Seismic retrofit of full-scale substandard rectangular RC columns through cfrp jacketing and external steel ties

Abstract

It is well known that Turkey is located in one of the most seismically active regions on earth. With this location, Turkey is subjected to large magnitude earthquakes, resulting in substantial loss of life and property damage. Moreover, a large portion of the existing reinforced concrete (RC) structures have not been constructed with sufficient seismic resistance and ductility. In addition, the brittle behavior of substandard columns during earthquakes is one of the most common reasons for structural failures. Since capacity design principles and ductile detailing concepts were not properly adopted during the construction of a large portion of the existing buildings, many of those buildings are in urgent need of seismic retrofitting, particularly in terms of earthquake resistance and the enhancement of ductility characteristics. Past earthquakes have demonstrated that the lack of sufficient transverse reinforcement, poor concrete quality, and plain bars used within columns comprise the main deficiencies of the typical RC structures. One important consequence of these deficiencies remains the lack of sufficient ductility, remarkably limiting the displacement capacity of the existing structures. While confinement is very effective for strength and ductility enhancement for circular columns, the methodology has significant weaknesses for rectangular members. Limited studies have been executed on the study of full-scale columns, and those studies are not sufficiently robust to offer realistic and reliable remediation solutions, particularly for typical, substandard rectangular RC columns. External fiber-reinforced polymer (FRP) jacketing of columns is a feasible and promising method to overcome the insufficient ductility problem of existing substandard structures through increasing their confinement efficiency. In this study, the seismic behavior of full-scale substandard rectangular RC columns were investigated, both experimentally and theoretically. Additionally, this study evaluated the various FRP confinement models, and the ability to improve the model's predictions based on experimental results. Moreover, this study critiques the current seismic design codes, and their limitations in terms of the required strength enhancements and recognition of FRP effective rupture strain. The findings of the study clearly demonstrated the efficiency of FRP jacketing for substandard columns, even in the case of the extended, rectangular cross-sectioned columns. FRP jacketing is particularly effective for ductility, and consequently in lateral displacement (drift) capacity. Furthermore, this study introduced an improved method, using carbon fiber reinforced polymers (CFRP) and external steel ties to accomplish the seismic retrofitting of rectangular RC columns loaded in their weak directions. In the experimental portion of the study, eleven cantilever columns with a cross-sectional aspect ratio of two (300 mm x 600 mm cross-section dimensions) were tested under constant axial load and reversed cyclic lateral loads. The columns are representative of existing substandard structural members, mimicking their characteristics, such as low concrete quality, incorrectly detailed and insufficient transverse reinforcement ratios, and construction using plain round bars. The specimens were tested under two axial load levels (approximately 20% and 35% of the axial load capacity of the columns) and two loading directions (strong and weak directions). In this manner, the specimens were classified into four categories based on axial load, loading direction, CFRP jacketing layers, and the use of external steel ties. The test results indicated that FRP jacketing and insertion of external steel ties significantly improved the seismic performance in terms of ductility and energy dissipation. The theoretical portion of the study consists of three parts. First, the reinforcement specimens were modeled separately using SeismoStruct software as a finite element platform. The spread of the material nonlinearity along the member length and across the section area was explicitly represented through the employment of a fiber modeling approach, implicit in the displacement-based formulation for the inelastic column elements employed in the analyses. Second, various FRP confinement models that were available in the literature were used for comparison with the predictions for this study's models and their experimental results. Additionally, the analytical findings were compared with the test results for both the strong and weak directions of the columns. Comparisons also revealed that the use of various models for the stress-strain relationship of FRP-confined concrete can yield significantly unrealistic or too conservative retrofit designs, in terms the determined yielding point, and the ultimate strain point of the lateral load-displacement relationships obtained from the analytical procedures. One of the stress-strain confinement models was slightly modified to obtain more accurate agreement with test data. Nonlinear analyses were executed for columns tested in this study, as well as columns tested by others. Finally, the current seismic design assessment guidelines were used for prediction of the behavior of retrofitted columns, and the test results were compared with this study's predictions. Such testing revealed that the retrofit design approach and seismic assessment documents, such as ACI 440.2R (2008) and the Turkish Seismic Design Code (2007), are quite conservative, leading to infeasible retrofit options, whereas the approach employing Eurocode 8 (2005) is not conservative and led to more feasible retrofit solutions.