Düzce Üniversitesi Bilim ve Teknoloji Dergisi (Feb 2015)
Prediction Of Compressive Strength Of Normal Weight Concrete Including Fly Ash By Using Artificial Neural Network
Abstract
The compressive strength of normal weight concretes which include fly ash have been predicted by artificial neural network (ANN) model. For the model 103 experimental results were used and trained. Cement content, water content, fly ash amount and slump values were used as inputs and compressive strength (MPa) was used as output while developing ANN model. Some of the architectures with different number of neurons studied here in hidden layer and their correlations with experimental results investigated. One hidden layer with seven neurons was the best model because of its high correlation with experimental results for testing set. Results have shown that ANN has strong potential for predicting compressive strength of concretes containing fly ash.
