Journal of Constructional Steel Research

Transcription

1 Journal of Constructional Steel Research 14 (215) Contents lists available at ScienceDirect Journal of Constructional Steel Research Seismic behavior of replaceable steel truss coupling beams with buckling restrained webs Xian Li a,b,,heng-linlv a,b, Guang-Chang Zhang b, Bei-Dou Ding b a State Key Laboratory for Geomechanics & Deep Underground Engineering, China University of Mining and Technology, Xuzhou, Jiangsu, China b Jiangsu Key Laboratory of Environmental Impact and Structural Safety in Civil Engineering, School of Mechanics and Civil Engineering, China University of Mining and Technology, Xuzhou, Jiangsu, China article info abstract Article history: Received 1 December 213 Accepted 24 September 214 Available online 3 October 214 Keywords: Seismic behavior Steel coupling beams Hybrid coupled wall systems Buckling restrained web Strip model In order to facilitate the constructability and meanwhile ensure desirable seismic behavior, an innovative type of replaceable steel truss coupling beam with a buckling restrained web was conceived and studied. The buckling restrained steel web is designed and detailed as a fuse and a damper of the beam in which all inelastic deformations and damage are concentrated. Bolted connections between steel webs and beam chords, as well as pin connections between beam chords and adjacent reinforced concrete shear walls, are employed to minimize the postevent repair/replacement difficulties and expenses. To evaluate the seismic behavior, three large scale coupling beam specimens were constructed and tested under cyclic loadings. The effects of some configurations of the steel webs including welded edge stiffeners and slits on the seismic behavior were highlighted. The test results indicate that all three specimens failed in a ductile manner with a concentration of inelastic deformation at steel webs and thus exhibited desirable deformation and energy absorption capacities. The strength and stiffness of the proposed coupling beam can be enhanced by welding edge stiffeners to steel webs while the steel web with slits is susceptive to suffer significant buckling of flexural links, resulting in a relatively lower strength and ductility of the beam. To predict the load carrying capacity of the proposed coupling beam, a modified strip model to account for the beneficial effects of the buckling bracing provided by precast reinforced concrete panels was developed. The analytical results were compared with the experimental results. 214 Elsevier Ltd. All rights reserved. 1. Introduction Coupled shear wall systems, which consist of two or more in-plane shear walls interconnected with coupling beams, are frequently used in medium and high-rise buildings [1]. Due to the favorable coupling actions created by coupling beams, coupled wall systems resist overturning moments induced by seismic events partially through axial compression tension couples across the wall systems rather than totally through the individual flexural actions of the walls, and thus they are efficient systems with great lateral stiffness and strength. Moreover, in contrast to conventional individual walls in which the dissipation of seismic energy usually concentrates at shear wall bases, properly designed coupled wall systems develop a means of dissipating energy mainly through large inelastic deformations of coupling beams over the entire height of the wall systems. Thus, the overall seismic responses of a coupled wall system largely depend on the type and details of the coupling beams used. However, previous earthquakes such as Wenchuan earthquake in China have demonstrated the vulnerability Corresponding author at: Jiangsu Key Laboratory of Environmental Impact and Structural Safety in Civil Engineering, School of Mechanics and Civil Engineering, China University of Mining and Technology, Xuzhou, Jiangsu, China. address: lixian@cumt.edu.cn (X. Li). of conventionally reinforced concrete coupling beams with longitudinal flexural and vertical shear reinforcements subjected to seismic attacks [2,3]. To improve the seismic behavior of conventional coupling beams, some efforts have been made by providing special reinforcement layouts [4,5] or using steel and steel concrete composite coupling beams [6 11] as substitutes in recent decades. Although pertinent experimental results showed that the aforementioned details offered coupling beams an improved performance over conventional beam types to some extent; however, each of them had some drawbacks to limit their wide use in practice. Diagonally reinforced concrete coupling beams proposed by Park and Paulay [4] have proved to be very complicated, if not impossible, to fabricate and construct. Steel and steel concrete composite coupling beams have gradually gained acceptance as a viable substitute for reinforced concrete coupling beams in high seismic regions and have been studied by some researchers [6 11].Their experimental results indicated that steel and steel concrete composite coupling beams, similar to steel link beams in eccentrically braced frames, can provide excellent ductility and energy dissipating characteristics. However, special configurations of beam-to-wall connections are usually required and significant damage of the connections and coupling beams leads to very costly post-event repair. To further enhance the advantages of steel coupling beams, Fortney [12] studied a type of built-up X/ 214 Elsevier Ltd. All rights reserved.

