You are on page 1of 36

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/327644589

Simplified Design Method for the Shear Capacity of Steel Plate Shear-
Strengthened Reinforced-Concrete Beams

Article  in  Journal of Bridge Engineering · November 2018


DOI: 10.1061/(ASCE)BE.1943-5592.0001310

CITATIONS READS

6 819

5 authors, including:

Wang Jiaji Xin Nie


University of Houston Tsinghua University
24 PUBLICATIONS   49 CITATIONS    88 PUBLICATIONS   326 CITATIONS   

SEE PROFILE SEE PROFILE

Mu-Xuan Tao
Tsinghua University
105 PUBLICATIONS   762 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Predicting cracking behavior of RC and composite members based on fracture mechanics approach View project

Performance Assessment of High-Performance Composite/Mixed Structural Systems View project

All content following this page was uploaded by Mu-Xuan Tao on 22 October 2018.

The user has requested enhancement of the downloaded file.


Manuscript
This manuscript was published at: Wang J J, Zhou M, Nie X, Fan J S, Tao M X*. Simplified design method for the shear
capacity of steel plate shear-strengthened reinforced-concrete beams. ASCE Journal of Bridge Engineering, 2018, 23(11):
04018089.
The final publication is available at the journal website. The researchers can also privately get the final publication version via
sending Email or ResearchGate message to Prof. Mu-Xuan Tao (taomuxuan@tsinghua.edu.cn).

1 A Simplified Design Method for the Shear Capacity of Steel Plate

2 Shear-strengthened Reinforced Concrete Beams

3 Jia-Ji Wang1, Meng Zhou2, Xin Nie3, Jian-Sheng Fan4 and *Mu-Xuan Tao5

4 ABSTRACT: Recently, a steel plate concrete composite technique has been developed for shear strengthening

5 of reinforced concrete (RC) bridges. The steel plate shear-strengthened (SPSS) beam is composed of an old

6 RC beam, new concrete and steel web. In order to investigate the shear behavior of SPSS beam strengthened in

7 the shear span, six SPSS beams and four RC beams are tested. The test results showed that the failure mode of

8 SPSS beams was the peeling failure at the interface between the old concrete and new concrete. The proposed

9 technique showed a significantly enhanced shear capacity even when the tensile strength of the interface

10 between the old and new concrete was relatively low. The initial interface failure appeared at the loading point,

11 support, and corner of the new concrete and then expanded to the entire shear span. The FE model of the SPSS

12 beam is established by the finite element software MSC.Marc, and the cohesive element is used to model the

13 interface between the old and new concrete. The modeling scheme, material constitutive law, and material

14 parameters are illustrated. The results of the proposed FE model agree fairly well with the test results in terms

15 of the overall load–displacement curve, ultimate shear capacity, and interface failure pattern. Based on the FE

16 model, the shear force contribution of the old concrete, new concrete, and steel web are decomposed with

17 sufficient accuracy. The effects of shear span to depth ratio, material strength and interface strength are

18 investigated by an FE parametric analysis, and the interface strength between the old and new concrete is found

19 to be the most important parameter. The shear contribution of the steel web is mainly affected by the interface

1 Jia-Ji Wang, PhD candidate, Key Laboratory of Civil Engineering Safety and Durability of China Education Ministry, Department
of Civil Engineering, Tsinghua University, Beijing, China 100084. E-mail: wangjiaji_tsinghua@163.com
2 Meng Zhou, PhD, Zhuhai Urban Planning Verifying & Informatization Center (Zhuhai Urban Planning Exhibition Hall), Zhuhai,
China 519000. E-mail: zhoum07@foxmail.com
3 Xin Nie, Postdoctoral Scholar, Key Laboratory of Civil Engineering Safety and Durability of China Education Ministry,
Department of Civil Engineering, Tsinghua University, Beijing, China 100084. E-mail: xinnie@tsinghua.edu.cn
4 Jian-Sheng Fan, Professor, Key Lab. of Civil Engineering Safety and Durability of China Education Ministry, Dept. of Civil
Engineering, Tsinghua University, Beijing 100084, China. E-mail: fanjsh@tsinghua.edu.cn
5 Mu-Xuan Tao, Associate Professor, Beijing Engineering Research Center of Steel and Concrete Composite Structures, Tsinghua
University, Beijing, China 100084. (corresponding author). E-mail: taomuxuan@tsinghua.edu.cn

1
20 strength. Finally, 1296 elaborate FE models of SPSS beams are established by MSC.Marc, and a practical

21 design formula considering peeling failure is developed with a reasonable level of accuracy.

22 Keywords: Reinforced concrete; Shear capacity; Steel plate shear strengthening; Peeling failure; Finite

23 element model; Cohesive element

24 Introduction

25 Because of the increased loading requirements, deterioration of existing materials, and development of

26 design codes, many existing reinforced concrete (RC) structures need retrofitting. To investigate the

27 strengthening effect, most research has been focused on conventional strengthening techniques. Among them,

28 the external steel plate method has been investigated by experimental, theoretical and numerical analysis

29 (Raoof et al. 2000; Barnes and Geoffrey 2006a; Barnes and Geoffrey 2006b; Arslan et al. 2008; Shi et al. 2012;

30 Feng et al. 2015). The existing literature showed that the interface peeling failure is the dominant failure mode

31 in this method, and the ultimate peeling capacity can be even lower than the original RC beam (Raoof et al.

32 2000) and many design recommendations have been proposed for enhanced design of external steel plate

33 method. Apart from that, many other strengthening technique have also been applied in retrofitting engineering

34 structures, such as fiber-reinforced plastic (FRP) plated method (Chen and Teng 2001; Aprile et al. 2001; Ding

35 et al. 2018), enlarged section method (Li et al. 2014), and prestressing method (Wu et al. 1999; Miyamoto et al.

36 2000). An extensive volume of research has supported the use of these design methods, and the design

37 guidelines have been included in many design codes (ACI 2008; CEB-FIP 2010).

38 Recently, Nie and Zhao (2008) proposed the steel plate concrete composite technique for flexural

39 strengthening of existing RC bridges as shown in Fig. 1(a), which is a combination of the external steel plate

40 method and the enlarged section method. The steel plate strengthened beam consists of old concrete, new

41 concrete, and a steel plate. The manufacturing process is divided into the following three stages. First, drill

42 holes at the soffit of existing RC beam. Second, post-installed anchors are embedded into the holes and

43 grouted mortar is used to fix the anchors. Finally, the steel plate with headed shear studs is used as the

44 bottom framework and new concrete is poured between the steel plate and RC beam. Nie and Zhao (2008)

45 showed that the proposed technique significantly enhanced the bending capacity and stiffness when compared

46 with original RC beam. The SPSS technique had also been used to retrofit many existing RC bridges,

2
47 including Wanliu Bridge, Yang Bridge, Fengyi Bridge, Madian Bridge and Zizhuyuan Bridge in Beijing in

48 2008. These bridges have been reported to work well after strengthening and the interface between old and

49 new concrete are in good condition based on visual inspections.

