Hysteretic behavior of tubular steel braces retrofitted with CFRP memebers

Steel braced frames have been used in structures located in seismic regions for many decades. Use of tubular sections (box or pipe) in structures as braces is widespread in construction industry due to their high compression capacity and ease of production. Reconnaissance investigations carried out following destructive earthquakes showed that most tubular braces failed from net sections of end connections under tension effects while some of them failed at mid-sections due to local buckling under repeated compression forces. In parallel with these findings, many previous studies suggest to retrofit net sections of tubular braces using steel plates and to decrease local buckling effects by selecting sections having local slenderness less than a certain limit. Specifically, brittle heat affected zones (HAZ) occur inevitably when existing tubular braces are retrofitted with on-site welded steel plates mainly due to inappropriate welding process. In addition, tubular braces in most of the existing older structures do not meet local slenderness requirements, leading to possibility of early local buckling. Moreover, updated international seismic codes and Specification for Buildings to be Built in Seismic Zones-2007 state that brace members shall never suffer from local buckling until the braces reach their axial compression capacities, given that brace sections meet local slenderness limits. However, many studies have indicated that even when these limit values were met, local buckling was inevitable under reversed cyclic loading. Application of carbon fiber reinforced polymer (CFRP) composites in steel bracings and their experimental investigations have been somewhat limited when compared to their use on steel bending members. One of the original aspects of this work is the application of such composite materials to buckling of compression members, and thus to delay possible local buckling and fracture from net sections. Consequently, only net section reinforcement for compact and slender tubular sections and, local reinforcements at mid-section and net section for noncompact and stocky braces are recommended. A total of 20 specimens was designed per AISC codes and produced; 5 specimens (box section) to investigate only net section reinforcements, 12 specimens (box section) and 3 specimens (pipe section) to investigate reinforcements of net sections and local buckling region were tested until fracture at Structural and Earthquake Engineering Laboratory (STEEL) of Istanbul Technical University (ITU) using a test set-up specially designed for this study. Specimens were divided into 7 groups of 3, only one group containing 2 specimens. Each group had one reference (unretrofitted or bare) specimen. Tension and compression forces were applied to all specimens with displacement controlled loading histories using a hydraulic actuator having a tension and compression capacities of 250 kN, and a total displacement capacity of ±300 mm. 18 specimens were tested following the ATC-24 (1992) cycling loading protocol, and two specimens were tested per SAC/BD-00/10 near fault loading protocol. Coupon tests were carried out to get actual material properties per ASTM-A370-08a (2008). Actual material properties were used to estimate yield points of the specimens. The proposed retrofitting method preserves its originality in terms of being local when the existing literature is reviewed. Experiments showed that fracture lives and displacement ductilities of box specimens having local slenderness of between 14 and 17 were effectively increased using local CFRP reinforcements at mid and net sections. End net sections were more critical when local slenderness of box sections was less than 14, therefore only end section reinforcements contributed effectively to fracture lives and ductilities of such specimens. The proposed retrofit method at mid section of braces was partially wrapping this region at least 1/3 of clear brace length with at least two layers of CFRP; note that all fibres in CFRP is to be in the direction of wrap. Factor of safety was taken at least 2.00 while calculating the bond strength of adhesive and load carrying capacity of CFRP plate for end net sections. Also, two layers of CFRP sheet wrap were applied after having CFRP plates on two sides of end net section. Mid-span retrofit does not provide any improvement on fracture life when local slenderness equals to 20 or greater. In addition, effectiveness of mid-span retrofit significantly decreased when local slenderness was between 17 and 20. Local retrofitting method was ineffective under near fault loading history. Moreover, two reference specimens were modeled and analyzed using ABAQUS 6.14 in order to understand the complex hysteretic behavior of such braces under cyclic loading. Stable hysteretic behaviors were acquired and were quite similar when compared with experimental results. Note that fracture is not modeled in this work. Contents of the main chapters of this thesis are given below: In the first chapter, aim and scope of study are explained and information on the methods followed are elucidated. Literature review on complex hysteretic behavior of braces and current situation of CFRP application on steel members are explained in Chapter 2. In addition, tubular brace design rules in current design codes are summarized. Design steps of L-shaped test setup and special details providing stability to setup are given in Chapter 3. Moreover, coupon tests, design and manufacturing of specimens, CFRP retrofitting details, and instrumentation of the test setup are explained in detail. Detailed information on loading histories and results from pushover analyses using actual material properties are given in Chapter 4. Furthermore, experimental observations for each specimen are presented herein chapter. Chapter 5 is devoted to the explanations on performance criteria to evaluate the results of bracing tests. In addition, test results are discussed for each specimen and comprehensive comparisons are carried out. Main findings derived from the test results are given in Chapter 6. Chapter 7 includes details of the theoretical study, finite element model (FEM), and assumptions used in the analyses. Moreover, hysteretic curves from tests and finite element analyses are compared and interpreted. In Chapter 8, general results are evaluated and suggestions for potential future studies are made. Results from this experimental study showed that local/partial retrofitting using CFRP members could increase cumulative energy dissipation capacity of steel tubular braces and extend their fracture lives. It is determined that there will be significant increases in ductility and the hysteretic energy dissipation provided that the produced tube specimens are kept below a certain slenderness limit. The proposed retrofitting method plays an effective role in retrofitting of existing tube braced steel buildings in terms of improving their seismic behavior. Besides that, it is believed the proposed method could be cost effective since it provides significant advantages as discussed and also much easier to implement in practice.