2 168 X. Li et al. / Journal of Constructional Steel Research 14 (215) steel coupling beam consisting of two outer beam sections embedded in wall piers and a replaceable central fuse section. However, some types of weld failures, block shear failure and bolt bearing failure were observed at the end of the Fortney's test. In this study, an innovative type of replaceable steel coupling beam for coupled wall systems, as shown in Fig. 1, was conceived and studied. The proposed coupling beam is designed as a steel truss with a replaceable buckling restrained web. The buckling restrained web, which is confined by two precast reinforced concrete panels to avoid elastic and inelastic out-of-plane buckling, serves as a fuse and a damper of the beam where all deformations and damage are concentrated during medium and serious seismic events. That is to say, both beam chords and shear tabs for connections are intended to remain elastic when the buckling restrained web reaches its ultimate load to dissipate energy input. In the buckling restrained web, some gaps are set around the precast reinforced concrete panels to make the reinforced concrete panels provide only out-of-plane bracing to the web rather than resist lateral applied loadings. To facilitate the construction and post-event repair, the web is bolted to both top and bottom chords of the beam through fish plates while the ends of the beam chords are connected to the shear tabs by means of real pins, which allow free rotation of the chord members inside the beam plane. Then the shear tabs are welded on face plates with multiple steel angles embedded in reinforced concrete shear walls. If the wall boundaries are reinforced with steel structural columns, the shear tabs can also be directly connected to the steel columns. Compared with conventional steel coupling beams, the proposed steel truss coupling beam has the following noteworthy merits: 1) the installation and post-event repair of the beams are simple since bolted connections between steel webs and beam chords, as well as pin connections between beam chords and adjacent reinforced concrete shear walls, are employed; 2) the replaceable buckling-restrained web can achieve large inelastic deformation and thus can contribute to the energy dissipation capacity of the coupling beam; and 3) the strength and stiffness of the coupling beams can be easily adjusted by changing the configurations of the steel webs, facilitating the seismic design of overall structures. 2. Experimental program The main objectives to be achieved in this research program are: 1) to verify the applicability of the proposed replaceable steel truss coupling beam with a buckling restrained web for coupled wall systems; 2) to investigate the seismic behavior of the proposed coupling beams in terms of strength, stiffness, ductility and energy absorption capacity; 3) to evaluate the effects of some configurations of the steel webs including edge stiffeners and slits on the seismic behavior; and 4) to provide a rational analytical model to predict the strength of the proposed coupling beam Specimen design Three replaceable steel truss coupling beam specimens called as SCB1 SCB3 were constructed and tested under cyclic loadings. Each specimen included two reinforced concrete shear walls connected using a steel truss coupling beam to represent a floor-level subassembly at a hybrid coupled wall structure. The test beam prototype has a span of 18 mm and a depth of 12 mm, which is common in reinforced concrete frame-core wall structures for tall office buildings with typical wall openings and story heights. Due to the geometric and strength constraints of existing laboratory conditions, tests were conducted on 2/3- scale coupling beam specimens having a clear span of 12 mm and a distance between centerlines of beam chords equal to 65 mm, and the design strength of the coupling beam specimens was controlled by their replaceable steel web with measured dimensions of 72 mm (length) 47 mm (depth) 3.8 mm (thickness). As stated previously, the strength of the proposed coupling beam can be easily adjusted by changing the thickness of the web if necessary. The details of the specimens are shown in Fig. 2. Since the proposed coupling beams can be replaced after test, reinforced concrete walls and beam chords were reused during the tests. The dimensions and reinforcement layouts of the reinforced concrete walls are included in Fig. 2c. The reinforced concrete walls, as well as the shear tabs with multiple steel angles for the pin-connections between beam chords and adjacent reinforced concrete shear walls, were designed and detailed according to the Chinese code for seismic design of buildings [13] to ensure the full development of capacities of coupling beams. Three steel angles with a flange width of 5 mm and a thickness of 6 mm were adopted in each beam-to-wall connection. Concrete filled steel tubes, which were composed of square steel tubes with a cross section of 15 mm (width) 15 mm (depth) 1 mm (thickness) and infill concrete with a target compressive strength of 55 MPa, were employed for the top and bottom chords of the coupling beams. The replaceable thin steel webs were bolted to the chords using 25 mm thick fish plates and slip-critical high strength bolts with a diameter of 16 mm, and the potential premature elastic and inelastic out-of-plane buckling of the webs was to be inhibited by a couple of precast reinforced concrete Buckling restrained web Top chord A Reinforced concrete shear wall Bolts Top chord Precast RC panels Steel embedment for connection A Bottom chord Fish plates Bottom chord A-A section Fig. 1. Proposed steel truss coupling beam with a buckling restrained web.