50 In retrofit of shear-critical RC bridges, it is more cost efficient to strengthen the shear span (Fig. 1(c))

51 instead of strengthening the full span (Fig. 1(b)). However, the RC bridges strengthened in the shear span using

52 the SPSS technique as shown in Fig. 1(c) has not been investigated in existing literature. In addition, the effect

53 of the interface strength between the old and new concrete in the SPSS beam has not been investigated in

54 literature. In this study, experimental and finite element (FE) analysis are performed on the SPSS beams

55 strengthened in the shear span and the interface strength between the old and new concrete were designed to be

56 relatively low. The objectives of this study are as follows: (1) to develop an FE model and precisely predict the

57 shear behavior and interface behavior of the SPSS beams and (2) to develop a practical design method for

58 calculating the shear strength of SPSS beams.

59 Experimental Program

60 Specimen design and measurements

61 Four RC beams and six SPSS beams strengthened in the shear span were designed to investigate the

62 failure mode of the proposed shear-strengthening technique. The design of the test specimens are as shown in

63 Fig. 2 and Table 1. The parameters in the test include the steel plate thickness and the shear span to depth

64 ratio. First, four RC beams numbered as RC1 to RC4 were manufactured to serve as reference beams with

65 shear span lengths of 450 mm, 750 mm, 950 mm, and 1450 mm, respectively. The shear span to depth ratio

66 ranged from 1.11 to 3.56. In addition, six shear-strengthened specimens numbered as SS1 to SS6 were

67 designed to investigate the effect of shear span to depth ratio and steel plate thickness. SS1 to SS4 were

68 strengthened with steel plate of 3 mm in thickness, while SS5 to SS6 had steel plate of 4 mm in thickness.

69 The RC and SPSS beams had a width of 300 mm and 460 mm respectively. Shear strengthening

70 technique was only applied at the shear span as shown in Fig. 2(a). In the test, the supports and loads were

71 only applied at the old concrete beam in order to simulate the most unfavorable boundary condition in

72 engineering practice. The longitudinal reinforcement of 32 mm in diameter was applied to avoid flexural

73 failure. All beams had round stirrups of 6 mm in diameter and 200 mm in spacing. For the interface between

3
74 the new concrete and old concrete, embedded rebar with a diameter of 12 mm, length of 120 mm, and

75 spacing of 80 mm were used. The holes bored on the RC beams had a depth of 60 mm, which is equal to the

76 effective anchorage depth of the embedded rebar. CGMJM microdilatancy mortar was used to fix the

77 embedded rebar. The embedded depth of the rebar was relatively small to allow investigation of the

78 performance of the proposed strengthening technique when the interfacial tensile strength between the old

79 and new concrete was limited. For the interface between the new concrete and steel web, shear studs with a

80 diameter of 6 mm, length of 45 mm, and spacing of 80 mm were used. The studs were designed to transfer a

81 sufficient shear force for the yield strength of the steel web.

82 For the SPSS beams, ten strain rosettes were applied at the steel web for each specimen as shown in Fig.

83 2(b). In addition, ten strain gauges were applied at the stirrups in the middle of shear span. The LVDTs were

84 installed at the bottom of the beam to record the deflection. Monotonic vertical loads were applied by an

85 actuator and were controlled by displacement. The loading actuator had a capacity of 20000 kN, which

86 provided sufficient stiffness in the test. The loading of actuator and supports were only applied on the old RC

87 beam, which accurately simulates the retrofit condition in engineering practice.

88 Fabrication and material properties

89 Fig. 3 shows the fabrication procedure of the specimen described as follows. First, the surface of the

90 concrete block is roughened. Then, holes are bored at the beam side to a depth of 60 mm and dust in the

91 holes are cleaned by an air compressor and acetone (Fig. 3(a)). Second, the bonded steel bars are installed

92 and fixed in the holes by mortar (Fig. 3(b)). The bonded steel bars serve as grouted anchor and transfer the

93 shear and tensile stresses at the interface between the old and new concrete. Finally, a steel web is used as

94 framework and concrete is poured between the steel web and RC beam (Fig. 3(c)). The mechanical

95 properties of the reinforcing materials are obtained from the material property tests as listed in Table 1. The

96 cylinder compressive strength of concrete is formulated as follows (Chen et al. 1992):

0.8 focu  focu  50 


97 foc   (1)
 f ocu  10  focu  50 

0.8 f ncu  f ncu  50 


98 f nc   (2)
 f ncu  10  f ncu  50 

4
b1 f oc  b2 f nc
99 f c  (3)
b1  b2

100 where f'oc and f'nc denote the cylinder strength of the old and new concrete respectively; focu and fncu denote the

101 cubic strength of the old and new concrete respectively, and fc' denotes the average cylinder strength of the

102 concrete. Because f'oc and f'nc are very close, their average values fc' are listed in Table 1; b1 and b2 denote the

103 thickness of old and new concrete, respectively. The yield strength of the embedded rebar is 479 MPa.

104 Finite Element Analysis

105 Modeling scheme

106 The nonlinear FE model is developed in MSC.Marc (2015) to simulate the test results of the SPSS beams.

107 The modeling scheme is as shown in Fig. 4. Only one-fourth of the model is established owing to the

108 symmetry of the boundary condition. The symmetry condition ux = 0 was applied at the mid-span section and

109 the symmetry condition uy = 0 was applied at the centerline of the beam. In order to simulate the test

110 boundary condition, the loads and supports were only applied at the old concrete beam. The old concrete and

111 new concrete were modeled by eight-node first-order brick elements (Element 7). The reinforcement was

112 modeled by first order truss elements (Element 9) and the steel web was modeled by four-node membrane

113 elements (Element 18). The studs were designed to transfer a sufficient shear force for the yield strength of

114 the steel web; therefore, this interface was neglected and the new concrete and steel were sharing nodes. The

115 large strain option is included in MSC.Marc (2015) to consider the possible large displacement of elements

116 due to peeling failure. The element mesh size is set as 50 mm and totally 2000 to 9000 elements are used for

117 the FE modeling of each SPSS beam. Mesh sensitivity tests showed that subdividing the mesh by eight times

118 caused merely 4% fluctuation in capacity.

119 In addition, based on the FE model, the shear force components of the steel web, new concrete, and old

120 concrete are decomposed as follows:

121 VFE  Vs  Vnc  Voc (4)

122 where VFE denotes the shear force in FE model, Vs the shear force contribution of the steel plate, Vnc the shear

123 force contribution of new concrete, and Voc the shear force contribution of old concrete. The force

124 components are integrated from the stress results at the vertical cross section at the middle section of the

5
125 shear span as shown in Fig. 4. In FE model, the shear force of each component is calculated by integrating

126 the shear stress results in the cross section as shown in Fig. 4.

