Modeling FRP-confined RC Columns using SAP2000

Retrofitting using Fiber Reinforcement Polymer (FRP) for buildings that perform poorly against seismic force has attracted the attention of researchers. The poor performance of structures in original and rehabilitated forms is assessed based on the response of the entire structure against forces such as gravity and earthquake. Although researchers have attempted to extract the responses of FRP-confined Reinforced Concrete (RC) columns using several finite-element software and experiments, modeling the whole retrofitted structure using FRP-confined columns rarely performed and if performed has not been verified [1]. This is due to the fact that, modeling the whole structure is almost impossible in Finite Element Modeling (FEM) and is conventionally done in lumped plasticity modeling procedure. SAP2000 is conventionally used to analyze the structure’s model through finite elements and lumped mass modeling. Extracting the exact response of a structure using the SAP2000 analyzer requires the precise modeling of structure members, otherwise the softwares response reliability could not be enumerated. To reach this aim, this research attempted to verify the response of SAP2000 in lumped mass plasticity modeling procedure for FRP-confined columns following the Lam and Teng [2, 3] theory for the wrapped concrete material definition. The study was implemented by comparing independent calculation of FRP-confined RC section ductility, such as moment-curvature and PMM interaction, versus analysis output of SAP2000. In next step, lumped mass modeling of FRP-confined columns was also developed and compared with experiments. Comparison of analysis software and independent calculation of section ductility parameters are resulted in less than 10 percent difference. Furthermore, assessment of FRP-confined columns in lumped mass plasticity modeling procedure compare to experiment is obtained in almost less than 25 percent difference; however same trend of analysis and experiments results were barely discernable. 25% difference is the software acceptable range of difference between experiment and analysis. Defining the retrofitted material properties using various methods had a remarkable effect on the accuracy of software response that is studied in this research.