ACI

STRUCTURAL J O U R N A L

A JOURNAL OF THE AMERICAN CONCRETE INSTITUTE

Prepublished Paper
This is a prepublished manuscript. The
final manuscript is tentatively scheduled
for V. 120, No. 4 and is subject to change.

The DOI for this paper is 10.14359/51738743

and will not change, but won’t be activated
until the issue has been published.
 1 Reinforced Concrete Coupling Beams with Different Layouts under Seismic and Wind Load

3 Tse-An Chou, Seung Heon Lee, Chunho Chang and Thomas H.-K. Kang

5 ACI Member Tse-An Chou is a PhD Student in the Department of Architecture and Architectural

6 Engineering at Seoul National University, Seoul, Korea. He received his BS and MS in civil and

7 construction engineering from the National Taiwan University of Science and Technology, Taipei,

8 Taiwan. His research interests include seismic behavior and the design of reinforced concrete

9 structures.

10 ACI Member Seung Heon Lee is a PhD Student in the Department of Architecture and

11 Architectural Engineering at Seoul Nation University, Seoul, Korea. He received his BS and MS in

12 the Department of Architecture and Architectural Engineering at the Seoul National University,

13 Seoul, Korea. His research interests include structural analysis and design of reinforced concrete

14 structures under combined loading conditions.

15 Chunho Chang is a Professor of civil engineering and Director of the Intelligent Construction

16 System Core Center at Keimyung University, Daegu, Korea. He received his BS, MS and PhD from

17 Keimyung University. His research interests include the design and behavior of textile reinforced

18 concrete and composite structures under earthquake loading.

19 Thomas H.-K. Kang, FACI, is a Professor of structural engineering and interdisciplinary artificial

20 intelligence at Seoul National University. He is a member of Joint ACI-PTI Committee 320, Post-

21 Tensioned Concrete Building Code; Joint ACI-ASCE Committees 352, Joints and Connections in

22 Monolithic Concrete Structures, and 423, Prestressed Concrete; Joint ACI-ASME Committee 359,

23 Concrete Containments for Nuclear Reactors; and ACI Subcommittee 318-T, Post-Tensioned

24 Concrete (Structural Concrete Building Code). His research interests include the design and

25 behavior of reinforced, prestressed, and post-tensioned concrete structures.

26
1
27 ABSTRACT

28 Reinforced concrete (RC) coupling beams can act as an efficient energy dissipating fuse and

29 force transfer element between RC shear walls in low to high-rise buildings. To investigate the

30 effect of different reinforcement layouts, amounts of confinement, and loading protocols on RC

31 coupling beams, eight RC coupling beams with a span-depth ratio of 2.5 were tested with three

32 parameters of: 1) longitudinal or diagonal reinforcement layout; 2) full, 2/3, or 1/2 amount of

33 confinement relative to ACI 318-19 requirements; and 3) seismic or wind loading protocols. The

34 test results showed that: first, the nominal shear and upper limit equations for diagonally RC

35 coupling beams in ACI 318-19 may need to be improved, and it is also recommended to consider

36 the contribution of confinement to shear strength; second, since only minor cracks were observed

37 under the wind with no significant damage, the experiment in this study can act as an example of

38 structural verification for performance-based wind design.

39

40 Keywords: reinforced concrete; coupling beam; confinement; longitudinal reinforcement; diagonal

41 reinforcement; seismic loading; wind loading.

42

43 INTRODUCTION

44 With a rapidly growing population, taller and taller buildings are built in this era. Coupling

45 beams are often utilized in high-rise buildings due to their advantages of opening windows or

46 doorways on core walls, and acting as an efficient energy dissipating system to resist lateral loads,

47 and they are typically designed with a span–depth ratio (ln/h) of 2.4 for residential, and 3.3 for

48 office usage in high-rise buildings (Naish et al. 2013).

49 In the ACI 318-19 design procedure, intermediate reinforced concrete (RC) coupling beams

50 with a span–depth ratio of (2−4) do not have specific regulations of reinforcement layout, whether

51 longitudinally or diagonally. Thus, the nominal shear strength (Vn,beam) for a longitudinally RC

2
52 coupling beam is estimated based on the nominal one-way shear strength as a normal beam by Eq.

53 (1):

Vn ,beam  Vc  Vs (1)
54 where, Vc is the nominal shear strength provided by concrete, and Vs is the nominal shear strength

55 provided by confinements. The equation of nominal shear strength for a diagonally RC coupling

56 beam can be calculated using Eq. (2):

Vn ,beam  2 Avd f y sin a (2)

57 where, Avd is the total area of diagonal rebars in each group in a diagonally RC coupling beam, and

58 α is the angle between the diagonal bars and the longitudinal axis of a coupling beam. However, in

59 ACI 318-19, the nominal shear strength (Vn) shall not be taken greater than the upper limit (Vn,upper),

60 as Eq. (3):

Vn ,upper  0.83 f c' (MPa) Acw (3)

61 where, Acw is the area of concrete section of a coupling beam resisting shear.

62 Diagonally RC coupling beams can provide better shear strength, deformation capacity, and

63 energy dissipation behavior, compared to longitudinally RC coupling beams, because diagonal

64 rebars simultaneously function as flexural and shear reinforcements. Their contribution to shear

65 strength typically results in well-rounded hysteretic loops, without pinching effect. However, many

66 previous outcomes (Naish et al. 2013; Lim et al. 2016a and 2016b; Cheng et al. 2019) show that the

67 actual shear strength of a diagonally RC coupling beam is notably larger than the nominal shear

68 strength and upper limit in ACI 318-19. If coupling beams do not sufficiently develop plastic hinges

69 due to over-conservative design, this could cause undesirable forces (such as flexural force), or

70 damage to adjacent structural elements. Based on this mechanism, it may be more acceptable to use

71 the nominal flexural strength (Mn) of a coupling beam to estimate shear strength, as shown in Eq.

72 (4). Park et al. (2020) also proposed an estimation model for diagonally RC coupling beams with

73 (ln/h) of (1 to 3), which additionally considered the contribution of longitudinal rebars and concrete

74 for diagonally RC coupling beams.

3
 VMn  2 M n ln (4)
75 Moreover, the amount of confinement had a significant influence on the cyclic behavior and

76 failure modes of diagonally RC coupling beams (Han et al. 2019), and this effect has also not been

77 incorporated into the equations of nominal shear strength and the upper limit in ACI 318-19. In

78 general, these two aforementioned equations for diagonally RC coupling beams should be updated,

79 preferably also considering the effects of confinement, because excessive conservatism in coupling

80 beams may be detrimental to achieving the desired behavior for performance-based design and

81 evaluation.

82 For high-rise buildings, wind loads are as critical as seismic loads. In particular, the design of

83 certain structural elements, including coupling beams, is generally controlled by wind demands

84 (Aswegan et al. 2017). In terms of seismic design, structural elements are allowed to reach inelastic

85 behavior, and numerous previous tests (Paulay and Binney 1974; Barney et al. 1980; Tassios et al.