127 For modeling of the interface between the old and new concrete, numerous FE models have been

128 proposed for grouted anchors in existing literature and they can be divided into micro models and macro

129 models. The micro models establish the rebar and grouts in the grouted anchor as in the real condition

130 (Eligehausen et al. 2006), while the macro models avoid modeling these features and consider the grouted

131 anchor as a thin layer of cohesive element (Carol et al. 1997; Maekawa et al. 2003). The micro models

132 appear to be elaborate and may require a huge amount of calculation time in implicit iteration process of the

133 FE software. On the contrary, the macro models are comparatively more efficient in computation and are

134 used in the FE model. The eight-node three-dimensional cohesive element (Element 188) in MSC.Marc

135 (2015) is used for modeling the interface. The relative displacements between the top and bottom face of the

136 element were used as input quantity and the normal and shear stresses were the output.

137 Constitutive law of grouted anchor

138 The spring element has been reported to well simulate the interface with adequate shear connectors (Liu et

139 al. 2014). In comparison, the cohesive element has been reported to well predict the brittle failure of interface

140 as recommended by Maekawa et al. (2003). In this research, the constitutive law of the grouted anchor was

141 simulated by cohesive elements with the cohesive material model as shown in Fig. 5. The cohesive element can

142 simulate the interface with different strengths in tensile and shear. The descending curve of the

143 stress–displacement curve was assumed to be linear as recommended by Maekawa et al. (2003). The

144 MSC.Marc software (2015) adopts the cohesive material law of the cohesive element as recommended by

145 Geubelle and Baylor (1998). The input quantities required are the tensile strength f1, shear strength f2, cracking

146 displacement uc, maximum tensile opening displacement um1, and maximum shear displacement um2 as shown

147 in Fig. 5.

148 In MSC.Marc (2015), the damage of the interface is defined as a quantity monotonically increasing from

149 0 (initiation of delamination) to 1 (full delamination) as follows:

2 2 2
uN  uc u  uc2 u u
150 D  M  L c2 (5)
um1  uc um 2  uc2 um 2  uc2

6
151 x  0.5( x  x ) (6)

f2
152 uc2  uc (7)
f1

153 where D denotes the damage of the cohesive element, uc2 the shear cracking displacement, uN the crack

154 opening displacement, uM and uL the shear shear displacement, and the Macaulay brackets.

155 The criteria for the peeling failure in FE analysis is as follows. For the peeling failure mode, the damage

156 initiation of the interface between old and new concrete happened at the ultimate capacity and the interface

157 damage significantly expanded to the whole shear span, causing significant decrease of loading capacity. In

158 the FE model, the material parameters are determined from the existing literature as follows. As for the

159 uniaxial tensile parameters, significant efforts have been devoted to determining the ultimate tensile capacity

160 of grouted anchors in existing literature (Fuchs et al. 1995; Eligehausen and Balogh 1995; Zamora et al. 2003;

161 Cook and Burtz 2003; ACI 2014). As recommended by Cook and Burtz (2003), the strength of the anchor

162 group is the lowest value of the three failure modes as follows:

163 N  min( N  , N c , N s ) (8)

164 where N denotes the tensile strength of the anchor group, Nτ the strength of uniform bond model, Nc the

165 concrete breakout strength, and Ns the rebar yield strength.

166 The strength of the uniform bond model Nτ is formulated as

AN
167 Nτ   τ,e  d 0 hef (9)
AN0

168 AN0  9hef2 (10)

c
169  τ,e  0.7  0.3 (11)
8d 0

170 where AN denotes the projected concrete failure area of the anchor group total area of the anchor group; and

171 AN0 is the projected concrete failure area of a single anchor. hef denotes the effective anchorage depth of the

172 anchor. ψτ,e denotes the coefficient for group anchors. τ = 13 MPa denotes the bond strength specified by the

173 quality report of the mortar manufacturer, c denotes the anchor spacing, and d0 denotes the rebar diameter.

174 The concrete breakout strength Nc is formulated as (in unit of N and mm)

7
AN
175 Nc   c,e  12.6 f chef1.5 (12)
AN0

c
176  c,e  0.7  0.3 (13)
1.5hef

177 where ψc,e denotes the coefficient for group anchors in concrete breakout failure.

178 The yield strength of the rebar is formulated as

179 Ns  f ys Asr (14)

180 where fys the yield strength of rebar, and Asr the area of rebar in the shear span.

181 As recommended by Cook and Burtz (2003), a concrete breakout failure only occurs for very small

182 values of hef and a yield failure of rebar only occurs for very large values of hef. For the SPSS beams in the

183 test, the calculation results show that all the grouted anchors failed as predicted by the uniform bond model.

184 After obtaining the tensile strength of the rebar group, the average tensile strength f1 of the interface between

185 the old and new concrete is formulated as

N
186 f1  (15)
Ac

187 where Ac denotes the total area of the interface between the old and new concrete.

188 As for the shear strength of the interface between the old and new concrete, the CEB-FIP code

189 (CEB-FIP 2010) formulates the interface shear strength as a combination of adhesive bonding, shear friction,

190 and dowel action of rebar. In the test, the shear friction force is neglected because no compression force was

191 exerted perpendicular to the interface. Thus, the shear strength formula of the CEB-FIP code (CEB-FIP 2010)

192 is formulated as:

193 f2   a   2   f ys fc  c vfc (16)

30 1/3
194 v = 0.55( ) (17)
f ck

195 f ck  f c  8 (18)

196 where f2 denotes the average shear strength of the interface; τa denotes the adhesive bonding strength (for

197 very rough surface in the test, τa is recommended as 2.5 MPa by the CEB-FIP code; κ2 = 1.6 denotes the

8
198 coefficient for the dowel action strength; fys denotes the yield strength of the rebar; βc denotes a coefficient

199 for different surface roughness (for very rough surface in the test, βc = 0.5); v denotes the effectiveness factor

200 for the concrete strength, and fck denotes the standard compressive strength of concrete.

