Using Artificial Neural Network Containing Two Hidden Layers for Predicting Carbonation Depth of Concrete

Abstract

AbstractCarbonation of concrete is one of the main causes of corrosion of reinforced concrete and consequently inducing deterioration of reinforced concrete structures. The carbonation depth of concrete depends on numerous factors including mix design and exposure conditions. Number of models for predicting carbonation depth including mathematical and analytical predictions are proposed in the literature. However, the models seem to not take into account the complexity of carbonation process and effect of mix design and exposure condition on carbonation depth. Therefore, determining the carbonation depth of concrete is a challenge for civil engineers. In this investigation, Artificial neural network (ANN) model is used to predict the carbonation depth. To develop the model, 300 experimental data were collected from the literature. The collected data is randomly divided into 2 groups with 6 inputs such as cement content, fly ash (FA) content, water content, relative humidity, carbonate concentration and time exposure. Performance evaluation of the models was performed and compared on training dataset (70% data) and testing data set (30% to remaining data) by criteria of coefficient of correlation (R2), root mean square error (RMSE), mean absolute error (MAE). The performance values show that the ANN model can accurately predict the carbonation depth of concrete.KeywordsCarbonation depthMachine learningArtificial neural networkConcreteFly ash