Magazine of Concrete Research Magazine of Concrete Research, 2014, 66(10), 514–530

Volume 66 Issue 10 http://dx.doi.org/10.1680/macr.13.00310

Paper 1300310
A new empirical model for shear strength of Received 05/11/2013; revised 21/01/2014; accepted 23/01/2014
reinforced concrete beam–column Published online ahead of print 04/04/2014

connections ICE Publishing: All rights reserved

Tran, Hadi and Pham

A new empirical model for shear

strength of reinforced concrete
beam–column connections
Tung M. Tran Thong M. Pham
PhD Scholar, School of Civil, Mining and Environmental Engineering, PhD Candidate, School of Civil, Mining and Environmental Engineering,
University of Wollongong, Wollongong, NSW, Australia; Lecturer, University of Wollongong, Wollongong, NSW, Australia; Lecturer, Faculty
Department of Civil Engineering, Ton Duc Thang University, Ho Chi Minh, of Civil Engineering, HCMC University of Technology, Ho Chi Minh,
Vietnam Vietnam
Muhammad N. S. Hadi
Associate Professor, School of Civil, Mining and Environmental
Engineering, University of Wollongong, Wollongong, NSW, Australia

A new empirical model to estimate the joint shear strength of both exterior and interior beam–column connections
is proposed. In the model, four parameters that have the most influence on joint shear strength are considered.
Among these four, a new parameter is introduced to consider the bond condition and the possibility of beam bars
transferring joint shear force into the columns. Consideration of this parameter in the model significantly improves
the accuracy of the predicted joint shear strength. To calibrate the model, a large database of 98 reinforced concrete
(RC) exterior and 73 RC interior beam–column connections displaying joint failure mode was compiled from the
literature. A parametric study was also carried out to evaluate the dependence of the predicted to tested joint shear
strength ratio on the four influence parameters using the database. The proposed model showed superior perform-
ance over existing models. Moreover, comparisons of the predicted joint shear strength with experimental results
and with four existing models showed the accuracy of the proposed model.

Notation nb maximum number of the top and the bottom beam

Ajh effective joint area bars
Asb greater area of top or bottom beam bars Vch contribution of concrete strut on joint shear strength
Asjh , Asjv total area of horizontal and vertical shear Vjh horizontal joint shear strength
reinforcement respectively Vjh,model predicted joint shear strength
Astr effective area of the diagonal strut Vjh,test experimental shear strength
ai influence coefficients Vsh contribution of concrete truss on joint shear strength
BI beam reinforcement index xi input influence parameters
bb , bc width of the beam and the column sections Æ inclination of diagonal strut from the column
respectively longitudinal direction
bj effective joint width ÆNZS factor reflecting the influence of joint geometry and
c intercept value column axial load
dsb average diameter of beam tensile reinforcement Æt factor describing in-plane geometry
f c9 cylinder compressive strength of concrete  factor equal to 1.0 and 0.8 for interior and exterior
fjhy , fjvy yield strength of horizontal and vertical shear joints respectively
reinforcement respectively t factor describing out-of-plane geometry
hb height of beam cross-section ªACI joint shear factor in ACI 352R-02 (ACI, 2002)
hc height of column cross-section ª1 , ª2 ª1 ¼ 0.81, ª2 ¼ 0.14 for interior joints; ª1 ¼ 0.34,
JI joint transverse reinforcement index ª2 ¼ 0.22 for exterior joints
K factor of horizontal and vertical joint shear  softening coefficient
reinforcement t parameter to account for the influence of beam
k total number of collected experimental results eccentricity
k1 , k2 joint shear factors Ł inclination of the diagonal compression strut from
N column axial load the beam longitudinal direction
n number of influence parameters rsb beam reinforcement ratio

514
Magazine of Concrete Research A new empirical model for shear strength
Volume 66 Issue 10 of reinforced concrete beam–column
connections
Tran, Hadi and Pham