201 The calculated results for the interface strengths are

202 f1 = 0.7 MPa (19)

203 f 2 = 5.0 MPa (20)

204 The maximum shear displacement um2 is formulated as recommended by the CEB-FIP code as

205 um 2 =0.15d 0 (21)

206 The cracking displacement uc and maximum tensile opening displacement um1 are reported by Cook and

207 Burtz (2003) as

208 uc =0.1d 0 (22)

209 um1 =0.5d 0 (23)

210 Constitutive law of steel and concrete

211 The confinement effect has been reported to show significant influence on specimens with large amount of

212 lateral rebar (Bai et al. 2008; Nie et al. 2008). In this research, the confinement was neglected because the test

213 specimens showed peeling failure mode. Therefore, the elastic-perfectly plastic material is adopted for

214 simulation of the reinforcement and steel plate. The von Mises yield surface is used for modeling of concrete

215 and the compressive constitutive law of concrete is proposed by Guo (2014) as:

 2 /  0    /  0 2
 (   0 )
216    / 0 (24)
 (   0 )
 d  /  0  1   /  0
2

217 where σ denotes the concrete compressive stress, ε the compressive strain, ε0 the yield strain and αd = 0.35

218 denotes a parameter for the compression failure of concrete

219 The following material properties of concrete are calculated from fc' as proposed by the Model Code

220 (CEB-FIP 2010):

Ec =21500  fc / 10


1/3
221 (25)

222  c =0.2 (26)

9
0.3  f c  8 
  fc  58
2/3

223 ft   (27)
2.12ln(1  0.1 f c)
  fc  58

224 GF  0.073( fc)0.18 (28)

225 where Ec denotes the Young’s modulus, vc the Poisson’s ratio, ft the tensile strength of concrete and GF the

226 fracture energy.

227 As recommended by Maekawa et al. (2003), the fixed-angle crack assumption has the capability to

228 simulate the condition where stress and strains are non-coaxial. Therefore, the smeared crack model with

229 fixed-angle crack assumption was used to simulate the cracking of concrete. The slope of the tension

230 descending curve of concrete Ets is formulated as

f t 2lr
231 Ets = (29)
2GF

232 where lr denotes the element length.

233 In fixed-angle crack assumption, Rots (1991) introduced the shear retention ratio to adjust the cracked

234 shear stiffness of concrete element:

235 G  G0 (30)

Ec
236 G0  (31)
2(1   c )

237 where G denotes the tangential shear modulus, G0 the elastic shear modulus of concrete, η the shear retention

238 ratio, and vc the Poisson’s ratio.

239 Tao and Nie (2016) recommended an exponential relation between the shear retention ratio η and cracking

240 strain based on test results, which is adopted in the FE modeling:

241  =0.25  exp(800 cr ) (32)

242 where εcr denote the total crack strain.

243 Results and Discussion

244 Test observations

245 The typical failure mode of test specimens are shown in Fig. 6. As for the RC beams, RC1 and RC2, the

246 failure is in shear compression mode (Fig. 6(a-b)), while RC3 and RC4 failed in shear tension mode (Fig.

10
247 6(c-d)). As defined by Kani (1966), beams with λ exceeding 2.4 are termed as slender beams, and beams

248 with λ lower than 2.4 are non-slender beams. Kani (1966) first introduced the famous shear valley and

249 concluded that slender beams with λ exceeding 2.4 fail in shear tension mode and non-slender beams fail in

250 shear compression mode. The test results of RC1 to RC4 conformed to Kani’s (1966) prediction. The

251 shear-strengthened beams all failed in peeling failure mode with the delamination displacement in the

252 interface between the old and new concrete. As shown in Fig. 6(e-f), the peeling failure was a combination of

253 relative shear displacement and tensile displacement at the interface between the old and new concrete. The

254 peeling failure initiated at the maximum load around the loading section and the support at the maximum

255 load. When the load decreased, the interface peeling failure significantly spread to the whole shear span, with

256 rapid decreasing of the shear load. On the contrary, the interface between the new concrete and steel web

257 didn’t fail because the shear studs were designed to transfer a sufficient shear force.

258 To compare the test strength with the bending capacity, Kani (1966) first compared the shear capacity

259 with the theoretical flexural capacity and defined the ratio β as

260   M u / M flex (33)

261 where Mu denotes the maximum bending moment in the loading history, Mflex the flexural capacity calculated

262 by the fiber beam–column model (Tao and Nie, 2015). The fiber beam-column model has been reported to

263 well predict the flexural capacity of steel-concrete composite beams (Tao and Nie, 2015). The uniaxial

264 skeleton curve of the fiber beam-column model is consistent with the continuum-based model reported in

265 this research. Furthermore, the fiber beam-column elements includes the effect of loading history on the

266 residual plastic strain as recommended by Sakai and Kawashima (1998). The uniaxial material model of steel

267 reflects the significant Bauchinger effect.

268 The β result of each test specimen is as shown in Fig. 7. The β of RC beams ranged from 39% to 48%,

269 while the SPSS beams ranged from 64% to 75%. Furthermore, the β ratio increased with respect to λ for the

270 SPSS beams, indicating that the strengthening effect for the slender SPSS beams are more efficient than that

271 of the non-slender SPSS beams.

272 Comparison of test and FE results

11
273 Fig. 8 illustrates the shear force components with respect to the mid-span deflection. In Fig. 8, V denotes

274 the test shear force and VFE the shear force in FE analysis; Voc, Vnc, and Vs denote the force components of old

275 concrete, new concrete, and steel web in the FE analysis respectively as formulated in Eq.(4) and VRC denotes

276 the shear force of the reference RC beam with the same λ. The yield load is calculated by dividing the ultimate

277 capacity by the initial stiffness of the load-displacement skeleton curve (Han and Li, 2010) and plotted in Fig.

278 8. The test results show that generally, the strengthened beam has a significantly enhanced initial stiffness

279 compared with the reference RC beam. The ultimate shear capacity of the strengthened beam was 1.56 to 2.27

280 times that of the reference RC beam, indicating that the proposed strengthening technique provided improved

281 shear capacity. In addition, the comparison between the FE and test results in Fig. 8 indicates that the

282 developed FE model with cohesive elements can accurately predict the load–displacement relationship of the

283 strengthened beam. Based on the FE results, upon initial loading, the shear force was mainly maintained by the

284 old concrete because it had relatively high stiffness and the loads and supports were applied on the old concrete

285 beam. After reaching the yield load, Vnc and Vs grew more significantly compared with Voc. The old concrete

286 reached maximum capacity prior to the new concrete and steel web. At the ultimate capacity, Voc was

287 decreasing while Vnc and Vs were increasing for most of the SPSS specimens.

288 The error of the proposed force decomposition method based on the FE model can be evaluated as

Vs  Voc  Vnc  VFE


289 Error  (34)
VFE

290 Fig. 9 shows the relative error of Eq.(34) based on the FE results. It can be observed that all errors in the FE

291 model were under 6%, indicating that the proposed force decomposition method had sufficient accuracy.

292 The failure pattern, test shear strength V, Kani’s ratio β, predicted shear strength VFE, and shear force

293 components (Vs, Vnc, and Voc) are listed in Table 2. The force components are calculated at the maximum load

294 in the FE model. Among the SPSS beams, the SS1 and SS5 had VFE values slightly lower than V. However, the

295 proposed FE model accurately predicts the shear strength of the SPSS beams in general.

