Fire resistance prediction for reinforced concrete columns: A comparative application of artificial neural network and random tree methods

Abstract

Abstract The most common index for passive fire protection systems of built environments is fire resistance. Existing models for determining the fire resistance of reinforced concrete (RC) columns apply to certain conditions. Meanwhile, accurate determination of fire resistance of RC columns will aid in designing sustainable RC columns, saves lives, and protect property. This study leveraged on the power of soft computing methods – artificial neural network (ANN) and random tree (RT) – in predicting the fire resistance of RC columns. Ninety-six observations were assembled from existing fire tests on RC columns. Nine parameters are the predictors used to predict the fire resistance of RC columns. With respect to model predictive performance, the ANN model slightly outperformed the RT model. Genetic algorithm was also used to generate useful mathematical model for fire resistance prediction. The results of the current study highlight the merit of using soft computing (SC) methods in structural fire engineering applications given their extraordinary ability to comprehend multidimensional phenomena with high prediction accuracy.