y column axial stress semi-empirical models have also been developed based on
b beam bar index experimental data.
j normalised joint shear reinforcement
jh normalised horizontal joint shear reinforcement Most of the theoretical and empirical models currently available
jv normalised vertical joint shear reinforcement in the literature were summarised by Lima et al. (2012). Their
summary showed that, in total, 11 parameters (including geo-
Introduction metric and mechanical parameters) were taken into account by
Beam–column connections are known to be one of the most the available capacity shear strength models. Interestingly, these
critical regions of reinforced concrete (RC) structures as failure 11 parameters did not include the number and diameter of beam
of the connections under seismic loading often leads to partial or longitudinal bars, although these factors are the most important
total collapse of the whole structure. To improve the safety of RC parameters that control the mechanism of bond forces transferring
structures under seismic load, designers have to carefully consider from the beam reinforcement to the concrete at the joint area.
the shear strength and the ductility performance of beam–column Analysis of a large experimental database of 171 beam–column
connections to ensure that brittle shear failure at the joint region connections introduced in the following section showed that, in
is avoided. addition to parameters such as concrete compressive strength,
joint shear reinforcement and column axial stress, the diameter
To understand the behaviour of beam–column connections, and number of beam bars are important factors for the shear
numerous experimental and analytical studies have been con- strength of a beam–column connection. Therefore, in this work,
ducted since the mid-1960s. The first studies on beam–column the number and diameter of beam bars were considered in order
connections were carried out by Hanson and Connor (1967). to develop an empirical shear strength model for predicting the
These were then developed by Zhang and Jirsa (1982), Sarsam shear strength of beam–column connections. Development of the
and Phipps (1985), Pantazopoulou and Bonacci (1992), Hwang model was based on regression analysis using a large database
and Lee (1999). The latest studies on beam–column connections collected from published works. The superiority of the proposed
include those of Kim et al. (2009), Choi and Kim (2011), Joyklad model was evaluated by comparing the predicted joint shear
et al. (2012), Kim and Yu (2012), Najafian et al. (2013) and Patel strengths with 171 test results from the literature and with four
et al. (2013). Although many efforts have been made, the existing analytical models.
research community has not yet understood the full behaviour of
RC beam–column connections because of the large variations in Experimental database
both geometry and distribution of forces that occur in a relatively A database of 171 experimental RC beam–column connections
small volume at the joint region (Pantazopoulou and Bonacci, (98 exterior and 73 interior) was compiled from the published
1992). This point of view is evident when considering the literature (Alva et al., 2007; Antonopoulos and Triantafillou,
inconsistencies in existing design standards for predicting the 2003; Attaalla, 2004; Au et al., 2005; Chalioris et al., 2008;
shear strength of RC beam–column connections (ACI, 2008; AIJ, Chun et al., 2009; Chutarat and Aboutaha, 2003; Clyde et al.,
1999; BSI, 2004; SANZ, 1995). 2000; Dhakal et al., 2005; Durrani and Wight, 1985; Ehsani and
Alameddine, 1991; Ehsani and Wight, 1985; Ehsani et al., 1987;
In general, to carry the joint shear forces, resisting mechanisms, El-Amoury and Ghobarah, 2002; Fisher and Sezen, 2011;
including diagonal struts and/or trusses, are developed. The truss Ghobarah and El-Amoury, 2005; Ghobarah and Said, 2001, 2002;
mechanism is justified when the bond between the concrete and Hwang et al., 2004, 2005; Ishibashi, 1993; Kaku and Asakusa,
the beam and column reinforcement is perfect. In this case, the 1991; Karayannis and Sirkelis, 2005, 2008; Karayannis et al.,
joint core is considered as a uniform plane zone subjected to 2008; Kitayama et al., 1991; Kuang and Wong, 2006; Le-Trung
shear stress and joint shear reinforcement is thus required to et al., 2010; Lee et al., 2010; Leon, 1990; Liu, 2006; Lu et al.,
prevent shear failure of the joint core by diagonal principal 2012; Megget, 1974; Meinheit and Jirsa, 1977; Morita et al.,
tension stress. The diagonal strut mechanism is developed by the 1999; Murty et al., 2003; Noguchi and Kashiwazaki, 1992;
internal forces generated in the concrete and thus joint shear Noguchi and Kurusu, 1988; Oka and Shiohara, 1992; Otani et al.,
reinforcement is required to provide sufficient confinement for 1984; Pantelides et al., 2002; Park and Paulay, 1974; Shrestha et
improving the compressive strength of the concrete diagonal al., 2009; Supaviriyakit and Pimanmas, 2008; Tsonos, 2007;
strut(s). Different from the truss mechanism, in the diagonal strut Tsonos et al., 1992; Vatani-Oskouei, 2010; Wang and Hsu, 2009;
mechanism, the bond between the beam bars and the concrete is Wong and Kuang, 2008).
allowed to deteriorate. Based on the above resisting mechanisms,
numerous models have been developed for predicting the shear All of the specimens were subjected to quasi-static cyclic lateral
strength of beam–column connections. These theoretical models loading and were at least one-third scale. The final failure modes
were developed based on either the average stress approach with of the collected specimens were either joint shear or joint shear
compatibility of strains and stress equilibrium, or the strut-and-tie with yielding of beam reinforcement. All of the specimens had
approach. In addition, because the mechanism of the joint is no out-of-plane members (slabs and/or transverse beams) and no
complicated and depends on many parameters, empirical and eccentricity between beams and columns. Specimens that failed