296 To further validate the applicability of the FE model, the damage of the interface D is defined by

297 MSC.Marc (2015) as formulated in Eq. (5). The damage contour of the interface between the old and new

298 concrete of SS1 specimen is extracted from the FE model and plotted in Fig. 10. The initial interface failure

299 appeared at the loading point, the support, and corner of the new concrete and then expanded to the entire

12
300 shear span. When the shear load had dropped by 20%, the damage rapidly developed throughout the shear

301 span as shown in Fig. 10(a)–(b). The FE results agree well with the interface failure pattern in the test. The

302 peeling failure was localized at the loading point and the bottom of the strengthened concrete. The equivalent

303 plastic strain of old concrete is shown in Fig. 10(c) to illustrate its compression damage. The maximum

304 equivalent plastic strain was 0.021 around the loading point, this was attributed to the shear force transferred

305 by the grouted anchor to the old concrete. Furthermore, the maximum equivalent plastic strain of most part

306 of the old concrete was under 0.002, indicating the damage of old concrete didn’t expand to the whole shear

307 span. Both the test results and the FE results show that the peeling failure is the dominant failure mode in the

308 test.

309 Effects of key parameters

310 Effect of shear span to depth ratio

311 In the parametric analysis, the strength of the interface between the old and new concrete is first set to a

312 large value to avoid peeling failure and the effect of λ is examined. The FE models are analyzed with shear

313 span to depth ratios increasing from 1.1 to 4.6. Furthermore, the dimensionless quantities were defined to

314 reflect the yield ratio of each component as follows:

Voc  Vnc
315 c  (35)
f ocb1h0  f ncb2 h0

Voc
316  oc  (36)
f ocb1h0

Vnc
317  nc  (37)
f ncb2 h0

Vs Vs
318 s   (38)
Vus 1
f y As
3

319 where αc, αoc, αnc, and αs denote the yield ratio of concrete, old concrete, new concrete, and steel web,

320 respectively; f'oc and f'nc denote the cylinder strength of the old and new concrete, respectively; b1 and b2

321 denote the width of the old and new concrete, respectively; h0 denotes the distance from the beam top to the

322 center of tension rebar; Vus denotes the shear capacity of the steel web per Chinese code (CMC 2010); fy

323 denotes the yield strength of the steel web and As denotes the cross-sectional area of the steel web.

13
324 The force components of the FE model are plotted in Fig. 11(a). The FE results show that the force

325 components decrease with λ rapidly. The yield ratio αc, αoc, αnc, and αs are plotted in Fig. 11(b). The yield ratio

326 of each component also declines with respect to λ significantly. The FE results showed that the peeling failure

327 of the interface was avoided by setting the interface strength f1 as 4.2 MPa and f2 as 6 MPa and no damage of

328 the interface could be found in the loading history. However, the yield ratio of the steel web αs is lower than

329 100% based on test results. The following two reasons can explain this phenomenon. First, the steel web was

330 only applied at the side of the RC beam; therefore, the steel web was subjected to bending–shear loading

331 condition and the bending moment decreased its shear capacity. Second, the drift ratio corresponding to the

332 maximum load ranged from 1% to 2% in the SPSS beam, which is not adequate for the steel web to completely

333 yield. Because of the reasons mentioned above, the contribution of the steel web need to be reduced in design

334 practice. Furthermore, the yield ratio αoc is slightly larger than αnc as shown in Fig. 11(b). The difference

335 between αoc and αnc was negligible when λ ranged from 2.3 to 3.6. Therefore, the difference between the old

336 and new concrete can be neglected in developing a simplified design formula.

337 Effect of steel and concrete strength

338 To investigate the effect of steel and concrete strength, six FE models are established with different steel

339 web strength and fixed concrete strength. The interface between the old and new concrete was modeled with a

340 high-strength cohesive element to avoid peeling failure. The force components of the FE results are shown in

341 Fig. 12(a). When fy/fc’ increases significantly, the ultimate shear capacity V changes slightly. The yield ratio of

342 each component is shown in Fig. 12(b). It indicated that a high-strength steel is not suitable for the

343 strengthening technique because the yield ratio of the steel web αs is relatively low. Moreover, the yield ratio of

344 concrete αc is not changed with varying steel web strength.

345 Effect of interface strength

346 The experimental and numerical results both showed that when the interface strength was low (f1 = 0.7

347 MPa and f2 = 5.0 MPa), brittle peeling failure occurred on the shear-strengthened beams. To investigate the

348 effect of the interface strength, a parametric study was investigated based on the SS1 specimen. The interface

349 tensile strength f1 ranged from 0.7 MPa to 4.2 MPa and the interface shear strength f2 ranged from 2 MPa to

350 7 MPa. Based on the force decomposition method, the yield ratio of each component is plotted in Fig. 13. As

14
351 for the yield ratio of the steel web αs (Fig. 13(a)), the FE results show that both the tensile strength f1 and

352 shear strength f2 have significant effects on αs. When the interface strength is large enough in both tensile and

353 shear direction, the αs value no longer increases with increasing interface strength. Therefore, it is

354 recommended to ensure both the tensile strength and shear strength of the interface in design. For the

355 concrete, the yield ratio of the new concrete αnc and old concrete changed slightly with increasing interface

356 strength (Fig. 13(c)-(d)). As shown in Fig. 13(b), when the interface strength changed significantly, the

357 change of yield ratio of concrete αc was negligible. These results together indicate that the effect on the

358 interface is mainly due to the contribution of the steel web instead of concrete. In addition, the old and new

359 concrete could be treated integrally in developing the simplified design formula.

360 Simplified Design Method

361 In this section, 1296 three-dimensional FE models are calculated in MSC.Marc (2015) to develop

362 simplified design formula and identify the effect of the aforementioned parameters. The parameters

363 investigated are the shear span to depth ratio λ, average concrete strength fc', interface tensile strength f1,

364 interface shear strength f2, reinforcement ratio ρl, and steel content ratio ρs. In FE parametric analysis, the

365 concrete strength considered ranged from 25 MPa to 60 MPa, the interface tensile strength ranged from 0.35

366 MPa to 2.1 MPa and the interface shear strength from 2.5 MPa to 7.5 MPa. As shown in Fig. 14, the range of

367 parameters covered a wide range in design practice. After comparing the capacity of the proposed elaborate

368 FE model and the bending capacity of the specimen, 394 models are identified as bending failure specimens

369 and 902 specimens are identified as shear-critical specimens. After using the regression method, the

370 predicted shear capacity is formulated as follows, which reflects the effects of key parameters in the previous

371 section:

0.213 0.0194 As
372 Vpre   ( f ocb1  f ncb2 )h0  (0.0746  )  f yl Al   min(60.9 f1  28.7 f 2  0.155 f y , f y ) (39)
  0.160    1.40

373 where f'oc and f'nc denote the cylinder strength of the old and new concrete, respectively; b1 and b2 denote the

374 width of the old and new concrete, respectively; fyl the rebar strength; f1 and f2 denote the interface tensile

375 and shear strength; fy denote the steel strength and As denotes the longitudinal rebar area.