86 1996; Xiao et al. 1999; Galano and Vignoli 2000; Kwan and Zhao 2001, Naish et al. 2013; Chen et

87 al. 2019; Park et al. 2020) regarding seismic loading have been performed.

88 For wind design, compared to the former provision of ASCE 7-16, which used equivalent

89 static wind load to keep buildings in the elastic phase, inelastic behavior under (700 to 3,000) year

90 mean recurrence interval (MRI) wind loads in ASCE/SEI Prestandard for Performance-Based Wind

91 Design (2019) is now permitted in the latest ASCE 7-22 provision. Among the components of wind

92 load, including along-wind, across-wind, and torsional wind, the response of across-wind is

93 generally larger than along-wind for a taller building, due to its large resonant response (Alinejad et

94 al. 2020a). According to analysis results from Jeong et al. (2021), reducing the design wind force

95 (resonant component only) by using a response modification factor (RWR) of two and three to

96 introduce inelastic behavior could significantly decrease the design demand of a coupling beam,

97 while also increasing the ductility of the system.

98 In terms of structural performance levels for wind hazard scenarios, the damage control (DC)

99 performance level for performance-based wind design is suggested by Alinejad et al. (2020), which
4
100 is defined as the midpoint between immediate occupancy (IO) and life safety (LS) performance

101 objectives in ASCE 41-17. To satisfy the DC performance level and check the safety margin against

102 low-cycle fatigue and ratcheting failures under extreme wind events, the performance of coupling

103 beams is likely to be confirmed through testing under moderate inelastic deformations with an

104 appropriate number of cycles for wind load. Therefore, Abdullah et al. (2020) established a wind

105 loading protocol to represent the inelastic response of a tall building by determining the number and

106 amplitude of the cycles from loading histories corresponding to (1,700 to 3,000) year MRI wind

107 loads. The test results show that the specimens satisfied the IO performance level with relatively

108 negligible damage observed, but testing various specimens with alternative wind protocols was

109 recommended, due to a lack of experimental data.

110 To summarize all the needs stated above, a total of eight specimens with a span–depth ratio

111 (ln/h) of 2.5 were tested, and the purposes of this study were to: 1) investigate the behaviors of both

112 longitudinally and diagonally RC coupling beams with three different amounts of confinement (full,

113 2/3, and 1/2) relative to ACI 318-19 requirements; 2) develop a wind loading protocol based on the

114 peak factor of across-wind in KBC 2016, and represent the behavior of RC coupling beams under

115 the wind loading protocol composed of a large number of linear cycles and limited non-linear

116 cycles.

117

118 RESEARCH SIGNIFICANCE

119 This study aims to provide useful experimental data for the development of future building

120 codes, and investigate the feasibility of extending performance-based design to wind engineering.

121 Data from eight large-scale tests of RC coupling beams with a span–depth ratio of 2.5 tested under

122 seismic and wind loading protocols are represented. The key parameters were reinforcement layout,

123 (longitudinal or diagonal reinforcement), the amount of confinement, and loading protocol. The

124 findings show that the different layouts and amounts of confinement did have a meaningful

5
125 influence on the RC coupling beams, and specimens tested under a simulated wind event only had

126 minor cracks, with no extensive damage observed.

127

128 EXPERIMENTAL PROGRAM

129 Among a total of eight 2/3-scale RC coupling beams, there are four series of layouts, and each

130 series was tested under seismic and wind loading protocol. Except for the test parameters, all other

131 conditions were designed the same, with a focus on investigating modeling parameters for the RC

132 coupled wall systems used in regions of moderate-to-high seismicity and/or high wind speed. The

133 following sections describe the design details, material properties, test setup, instrumentation, and

134 loading protocols in this study.

135

136 Test specimens

137 Eight test specimens were designed based on common coupling beams in residential buildings

138 with a span–depth ratio (clear length/depth, ln/h) of approximately 2.5. Due to the laboratory space

139 and strength constraints, coupling beams were scaled down to 2/3-scale of the prototype coupling

140 beams. Thus, the cross-sectional dimensions (width × depth × span, bw × h × ln) were 300 mm × 500

141 mm × 1,250 mm (11.8 in × 19.7 in × 49.2 in). All details were designed according to the design

142 procedure specified in ACI 318-19, except for the reduction of confinements for study purpose. The

143 specified concrete strength (f’c) was 30 MPa (4.4 ksi), and yield strength (fy) was 400 MPa (58.0 ksi)

144 for all rebars. The longitudinal reinforcement ratio () of all specimens was approximately 1.9 %.

145 Detailed information is provided in subsequent paragraphs, and Fig. 1 illustrates the specimen

146 layouts.

147 For longitudinally RC coupling beams, specimen L100 series and L67 series were designed

148 using a traditional reinforcement layout with three D32 (No. 10) headed rebars as tensile and

149 compressive reinforcements with a sufficient development length of 500 mm (19.75 in), and being

150 confined by full and two-thirds amount of D13 rebars (No. 4) with a spacing of 105 mm (4.1 in) and
6
151 165 mm (6.5 in) respectively. With regard to the diagonally RC coupling beams, four D25 (No. 8)

152 headed rebars for each group were placed diagonally with an angle () of approximately 15 degrees

153 and an anchorage length of 500 mm (19.75 in) for specimen D67 series and D50 series. Since

154 diagonal coupling beams have notably higher shear strength than longitudinal coupling beams,

155 instead of designating the full amount of confinement, two-thirds amount with a spacing of 165 mm

156 (6.5 in) for D67 series and half amount with a spacing of 230 mm (9.1 in) for D50 series were

157 arranged. As shown in Fig. 1, four D10 (No. 3) skin rebars were placed with only 100 mm

158 embedment in the middle of each beam. For D67 and D50 series, three top and bottom D10 (No. 3)

159 longitudinal bars were provided with a development length of 350 mm (13.75 in), which is longer

160 than required.

161

162 Material properties

163 A normal-weight concrete with a design 28-day concrete compressive strength (f’c) of 30 MPa

164 (4.4 ksi) was specified for all specimens. The maximum aggregate size of 25 mm (1 in), and a

165 slump of 150 mm (5.9 in), were requested. Concrete strength was determined based on the average

166 of three standard 100 mm × 200 mm (4 in × 8 in) cylinders for each series. All cylinders were cast

167 on the same day along with casting specimens from each concrete truck at a local concrete plant.

168 Korean Standard (KS) SD400 deformed bars with a nominal yield strength (fy) of 400 MPa (58 ksi)

169 were specified for steel rebars. The average yield strength and ultimate strength (fu) for each size

170 were determined by submitting three 500 mm (19.7 in) long specimens to the direct tensile test.

171 Table 1 shows the test results of the concrete and steel rebars.

172

173 Test setup and instrumentation

174 When a building oscillates, a coupling beam is subject to a slight axial compressive

175 deformation, as well as prominent lateral deformation. To replicate this mechanism, the test setup

176 was arranged as shown in Fig. 2(a), where coupling beams were set in a vertical direction, and
7
177 embedded in two adjacent stiff RC blocks, which were taken as structural wall elements. The

178 bottom block was enlarged to avoid overturning, and bolted to the laboratory strong floor, whereas

179 the top block was bolted to the upper steel frame, which was connected to a 100 tonf (220 kips)

180 hydraulic actuator and two steel links including two vertical steel frames and four pin connections.

