Estimating the fire resistance of reinforced concrete column subjected to compression using an artificial neural network

Abstract

This study aims to develop an artificial neural network (ANN) model for predicting the fire resistance of reinforced concrete columns subjected to compression. A total of 374 experimental test results, incorporating 12 input parameters, were collected. These parameters include compressive strength, concrete cover, column length, crosssection width, cross-section height, boundary conditions, compressive load, load eccentricity, longitudinal reinforcement ratio,  number of reinforcements, diameter of reinforcements, and yield strength of reinforcements. Additionally, a software tool was developed to predict the fire resistance of reinforced concrete columns by integrating the ANN model's estimated dataset. The results showed that the ANN model effectively predicted fire resistance in the training set, achieving an R value of 0.94 and an RMSE of 23 minutes. In the testing set, the model demonstrated acceptable accuracy, with an R value of 0.78 and an RMSE of 59 minutes. Sensitivity analysis revealed that the cross-section height was the most important parameter influencing fire resistance. Furthermore, the developed software was found to be simple, userfriendly, and efficient.

keywords: Reinforced concrete column; compression; fire resistance; artificial neural network; sensitivity analysis.