Seismic retrofit of rectangular rc columns with crosssectional aspect ratio of three using frp sheet

It is believed that earthquake is the most damaging natural calamity in the world. It causes both loss of property and life due to structural collapse. High percentage of the territory of Turkey is at high earthquake risk. In 1999 Kocaeli earthquake, 17,127 people lost their lives, 43,953 people injured. In 1939 Great Erzincan earthquake, 32,962 people lost their lives. This shows that earthquake resistant buildings have to be designed and constructed and existing non-resistant old buildings have to be retrofitted or rebuilt in our country. It is well known that a great majority of existing structures in Turkey has not been constructed properly. Insufficient transverse reinforcement is among most common weaknesses of columns, which are vital structural members, against earthquakes. This weakness causes brittle behavior of structural members during earthquakes. There are limited number of research projects about 2, 3 or more than 3 aspect ratio columns' seismic performance. In particular, it could not be come across a research project that investigate seismic performance of columns with structural deficiencies like high aspect ratio, high axial load, low concrete quality, low volumetric ratio of transverse bars, 90◦ transverse hoop's hook angle etc. at the same time as the buildings of our country have. In this study, reinforced concrete columns with cross-sectional aspect ratios of 3 which are retrofitted by using lateral carbon fiber reinforced polymers sheets and/or laterally applied external steel rods are tested under constant axial load and reversed cyclic lateral incremental displacements. Tests are taken place as displacement-controlled. Lateral displacements are performed by using hydraulic actuator. This loading system is for simulating seismic scenario of structural columns. Specimens are tested before and after retrofitting with carbon fiber reinforced polymer composites and/or lateral steel rods applied in the direction of short edge of the columns. A reference column and three retrofitted columns are tested within this study. For representing the typical construction method for existing relatively old building in our country, poor concrete quality and plain round reinforcing bars are intentionally selected for design and construction of the specimens. Concrete compressive strength in this study is about 11 MPa at the times of tests. Reinforcing steel yield strength is 310 MPa and ultimate tensile stress is 430 MPa. 20 cm transverse bar spacing is selected to represent low confining of columns of existing old buildings. Hook length and hook angle of transverse hoops are 10 cm and 90◦, respectively. All columns are subjected to high constant vertical load (approximately 40% of the axial load capacity of the columns, 0,4f_co^' bh) throughout application of reversed cyclic lateral displacements representative of seismic actions. The reference specimens exhibited premature failure through crushing of concrete in compression zone followed by buckling of longitudinal bars in compression. On the other hand, the retrofitted columns exhibited a much superior seismic performance, particularly in terms of ductility. While the reference column lost high percent of its strength at around the drift ratio of 2%, the retrofitted specimens could reach drift ratios between 4 to 8% depending on the application details of retrofitting. Even specimen retrofitted with FRP strips is more ductile than reference specimen. It should be noted that based on unavailability of sufficient experimental data, existing technical documents limit the use of external fiber reinforced polymer jacketing to columns with maximum cross-sectional aspect ratio of 1,5~2,0. This study clearly demonstrates that columns with cross-sectional aspect ratios of 3 can significantly benefit from external confinement with fiber reinforced polymer jacketing. In order to compose a design algorithm, analytical studies were performed. For the analytical calculation, existing material model and a plastic hinge model proposed in this study. For reference specimen, which has no FRP jacket, unconfined and confined concrete model proposed by Mander et al. (1988) is used. In order to determine stress-strain relationship of column confined with FRP strips, compressive behavior of carbon fiber composite jacketed concrete model proposed by Ilki et al.(2004) is used. For modelling the behavior of specimens strengthened by using both fiber reinforced polymers and steel rods in the direction of short edge of the section, a confined concrete model suggested by Galala, Arafa ve Ghobarah (2005) is used. Because of the fact that this model takes contributions of both FRP and rods into account, it is preferred. Steel stress-strain model proposed by Yalcin and Saatcioglu (2000) which includes the effects of buckling for compressive behavior of steel is used for determining steel behavior of the specimens. By using evaluated material models, cross-sectional analysis is performed with computer software called XTRACT. This software uses the fiber approach to solve equilibrium equations of the section and incremental parameter is moment. The software keep axial load constant and increase moment monotonically. By using this software, moment-curvature relationship of the sections are obtained to calculate plastic hinge lengths and hence top displacements of the specimens. Plastic hinge model proposed in this study is based on yield moment capacity of the specimens. Under the assumption that longitudinal bars buckle at where cover concrete is crushed and plastic deformations are formed at that zone. In order to calculate the top displacement-lateral force relationship, a method based on superposition of elastic and plastic top displacement is used. Gross concrete section inertia moment and elasticity modulus of the concrete are used for the calculations. Determining the elastic top displacements is performed under the assumptions that confinement effects is negligible, contribution of steel to inertia moment is neglected and elastic deformations are distributed to whole length of the column. Determining the plastic top displacement of the specimens is performed under the assumptions that plastic deformations are formed at only plastic hinge zone uniformly and rest of the column is not deformed. Analytical studies showed that calculated results and test results are consistent in acceptable range. Therefore, calculation algorithm represented in this study can be executed by using some safety coefficients to design retrofitted structures. Furthermore, it is observed after analytical calculation that strength and ductility characteristics of reference and retrofitted by fiber reinforced polymers strips jacketed columns can be predicted quite accurately using available analytical models. As a conclusion, experimental and analytical results show that there is necessity to research seismic performance of the columns which are subjected to high axial load. Plus, it is clearly seen in this study that deformation limits for life safety state in relevant codes are quite conservative for not only non-retrofitted but also retrofitted columns. Detailed after studies may provide chance to revise codes, guides or other technical documents. As another conclusion, the retrofit design of high aspect ratio rectangular columns using external jacketing with fiber reinforced polymer composites can be executed using the analytical approach presented in this study. It is important to emphasize that the obtained results are valid for the range of parameters tested within this study. Therefore, the conclusions may not be valid for columns with cross-sectional aspect ratios more than 3. Similarly, it is important to highlight that the tests and theoretical analyses are carried out for the seismic actions in the strong direction of the columns, which are normally expected to be subjected to higher seismic demands during earthquakes. Therefore, the examination of the behavior around weak axis is outside the scope of this study.