181 However, if a coupling beam is tested with a consistent height, it will be subjected to axial

182 extension at large drift demands, and the resulting axial force may significantly impact the coupling

183 beam’s performance (Lequesne et al. 2013). Thus, two steel links were set to descend slightly while

184 keeping the upper steel frame horizontal, to restrain any end rotation and axial elongation that might

185 occur while the hydraulic actuator applied lateral displacement to the upper steel frame (Fig. 2b).

186 Two lower steel frames were set for fixing steel links to the laboratory’s strong floor, and two

187 gusseted angle brackets were inserted between the top block and upper steel frame to prevent

188 sliding.

189 The external deformation of each specimen was measured by 12 linear variable differential

190 transformers (LVDTs), and four string potentiometers were used to measure global displacements,

191 as shown in Fig. 3. A total of 30 strain gauges were installed to measure strains in longitudinal or

192 diagonal rebars, confinements, and longitudinal skin rebars. Crack widths were manually measured

193 at the end of each loading stage.

194

195 Loading protocols

196 In this study, each series was tested under both seismic and wind loading protocols. To date,

197 many guidelines have applied performance-based design to earthquake engineering, such as

198 PEER/ATC 72-1 (2010) and TBI 2017 (2017). ACI 374.2R-13 recommended a cyclic loading

199 protocol to perform seismic behavior for structural component tests, which was also applied to this

200 study with two displacement-controlled cycles at each stage, as shown in Fig. 4. In terms of wind

201 engineering, the application of performance-based wind design in ASCE 7-22 is still in its infancy,

202 and a wind loading protocol for testing structural components has not been established so far.
8
203 Therefore, considering that across-wind is the key factor of wind load for certain structural elements

204 in tall buildings, a displacement-controlled wind loading protocol was developed with a zero-mean

205 process. Buildings of (35−70) story of (150−300) m or (500−1,000) ft height, the general range of

206 the 300 tallest buildings in Korea) with fundamental periods between (3 and 6) s were considered,

207 and the following steps describe the procedure in detail:

208 Amplitude of cyclesThe expected maximum ductility demand for coupling beams in an

209 extreme wind event was assumed as 1.5 times the yield ratio (y) in this study, where the yield ratio

210 (θy) is the specified yield drift ratio. The ductility factor of 1.5 was considered to be an adequate

211 number, which is the prescribed ductility factor for deformation-controlled elements in the

212 evaluation process in ASCE (2019), and whose value is much less than the inherent ductility factor

213 of approximately (2.5 − 5) from the seismic response of RC coupling beams. The amplitude set for

214 each stage was increased from (0.25 to 1.5)y in increments of 0.25y, and then symmetrically

215 stepped down to 0.25y. The yield ratio was taken from the result of the seismic test, which was

216 performed prior to the wind test for the same series specimen.

217 Corresponding design forceThe maximum design force (Fmax) was set as gL, where gL (Eq.

218 [5]) is the peak factor in the across-wind direction in KBC 2016,  is the standard deviation of

219 equivalent static wind load, and nL is the natural frequency of the first mode in the across-wind

220 direction. The corresponding design force (F) in the elastic response is equal to the maximum

221 design force × force ratio (F/Fmax) derived from the equal energy principle, which resulted in the

222 force reduction factor (R = Fmax/Fy) of 1.41, where, Fmax is the maximum design force for an elastic

223 system, and Fy (= 0.71Fmax) is the specified yield strength, as shown in Fig. 5 and Table 2.

g L  2 ln  600 nL   1.2 (5)

224 Number of cyclesThe total amount of cycles was determined by considering tall buildings

225 oscillating under a one hour wind event, which is frequently taken in ASCE 7-16. Consequently, the

226 number of cycles at each stage could be estimated from the cumulative probability density of the
9
227 corresponding design force based on Gaussian distribution (Fig. 6). Table 2 shows the step-by-step

228 calculation result, and Fig.7 shows the result of the wind loading protocol that was developed.

229 Based on the fundamental period, it resulted in a total of 915 cycles in the wind loading protocol,

230 which was composed of 900 linear cycles and 15 non-linear cycles (1.25y and 1.5y). The expected

231 testing time for simulated wind events was about 10 h, whereas seismic events needed only 4 h.

232

233 TEST RESULTS

234 Cracking progression and hysteretic behavior

235 Figure 8 presents the crack pattern and maximum crack width (CW) at different drift ratios

236 (DR) or chord rotations () for all test specimens, where the DR and  were defined as the lateral

237 deflection of a specimen measured from the LVDT divided by the beam clear span, and adjusted by

238 extra rotations from the top and bottom blocks. Figure 9 presents the lateral load versus chord

239 rotation curves for all specimens, where the ductility demand (test/y,test) is also shown in the upper

240 axis, defined as the rotation demand divided by yield rotation (y,test). The yield rotation (y,test) was

241 obtained when the first main rebar reached yield strain from the test result of the seismic event.

242 Initial cracks were observed in the first stage for all specimens except specimen L100 Series (L100-

243 S and L100-W), which were observed in Stage 2 and Stage 3, respectively, due to smaller lateral

244 displacement caused by unexpected out-of-plane displacement. Horizontal cracks firstly developed

245 at the beam for all specimens, and inclined cracks developed subsequently with a maximum crack

246 width of 0.05 mm (0.002 in), as shown in the first figures of Figs. 8(a)–(h). All specimens showed

247 similar inclined crack patterns, which mainly appeared on the lower part of beams with an angle of

248 approximately 45 degrees in the final state (FS), as shown in the middle figures of Fig. 8. However,

249 compared to diagonal coupling beams, only longitudinal coupling beams developed vertical cracks

250 along the line of longitudinal rebars after the inclined cracks occurred. The follow-up progressions

251 of cracking and hysteretic behavior for each specimen are elaborated below.

10
252 L100 Series In specimen L100-S, a crack with 2.0 mm (0.079 in) width occurred at the

253 right side bottom at a drift ratio of 2.9 %, and concrete deteriorated in the lower right corner at the

254 reverse side of the beam at a drift ratio of 3.5 % as its peak shear strength (Vtest) reached 684.0 kN

255 (153.8 kips), where the test was stopped, due to unexpected out-of-plane displacement occurring.

256 Because the unexpected out-of-plane displacement limited yield rotation (y,test) to a small value,

257 specimen L100-W only had maximum crack width of 0.1 mm (0.004 in) at Stage 4 (1.0y target

258 drift), and obvious cracks mainly developed until Stage 7 (1.25y target drift). After Stage 7, the

259 specimen only had some extended slight cracks. The problem of unexpected out-of-plane

260 displacement was solved after finishing testing the L100 Series.