376 Fig. 14 shows a comparison between the predicted Vpre values using Eq. (39) and FE results. The results

377 of the proposed equation agree well with the FE results. The proposed formula estimates the mean value of

15
378 the shear capacity instead of the lower bound. It satisfactorily predicts the 902 FE results in the parametric

379 analysis with standard deviation of only 7.7%, which meets the precision requirement for engineering design.

380 The Eq. (39) well reflects the following two mechanisms. First, the proposed formula treats the old and new

381 concrete integrally, because the yield ratio of the new concrete and old concrete were very close when λ

382 ranged from 2.3 to 3.6 as shown in Fig. 11(b). Second, Eq. (39) shows that the interface strength is related to

383 the contribution of steel web instead of concrete. When the interface strength is large enough, the ultimate

384 strength will remain constant as shown in Fig. 13(a). Based on data regression analysis, if the following Eq.

385 (40) is realized, the peeling failure was not observed in the FE results in the loading history and the interface

386 strength no longer affects the shear capacity of the SPSS beam.

387 60.9 f1  28.7 f 2  0.845 f y (40)

388 For a proper design of the interface between old and new concrete, based on the FE parametric analysis,

389 it is recommended to calculate the interface strength and ensure the Eq.(40) is satisfied.

390 Conclusions

391 In this study, the peeling failure is found to be the most dominant behavior in the SPSS beam based on test

392 observations. Therefore, the three dimensional FE model with cohesive elements is developed to simulate the

393 peeling failure behavior. The conclusions are as follows:

394 (i) The proposed strengthening technique can effectively improve the shear strength and stiffness of the

395 shear-critical RC beams. The test results show that the shear capacity was improved from 1.56 to 2.27 times

396 that of the reference RC beam.

397 (ii) The peeling failure of SPSS beam can be simulated through the developed three-dimensional

398 elaborate FE model in MSC.Marc (2015). The load–displacement curve and the shear capacity are

399 satisfactorily predicted in the test. For the peeling failure in the test, the developed FE model with cohesive

400 element is capable of capturing the failure mode of the interface between the old and new concrete

401 throughout the loading history.

402 (iii) Based on the developed FE model, the effect of key factors including shear span to depth ratio,

403 material strength, and interface strength are investigated individually. The interface strength influences the

404 ultimate capacity significantly.

16
405 (iv) Based on the FE analysis, a practical design formula for the ultimate shear capacity is proposed with

406 acceptable accuracy and the recommendation for the interface design is proposed.

407 Acknowledgements

408 The writers gratefully acknowledge the financial support of the National Natural Science Foundation of

409 China (grant number 51722808) and the Beijing Natural Science Foundation (grant number 8162023).

410 References

411 ACI (American Concrete Institute). (2008). “Guide for the design and construction of externally bonded FRP systems for

412 strengthening concrete structures.” 440.2 R-08, Farmington Hills, MI.

413 ACI (American Concrete Institute). (2014). “Building code requirements for structural concrete and commentary.” ACI

414 318M-14, Farmington Hills, MI.

415 Aprile, A., Spacone, E., and Limkatanyu, S. (2001). “Role of bond in RC beams strengthened with steel and FRP plates.”

416 ASCE J Struct Eng., 127(12), 1445-1452.

417 Arslan, G., Sevuk, F., and Ekiz, I. (2008). “Steel plate contribution to load-carrying capacity of retrofitted RC beams.”

418 Constr Build Mater., 22(3), 143-153.

419 Bai, Y., Nie, J. G., and Cai, C. S. (2008). “New connection system for confined concrete columns and beams. II:

420 theoretical modeling.” ASCE J Struct Eng., 134 (12): 1800-1809.

421 Barnes, R. A., and Geoffrey, C. M. (2006a) “Strengthening of reinforced concrete beams in shear by the use of externally

422 bonded steel plates: Part 1–Experimental programme.” Constr Build Mater., 20(6), 396-402.

423 Barnes, R. A., and Geoffrey, C. M. (2006b) “Strengthening of reinforced concrete beams in shear by the use of externally

424 bonded steel plates: Part 2–Design guidelines.” Constr Build Mater., 20(6), 403-411.

425 Carol, I., Pere, P. C., and López, C. M. (1997) “Normal/shear cracking model: application to discrete crack analysis.”

426 ASCE J Eng Mech., 123(8), 765-773.

427 Chen, J. F., and Teng, J. G. (2001) “Anchorage strength models for FRP and steel plates bonded to concrete.” ASCE J

428 Struct Eng., 127(7), 784-791.

429 CMC (China Ministry of Construction). (2010). “Code for design of concrete structures”. CMC 2010, Beijing. (in

430 Chinese).

431 Comité Euro-International du Béton-Fédération International de la Précontrainte (CEB-FIP). (2010). “CEB-FIP model

432 code 2010.” design code, Thomas Telford, London.

17
433 Cook, R. A., and Burtz, J. L. (2003). “Design guidelines and specifications for engineered grouts.” Final Report,

434 University of Florida, Gainesville, Florida.

435 Ding, F. X., Lu, D. R., Bai, Y., Gong, Y. Z., Yu, Z. W., Ni, M. and Li, W. (2018). “Behaviour of CFRP-confined

436 concrete-filled circular steel tube stub columns under axial loading.” Thin-Walled Structures., 125: 107-118.

437 Eligehausen, R., and Balogh, T. (1995) “Behavior of fasteners loaded in tension in cracked reinforced concrete.” ACI

438 Struct J., 92(3), 365-379.

439 Eligehausen, R., and Ronald, A. C., and Appl, J. (2006) “Behavior and design of adhesive bonded anchors.” ACI Struct

440 J., 103(6), 822-831.

441 Feng, P., Cheng, S., Bai, Y., and Ye, L. P. (2015). “Mechanical behavior of concrete-filled square steel tube with

442 FRP-confined concrete core subjected to axial compression.” Composite Structures., 123: 312-324.

443 Fuchs, W., Eligehausen. R., and Breen, J. E. (1995). “Concrete capacity design (CCD) approach for fastening to

444 concrete.” ACI Struct J., 92(1), 73-94.