3 X. Li et al. / Journal of Constructional Steel Research 14 (215) b) A-A section a) Elevation of specimens c) Reinforcement layouts of reinforced concrete shear walls d) Details of webs for specimens SCB1 and SCB3 e) Details of web for specimen SCB2 Fig. 2. Details of steel truss coupling beam specimens (unit: mm). panels, as shown in Fig. 2b. The precast reinforced concrete panels had dimensions of 66 mm (length) 29 mm (width) 5 mm (thickness) and were transversely and longitudinally reinforced by one layer of 6 mm-diameter deformed bars with spacing of 1 mm. In order to ensure that the precast reinforced concrete panels provide only transverse bracing to the webs and do not resist the applied shear forces during tests, 2 mm-wide gaps between the reinforced concrete panels and fish plates/washers were set and the perforations in the reinforced concrete panels were oversized with respect to the bolt diameter. The main test variable in this research was the configurations of the replaceable steel webs. Specimen SCB1 was designed as a benchmark and two free edges of its steel web were each stiffened by a 15 mm wide and 12 mm thick stiffener. This specimen was intended to dissipate input energy by large shear deformations of the stiffened web. The web of specimen SCB2 was identical to that of specimen SCB1 but was divided by two 1 mm wide slits into three sections, as shown in Fig. 2e. To minimize stress concentrations, the ends of the slits had circular arcs. The purpose of using slits was for the steel plate between the slits to behave as a series of flexural links with a width-tothickness ratio of 6.5, which would undergo large flexural deformations relative to their shear deformations. Thus the effects of the different failure modes of the steel webs on the seismic behavior of the coupling beams can be compared by testing these two specimens. In specimen SCB3, no edge stiffeners and slits were utilized at the web. The details of all the three webs are plotted in Fig. 2d ande Material properties The concrete cubic compressive strength f cu was obtained by testing 15 mm concrete cubes. The average concrete compressive strengths of reinforced concrete walls, precast reinforced concrete panels and concrete in concrete filled steel tube chords at the time of model testing were 56.9 MPa, 53.2 MPa and 56.9 MPa, respectively. The cylinder strength f c can be generally calculated as.8 times the cubic strength

4 17 X. Li et al. / Journal of Constructional Steel Research 14 (215) f cu. The mechanical properties of steel components were obtained from standard tests of three tensile coupons, and the resulting average yield stress f y and ultimate stress f u were given in Table Test methods The experimental setup shown in Fig. 3 was designed to simulate the double curvatures expected in a coupling beam under seismic loadings. One of the specimen's wall piers, referred to as the reaction wall,was fixed to the laboratory's strong floor by post-tensioned high strength bolts. The other wall pier, referred to as the load wall, was clamped by two very stiff I-shape beams through eight post-tensioned rods to transfer the applied loads. The loads were applied by two parallel actuators that were connected to a rigid reinforced concrete reaction wall and the applied displacements of the actuators were identical and synchronous to prevent in-plane rotation of the load wall. Two steel frames were used to avoid the potential accidental out-of-plane deformation of the whole test specimens. The loading sequence of all the specimens was controlled by the displacement at shear force application point, corresponding to multiples of the displacement δ = 9 mm. One cycle each corresponding to a peak displacement of.25δ,.5δ and.75δ was applied. Then three complete loading cycles were attempted corresponding to a peak displacement of 1.δ, 1.5δ, 2.δ and 3.δ. Lastly, two cycles corresponding to an incremental peak displacement of δ were applied till the coupling beam specimens failed or developed any major physical deterioration. The displacements of reaction and load walls and several locations along the beam were monitored using displacement transducers. Diagonal displacement transducers were also installed in coupling beams to obtain information related to the shear deformation. Electrical resistance strain gauges were affixed on steel webs and beam-to-wall connections at predefined locations. 