261 L67 Series A maximum 3.5 mm (0.14 in)-width vertical crack appeared on specimen L67-S

262 at the first cycle of Stage 7 (3.0 % target drift ratio). The peak shear strength (Vtest) of 632.1 kN

263 (142.1 kips) was reached at a 2.0 % drift ratio, and dropped drastically to approximately 60 % for

264 the sequential cycle. Concrete significantly spalled out along the line of longitudinal rebars at the

265 final stage, with roughly 30 % of peak shear load and an ultimate drift ratio u of 3.94 %. For

266 specimen L67-W, a vertical 1.6 mm (0.063 in)-width crack developed along with longitudinal

267 rebars at Stage 6 (1.5y target drift) with a maximum drift ratio of 2.14 %, and the crack width got

268 wider to 2.0 mm (0.079 in) at Stage 7. After that, while the drift ratio decreased progressively, only

269 some slight extended cracks could be observed.

270 D67 Series The specimen D67-S had a 0.65 mm (0.026 in)-width crack at a drift ratio of

271 2.97 % in the lower right corner at the left side of the beam. The crack continuously widened up to

272 6.0 mm (0.24 in) as the peak shear strength Vtest of 809.0 kN (181.9 kips) was reached at a drift ratio

273 of 6.43 %, and then its lateral strength declined by about 10 % of the peak lateral strength, while

274 concrete spalled off of the four corners. This crushing pattern of concrete was most likely caused by

275 flexural compression. The maximum crack width of specimen D67-W developed from (0.15 to 0.45)

276 mm ((0.006 to 0.018) in) from Stage 4 to 6. Except for small sections in lower corners chipping off

11
277 at Stages 8 and 9 (1.0y and 0.75y target drift) due to flexural compression, the coupling beam did

278 not have any wider cracks.

279 D50 Series The crack width of D50-S developed to 5 mm (0.197 in) at a drift ratio of 3.8 %

280 (Stage 8), and then the coupling beam reached a maximum lateral strength Vtest of −761.4 kN (171.2

281 kips) at Stage 9. However, after concrete in three corners of the beam spalled off at Stage 8

282 (maximum drift ratio 6.3 %), the strength was reduced to around 90 % of peak lateral strength at the

283 second cycle. Finally, owing to the presence of rebars, the test was stopped when the shear strength

284 had dropped to about 50 % of peak lateral strength with an ultimate drift ratio u of 8.69 %. For

285 specimen D50-W, the maximum crack width increased from (0.2 to 0.35) mm between Stage 4 and

286 6. A small part of facial concrete in the lower-right corner of the beam fell off at Stage 8 due to

287 flexural compression; however, no new crack or large extended crack was observed.

288 As can be seen from the test results in Fig. 9, similar to previous findings (Barney et al. 1980;

289 Lim et al. 2016), this study reconfirms that diagonally RC coupling beams do indeed provide higher

290 peak shear strengths (Vtest), and better capacities of deformation and energy dissipation than

291 longitudinally RC coupling beams, due to the well-rounded curves, which indicate the absence of

292 pinching effects. Therefore, based on the ACI 318 design process of coupling beams with a span–

293 depth ratio of 2.5, under a similar ratio of main reinforcements, a diagonally RC coupling beam has

294 superior seismic behavior to a longitudinally RC coupling beam.

295 The specimens tested under simulated wind event (blue curves) presented more pinching

296 effect due to an increased number of cycles after Stage 6, as shown in Fig. 10, but there was no

297 occurrence of extensive damage ,and only minor cracks were observed, with widths ranging (0.35

298 to 2.0) mm ((0.014 to 0.079) in). The specimens satisfy the DC performance level proposed by

299 Alinejad et al. (2020) for the extreme conditions in the performance-based wind design framework.

300 In this possible scenario of (1,700 to 3,000) year MRI wind load for DC performance level, which is

301 also within the applicable range of ASCE (2019), the drift limit is to be set such that a coupling
12
302 beam shall not show severe damage after the applied force exceeds the yield force and

303 corresponding deformation limit.

304 Overall, the test results of hysteretic behavior signal that diagonally RC beams with half of

305 the required confinement and longitudinally RC beams with two-thirds of the required could be

306 applied to the case of low-to-moderate seismicity and high wind hazard. Moreover, the experiments

307 performed in this study can be seen as an example of the application of the performance-based wind

308 design framework.

309

310 Drift contribution

311 To understand the behavior of coupling beams during the test, the contributions of four

312 components to the total chord rotation were investigated in this study, including shear deformation

313 of the beam, flexure or curvature deformation of the beam, bar slip or extension at the beam ends,

314 and sliding at the beam ends. Each contribution was determined using the measured data and the

315 same approach taken by Cheng et al. (2019) and Abdullah (2020). As seen in Fig. 11, the vertical

316 axis represents the accumulation of contribution from the four components at peak force in each test

317 stage, and the horizontal axis shows the corresponding drift ratio for the eight test specimens.

318 The results show that shear and flexure deformations were the main contributors to the initial

319 total chord rotation for all specimens. The most rapid increase among all the four components was

320 shear deformation, which followed an increase in drift ratio, and accounted for the largest

321 contribution with more than 70 % in the final stage subjected to seismic load, and 60 % to wind

322 load in Stage 6; in the meantime, there was a very noticeable trend of the proportion of flexure

323 deformation sharply decreasing to approximately 10 % in the final stage. This is because inclined

324 shear cracks grew faster than flexural cracks as deformation demand increased until the coupling

325 beams eventually failed.

326 However, a different trend emerged for specimens tested under wind load in ramp-down

327 stages (with smaller deformation demands), where the contribution of shear deformation narrowed
13
328 down again, except for specimen L67-W. The reason for this phenomenon appears to be that

329 specimen L67-W became susceptible to shear distortion, after conspicuous cracks occurred along

330 the line of longitudinal rebars. Specimen L100-W did not have evident cracks, due to the small

331 demand of yield drift ratio taken from L100-S, and the diagonal rebars in specimens of D67-W and

332 D50-W could restrain shear distortion. Therefore, only the shear contribution of L67-W kept rising

333 until failure. Although the contribution of bar slip/extension showed a few fluctuations, it generally

334 remained under 25 %, except for specimen D67-S. Its bar slip/extension almost reached 40 % in the

335 final stage, likely due to the flexural mechanism resulting in the deterioration of corner concrete,

336 and the consequent bending of diagonal rebars. For specimen L100-S, due to unexpected out-of-

337 plane displacement occurring in the pushing direction (positive drift ratio), radial cracks developed

338 in the lower corner, and caused the contribution of bar slip/extension to be asymmetric. In terms of

339 sliding contribution, it remained steady at under 10 % for all test specimens.

340 When coupled with the graphic information, a possible conclusion can be derived that the

341 shear contribution increased as deformation increased, but different test mechanisms and

342 reinforcement layouts could cause the different composition of drift contribution to RC coupling

343 beams. Although the bar slip/extension contribution was larger at lower drift ratios under wind load

344 than under seismic load, the general compositions were similar for all the specimens with a span–

345 depth ratio (ln/h) of 2.5.

