Predicting Displacement Data of Three-Dimensional Reinforced Concrete Frames with Different Strengthening Applications Using ANN

  • Fatih Bahadır Construction Technology, Necmettin Erbakan University, Eregli Kemal Akman High Vocational School, Eregli, Konya

Abstract

In this study, the artificial neural network (ANN) methodwas used to estimate unavailable displacement data of threedimensional(3D) reinforced concrete (RC) frames with differentstrengthening applications. Four 3D-RC frames wereproduced two storeys and one bay in 1/6 geometric scale withthe deficiencies commonly observed in residential buildingsin Turkey. The first specimen was a bare frame containing nobrick walls and no strengthening. The second specimen wasall brick walls and no strengthening. The third specimen wasstrengthened with an internal steel panel. The fourth specimenwas strengthened with an infilled RC shear wall. The specimenswere tested under reverse cyclic lateral loading and constantvertical loading until failure. This study investigated the estimationof displacement data when the linear variable differentialtransformer of 104 numbers is corrupted and some hystereticloop data are missing. Using the method proposed the unavailableor incorrect displacement data can be predicted by ANNwithout performing any additional experiments. Root meansquared error, coefficient determination, mean absolute error,mean squared error and normalised mean absolute error statisticalvalues were used to compare experimental results withANN model results. These statistical values usually exhibit verylow error rate until a cycle of maximum load is reached.

Keywords

three-dimensional, reinforced concrete, artificial neural network, hysteretic loops, displacement data
Published
08-05-2017
How to Cite
BAHADIR, Fatih. Predicting Displacement Data of Three-Dimensional Reinforced Concrete Frames with Different Strengthening Applications Using ANN. Periodica Polytechnica Civil Engineering, [S.l.], may 2017. ISSN 1587-3773. Available at: <https://pp.bme.hu/ci/article/view/9652>. Date accessed: 26 sep. 2017. doi: https://doi.org/10.3311/PPci.9652.
Section
Research Article