Artificial Neural Network Model For The Prediction Of Elastic Modulus Of Concrete

Subscribe to access this work and thousands more

ABSTRACT

This research presents Artificial Neural Network Model for the prediction of the Modulus of Elasticity of Concrete. Egbulonu (2011) equation derived from Scheffe’s (4, 2) simplex equation for predicting the Modulus of Elasticity (MOE) was used to generate 800 values. These data represent different values of Modulus of Elasticity (MOE) out of which, 571 values were selected randomly by the artificial neural network. From the selected values, 400 were used for training the network, 89 for testing the network and the remaining 82 for validating the network. The network was used to predict the Modulus of Elasticity (MOE) of the concrete. Its result was compared to the experimental results and was found to be 1.49% which is very close. The network was tested for good fit. It was found that the coefficient of correlation, R values for training, testing and validating data were 0.95237, 0.93731 and 0.91905 respectively, which showed that the data used for training, testing and validating the network have good fit since their R values is greater than 0.9. The network was also tested for adequacy at 0.05 significance level using statistical student’s test and was found to be adequate. Thus, the model can be used to predict the Modulus of Elasticity of concrete for any given mix ratio or vice versa.

Keywords: Concrete, Elasticity, Artificial Neural Network, Training, Testing, Validating, Mix Ratios, Optimization.  

Subscribe to access this work and thousands more