346

347 DISCUSSION

348 ACI 318-19 shear design equations

349 Table 3 shows the test results and evaluated shear strength from ACI 318-19 of four

350 specimens subjected to seismic load. All the notations were defined as follows. The shear strength

351 (Vn,beam) was estimated by using Eq. (1) for a longitudinally coupling beam, and Eq. (2) for a

352 diagonally RC coupling beam, while the upper limit (Vn,upper) was calculated by Eq. (3). As can be

353 seen in column 4 of Table 3, the nominal shear strength (Vn) of specimen L100-S was controlled by
14
354 the limitation equation, while the others were equal to Vn,beam. Column 6 of Table 3 shows that all

355 coupling beams in this study had higher normalized shear strength (Vtest/√f’c (MPa)Acw) than the

356 parameter 0.83 in Eq. (2). The rest results of 20 longitudinal coupling beams and 29 diagonal

357 coupling beams with span–depth ratios between two and four from previous findings (Barney et al

358 1980; Kanakubo et al. 1996; Shimazaki 2004; Breña and Ihtiyar 2011; Fortney et al. 2008; Lim et al

359 2016a and 2016b; Han et al. 2019; Naish et al 2013; Cheng et al. 2019; Park et al. 2020; and

360 Abdulla et al. 2020) were collected, and compared with the results of this study in Figs. 12−15.

361 Figure 12 reflects that the limitation equation covered most longitudinally RC coupling beams

362 (Longi.); however, two-thirds of diagonally RC coupling beams (Diag.) exceeded the strength limit

363 of ACI 318-19. Additionally, Fig. 13 reveals that the normalized shear strengths (Vtest/Avfysinα) of

364 29 beams mostly ranged (2.5 to 4.0), which is higher than the parameter of 2.0 in Eq. (2). Those

365 results indicate that the nominal shear strength (Vn) in ACI 318-19 significantly underestimates the

366 shear strength for diagonally RC coupling beams, which may incur damage at the adjacent walls

367 before the diagonally RC coupling beams develop sufficient plastic hinges. Rows 9, 10, and 11 of

368 Table 3 and Fig. 14 compare the ratios of Vn, VMn (Eq. [4]) and Vn,Park. to Vtest. The results in Fig.

369 14(b) show that VMn/Vtest and Vn,Park/Vtest have closer-to-one average values (μ) and smaller standard

370 deviation (σ) than Vn/Vtest, which indicate that the shear strength estimated by using the nominal

371 flexural strength (Mn) or Park et al. model show better prediction than the ACI 318 nominal shear

372 strength (Vn) for diagonally RC coupling beams.

373 Therefore, using the nominal flexural strength or the Park et al. model may be a relatively

374 efficient and accurate way to evaluate a diagonally RC coupling beam with a span–depth ratio

375 between (2 and 4) and a ratio of confinement of at least 0.35, and it is desirable for the ACI 318

376 nominal shear strength equation (Eq. [2]) and upper limit (Eq. [3]) to be revised to achieve an

377 efficient design. Here, the ratio of confinement (t) is defined as the total cross-sectional area of

378 confinement of each layer perpendicular to the beam axis divided by the gross concrete area

15
379 perpendicular to the confinement (t = At/bws), where s is the spacing between the layers of

380 confinements.

381

382 Impact of confinements

383 The test results of two longitudinally RC coupling beams with full and two-thirds amounts of

384 confinement and two diagonally RC coupling beams with two-thirds and half amounts of

385 confinement were compared with a total of 62 longitudinally and diagonally RC coupling beams

386 from previous findings in Fig. 15. The coupling beams with more confinement presented higher

387 shear strength for both longitudinally and diagonally RC coupling beams in this study, which is

388 within the trend of test results from previous findings for longitudinally RC coupling beams in Fig.

389 15(a). In Fig. 15(b), although the relationship between shear strength and the ratio of confinement

390 could not be clearly recognized, the results of this study showed that the maximum shear strength

391 diminished by about 32 kN (7.2 kips), while the total number of confinements was reduced from

392 eight to six (D67-S to D50-S).

393 In addition, Han et al. (2019) reported that the shear strength of diagonal coupling beams was

394 strongly affected by the amount of confinement, and the test results (triangle green marks in Fig.

395 15(b)) also show the same trend in this study – shear strength increases as the ratio of confinement

396 rises. Therefore, just as the contribution of transverse reinforcement is considered for estimating

397 nominal shear strength as a conventional beam for longitudinally RC coupling beams, the impact of

398 confinements on diagonally RC coupling beams may also need to be implemented in their design

399 procedure.

400

401 Stiffness

402 The effective flexural stiffness (Kf) and effective shear stiffness (Ks) of the test specimens

403 were evaluated as follows: 1) The secant stiffness at the value of 0.6 times yield force (0.6Vy,test)

404 was obtained, where Vy,test was obtained when the first main rebar reached yield strain from the test
16
405 result; and 2) this value was divided by flexural and shear deformation, respectively, where the

406 flexural deformation including flexure and bar slip/extension was taken as the deformation

407 (corresponding to 0.6Vy,test) times the percentage of the two components, and shear deformation was

408 taken as the corresponding deformation times the percentage of shear and sliding components from

409 drift contribution. The flexural rigidity (EcIeff) was defined as Kf × ln3/12, and the shear rigidity

410 (GcAeff) was determined by multiplying Ks by ln, where the modulus of elasticity of concrete was Ec

411 = 4,700√fc′ (MPa) (57,000√fc′ [psi]), and the shear modulus of concrete was Gc ≈ Ec/2.4. As can be

412 seen in Fig. 16, in this study, the test results of effective flexural stiffness values (EcIeff/EcIg) range

413 (0.09 to 0.14) and the effective shear stiffness values (GcAeff/GcAg) are about 0.05, where Ig and Ag

414 are the moment of inertia and area of the gross concrete section, respectively. Figure 16 also

415 indicates the test result data of stiffness subjected to seismic protocols collected from Vu et al.

416 (2014), Cheng et al. (2019), and Abdullah (2020). While coupling beams with span–depth ratios

417 ln/h < 2 are primarily governed by shear behavior, and those with ln/h > 4 could be seen as a flexural

418 beam in the ACI 318-19 design process, the behavior of intermediate coupling beams (2 ≤ ln/h ≤ 4)

419 could be indefinite and unpredictable. According to the contribution drift results from Lequesne

420 (2011), Cheng et al. (2019), and Abduallah (2020), the shear deformation accounted for

421 approximately 40 % for deep coupling beams (ln/h < 2), and 20 % for coupling beams with a span–

422 depth ratio of around 3.5 in the initial state. Conversely, the deformation contributed by flexure was

423 assumed to be 60 % for ln/h < 2, (1 + ln/h) × 20 % for 2 ≤ ln/h ≤ 4, and 80 % for ln/h > 4 in the

424 calculation process for data of Vu et al. (2014).

