Prediction of ultimate shear capacity of FRP-reinforced concrete beams without stirrups using neural networks
Author
Source
ZANCO Journal of Pure and Applied Sciences
Issue
Vol. 27, Issue 4 (30 Jun. 2015), pp.1-19, 19 p.
Publisher
Salahaddin University-Erbil Department of Scientific Publications
Publication Date
2015-06-30
Country of Publication
Iraq
No. of Pages
19
Main Subjects
Topics
Abstract EN
This paper explores the use of artificial neural networks (ANNs) with feed forward back propagation in constructing a model for predicting the ultimate shear capacity of slender beams reinforced longitudinally with fiber reinforced polymer (FRP) bars as the ANNs techniques have shown promise for modeling complex relationships.
The ANN model was trained and tested using experimental results of 166 beams under concentrated loads without shear reinforcement.
The input layer of the ANN model comprised beam geometry, concrete and FRP reinforcement properties.
The output layer of the model contained one parameter representing the ultimate shear strength of the beam.
The ANN model successfully predicted the ultimate shear strength of FRP-reinforced concrete beams within the range of the considered input shear parameters.
The mean and the standard deviation of the experimental to the predicted shear strength (Vexp/Vpred) were 1.01 and 0.18 for the training data and 1.02 and 0.13 for the testing data, respectively.
The close correlation between predicted and experimental shear strength shows that the ANN-based modeling is a very reasonable method for predicting the ultimate shear capacity of FRP-reinforced beams.
The predicted shear strength capacities were also compared with the predictions of ACI 440-06, BISE-99, CSA S806-02, JSCE-97 guidelines and a proposed method.
The comparison showed that the ANN model has a higher potential in predicting the ultimate shear strength of FRP-reinforced concrete beams within the range of input shear parameters.
Also the shear strength ratio (Vexp/Vpred) of the constructed ANN model, When compared with the other design methods, provides a uniform level of safety across the entire range of the input shear parameters of the testing data followed by the proposed method.
The trained ANN model was also used to perform parametric studies to evaluate the effect of different input shear parameters on the ultimate shear capacity of FRP-reinforced concrete beams.
One of the interesting finding was that the normalized shear strength decreases with increasing beam depth which agrees well with the test results and the theoretical size-effect analysis.
American Psychological Association (APA)
Yusuf, Ali Ramadan. 2015. Prediction of ultimate shear capacity of FRP-reinforced concrete beams without stirrups using neural networks. ZANCO Journal of Pure and Applied Sciences،Vol. 27, no. 4, pp.1-19.
https://search.emarefa.net/detail/BIM-667653
Modern Language Association (MLA)
Yusuf, Ali Ramadan. Prediction of ultimate shear capacity of FRP-reinforced concrete beams without stirrups using neural networks. ZANCO Journal of Pure and Applied Sciences Vol. 27, no. 4 (2015), pp.1-19.
https://search.emarefa.net/detail/BIM-667653
American Medical Association (AMA)
Yusuf, Ali Ramadan. Prediction of ultimate shear capacity of FRP-reinforced concrete beams without stirrups using neural networks. ZANCO Journal of Pure and Applied Sciences. 2015. Vol. 27, no. 4, pp.1-19.
https://search.emarefa.net/detail/BIM-667653
Data Type
Journal Articles
Language
English
Notes
Includes appendices : p. 16-19
Record ID
BIM-667653