MODELING AND ESTIMATION OF PLASTIC HINGE LENGTH OF RC COLUMNS USING ARTIFICIAL NEURAL NETWORKS

Artificial neural networks are computing systems that simulate the biological neural systems of human brain. ANNs are structures deliberately designed to mimic and utilize the organizational principles observed in the human brain. Although three dimensional nonlinear dynamic analyses of RC buildings provide valuable information about their behavior, they are expensive and time-consuming. Using ANN based models it is possible, at comparatively low cost and effort, to predict the response of complex RC structures provided that adequate input layers with correct input parameters are chosen and trained. They also enable the designer to rapidly compute the three dimensional response of buildings during the preliminary design stage. In these models the goal is to drastically reduce the computational effort. In this paper, ANN based models were employed as an alternative to determine the plastic hinge length of reinforced concrete columns. This study has shown the feasibility of the potential use of ANN models in determining the response of columns subjected to earthquakes. The promising results observed in the dynamic analysis of RC columns indicate that the ANN models enable the designers to rapidly evaluate the response of columns during the preliminary design stage.