Simplified predictive ANN model for ultimate compressive strength of circular RC columns using CFRP as confinement

• Main Author: Ongpeng, Jason Maximino C.
• Format: text
• Published: Animo Repository 2008
• Subjects: Carbon fiber-reinforced plastics—Testing; Civil Engineering
• Online Access: https://animorepository.dlsu.edu.ph/faculty_research/8872
• Institution: De La Salle University

Today, retrofitting structures became dominant due to the changes in codes and deterioration of structures. One way of retrofitting reinforced concrete (RC) structure columns is by the application of Carbon Fiber Reinforced Polymer (CFRP). This method uses epoxy resin as the binding materials together with carbon fiber sheets to form a fiber-reinforced polymer matrix termed as CFRP wrapped arounc RC columns. The use of CFRP is found to be an effective technique of providing structural strengthening that is lighter, non-corrosive, and less labor intensive than the application of steel plate. At De La Salle University year 2006, experiments and testing of fabricated circular RC columns reinforced with CFRP were done to acquire numerous local data together with the author as the adviser of the researches. With these experimental data, conjunction of foreign published data and local data were used to conduct and analyze the strength performance of CFRP in concrete. From the Self Organized Map (SOM), two classifications were derived from the foreign and local data available. It was observed that the CFRP as a retrofitting material behaved in two ways: active and passive. If the volumetric ratio of steel is low, then the tendency of the CFRP will be active that causes higher development in ultimate confined compressive strength. If the volumetric ratio of steel is high, then the tendency of the CFRP will be passive that causes lower development in ultimate confined compressive strength. In Back-Propagation (BP), a simplified version of a model can be achieved having only 2 input parameters pcfrp (volumetric ratio of CFRP to concrete) and ps (volumetric ratio of steel ties to concrete). With these models, parametric studies were made to investigate the differences in the behavior of both classes. The combination of SOM and BP proves to be effective in classifying and promoting good preditions in the ultimate confined compressive strength of circular RC columns.