ESTIMATING CONCRETE COMPRESSIVE STRENGTH PRODUCED BY GFRP AND POZZOLANIC MATERIALS EXPOSED TO FIRE USING ANN METHOD

In this article, in order to estimate concrete strength reinforced by GFRP and pozzolanic material or non-reinforced concrete, ANN method has been used. Hence, applied materials specifications, produced concrete properties before and after the fire have been considered as system inputs. Some of the mentioned properties are the ratio of w/c, using or non-using GFRP, amount of applied pozzolanic materials and concrete water absorption. Utilized w/c ratios are 0.3, 0.35, 0.4 and the quantity of pumice or expanded perlite is 10%, 20%, 30% of fine aggregates volumes. The temperatures of experiments have been 25 C or 600 C. The result of article shows the reasonable coincidence between the output of the program and outcome of the experiment. Consequently, Artificial Neural Network (ANN) can be used as an effective method to predict the concrete strength exposed/unexposed to fire.