3. Experimental results and discussion The seismic behavior of the specimens is evaluated through the observed failure modes and by using the measured responses. The damage patterns of specimens SCB1 SCB3 are presented in Figs. 4 to 6, respectively, and their corresponding shear load versus beam rotation angle hysteresis loops, along with envelope curves, are plotted in Fig. 7. According to AISC seismic provisions for structural steel buildings [15], the total beam rotation angle herein is computed by simply taking the relative displacement of one end of the beam chord with respect to the other end, and dividing by the distance between the centers of the pin connections Experimental observations Specimen SCB1 Specimen SCB1 behaved in a ductile manner with a concentration of inelastic shear deformation and damage at the steel web. The specimen was almost elastic prior to a peak beam rotation angle of.135 rad, where an abrupt but slight drop of the load carrying capacity occurred possibly due to the onset of local buckling of the web. Since the precast reinforced concrete panels provided out-of-plane bracing to the web, the capacity of the specimen recovered gradually and further increased Table 1 Properties of steel. Category Call names f y (MPa) f u (MPa) Web t = 3.8 mm Steel angle t = 6 mm Steel reinforcement HRB Steel reinforcement HRB Steel reinforcement HRB as the applied displacement continued. Sounds induced by the friction between the web and reinforced concrete panels were heard at following cycles, indicating the contribution of the bracing of precast reinforced concrete panels to the web. The initial flexural cracking of reinforced concrete panels was observed at the peak beam rotation angle of.54 rad. However, the load carrying capacity of the specimen kept increasing until to a peak beam rotation angle of.72 rad where one of the reinforced concrete panels formed an obvious X crush shape while the other one was significantly bent and fractured around the middle two holes for bolts. The crushing of the reinforced concrete panels was largely due to the formation of large buckling waves at the web, which caused extrusion forces to the reinforced concrete panels and tension forces to the bolts to balance the extrusion forces. Under the composite actions of the above-mentioned forces, the reinforced concrete panels cracked and eventually crushed. The capacity of the specimen started to drop at the peak beam rotation angle of.99 rad, where the reinforced concrete panels almost completely lost their bracing effects. At the peak beam rotation angle of.117 rad, the specimen was considered to have been failed since the web suffered significant inelastic shear deformation and even was torn off around the holes for bolts. The photo of the specimen at the final condition is shown in Fig. 4a. After the test, the reinforced concrete panels were removed and the failure mode of the edge stiffened steel web was shown in Fig. 4b. The web formed obvious X buckling shapes and the edge stiffeners were bent significantly due to the tension field action developed in the web. Generally speaking, the failures of specimen SCB1 started by the yielding of the steel web followed by the cracking and damage of reinforced concrete panels lastly by significant inelastic buckling and fracture of the steel web. During the test, the chords remained approximately in the elastic range, as indicated by strain monitoring. Only some cracking of concrete cover around the face plates of the connections occurred in the reinforced concrete walls Specimen SCB2 As described previously, the web of specimen SCB2 was divided into three flexural links by two slits and the specimen was expected to dissipate input energy through large flexural inelastic deformations of the links. However, the final failure of specimen SCB2 was dominated by the significant out of plane deformation of the flexural links after crushing of the precast reinforced concrete panels. Prior to the test, the cracks of reinforced concrete shear walls that formed during testing of specimen SCB1 were repaired using epoxy and CFRP sheets. During this test, no obvious physical observations were found until to a peak beam rotation angle of.45 rad, where some cracks in the reinforced concrete panels were observed at the locations corresponding to the positions of the slits in the steel web. The formation of these cracks was mainly due to the bending actions induced by the large out-of-plane deformation of the web's free edges. As the loads increased, cracking of the reinforced concrete panels was concentrated primarily in the vicinities where the panels provided confinement to the middle flexural link, and eventually the reinforced concrete panels there were crushed heavily to rubbles and started to spall off at the peak beam rotation angle of.81 rad. The different failure modes of the reinforced concrete panels observed during testing of specimens SCB1 and SCB2 implied that slits in webs have significant impacts on the buckling mode of the web and consequently on the loading conditions of the reinforced concrete panels. The applied loads dropped below 8% of the maximum loads at a beam rotation angle corresponding to.9 rad due to the pronounced transverse buckling of the flexural links after yielding and then the test stopped. The overall failure pattern of specimen SCB2 at the final condition is plotted in Fig. 5a. The failure mode of the steel web, as shown in Fig. 5b, was also observed after the test by removing the damaged reinforced concrete panels. The web fractured at the ends of slits owing to the severe twisting of the flexural links and the welded edge stiffeners in specimen SCB2 suffered a less serious bending deformation compared with that in specimen SCB1.