445 Geubelle, P. H., and Baylor, J. S. (1998). “Impact-induced delamination of composites: a 2D simulation.” Compos Part

446 B-Eng., 29(5), 589-602.

447 Guo, Z. H. (2014). Principles of reinforced concrete, London, Butterworth-Heinemann., U.K.

448 Han, L. H., and Li, W. (2010). "Seismic performance of CFST column to steel beam joint with RC slab: Experiments."

449 ASCE J Struct Eng., 66(11), 1374-1386.

450 Kani, G. N. J. (1966). “Basic facts concerning shear failure.” ACI Journal Proceedings., 63(6), 675-692.

451 Li, Y. Y., Guo, B., and Liu, J. (2014). “Research on reinforced concrete beam enlarged cross section method experiment

452 and finite element simulation.” Applied Mechanics and Materials., 638, 208-213.

453 Liu, X. G., Fan, J. S., Bai, Y., Tao, M. X. and Nie, J. G. (2014). “Stress increment of unbonded prestressing tendons in

454 prestressed concrete girders with corrugated steel webs.” ASCE J Bridge Eng., 04014094.

455 MSC. Marc. (2015). [Computer software]. MSC. Software Corp, Santa Ana, CA.

456 Maekawa, K., Okamura, H., and Pimanmas, A. (2003). “Non-linear mechanics of reinforced concrete.” London; Spon

457 Press, U.K.

458 Miyamoto, A., Tei, K., Nakamura, H., and Bull, J. W. (2000). “Behavior of prestressed beam strengthened with external

459 tendons.” ASCE J Struct Eng. 126(9), 1033-1044.

460 Nie, J. G., Bai, Y. and Cai, C. S. (2008). “New connection system for confined concrete columns and beams. I:

461 experimental study.” ASCE J Struct Eng., 134 (12): 1787-1799.

18
462 Nie, J. G., and Zhao, J. (2008). “Experimental study on simply supported RC beams strengthened by steel plate-concrete

463 composite technique”. Journal of Building Structures., 29(5), 50-56. (in Chinese)

464 Raoof, M., El-Rimawi, J. A., and Hassanen, M. A. (2000). “Theoretical and experimental study on externally plated RC

465 beams." Eng Struct. 22(1), 85-101.

466 Rots, J. G. (1991). “Smeared and discrete representations of localized fracture”. Int J Fracture., 51(1), 45-59.

467 Sakai, J., and Kawashima, K. (2006). “Unloading and reloading stress–strain model for confined concrete.” ASCE J.

468 Struct. Eng. 132(1), 112-122.

469 Shi, G., Liu, Z., Bai, Y., Shi, Y. J., and Wang Y. Q. (2012). “In-plane bending of laminated glass fin strengthened through

470 external bonding.” Advances in Structural Engineering, 15(1): 55-64.

471 Tao, M. X., and Nie, J. G. (2015). “Element mesh, section discretization and material hysteretic laws for fiber

472 beam–column elements of composite structural members.” Mater Struct., 48(8): 2521-2544.

473 Tao, M. X., and Nie, J. G. (2016). “Multi-scale modeling for deformation mechanism analysis of composite joint

474 substructures.” Eng Struct. 118, 55-73.

475 Wu, Z., Matsuzaki, T., Yokoyama, K., and Kanda, T. (1999). “Retrofitting method for reinforced concrete structures

476 with externally prestressed carbon fiber sheets.” ACI special publication., 188: 751-766.

477 Zamora, N. A., Cook, R. A., Konz, R. C., and Consolazio, G. R. (2003). “Behavior and design of single, headed and

478 unheaded, grouted anchors under tensile load”. ACI Struct J., 100(2): 222-230.

19
Table

1 Table

2 Table 1 Test design and material properties (in units of MPa and mm)

No Specimen λ a B H ts Stud Grouted anchor fyl fyv fy fcu fc'

1 SS1 1.11 450 460 480 3 D6L45@80 D12L120@80 432 285 338 44.0 35.2
2 SS2 1.84 750 460 480 3 D6L45@80 D12L120@80 432 285 338 45.1 36.1
3 SS3 2.33 950 460 480 3 D6L45@80 D12L120@80 432 285 338 46.6 37.3
4 SS4 3.56 1450 460 480 3 D6L45@80 D12L120@80 432 285 338 48.1 38.6
5 SS5 1.11 450 460 480 4 D6L45@80 D12L120@80 432 285 303 42.7 34.2
6 SS6 1.84 750 460 480 4 D6L45@80 D12L120@80 432 285 303 48.0 38.4
7 RC1 1.11 450 300 480 / / / 432 285 / 34.7 27.7
8 RC2 1.84 750 300 480 / / / 432 285 / 34.7 27.7
9 RC3 2.33 950 300 480 / / / 432 285 / 34.7 27.7
10 RC4 3.56 1450 300 480 / / / 432 285 / 34.7 27.7

3 Note: λ = shear span to depth ratio; a = shear span length; B = beam width; H = beam height; ts= steel plate thickness;
4 fyl = yield strength of longitudinal rebar; fyv= yield strength of stirrup; fy = yield strength of steel web; fcu = cubic strength
5 of concrete; fc' = cylinder strength of concrete. For the stud and grouted anchor, “D” denotes the diameter, “L” the length,
6 and “@” the spacing.

8 Table 2 Test strength and force components in FE model

No Specimen λ Failure pattern V(kN) β V/VRC VFE(kN) VFE/V Vs(kN) Vnc(kN) Voc(kN)

1 SS1 1.11 Peeling failure 1810 64% 1.69 1631 0.90 80 331 1254

2 SS2 1.84 Peeling failure 1129 66% 1.63 1107 0.98 87 252 780

3 SS3 2.33 Peeling failure 937 70% 1.80 890 0.95 87 214 587

4 SS4 3.56 Peeling failure 657 75% 2.27 631 0.96 65 138 426

5 SS5 1.11 Peeling failure 1855 65% 1.74 1645 0.89 72 327 1293

6 SS6 1.84 Peeling failure 1082 64% 1.56 1107 1.02 74 235 741

7 RC1 1.11 Shear compression 1069 44% / / / / / /

8 RC2 1.84 Shear compression 693 48% / / / / / /

9 RC3 2.33 Shear tension 522 46% / / / / / /

10 RC4 3.56 Shear tension 289 39% / / / / / /

9 Note: V = shear strength; β =Mu/Mflex; Mu =maximum moment of test; Mflex = bending capacity; VRC = shear strength of
10 reference RC beam; VFE = FE results of shear strength; Vs = shear force contribution of steel web; Vnc = shear force
11 contribution of new concrete; Voc = shear force contribution of old concrete.