515
Magazine of Concrete Research A new empirical model for shear strength
Volume 66 Issue 10 of reinforced concrete beam–column
connections
Tran, Hadi and Pham

in terms of weak column–strong beam were excluded from the on the joint shear strength are evaluated in the following
collected data because, in these cases, the column flexural sections.
capacity is relatively low and thus failure of the column caused
by flexural moment can occur before the shear strength of the Parameters of influence in beam–column
connection is reached. For interior connections, the column and connections
beam reinforcement continuously passed through the joint panel. The ‘key’ influence parameters for the joint shear behaviour of
For exterior joints, only those specimens with conventional rein- beam–column connections were investigated by Kim and LaFave
forcement anchorage were included (i.e. the longitudinal bars of (2007). They found that concrete compressive strength f c9 , in-
the beams were anchored by hooks towards the core of the plane geometry (interior, exterior or knee connections), dimen-
exterior joints). In the collected database, the test shear forces, sions of the beams and columns (hb , bb , hc , bc ), joint transverse
Vjh,test , were either collected from the reported values or derived reinforcement and beam reinforcement were among the influence
using the maximum applied load measured from the test. In this parameters for the shear strength of beam–column connections.
calculation, the moment arm of the beam cross-section was Besides the above parameters, bond condition – which is strongly
assumed to be 80% of the total beam height hb : influenced by the number and diameter of reinforcement bars –
and column axial stress are also known to affect joint shear
The collected database covered a broad range of various para- strength.
meters, including joint reinforcement ratio, reinforcement yield-
ing stress, concrete strength, column axial average stresses and Beam longitudinal bars passing or anchored in a joint core
beam height to column height ratio (hb /hc ), as summarised in transfer a fraction of bond stress into the joint core and the
Table 1 for exterior joints and Table 2 for interior joints. Of the remainder into the upper and lower columns. When the bond
exterior connections, 46 specimens failed through joint shear and force of the beam bars is low, the fraction of bond force
52 specimens failed in joint shear with yielding of beam rein- transferred into the columns is low, thus most of the shear force
forcement; the corresponding numbers for the interior connec- is claimed by the joint and slippage of the beam bars also occurs.
tions were 41 and 32 respectively. The combination of these two unfavourable factors leads to a
reduction in joint shear strength. Conversely, if the beam bars’
The definitions and the ranges of the main parameters in the bond force is high, the fraction of bond force transferred into the
collected database are summarised in Table 3, in which N is the columns is significant; thus, a relatively small fraction of shear
column axial load, Asjh and fjhy are the area and yield strength of force is claimed by the joint and slippage of the beam bars would
the horizontal joint shear reinforcement placed between the top not occur, leading to an improvement in joint shear strength.
and the bottom beam reinforcement respectively, Asjv and fjvy are
the area and yield strength of the intermediate vertical reinforce- Among the parameters of influence, concrete compressive
ment passing through the joint respectively, bj ¼ (bc + bb )/2 is strength, in-plane geometry and the dimensions of the beam and
the effective joint width (bc and bb are the section widths of the column are the strongest, so their roles in joint shear strength
beam and the column respectively), nb is the maximum number have been mostly evaluated. The remaining parameters are still
of the top and the bottom beam bars and dsb is the corresponding being debated. Some researchers (Bakir and Boduroğlu, 2002;
average beam bar diameter. Hegger et al., 2003; Kim and LaFave, 2008; Parker and Bullman,
1997; Paulay and Priestley, 1992; Sarsam and Phipps, 1985;
A new parameter b , referred to as the beam bar index, is Vollum and Newman, 1999) have proposed inconsistent contribu-
proposed. From its definition in Table 3, it can be seen that b is tions of joint shear reinforcement in the shear strength of beam–
a function of the number of beam bars nb and their average column joints while some (Marques and Jirsa, 1975; Meinheit
diameter dsb : In addition, geometrical properties are also inte- and Jirsa, 1977; Pantazopoulou and Bonacci, 1992) have indi-
grated into b to form a dimensionless parameter reflecting the cated that column axial load has no coherent effect on joint shear
normalised contact area between the beam reinforcement and the strength. Clyde et al. (2000) reported that column axial load
surrounding concrete. As the concrete–reinforcement contact helped to improve the joint shear strength while Park and
area has a direct influence on the magnitude of the bond forces Mosalam (2012) showed that joint shear strength is not clearly
transferring from the beam reinforcement into the concrete at affected by column axial stress up to 0:2 f c9 :
the joint core, b is proposed as a parameter that can affect the
joint shear strength. By considering its definition, it is easy to Models for shear strength of beam–column
see that the joint aspect ratio (hb /hc ), which is considered as an connections
influence factor of the joint shear strength, is integrated in b : As mentioned earlier, strut and truss mechanisms are developed
Moreover, the conventional influence factor, the beam reinforce- to resist joint shear forces, thus the joint shear strength Vjh is
ment ratio (rsb ¼ Asb /bb hb , where Asb is the greater of area of usually proposed to be composed of two components
the top or bottom beam bars) is also incorporated to a certain
degree via the term nb dsb /bb hb in this parameter. Details about 1: V jh ¼ V ch þ V sh
the influence of both the beam bar index and other parameters