5 X. Li et al. / Journal of Constructional Steel Research 14 (215) Steel Beams Loading Beams 1-ton Actuator Post-tensioned Rods Specimen RC Reaction Wall High Strength Bolts Hydraulic Jacks Post-tensioned Rods Fig. 3. Experimental setup (unit: mm) Specimen SCB3 The failure of specimen SCB3 was caused by the large shear deformation of the replaceable steel web confined by the precast reinforced concrete panels. During the test, the out-of-plane deformation of the steel web at the free edges was observed at the first loading cycle corresponding to a peak beam rotation angle of.27 rad due to the lack of welded edge stiffeners. However, the out-of-plane deformation did not result in obvious deterioration of the load carrying capacity. The cracking of one reinforced concrete panel under bending occurred at the peak beam rotation angle corresponding to.36 rad. Meanwhile, some inclined cracks were found in the other reinforced concrete panel, signifying the formation of large buckling waves in the web under shear. As loading continued, the number and the width of the cracks in reinforced concrete panels increased, and serious crushing of the reinforced concrete panels was observed at their corners during the cycles corresponding to a beam rotation angle of.54 rad owing to the large out-of-plane buckling at the web edges. At a beam rotation angle of.72 rad, serious cracking and crushing of the reinforced concrete panels along the inclined buckling waves of the steel web underneath were observed. Therefore, the bracing effects of the panels on the web started to deteriorate, resulting in a gradual decreasing of the load carrying capacity. The specimen was considered to have failed at a beam rotation angle of.99 rad where the applied loads degraded to a level below 8% of the maximum load. The overall damage of specimen SCB3 at the final condition is shown in Fig. 6a. It can be found that the damage patterns of the reinforced concrete panels in specimens SCB1 SCB3 differed much from each other due to the different buckling modes of the steel webs with different configurations. The buckling shape of the steel web after testing is presented in Fig. 6b. It can be found that the web of specimen SCB3 developed a series of corner-tocorner buckling waves under shear loads. No fracture of the steel web was observed but the bolt holes at the corners for the web-to-chord connections were significantly deformed due to the tension field actions developed in the web Load-deformation responses Fig. 7 plots the measured shear load versus beam rotation angle hysteresis loops and the measured load versus beam rotation angle envelope curves of all three specimens. As shown in the figure, all the specimens exhibited excellent energy dissipation capacity with large and stable hysteresis loops. The specimens could continue to sustain an appreciable level of load even at near ultimate limit state where the specimens experienced a rather large applied displacement, and Fig. 4. Failure mode of specimen SCB1 at final condition: a) overall damage and b) steel web.

6 172 X. Li et al. / Journal of Constructional Steel Research 14 (215) Fig. 5. Failure mode of specimen SCB2 at final condition: a) overall damage and b) steel web. the rate of load degradation beyond the maximum load was rather slow. All those validated the inherent desirable seismic behavior of the proposed steel truss coupling beams with a buckling restrained steel web. From Fig. 7, it also can be found that the hysteretic behavior of specimen SCB2 showed a relatively more pronounced level of pinching, which was mainly attributed to the above-described significant out-of- plane buckling of the flexural links after severe damage of reinforced concrete panels. Specimens SCB1, SCB2 and SCB3 developed a peak shear force P max equal to kn, kn and kn at a corresponding beam rotation angle of.99 rad,.67 rad and.81 rad, respectively. Due to the presence of slits in the web, specimen SCB2 resisted a peak shear force corresponding to only 57.3% of the capacity of specimen SCB1, indicating that the attempt of using slits to change the failure modes from shear to flexural hinging might result in strength deterioration. Specimen SCB3 without edge stiffeners carried a peak load only equal to 51.4% of that of specimen SCB1. This is because the edge stiffeners expanded the tension fields in the web to develop the full shear capacity of the web. The contribution of the edge stiffeners to the strength was also evidenced by the significant residual bending deformation of specimen SCB1 after testing Stiffness degradation Stiffness degradation in all three specimens under cyclic loadings was evaluated by means of peak-to-peak stiffness and normalized stiffness as illustrated in Fig. 8. The peak-to-peak stiffness is defined as the ratio of the peak shear load at each load level to their corresponding displacement. To account for the variations in the configurations of the steel webs, the peak-to-peak stiffness values were further normalized with respect to the peak-to-peak stiffness at.25 rad for each specimen and the normalized stiffness is plotted in Fig. 8b. As illustrated in Fig. 8, the stiffness of all the specimens decayed gradually prior to the beam rotation angle corresponding to.135 rad, and then got slight recovery due to the contribution of edge stiffeners and reinforced concrete panels, and lastly decreased gradually owing to the significant inelastic deformation of the steel webs and the damage of reinforced concrete panels. The rate of stiffness degradation of specimens SCB1 and SCB3 after.3 rad was similar and essentially linear up to failure. Fig. 8b also indicated that the recovery of the stiffness in specimen SCB2 during the beam rotation angle from.135 rad to.5 rad was relatively more obvious. This may be attributed to that the links without edge stiffeners are more prone to suffer out-of-plane deformation and thus the closing of the gap between the reinforced concrete panels and the steel web is relatively earlier. The participation of the reinforced concrete panels can inhibit the further buckling of the web and bring in extra stiffness in compensation for the stiffness loss of the steel web due to buckling. When the peak-to-peak stiffness values of all the three specimens at the beam rotation angle corresponding to.1 rad are compared, it can be found that the stiffness of specimen SCB1 is about 18% larger than that of specimen SCB3 owing to the beneficial contribution of welded edge stiffeners, whereas specimen SCB2 got the stiffness only approximately half of that of specimen SCB1 due to the adverse effects of slits and previous slight cracking of beam-to-wall connections. Note that the stiffness of the specimens is affected by the thickness and configurations of the steel web as well as the types of beam-to-wall connections. The test results also indicated that the coupling beams tested, compared with other types of steel coupling beams with rigid beamto-wall connections investigated by Harries [6] and Park [1], exhibited a relatively lower elastic stiffness due to their thin steel web and pin connections between beam chords and shear walls. Fortunately, according to the study conducted by Weldon [14], the stiffness of the proposed coupling beam can be significantly improved by adoption of unbonded post-tensioned tendons and its effectiveness is being proved by the Fig. 6. Failure mode of specimen SCB3 at final condition: a) overall damage and b) steel web.