425 It can be observed from Fig. 16(a) that effective flexural stiffness values that consider only

426 the flexural deformation show a similar trend to TBI 2017 (EcIeff/EcIg = 0.07ln/h ≤ 0.3) for deep

427 coupling beams, while intermediate coupling beams seem more consistent with a constant of 0.15

428 defined by PEER/ATC 72-1. In terms of effective shear stiffness values, the average value of the

429 coupling beams is 0.12, and the majority show values between (0.05 and 0.1) in Fig. 16(b), which

430 are significantly less than the values proposed in TBI 2017 (2017) and PEER/ATC 72-1 (2010).
17
431 Figure 17 shows the secant stiffness (EcIsec) at each stage divided by the initial secant stiffness

432 ([EcIsec]initial), where the secant stiffness is defined as the ratio of the shear strength to the maximum

433 displacement. Specimens tested under wind loading protocols showed a sharp drop at Stage 2,

434 because of the large number of testing cycles. For ramp-down stages, even though specimens had

435 smaller displacement demands after yielding, the value of EcIsec/(EcIsec)initial showed a steady decline

436 until the end, where the pinching behavior becomes more prevalent, and low-cycle fatigue failure

437 may occur due to cyclic softening. In contrast, for coupling beams subjected to seismic loading

438 protocol, the secant stiffness dropped sharply after Stage 4, due to larger displacement demands.

439

440 CONCLUSIONS

441 In this study, eight reinforced concrete (RC) coupling beams with a span–depth ratio of 2.5

442 and four series of layouts were tested under seismic and wind load. Based on the test results, the

443 following conclusions and recommendations can be drawn:

444 1. Since the underestimation of nominal shear strength and upper limit for diagonally RC

445 coupling beams in ACI 318-19 might cause undesirable forces or damage to the wall system, it is

446 advisable to improve the two formulae. Additionally, considering the contribution of confinements

447 for better shear strength prediction is recommended.

448 2. The test results show that the effective flexural stiffness (EcIeff/EcIg) of approximately 0.12

449 is more comparable to PEER/ATC 72-1, and the values of effective shear stiffness (GcAeff/GcAg)

450 around 0.05 are significantly less than both TBI 2017 and PEER/ATC 72-1.

451 3. This experiment can act as an example of structural verification for performance-based

452 wind design, because specimens subjected to the simulated wind event satisfied the DC

453 performance objective with no extensive damage, and only minor cracks were observed. However,

454 as the RC coupling beams presented more pinching behavior and less stiffness after yielding, the

455 influence of low-cycle fatigue requires further investigation.

456
18
457 Acknowledgement

458 This research was supported by Korea Basic Science Institute (National Research Facilities

459 and Equipment Center) grant funded by the Ministry of Education (No. 2020R1A6C101B189), and

460 by the Institute of Construction and Environmental Engineering at Seoul National University. The

461 views expressed are those of the authors, and do not necessarily represent those of the sponsors or

462 discussants.

463

19
464 References

465 ACI Committee 318, 2019, “Building Code Requirements for Structural Concrete (ACI 318-
466 19) and Commentary (ACI 318R-19),” American Concrete Institute, Farmington Hills, MI, 623 pp.
467 ACI Committee 374, 2013, “Guide for Testing Reinforced Concrete Structural Elements
468 under Slowly Applied Simulated Seismic Loads (ACI 374.2R-13),” American Concrete Institute,
469 Farmington Hills, MI, 18 pp.
470 Abdullah, S. A.; and Wallace, J. W., 2020, “Experimental Study of Concrete Coupling Beams
471 subjected to Wind and Seismic Loading Protocols,” Final Report, University of California, Los
472 Angeles, 276 pp.
473 Abdullah, S. A.; Aswegan, K.; Klemencic, R.; and Wallace, J. W., 2020, “Performance of
474 Reinforced Concrete Coupling Beams Subjected to Simulated Wind Loading,” ACI Structural
475 Journal, V. 117, No.3, May, pp. 283-295.
476 Alinejad, H.; and Kang, T. H.-K., 2020a, “Engineering Review of ASCE 7-16 Wind-Load
477 Provisions and Wind Effect on Tall Concrete-Frame Buildings,” Journal of Structural Engineering,
478 V. 146, No. 6, June, pp. 13.
479 Alinejad, H.; Jeong, S. Y.; and Kang, T. H.-K., 2020b, “Performance-Based Design of Tall
480 Buildings for Wind Load and Application of Response Modification Factor,” Wind and Structures,
481 V. 31, No. 2, April, pp. 153-164.
482 Alinejad, H.; Kang, T. H.-K.; and Jeong, S. Y., 2021, “Performance-based Wind Design
483 Framework Proposal for Tall Buildings,” Wind and Structure, V. 32, No. 4, April, pp. 283-292.
484 ASCE, 2017, “Minimum Design Loads and Associated Criteria for Buildings and Other
485 Structures (ASCE 7-16),” American Society of Civil Engineers, Reston, VA.
486 ASCE, 2017, “Seismic Evaluation and Retrofit of Existing Buildings (ASCE 41-17),”
487 American Society of Civil Engineers, Reston, VA.
488 ASCE, 2019, “Prestandard for Performance-Based Wind Design,” American Society of Civil
489 Engineers, Reston, VA.
490 ASCE, 2022, “Minimum Design Loads and Associated Criteria for Buildings and Other
491 Structures (ASCE 7-22),” American Society of Civil Engineers, Reston, VA.
492 Aswegan, K.; Larsen, R.; Klemencic, R.; Hooper, J.; and Hasselbauer, J., 2017,
493 “Performance-Based Wind and Seismic Engineering: Benefits and Considering Multiple Hazards,”
494 Structures Congress 2017, pp. 473-484.
495 Barney, G. B.; Shiu, K. N.; Rabbat, B. G.; Fiorato, A. E.; Russell, H. G.; and Corley, W. G.,
496 1980, “Behavior of Coupling Beams under Load Reversals (RD068.01B),” Portland Cement
497 Association, Skokie, IL.
498 Breña, S. F.; and Ihtiyar, O., 2011, “Performance of Conventionally Reinforced Coupling
499 Beams Subjected to Cyclic Loading,” Journal of Structural Engineering, V. 137, No. 6, June, pp.
500 665-676.
501 Cheng, M.-Y.; Gitomarsono, J.; and Zeng, H.-Y., 2019, "Cyclic Test of Diagonally
502 Reinforced Concrete Coupling Beam with Different Shear Demand," ACI Structural Journal, V.
503 116, No. 6, Nov., pp. 241-250.
504 Fortney, P. J.; Rassati, G. A.; and Shahrooz, B. M., 2008, “Investigation on Effect of
505 Transverse Reinforcement on Performance of Diagonally Reinforced Coupling Beams,” ACI
506 Structural Journal, V. 105, No. 6, Nov.-Dec., pp. 781-788.