516
Magazine of Concrete Research A new empirical model for shear strength
Volume 66 Issue 10 of reinforced concrete beam–column
connections
Tran, Hadi and Pham

Specimen bc : mm hc : mm bb : mm hb : mm f c9 : MPa N nb dsb : mm Vjh,test : Vjh,model : V jh,model Failure

bc hc f c9 kN kN V jh,test modea

Ehsani and Alameddine (1991)

HL8 356 356 318 508 56 0.07 4 29 986 869 0.88 J
LL11 356 356 318 508 74 0.03 4 25 768 865 1.13 J
HL11 356 356 318 508 74 0.06 4 29 967 954 0.99 J
Wong and Kuang (2008)
BS-L-300 300 300 260 300 34 0.15 3 20 505 374 0.74 J
BS-L-450 300 300 260 450 31 0.15 3 20 316 313 0.99 J
BS-L-600 300 300 260 600 36 0.15 3 20 284 316 1.11 J
BS-L-V2 300 300 260 450 33 0.15 3 20 399 355 0.89 J
BS-L-V4 300 300 260 450 28 0.15 3 20 403 367 0.91 J
BS-L-H1 300 300 260 450 33 0.15 3 20 389 342 0.88 J
Tsonos (2007)
E1 200 200 200 300 22 0.18 3 14 232 195 0.84 J
G2 200 200 200 300 22 0.18 3 14 222 186 0.83 J
Clyde et al. (2000)
Test 2 305 457 305 406 46 0.10 4 29 947 946 1.00 J
Test 4 305 457 305 406 41 0.25 4 29 982 1029 1.05 J
Test 5 305 457 305 406 37 0.25 4 29 941 982 1.04 J
Test 6 305 457 305 406 40 0.10 4 29 927 886 0.96 J
Kuang and Wong (2006)
BS-L 300 300 260 450 31 0.14 3 20 316 307 0.97 J
BS-U 300 300 260 450 31 0.14 3 20 341 308 0.90 J
Tsonos et al. (1992)
S6’ 200 200 200 300 29 0.40 4 14 202 239 1.18 J
Pantelides et al. (2002)
Test Unit 1 406 406 406 406 33 0.10 4 29 872 733 0.84 J
Test Unit 2 406 406 406 406 30 0.25 4 29 833 836 1.00 J
Test Unit 3 406 406 406 406 34 0.10 4 29 826 743 0.90 J
Test Unit 4 406 406 406 406 32 0.25 4 29 927 855 0.92 J
Test Unit 5 406 406 406 406 32 0.10 4 29 770 717 0.93 J
Test Unit 6 406 406 406 406 31 0.25 4 29 851 847 1.00 J
Chalioris et al. (2008)
JA-0 200 300 200 300 34 0.05 4 12 218 255 1.17 J
JB-0 200 300 200 300 32 0.05 6 10 201 251 1.25 J
JB-s1 200 300 200 300 32 0.05 6 10 219 258 1.18 J
JCa-0 100 200 100 200 21 0.10 3 8 66 66 1.01 J
JCb-0 100 200 100 200 23 0.10 3 10 84 76 0.90 J
JCb-s1 100 200 100 200 23 0.10 3 10 97 83 0.86 J
JCb-s2 100 200 100 200 23 0.10 3 10 88 91 1.03 J
Karayannis and Sirkelis (2008)
B0 200 200 200 300 32 0.05 6 10 199 141 0.71 J
B1 200 200 200 300 32 0.05 6 10 215 153 0.71 J
C0 200 200 200 300 32 0.05 4 12 209 150 0.72 J
Alva et al. (2007)
LVP4 200 300 200 400 25 0.15 4 16 327 360 1.10 J
Antonopoulos and Triantafillou (2003)
C1 200 200 200 300 16 0.12 3 14 116 98 0.85 J
C2 200 200 200 300 19 0.10 3 14 115 105 0.91 J
S-C 200 200 200 300 15 0.12 3 14 123 103 0.84 J

Table 1. Database of 98 exterior joints (continued on next page)

517
Magazine of Concrete Research A new empirical model for shear strength
Volume 66 Issue 10 of reinforced concrete beam–column
connections
Tran, Hadi and Pham