7 X. Li et al. / Journal of Constructional Steel Research 14 (215) Shear Load(kN) Shear Load(kN) 5 Specimen SCB Specimen SCB Shear Load(kN) Shear Load(kN) Specimen SCB SCB1 SCB2 SCB3 Fig. 7. Hysteresis loops and envelope curves of specimens. finite element analysis done by the authors. Therefore, the stiffness of the coupling beams can be reasonably adjusted by changing the above parameters if necessary Deformation capacity and ductility The main experimental results and the key parameters characterizing the behavior of test specimens are presented in Table 2. The yield and ultimate beam rotation angles in the table are the averages of the corresponding values at the positive and negative loading directions. The ductility factor μ is defined as the ratio of the beam rotation angle at ultimate stage θ u to that at the yield stage θ y,wheretheformerisobtained at 85% of the maximum capacity after reaching the peak load, while the latter is obtained based on the concept of equivalent energy method such that the area enclosed by the idealized elastic plastic envelope curve is equal to that enclosed by the actual envelope curve. As described in the AISC seismic provisions for structural steel buildings [15], coupling beam response in coupled wall systems is intended to be similar to shear link response in eccentrically braced steel frames and a total beam rotation angle larger than.8 rad is conservatively required to assure satisfactory behavior of the systems. As illustrated in Table 2, all the three specimens achieved an ultimate beam rotation angle larger than.8 rad, satisfying the requirement recommended by AISC (21) [15] for hybrid coupled wall systems. Moreover, specimens SCB1 and SCB3 sustained an inelastic beam rotation angle greater than.4 rad. As for the ductility, though all the specimens suffered a ductile failure and achieved a large inelastic deformation, the ductility factor of the three specimens, which ranged from 1.73 to 2.83, was not Stiffness(kN/mm) SCB1 SCB2 SCB3 Normalized Stiffness SCB1 SCB2 SCB a) Peak-to-peak stiffness versus beam rotation b) Normalized stiffness versus beam angle response rotation angle response Fig. 8. Degradation of the stiffness of specimens SCB1 SCB3.

8 174 X. Li et al. / Journal of Constructional Steel Research 14 (215) Table 2 Summary of test results. Specimens P y (kn) P max (kn) θ y (1 2 rad) θ u (1 2 rad) μ = Δ u /Δ y V n V n /P y SCB SCB SCB very high compared with the test results of Lam [9] and Park [1]. This can be attributed to two reasons: firstly, the yield beam rotation angles of the three specimens were relatively large owing to the low stiffness of the specimens; secondly, the width of the tension fields developed in the webs increased as the applied displacements increased, as a result the yield strength of the specimens enhanced. Due to the presence of edge stiffeners, specimen SCB1 experienced a yield beam rotation angle of.495 rad and an ultimate bean rotation angle of.117 rad, which were larger than those of specimen SCB3. However, the ductility factors of the two specimens were comparable since the edge stiffeners increased both the yield and ultimate beam rotation angles of specimen SCB1. The significant out-of-plane deformation of the links in the steel web deteriorated the ductility of specimen SCB2 and resulted in a ductility factor equal to 1.73, which was only 73.3% of the corresponding value of specimen SCB1. Therefore, stiffer reinforced concrete panels are required to avoid the significant twisting of the flexural links with free edges. inferior to that of specimen SCB1 since the buckling of the flexural links hindered the full formation of plastic hinges at the links. The energy dissipation index of specimen SCB3 was comparable to that of specimen SCB1; however, the cumulative dissipated energy of specimen SCB3 at the beam rotation angle of.99 rad was only 33.2% of the corresponding value of specimen SCB1 due to the relatively lower strength. 4. Analytical model for load-carrying capacity prediction To properly design the proposed steel truss coupling beams, it is necessary to have a simplified and reliable analytical model to predict the load carrying capacity of the beams. Since the steel web serves as a replaceable fuse of the coupling beam, the load carrying capacity of the beam is determined by that of the steel web. For the specimen SCB1, due to the combined actions of edge stiffeners and reinforced concrete panels, the steel web developed a full shear yielding capacity and therefore the nominal capacity V n of specimen SCB1 can be calculated as 3.5. Energy dissipation capacity V n ¼ :6f y ht w ð1þ The energy dissipation capacity of the specimens under cyclic loadings was evaluated by two indices: energy dissipation index E h and cumulative dissipated