Utilizing Experimental Model Tests and Artificial Neural-Nets to Estimate the Strength Loss of Heated Reinforced Concrete Columns

An artificial neural network-based model is developed to predict the loss in capacity ofreinforced concrete columns subjected to elevated temperature. A series of RC columnmodels have been tested. The process of increasing the temperature is performed while themodel columns carrying the service loads, thus simulating the actual condition taking place during real fire event. Different column sections; and aggregate, plaster and admixture types are used. To study the effect of these factors on the residual strength. Results of experimental model tests are then analyzed, clustered and used to train a specially designed artificial neural network (ANN) to be capable of predicting reduced concrete strength. ANN estimations, when compared to model test results, showed very good agreement. Such observation indicates that ANN could be effectively used to accurately predict strength reduction due to exposed to elevated temperature.