Specimen bc : mm hc : mm bb : mm hb : mm f c9 : MPa N nb dsb : mm Vjh,test : Vjh,model : V jh,model Failure

bc hc f c9 kN kN V jh,test modea

El-Amoury and Ghobarah (2002)

T0 250 400 250 400 31 0.20 4 20 420 536 1.28 J
Ghobarah and Said (2001)
T2 250 400 250 400 31 0.10 4 20 502 483 0.96 J
Ghobarah and El-Amoury (2005)
T-BS3 250 400 250 400 30 0.20 4 20 421 534 1.27 J
Shrestha et al. (2009)
UC1 300 300 300 450 26 0.08 4 24 293 304 1.04 J
Fisher and Sezen (2011)
B-1-RC 152 152 152 203 30 0.00 3 13 130 109 0.84 J
Murty et al. (2003)
Q1 200 250 200 400 26 0.00 2 20 156 156 1.00 J
R1 200 250 200 400 30 0.00 2 20 173 167 0.97 J
S1 200 250 200 400 28 0.00 2 20 163 161 0.99 J
Ehsani and Wight (1985)
1B 300 300 259 480 34 0.06 6 21 575 510 0.89 BJ
2B 300 300 259 439 35 0.07 6 21 587 540 0.92 BJ
5B 340 340 300 480 24 0.13 6 22 679 625 0.92 BJ
Ehsani et al. (1987)
4 300 300 259 439 67 0.05 5 20 736 691 0.94 BJ
Ehsani and Alameddine (1991)
LL8 355.6 355.6 317.5 508 56 0.04 4 25 860 790 0.92 BJ
LH8 355.6 355.6 317.5 508 56 0.04 4 25 837 861 1.03 BJ
HH8 355.6 355.6 317.5 508 56 0.07 4 29 985 939 0.95 BJ
HH11 355.6 355.6 317.5 508 74 0.06 4 29 1020 1027 1.01 BJ
LL14 355.6 355.6 317.5 508 94 0.02 4 25 877 936 1.07 BJ
LH14 355.6 355.6 317.5 508 94 0.02 4 25 890 1005 1.13 BJ
HH14 355.6 355.6 317.5 508 94 0.04 4 29 1032 1092 1.06 BJ
Hwang et al. (2005)
0T0 420 420 320 450 67 0.02 4 25 997 1048 1.05 BJ
3T3 420 420 320 450 69 0.02 4 25 1132 1124 0.99 BJ
2T4 420 420 320 450 71 0.02 4 25 1080 1155 1.07 BJ
1T44 420 420 320 450 73 0.02 4 25 1039 1166 1.12 BJ
Wong and Kuang (2008)
BS-L-H2 300 300 260 450 42 0.15 3 20 479 399 0.83 BJ
Chalioris et al. (2008)
JCa-s1 100 200 100 200 21 0.10 3 8 71 73 1.04 BJ
JCa-s2 100 200 100 200 21 0.10 3 8 71 81 1.14 BJ
Chun et al. (2009)
JC-2 500 500 350 500 60 0.03 8 22 1320 1635 1.24 BJ
JC-No.11 650 520 450 505 31 0.00 3 36 1179 1279 1.08 BJ
Karayannis and Sirkelis (2008)
A1 200 200 200 300 36 0.05 2 10 76 112 1.48 BJ
A2 200 200 200 300 36 0.05 2 10 74 112 1.51 BJ
Karayannis et al. (2008)
A0 200 200 200 300 32 0.05 2 10 83 105 1.26 BJ
C2 200 200 200 300 32 0.05 4 12 209 175 0.84 BJ
Hwang et al. (2004)
28–0T0 550 550 380 500 33 0.02 4 25 1138 1356 1.19 BJ

Table 1. Database of 98 exterior joints (continued on next page)

518
Magazine of Concrete Research A new empirical model for shear strength
Volume 66 Issue 10 of reinforced concrete beam–column
connections
Tran, Hadi and Pham

Specimen bc : mm hc : mm bb : mm hb : mm f c9 : MPa N nb dsb : mm Vjh,test : Vjh,model : V jh,model Failure

bc hc f c9 kN kN V jh,test modea

Alva et al. (2007)