20
507 Fortney, P. J.; Shahrooz, B. M.; and Rassati, G. A., 2012, “The Next Generation of Coupling
508 Beams,” Composite Construction in Steel and Concrete V, April, pp. 619-630.
509 Galano, L.; and Vignoli, A., 2000, “Seismic Behavior of Short Coupling Beams with
510 Different Reinforcement Layouts,” ACI Structural Journal, V. 97, No. 6, Nov.-Dec., pp. 876-885.
511 Han, S. W.; Kim S. B.; and Kim T., 2019, “Effect of Transverse Reinforcement on the
512 Seismic Behavior of Diagonally Reinforced Concrete Coupling Beams,” Engineering Structure, V.
513 196, October, pp. 109307.
514 Jeong, S. Y.; Alinejad, H.; and Kang, T. H.-K., 2021, “Performance-Based Wind Design of
515 High-Rise Buildings Using Generated Time-History Wind Loads,” Journal of Structural
516 Engineering, V. 147, No. 9, pp.17.
517 Kanakubo, T.; Fujisawa, M.; Sako, N.; and Sonobe, Y., “Ductility of Shor Span RC Beams
518 (Paper No. 1369),’ Proceedings, Eleventh World Conference on Earthquake Engineering, Acapulco,
519 Mexico, pp. 1-8.
520 KBC 2016 (Korean Building Code), 2016, “Korean Building Code (KBC 2016)”, Ministry of
521 Land, Infrastructure and Transport of Korea, Seoul, Korea.
522 Kwan, A. K. H., and Zhao, Z. Z., 2001, “Testing of Coupling Beams with Equal End
523 Rotations Maintained and Local Joint Deformation Allowed,” Structures and Buildings, V. 152, No.
524 1, pp. 67-78.
525 Lequesne, R. D., 2011, “Behavior and Design of High-Performance Fiber-Reinforced
526 Concrete Coupling Beams and Coupled-Wall Systems,” Ph.D. Dissertation, University of Michigan,
527 Ann Arbor, MI, 277pp.
528 Lequesne, R. D.; Parra-Montesinos G. J.; and Wight, J. K., 2013, “Seismic Behavior and
529 Detailing of High-Performance Fiber-Reinforced Concrete Coupling Beams and Coupled Wall
530 Systems,” Journal of Structural Engineering, V. 139, No. 8, pp. 1362-1370.
531 Lim, E.; Hwang, S.-J.; Wang, T.-W; and Chang, Y.-H., 2016a, “An Investigation on the
532 Seismic Behavior of Deep Reinforced Concrete Coupling Beams,” ACI Structural Journal, V. 113,
533 No. 2, pp. 1-10.
534 Lim, E.; Hwang, S.-J.; Cheng, C.-H; and Lin, P.-Y., 2016b, “Cycling Tests of Reinforced
535 Concrete Coupling Beams with Intermediate Span-depth Ratio,” ACI Structural Journal, V. 113,
536 No. 3, pp. 515-524.
537 Naish, D.; Fry, A.; Klemencic, R.; and Wallace, J., 2013, “Reinforced Concrete Coupling
538 Beams. Part I: Testing,” ACI Structural Journal, V. 110, No. 6, Nov.-Dec., pp. 1057-1066.
539 Park, W.-S.; Kang, T. H.-K.; Kim, S.; and Yun H.-D., 2020, “Seismic Performance of
540 Moderately Short Concrete Coupling Beams with Various Reinforcements,” ACI Structural Journal,
541 V. 117, No. 3, May, pp. 141-154.
542 Paulay, T., and Binney, J. R., 1974, “Diagonally Reinforced Coupling Beams of Shear Walls,”
543 Shear in Reinforced Concrete, SP-42, American Concrete Institute, Farmington Hills, MI, pp. 579-
544 598.
545 PEER/ATC 72-1, 2010, “Modeling and Acceptance Criteria for Seismic Design and Analysis
546 of Tall Buildings (PEER/ATC 72-1),” Pacific Earthquake Engineering Research Center
547 (PEER)/Applied Technology Council (ATC), Berkeley, CA.
548 Shimazaki, K., 2004, “De-bonded Diagonally Reinforced Beam for Good Repairability (Paper
549 No. 2831),” Proceedings, Thirteenth World Conference on Earthquake Engineering, Vancouver, BC,
550 pp. 1-14.
21
551 Tassios, T. P.; Moretti, M.; and Bezas, A., 1996, “On the Coupling Behavior and Ductility of
552 Reinforced Concrete Coupling Beams of Shear Walls,” ACI Structural Journal, V. 93, No. 6, Nov.-
553 Dec., pp. 711-720.
554 TBI 2017, 2017, “Guidelines for Performance-Based Seismic Design of Tall Buildings,” Tall
555 Building Initiative, Pacific Earthquake Engineering Research Center, Berkeley, CA.
556 Vu, N. S.; Li, B.; and Beyer, K., 2014, “Effective Stiffness of Reinforced Concrete Coupling
557 Beams,” Engineering Structures, V. 76, October, pp. 371-382.
558 Xiao, Y.; Esmaeily-Ghasemabadi, A.; and Wu, H., 1999, “High-Strength Concrete Beams
559 Subjected to Cyclic Shear,” ACI Structural Journal, V. 96, No. 3, May-June, pp. 392-399.
560

561

562

563

564

565

566

567

568

569

570

571

572

573

574

575

576

577

578

579

22
580 List of tables
581 Table 1Measured material properties

582 Table 2Calculation for wind loading protocol

583 Table 3Strength of test specimens

584

585 List of figures

586 Fig. 1Specimen layout.

587 Fig. 2Test setup.

588 Fig. 3Instrument layout.

589 Fig. 4Seismic loading protocol.

590 Fig. 5Equal energy principle.

591 Fig. 6Gaussian distribution (based on a 200-meter-tall building).

592 Fig. 7Wind loading protocol.

593 Fig. 8Cracking progression of specimens.

594 Fig. 9Lateral load versus chord rotation curves.

595 Fig. 10 Lateral load versus chord rotation curves at 1.0θy.

596 Fig. 11 Drift contribution.

597 Fig. 12 Shear strength normalized by √f’c (MPa)Acw.

598 Fig. 13 Shear strength normalized by Avdfysinα for diagonally RC coupling beams.

599 Fig. 14 Comparison of Vn/Vmax, VMn/Vmax and Vn,Park/Vmax.

600 Fig. 15Shear strength versus ratio of confinement.

601 Fig. 16Initial effective stiffness.

602 Fig. 17Secant stiffness at each stage.

603

604
23
605 Table 1Measred material properties
Specimen series L100 L67 D67 D50
f’c, MPa 32.2 32.5 32.2 28.9
fy 461.7 -
D32
fu 645.1 -
Steel fy - 427.6
D25
reinforce fu - 668.6
-ment, fy 465.3
MPa D13
fu 702.3
fy 433.7
D10
fu 685.6
606 Note: 1 MPa = 145 psi.