energy. The energy dissipation index E h is defined as the ratio of total energy dissipation (area S ABCDE )atthefirst cycle of each load level to the elastic potential energy (areas S OBF + S ODG ), as shown in Fig. 9a. The total energy dissipation is the area enclosed by load deformation hysteresis loops. A larger energy dissipation index reflects a greater ability to dissipate energy. As seen from Fig. 9, the energy dissipation capacity of all the specimens dropped slightly when the steel webs buckled. After that, the energy dissipation capacity essentially linearly increased with the applied beam rotation angles. Specimen SCB1 exhibited an excellent energy dissipation capacity, and the energy dissipation index and the cumulative dissipated energy value at the beam rotation angle of.117 rad were equal to 1.66 and kj, respectively. The energy dissipation capacity of specimen SCB2 was relatively where f y is the yield strength of the steel web; t w is the thickness of the steel web; and h is the height of the steel web, which is calculated as the clear distance of chords for specimen SCB1. For specimen SCB3 without edge stiffeners, some areas of the steel web did not develop their full yielding capacity as indicated by the experimental observations and therefore a strip model as illustrated in Fig. 1 was proposed to calculate the load carrying capacity herein. Nowadays, various strip models have been proposed by [16 18] to simulate the behavior of thin steel plates by a series of pin-ended tension strips inclined at angle α. However, few strip models for two-side connected buckling restrained steel plates can be found to date. In the proposed model, RC walls and chords of coupling beams are assumed to be rigid elements with pin connections. Since the free edges cannot provide sufficient anchor to the strips, only strips connected to chords are considered. Owing to the presence of reinforced concrete panels, the E h P SCB1.4 SCB2.2 SCB Cumulative dissipated energy(kn.m) SCB1 SCB2 SCB a) Energy dissipation index b) Cumulative dissipated energy Fig. 9. Energy dissipation capacity of specimens.

9 X. Li et al. / Journal of Constructional Steel Research 14 (215) Summary and conclusions A new type of replaceable steel truss coupling beam with a buckling restrained web was developed to improve the energy absorption capacity of conventional coupling beams. The seismic behavior of the proposed coupling beams was evaluated by testing of three large-scale coupling beam specimens under cyclic loadings and the effects of some configurations of the steel webs including welded edge stiffeners and slits on the seismic behavior were investigated. The following conclusions can be drawn from the studies: Fig. 1. Strip model for proposed coupling beams with buckling restrained webs. strips can carry both tension and compression forces in this model. The inclination angle α of the strips to the vertical free edges can be determined using the principle of energy conservation. For a given shear force V, the axial stress σ s of the strips can be calculated as Vlþ ð 2aÞ σ s ¼ t w h e ðl h e tan αþ sin 2α where l is the effective length of the web, it takes as the distance between the outer edges of the bolts at the ends of a fish plate; a is the distance between the hinges of chords and the edge bolts; and h e is the effective height of the web, which is defined as centerline-tocenterline distance between the bolts in fish plates welded to the top and bottom chords. α is the degree of inclination angle. The strain energy of strips can be accordingly computed as W s ¼ σ 2 s 2E t V 2 ðl þ 2aÞ 2 wda ¼ Et w h e ðl h e tanαþsin 2 2α : ð3þ According to the principle of least work ( W s / α =)ineq.(3),the relationship between the inclination angle of the strip α and the aspect ratio of the steel web (l/h e ) can be expressed as tan 3 α 2l=h e tan 2 α 3 tanα þ 2l=h e ¼ : It can be found that Eq. (4) is still cumbersome and the exact inclination angle α cannot be easily obtained. Therefore, a simplified but highquality fitting curve determined by Eq. (5) is proposed to calculate the inclination angle α herein. α ¼ 4:57ðl=h e Þ 2 þ 22:94l=h e þ 9:23: ð5þ Assuming that all the steel strips reach their yielding strength f y at the final condition, the nominal load carrying capacity of the coupling beam with a buckling restrained web V n can be obtained from Eq. (2). V n ¼ f y t w h e ðl h e tan αþ sin 2α= ðl þ 2aÞ ð6þ The analytical results V n coupled with the comparison between the analytical and experimental results are presented in Table 2. Theratios of the predicted results of specimens SCB1 and SCB3 to the corresponding test results were.93 and.94, respectively, indicating that the predictions by the analytical model agreed well with the experimental results. Note