LVP2 200 300 200 400 44 0.15 4 16 514 438 0.85 BJ
LVP3 200 300 200 400 24 0.15 4 16 364 383 1.05 BJ
LVP5 200 300 200 400 26 0.15 4 16 380 392 1.03 BJ
Liu (2006)
Unit RC-1 230 230 200 330 19 0.07 6 10 140 143 1.03 BJ
Karayannis and Sirkelis (2005)
AJ1s 200 200 200 300 33 0.05 2 10 87 120 1.39 BJ
Le-Trung et al. (2010)
NS 167 167 134 200 37 0.00 4 10 120 96 0.80 BJ
SD 167 167 134 200 37 0.00 4 10 114 101 0.88 BJ
Ghobarah and Said (2002)
T1 250 400 250 400 31 0.19 4 20 527 537 1.02 BJ
Vatani-Oskouei (2010)
1st 350 350 350 400 24 0.00 4 18 400 463 1.16 BJ
2nd 350 350 350 400 20 0.00 4 18 425 429 1.01 BJ
Fisher and Sezen (2011)
C-2-RC 152 152 152 203 30 0.00 3 10 102 102 1.00 BJ
E-1-RC 152 152 152 203 30 0.00 3 12 113 107 0.95 BJ
Megget (1974)
UNIT A 330 380 255 460 22 0.07 3 28 547 548 1.00 BJ
Murty et al. (2003)
Q3 200 250 200 400 27 0.00 2 20 211 222 1.05 BJ
S3 200 250 200 400 30 0.00 2 20 198 229 1.16 BJ
Chutarat and Aboutaha (2003)
Specimen I 406 406 356 457 28 0.00 4 25 1040 846 0.81 BJ
Kaku and Asakusa (1991)
NO 3 220 220 160 220 42 0.00 4 13 217 208 0.96 BJ
NO 4 220 220 160 220 45 0.17 4 13 239 254 1.06 BJ
NO 5 220 220 160 220 37 0.09 4 13 221 210 0.95 BJ
NO 6 220 220 160 220 40 0.00 4 13 209 196 0.94 BJ
NO 9 220 220 160 220 41 0.00 4 13 236 232 0.98 BJ
NO 11 220 220 160 220 42 0.08 4 13 231 248 1.08 BJ
NO 12 220 220 160 220 35 0.00 4 13 207 210 1.01 BJ
NO 13 220 220 160 220 46 0.04 4 13 209 236 1.13 BJ
NO 14 220 220 160 220 41 0.08 4 13 226 226 1.00 BJ
NO 15 220 220 160 220 40 0.08 4 13 230 229 0.99 BJ
Park and Paulay (1974)
S4 330 381 254 457 21 0.00 2 29 317 369 1.17 BJ
Average 1.00
CoV 0.147
a
J, joint shear failure; BJ, joint shear failure with yielding of beam reinforcement

Table 1. (continued)

where Vch and Vsh are the contributions of the concrete strut and ACI 352R-02 (ACI, 2002) and Architectural Institute of Japan
truss mechanisms respectively. The first component is related to design guidelines AIJ (1999) ignore the contribution of Vsh , thus the
the concrete strength f c9 and the second term relates to the joint shear strength is expressed as a function of concrete compres-
horizontal and vertical joint reinforcement. sive strength and the joint geometry. The form of the expression is

519
Magazine of Concrete Research A new empirical model for shear strength
Volume 66 Issue 10 of reinforced concrete beam–column
connections
Tran, Hadi and Pham

Specimen bc : mm hc : mm bb : mm hb : mm f c9 : MPa N nb dsb : mm Vjh,test : Vjh,model : V jh,model Failure

bc hc f c9 kN kN V jh,test modea

Meinheit and Jirsa (1977)