607

608 Table 2Calculation for wind loading protocol

Force ratio Expected No. of Design No. of
Rotation Design force Probability
(F/Fmax) cycles cycles
1.5y 1.00 3.23 ~ 3.44 (= gL) 0.002 ~ 0.004 2~3 3
1.25y 0.87 2.80 ~ 2.98 0.012 ~ 0.017 11 ~ 15 12
1.0y 0.71 2.28 ~ 2.43 0.054 ~ 0.065 39 ~ 65 50
0.75y 0.53 1.71 ~ 1.82 0.158 ~ 0.167 101 ~ 191 150
0.5y 0.35 1.14 ~ 1.22 0.314 ~ 0.322 191 ~ 391 300
0.25y 0.18 0.57 ~ 0.61 0.431 ~ 0.451 261 ~ 547 400
609

610 Table 3Strength of test specimens

Specimen Row L100-S L67-S D67-S D50-S
Vtest, kN (1) 684.0 632.1 809.0 -761.4
Vn,beam, kN (2) 848.6 585.0 448.6 448.6
Vn,upper, kN (3) 609.2 612.0 586.9 556.0
Vn, kN (4) 609.2 585.0 448.6 448.6
VMn, kN (5) 634.0 635.1 542.3 532.0
Vn,Park, kN (6) 703.5 704.3 666.9 657.9
Vtest/√ f’c(MPa)Acw (7) 0.932 0.857 1.144 1.137
Vtest/Avdfysinα (8) - - 3.606 3.395
(1)/(4) (9) 1.123 1.080 1.803 1.697
(1)/(5) (10) 1.079 0.995 1.492 1.431
(1)/(6) (11) 0.972 0.897 1.213 1.157
611 Note: 1 MPa = 145 psi; √f’c (MPa) = 12√f’c (psi)
612 All data were calculated by tested material properties.
613

24
 1250 mm 1250 mm

300 mm 300 mm
Stirrups D13@105 mm Stirrups D13@165 mm
3@D32 3@D32
500 mm

500 mm
100 mm 2@D10 100 mm 2@D10
2@D10 2@D10
3@D32 3@D32

(a) L100 Series (b) L67 Series

1250 mm 1250 mm

300 mm 300 mm
Stirrups D13@165 mm 4@D25 Stirrups D13@230 mm 4@D25
3@D10 3@D10
15° 15°
500 mm

500 mm
100 mm 2@D10 100 mm 2@D10
2@D10 2@D10
3@D10 3@D10
4@D25 4@D25
350 mm (Longi.) 350 mm (Longi.)
(c) D67 Series (d) D50 Series
614

615 Fig. 1Specimen layout (Note: 1 mm = 0.039 in.)

616

Upper Steel Frame Gusseted Angle Bracket

Top Block
Actuator 500 mm
2200 mm

Pin
Reaction Wall Connection
1250 mm

Vertical Steel Coupling

Frame Beam

Lower Steel Frame

Bottom Block

3000 mm

(a) Test setup (b) Mechanism

617

618 Fig. 2Test setup.

619

620

621

622

623

624

25
 LN13
LN01 LN02
LN16 LN16
LN14
LN05 LN07

LN09 LN10

LN06 LN08
LN11 LN12
LN15
LN04 LN03

625

626 Fig. 3Instrument layout.

Stage 1 2 3 4 5 6 7 8 9 10 11
Cycles 2 2 2 2 2 2 2 2 2 2 2
10
8
6
Drift Ratio [%]

4
2
0
±0.25 ±0.5 ±0.75
-2 ±1.0 ±1.5
±2.0
-4 ±3.0
±4.0
-6
±6.0
-8
±8.0
-10
±10.0
627

628 Fig. 4Seismic loading protocol.

629

1.20

1.0 Elastic response

1.00
0.87
0.80
0.71 Elasto-plastic
response
F/Fmax

0.60

0.40
1.41
1.22

0.20

0.00
0.00 0.25 0.50 0.75 1.00 1.25 1.50
θ/θy
630

631 Fig. 5Equal energy principle.

26
632

0.5

0.4

Probability Density
0.3
1.19
0.2

0.1 1.78
0.16/2

0
0 1 2 3
Peak factor, gL
633

634 Fig. 6Gaussian distribution (based on a 200-meter-tall building).

635

Stage 1 2 3 4 5 6 7 8 9 10 11
Cycles 200 150 75 25 6 3 6 25 75 150 200
1.5

0.5
θ / θy

0
≈

±0.25 ≈
±0.25
-0.5
±0.5 ±0.5
±0.75 ±0.75
-1
±1.0 ±1.0
±1.25 ±1.25
-1.5
±1.5
636

637 Fig. 7Wind loading protocol.

638

639

640

641

27
642

643 Fig. 8Cracking progression of specimens. (Note: DR is the drift ratio, CW is the maximum

644 crack width, and FS is the final state of the specimen.)

645
28
 θtest/θy,test θtest/θy,test
-4 -2 0 2 4 -3 -2 -1 0 1 2 3
800 800
(684.2kN, 3.5%)
150 (632.1 kN, 2.0%) 150
600 600

100 100
400 400

200 50 200 50

Force [kips]
Force [kN]

0 0 0 0

-200 -50 -200 -50

-400 -400
-100 -100

-600 -600
-150 -150
-800 -800
-4 -2 0 2 4 -4 -2 0 2 4
Drift Ratio [%] Drift Ratio [%]

θtest/θy,test
-6 -4 -2 0 2 4 6
1000
200
800
150
600
Vn=448.6 kN 100
400
50
Force [kips]

200
Force [kN]

0 0

-200 -50

-400 -100
-600
-150
-800 (-761.4 kN, -5.7%)
-200
-1000
-10 -8 -6 -4 -2 0 2 4 6 8 10
Drift Ratio [%]
646

647 Fig. 9Lateral load versus chord rotation curves.

648

29
 θtest/θy,test θtest/θy,test
-2 -1 0 1 2 -2 -1 0 1 2
800 800
150 150
600 600

100 100
400 400

200 50 200 50

Force [kips]
Force [kN]

0 0 0 0

-200 -50 -200 -50

-400 -400
-100 -100

-600 -600
-150 -150
-800 -800
-4 -2 0 2 4 -4 -2 0 2 4
Drift Ratio [%] Drift Ratio [%]
649

650 Fig. 10 Lateral load versus chord rotation curves at 1.0θy.

651

652

653 Fig. 11 Drift contribution.

654

30
 2

1.8

1.6

1.4

1.2

0.8

0.6

0.4

0.2
2 2.5 3 3.5 4
655 Span-depth Ratio [ ln/h]

656 Fig. 12 Shear strength normalized by √f’c (MPa)Acw.

657

0
2 2.5 3 3.5 4
658 Span-depth Ratio [ln/h]

659 Fig. 13 Shear strength normalized by Avdfysinα for diagonally RC coupling beams.

660

31
 Vtest/(Vn or VMn or Vn,Park)

Vtest/(Vn or VMn or Vn,Park)

661

662 Fig. 14 Comparison of Vtest/Vn, Vtest/VMn and Vtest/Vn,Park.

663
Shear Strength [Vtest/√fc'(MPa)Acw]

Shear Strength [Vtest/√fc'(MPa)Acw]

664

665 Fig. 15Shear strength versus ratio of confinement.

666

667

32
 0.5

0.4

0.05

0.3 0.045
EcIeff /EcIg

0.04

0.2

0.1 0.12 (Average)

0
0 1 2 3 4 5
Span-depth Ratio [ln/h]
668 (b) Effective Shear Stiffness

669 Fig. 16Effective stiffness.

670
EcIsec / (EcIsec)initial

671

672 Fig. 17Secant stiffness at each stage.

673

674

33