1
Fig.1
Grouted anchor
Beam bottom Interface
New concrete

Rebar Old RC beam


Stud Stud
Steel plate
Grouted
anchor
Steel web Steel web

(a) (b) (c)


Fig.2
Actuator
S N
Steel plate Spreader beam Steel plate

334 41
480
ϕ32

41 64
ϕ6@50 ϕ6@200
200 a=450;750; 100 1300 100 a=450;750; 200
950;1450 950;1450
(a)
ε90 Strain rosette on steel
ε45 Actuator
Strain gauge on stirrup
ε0 Spreader beam Displ. meter

S1-S5 N1-N5

480
Steel plate

200 a=450;750; 100 1300 100 a=450;750; 200


950;1450 950;1450
(b)
ts=3 ts=4

480
480
480

300 80 300 80 80 300 80


RC1-4 SS1-4 SS5-6
(c)
Fig.3

(a) (b) (c)


Fig.4
Actuator
z (Loading on old concrete)

Symmetry
x y boundary: uy=0
Voc
Vnc

Vs

Symmetry
boundary: ux=0
Old concrete

Steel plate

Section for shear Vs:shear force on steel web


Cohesive element force components Vnc:shear force on new concrete
Support: uz=0
New concrete (Only on old concrete) Voc:shear force on old concrete
Fig.5
σ

f2

Shear
f1
Tensile

uc uc2 um1 um2


Fig.6

Shear
compression

(a)

Shear
compression

(b)
Diagonal
Cracking

(c)
Diagonal
Cracking

(d)

Peeling
failure

(e)

Peeling
failure

(f)
Fig.7
80%

70%
RC1-RC4
SS1-SS6
60%

β =Mu/Mflex
50%

40%

30%
1 1.5 2 2.5 3 3.5 4
λ
Fig.8

2000 1200 1000


V V
VFE V yield
Voc 1000 yield VFE 800 VFE
1500 Vnc Voc Voc
Vs 800 Vnc Vnc
VRC Vs 600 Vs
V/kN

V/kN

V/kN
1000 600 VRC VRC
yield 400
400 yield
500 yield yield
200 200

0 0 0
0 5 10 15 0 10 20 30 0 10 20 30
Δ/mm Δ/mm Δ/mm
(a) (b) (c)
800 2000 1200
V V V
VFE yield yield VFE yield VFE
Voc 1000
600 Voc Voc
Vnc 1500 Vnc
Vs Vs 800 Vnc
VRC VRC Vs

V/kN
VRC
V/kN
V/kN

400 1000 600


yield 400 yield
200 500
yield
200
0 0 0
0 10 20 30 0 5 10 15 20 0 10 20 30
Δ/mm Δ/mm Δ/mm
(d) (e) (f)
Fig.9
10%
SS1
SS2
SS3

Error=(Vs+Voc+Vnc-VFE)/VFE
5%
SS4
SS5
SS6
0%

-5%

-10%
0 5 10 15 20 25 30
Δ(mm)
Fig.10
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
(a) (b)

0.018
0.016
0.014
0.012
0.01
0.008
0.006
0.004
0.002
0
(c)
Fig.11
2000 50%
FE:
VFE αc
f'oc=f'nc=37.2 Mpa
1600 Voc αoc
λ=1.1~4.6 40%
f1=4.2 Mpa αnc
Vnc αs
f2=6 MPa
1200

Yield ratio
fy=338 MPa Vs 30%

V/kN 800 20%

400 10%

0%
0 1 1.5 2 2.5 3 3.5 4 4.5 5
1 1.5 2 2.5 3 3.5 4 4.5 5
λ λ
(a) (b)
Fig.12
2000 70%

60% FE:
1600 V αc
f'oc=f'nc=37.2 MPa
Voc
Vnc f1=4.2 MPa αoc
50% f2=6 MPa
Vs αnc

Yield ratio
1200
fy/f'c1=4~9

V/kN
40% αs
λ=1.1
800
30%

400 20%

0 10%
4 4.5 5 5.5 6 6.5 7 7.5 8 8.5 9 4 4.5 5 5.5 6 6.5 7 7.5 8 8.5 9
r=fy/f'c r=fy/f'c
(a) (b)
Fig.13
0.50
4 4
0.22
3.5 0.45 3.5
0.218
3 3
0.40
f1/MPa

f1/MPa
0.216
2.5 2.5
0.214
2 0.35 2
0.212
1.5 1.5
0.30
0.21
1 1

2 3 4 5 6 7 2 3 4 5 6 7
f2/MPa f2/MPa
(a) (b)
0.18 0.256
4 4
0.17 3. 0.254
3.5
5
0.252
3 0.16 3
0.250

f1/MPa
f1/MPa

0.15 2.
2.5
5 0.248
2 0.14 2
0.246
1.
1.5 0.13
5 0.244
1 0.12 1 0.242
2 3 4 5 6 7 2 3 4 5 6 7
f2/MPa f2/MPa
(c) (d)
Fig.14
3000

-15% error
2500

2000 +15% error

VFE/kN
1500

902 FE models
1000 λ=1.1;1.8;2.3;3.6
f'nc=f'oc=25;40;60MPa
fl=0.35;0.7;1.4;2.1MPa
500 f2=2.5;5;7.5MPa
ρl=3%;6%;9%
ρs=0.67%;1.33%;2%
0
0 500 1000 1500 2000 2500 3000
VPre/kN
Figure Captions List

Figure Caption List

Fig. 1 Schematic plot of the steel plate concrete composite strengthening method: (a) bending strengthening; (b) shear

strengthening of full span; (c) shear strengthening of shear span

Fig. 2 Design and dimensions (in mm) of RC beams and strengthened beam: (a) elevation view; (b) test measurement;

(c) cross-section view

Fig. 3 Photos of manufacturing the specimens: (a) drill holes; (b) embed anchors; (c) steel plates as framework

Fig. 4 Modeling scheme and boundary conditions in MSC.Marc

Fig. 5 Uniaxial stress–displacement skeleton curve of grouted anchor

Fig. 6 Comparison of test failure modes: (a) RC1; (b) RC2; (c) RC3; (d) RC4; (e) SS3; (f) SS4

Fig. 7 β results of SPSS beams and RC beams

Fig. 8 Shear force-displacement curve of test results and FE model: (a) SS1; (b) SS2; (c) SS3; (d) SS4; (e) SS5; (f) SS6;

Fig. 9 Errors of Eq. (34) for the FE results

Fig. 10 Failure mode of SS1 specimen at 0.8Pu in descending branch: (a) observed interface failure; (b) FE obtained

interface damage; (c) equivalent plastic strain of old concrete

Fig. 11 Effect of λ on the shear force contribution of concrete and steel: (a) force components; (b) yield ratio

Fig. 12 Effect of material strength on the shear contributions of concrete and steel: (a) force components; (b) yield ratio

Fig. 13 Effect of interface strength f1 and f2 by FE analysis: (a) yield ratio of steel web αs; (b) shear ratio of concrete αc; (c)

shear ratio of new concrete αnc; (d) shear ratio of old concrete αnc

Fig. 14 Comparison of predicted capacity by Eq. (39) and FE results

View publication stats

You might also like