I 330 457 279 457 26 0.40 3 25 1090 1223 1.12 J
II 330 457 279 457 42 0.25 3 25 1597 1477 0.92 J
III 330 457 279 457 27 0.39 3 25 1228 1448 1.18 J
IV 457 330 406 457 36 0.30 3 25 1455 1197 0.82 J
V 330 457 279 457 36 0.04 3 25 1530 1210 0.79 J
VII 457 330 406 457 37 0.47 3 25 1646 1594 0.97 J
XIII 330 457 279 457 41 0.25 3 25 1468 1361 0.93 J
XIV 457 330 406 457 33 0.32 3 25 1948 1528 0.78 J
Oka and Shiohara (1992)
J-2 300 300 240 300 81 0.11 8 13 1557 1528 0.98 J
J-10 300 300 240 300 39 0.12 9 13 1539 1199 0.78 J
J-11 300 300 240 300 39 0.12 9 19 516 506 0.98 J
Attaalla (2004)
SHC1 127 178 127 203 57 0.05 2 12 536 502 0.94 J
SHC2 127 178 127 203 60 0.04 2 12 576 517 0.90 J
SOC3 127 178 127 203 47 0.05 2 12 503 564 1.12 J
Kitayama et al. (1991)
J1 300 300 200 300 26 0.08 8 13 491 497 1.01 J
A1 300 300 200 300 31 0.06 8 13 840 911 1.08 J
Au et al. (2005)
E-0.0 300 300 250 300 41 0.00 4 16 853 931 1.09 J
H-0.0 300 300 250 300 41 0.00 4 16 629 807 1.28 J
E-0.3 300 300 250 300 40 0.35 4 16 1042 1190 1.14 J
Lee et al. (2010)
J10 400 400 300 600 27 0.19 3 25 1119 1306 1.17 J
Noguchi and Kashiwazaki (1992)
OKJ3 300 300 200 300 118 0.12 10 13 1075 1166 1.08 J
OKJ5 300 300 200 300 78 0.12 10 13 1204 1178 0.98 J
OKJ6 300 300 200 300 59 0.12 8 13 1111 1137 1.02 J
Dhakal et al. (2005)
C1PD 350 500 300 550 32 0.12 5 32 888 1139 1.28 J
C4PD 400 400 300 550 33 0.12 6 32 1261 1180 0.94 J
Morita et al. (1999)
No 1 350 350 250 350 22 0.31 4 25 795 873 1.10 J
No 2 350 350 250 350 22 0.31 4 25 941 857 0.91 J
No 4 350 350 250 350 23 0.30 4 25 156 186 1.19 J
No 5 350 350 250 350 22 0.31 7 16 164 193 1.18 J
No 6 350 350 250 350 22 0.31 7 16 163 174 1.07 J
Ishibashi (1993)
D19-S1 400 400 260 400 44 0.00 5 19 493 573 1.16 J
D19-S2 400 400 260 400 43 0.00 3 19 429 573 1.34 J
D25-S3 400 400 260 400 48 0.00 3 25 628 575 0.92 J
D25-S4 400 400 260 400 48 0.00 3 25 545 568 1.04 J
D29-S5 400 400 260 400 48 0.00 3 29 826 662 0.80 J
Wang and Hsu (2009)
Ko-JI1 300 300 300 500 32 0.14 4 25 750 679 0.91 J
Ho-JI1 400 400 300 400 26 0.00 4 19 869 763 0.88 J
Au et al. (2005)
E-0.0 300 300 250 300 37 0.00 4 16 814 855 1.05 J

Table 2. Database of 73 interior joints (continued on next page)

520
Magazine of Concrete Research A new empirical model for shear strength
Volume 66 Issue 10 of reinforced concrete beam–column
connections
Tran, Hadi and Pham

Specimen bc : mm hc : mm bb : mm hb : mm f c9 : MPa N nb dsb : mm Vjh,test : Vjh,model : V jh,model Failure

bc hc f c9 kN kN V jh,test modea

E-0.3 300 300 250 300 39 0.26 4 16 1038 991 0.96 J

H-0.0 300 300 250 300 43 0.00 4 16 1081 974 0.90 J
H-0.3 300 300 250 300 38 0.26 4 16 378 384 1.02 J
Meinheit and Jirsa (1977)
VI 330 457 279 457 37 0.48 3 25 414 451 1.09 BJ
XII 330 457 279 457 35 0.30 3 25 876 940 1.07 BJ
Otani et al. (1984)
J1 300 300 200 300 26 0.08 3 19 1180 1086 0.92 BJ
J2 300 300 200 300 24 0.08 3 19 1458 1328 0.91 BJ
J3 300 300 200 300 24 0.08 3 19 1180 1086 0.92 BJ
J4 300 300 200 300 26 0.23 3 19 1180 1113 0.94 BJ
J5 300 300 200 300 29 0.07 3 19 1031 964 0.94 BJ
Durrani and Wight (1985)
X1 362 362 279 419 34 0.05 4 22 798 635 0.80 BJ
X2 362 362 279 419 34 0.06 4 22 660 635 0.96 BJ
X3 362 362 279 419 31 0.05 3 22 844 630 0.75 BJ
Oka and Shiohara (1992)
J-1 300 300 240 300 81 0.11 9 13 737 596 0.81 BJ
J-4 300 300 240 300 73 0.13 10 13 1273 1436 1.13 BJ
J-5 300 300 240 300 73 0.13 9 13 1045 1203 1.15 BJ
J-6 300 300 240 300 79 0.12 9 13 871 988 1.14 BJ
J-7 300 300 240 300 79 0.12 7 13 822 683 0.83 BJ
J-8 300 300 240 300 79 0.12 9 19 920 993 1.08 BJ
Kitayama et al. (1991)
C1 300 300 200 300 26 0.08 12 10 865 991 1.15 BJ
B1 300 300 200 300 25 0.08 8 13 810 686 0.85 BJ
B3 300 300 200 300 25 0.08 8 13 1509 1318 0.87 BJ
Lu et al. (2012)
J1–1 400 400 250 400 30 0.04 4 20 973 1151 1.18 BJ
J1–4 400 400 250 400 30 0.04 3 25 1542 1272 0.82 BJ
Leon (1990)
BCJ2 254 254 203 305 30 0.00 4 13 1053 1272 1.21 BJ
BCJ3 254 305 203 305 27 0.00 4 13 1309 1322 1.01 BJ
Noguchi and Kashiwazaki (1992)
OKJ1 300 300 200 300 78 0.12 9 13 411 501 1.22 BJ
OKJ4 300 300 200 300 78 0.12 9 13 467 537 1.15 BJ
Noguchi and Kurusu (1988)
NO.1 300 300 200 300 34 0.06 12 10 484 560 1.16 BJ
NO.2 300 300 200 300 34 0.06 10 10 759 606 0.80 BJ
NO.3 300 300 200 300 34 0.06 6 13 893 868 0.97 BJ
NO.4 300 300 200 300 34 0.06 5 13 750 644 0.86 BJ
Supaviriyakit and Pimanmas (2008)
J1 200 350 175 300 26 0.11 6 12 814 803 0.99 BJ
J2A 200 350 175 300 29 0.10 6 12 869 770 0.89 BJ
J3B 200 350 175 300 24 0.12 6 12 744 854 1.15 BJ
Average 1.00
CoV 0.140
a
J, joint shear failure; BJ, joint shear failure with yielding of beam reinforcement

