 471 experimental data were extracted from valid sources and parameters such as cement, fly ash, water, superplasticizer, fine and coarse aggregate and age of the samples were considered as input parameters and compressive strength of samples was considered as output parameters. Among the networks with different number of neurons, a network which has the best correlation coefficient values obtained from training, evaluation and testing data and also has the lowest mean square error (MSE) value was selected as the optimal network. In this study, the network with the number of 6 neurons provided the best results. Also, the effect of each parameter in different numerical ranges on the compressive strength of concrete was investigated and presented, and it was found that the age of the samples and the amount of cement, fly ash and water have the greatest relative importance.