Conference Proceedings

Abstract

In this study, artificial neural network (ANN) method is used to predict displacement data of 3D-reinforced concrete frames and compared with the experimental results of a testing series. Three reinforced concrete frames were produced two storey and 3D in 1/6 geometric scales which contained the deficiencies commonly observed problems in residential buildings in Turkey were tested under reverse-cyclic lateral loading as well as constant vertical loading until failure. These experimental studies are 3-D and having different window opening in brick wall. This study is concerned with the problem of estimation of displacement data when the LVDT of 103 numbers are corrupted and some data of hysteretic cycles are missed. As a result, the values are very closer to the experimental results obtained from training and testing for artificial neural networks model. RMSE, R2 and MAE statistical values that calculated for comparing experimental results with artificial neural networks model results have shown this situation.