An artificial neural network modeling of the flexural behavior of an RC beam strengthened using carbon fiber-reinforced polymer

• Main Authors: Arcilla, Chelsea Marguerite V., Vilaga, Frances Ira B.
• Format: text
• Language: English
• Published: Animo Repository 2001
• Subjects: Concrete beams; Flexure > Testing; Strength of materials > Testing; Civil Engineering
• Online Access: https://animorepository.dlsu.edu.ph/etd_bachelors/5156
• Institution: De La Salle University

The use of externally bonded and near-surface mounted reinforcement of CFRP on flexure beams with the use of epoxy or resin is an effective strengthening and repairing technique for RC beams. For our study 158 beams specimens were gathered for the development of flexural strength analysis through the use of artificial neural network modeling. The 158 beam specimens are divided into three groups 70% of the data were used for the training, 15% were tested, and the remaining 15% were used for verification. The input parameters used are the easily available variables within the journals, namely: base (b), height (h), length (L), cross-sectional area of the tension enforcement bar (As), cross-sectional area of the CFRP plate/sheet (Af), and the yield strength of CFRP (ftf) the output parameter is the ultimate flexure moment. The developed ANN model is then compared with other existing models to determine its efficiency and accuracy. Lastly, a parametric study was done in order to determine the importance, effect, and behavior of each parameter. Through this, the proponents of this study were able to produce a reliable flexural behavior model of a RC beam strengthened with CFRP even more efficient and effective than that of other studies of the same topic.