that the applicability of the proposed strip model should be further verified by more test data. ð2þ ð4þ (i) The proposed replaceable steel truss coupling beam with a buckling restrained web exhibited good energy-dissipation capacity. Hence, it is a viable alternative for the coupling beams for coupled wall systems in regions of medium and high seismicity. (ii) The edge stiffeners of a steel web can enhance the strength, stiffness and energy-dissipation capacity of the proposed coupling beam. Owing to the beneficial effects of edge stiffeners and reinforced concrete panels, a thin steel web without slits can develop full yield capacity under shear and thus the load carrying capacity of the coupling beams similar to specimen SCB1 can be calculated based on the full shear yielding strength of the web. (iii) The slits in a steel web have adverse effects on the strength and stiffness of the proposed coupling beams, which are prone to result in an out-of-plane deformation of the free edges around the slits. To prevent premature buckling of the web and improve the ductility and energy-dissipation capacity of the beams, higher strength and rigidity of the reinforced concrete panels are required. (iv) An analytical model was developed to predict the load carrying capacity of the proposed coupling beams without slits in the web, and the analytical model was shown to predict the capacity well. Acknowledgments The authors would like to express their sincere appreciation for the partial financial support from the National Natural Science Foundation of China (grant numbers: 5183 and ), the Jiangsu Provincial Science and Technology Department (grant number: BK211221), and the Fundamental Research Funds for the Central Universities (grant number: 212QNA56). The experimental work described in this paper was conducted at the Jiangsu Key Laboratory of Environmental Impact and Structural Safety in Civil Engineering in the China University of Mining and Technology, which is supported by the Priority Academic Program Development of Jiangsu Higher Education Institutions (PAPD) (Project number: Suzhengbanfa(214)-No. 37). Help during the testing from staffs and students at the Laboratory is greatly acknowledged. References [1] Hassan M, EI-Tawil S. Inelastic dynamic behavior of hybrid coupled walls. J Struct Eng 24;13(2): [2] Wang YY. Lessons learnt from building damages in Wenchuan earthquake: seismic concept design of buildings. J Build Struct 28;29(4):2 5 (In Chinese). [3] Brena SF, Ihtiyar O. Performance of conventionally reinforced coupling beams subjected to cyclic loading. J Struct Eng 211;137(6): [4] Park R, Paulay T. Reinforced concrete structures. New York: John Wiley & Sons, Inc.; [5] Galano L, Vignoli A. Seismic behavior of short coupling beams with different reinforcement layouts. ACI Struct J 2;97(6): [6] Harries KA, Mitchell D, Cook WD. Seismic response of steel beams coupling concrete walls. J Struct Eng 1993;119(12): [7] Gong BN, Shahrooz BM. Concrete steel composite coupling beams. I: component testing. J Struct Eng 21;127(6): [8] Gong BN, Shahrooz BM. Concrete steel composite coupling beams. II: subassembly testing and design verification. J Struct Eng 21;127(6): [9] Lam WY, Su RKL, Pam HJ. Experimental study on embedded steel plate composite coupling beams. J Struct Eng 25;131(8):

10 176 X. Li et al. / Journal of Constructional Steel Research 14 (215) [1] Park WS, Yun HD. Seismic behaviour of steel coupling beams linking reinforced concrete shear walls. J Constr Steel Res 25;27(7): [11] Park WS, Yun HD. Panel shear strength of steel coupling beam wall connections in a hybrid wall system. J Constr Steel Res 26;62(1): [12] Fortney PJ, Shahrooz BM, Rassati GA. Large-scale testing of a replaceable fuse steel coupling beam. J Struct Eng 27;133(12): [13] GB Code for seismic design of buildings. Beijing (China): China Architecture and Industry Press; 22 (in Chinese). [14] Weldon BD, Kurama YC. Experimental evaluation of posttensioned precast concrete coupling beams. J Struct Eng 21;136(9): [15] AISC. Seismic provisions for structural steel buildings. Chicago: American Institute of Steel Construction; 21. [16] Thorbum LJ, Kulak GL, Montgomery CJ. Analysis of steel plate shear walls. Structural engineering report no. 17. Edmonton: Department of Civil Engineering, University of Alberta; [17] Shishkin JJ. Analysis of steel plate shear walls using the modified strip model. [Master thesis] Edmonton: Department of Civil and Environmental Engineering, University of Alberta; 26. [18] Choi IR, Park HG. Steel plate shear walls with various infill plate designs. J Struct Eng 29;135(7):

11 本文献由 学霸图书馆 - 文献云下载 收集自网络, 仅供学习交流使用 学霸图书馆 ( 是一个 整合众多图书馆数据库资源, 提供一站式文献检索和下载服务 的 24 小时在线不限 IP 图书馆 图书馆致力于便利 促进学习与科研, 提供最强文献下载服务 图书馆导航 : 图书馆首页文献云下载图书馆入口外文数据库大全疑难文献辅助工具