Table 2. (continued)

521
Magazine of Concrete Research A new empirical model for shear strength
Volume 66 Issue 10 of reinforced concrete beam–column
connections
Tran, Hadi and Pham

Definition Exterior joints Interior joints

Minimum Maximum Minimum Maximum

Concrete compressive strength f c9 : MPa 15 94 22 118

N
Normalised column axial stress 0.04 0.40 0.31 0.48
bc hc f c9
Asjh f jhy
Normalised horizontal joint shear reinforcement jh ¼ : 0.00 1.11 0.00 1.23
bj hc (f c9 )0 5

Asjv f jvy
Normalised vertical joint shear reinforcement jv ¼ : 0.00 1.63 0.00 3.20
bj hc (f c9 )0 5
Normalised joint shear reinforcement j ¼ jh þ jv 0.00 2.45 0.00 3.50
Column depth to beam height ratio hc /hb 0.50 1.13 0.67 1.00
Beam bar diameter to column depth ratio dsb /hc 0.03 0.09 0.03 0.08
nb d sb hc
Beam bar index b ¼ 0.07 0.51 0.13 0.71
bb hb
Table 3. Definitions and ranges of parameters in the collected
database

2: V jh ¼ k 1 k 2 ( f c9 )m Ajh Bakir and Boduroğlu (2002) proposed an empirical model in

which Vch depends on concrete strength, details of the beam
reinforcement and joint aspect ratio hb /hc : In their model, the
where k1 and k2 are joint shear factors that depend on the joint contribution of the truss mechanism is determined as Vsh ¼ Æj Asjh
geometry (in ACI 352R-02, k1 3 k2 is equal to the joint shear fjhy , where Æj depends on the joint shear reinforcement ratio. A
strength factor, ªACI ), Ajh is the effective joint area and m ¼ 0.5 similar form was proposed by Hegger et al. (2003) for determin-
in the ACI model and m ¼ 0.7 in the AJI model. ing Vsh , but the parameter Æj was supposed to be influenced by
the pattern of the beam bars being anchored and details of the
In NZS-3101 (SANZ, 1995), the contribution of the truss joint shear reinforcement. Different from other models, this
mechanism is somehow considered but a clear contribution of Vsh model accounted for the influence of the column reinforcement
on joint shear strength is not available. The formulation of joint ratio.
shear strength in NZS-3101 is
Using the Bayesian parameter estimation method, Kim et al.
f jhy Asjh (2009) developed an empirical joint shear strength model. The
V jh ¼ ( f 9 )m Ajh < 0:2( f c9 )Ajh
3: 6ÆNZS f by Asb c distinctive feature of this model is that the joint shear strength is
given in terms of multiplication of the influence parameters

in which m ¼ 1, fjhy and Asjh are the yield strength and the total
: : :
cross-sectional area of the horizontal joint transverse reinforce- 4: V jh ¼ 1:31Æt t t (JI)0 15 (BI)0 3 ( f c9 )0 75 Ajh
ment respectively, fby and Asb are the yield strength and the
greater of the area of top or bottom beam bars respectively and
ÆNZS is a factor that reflects the influence of both joint geometry in which Æt and t are parameters for describing in-plane and
and column axial load. out-of-plane geometry respectively, t is a parameter to account
for the influence of beam eccentricity, JI is the joint transverse
Paulay and Priestley (1992) proposed a theoretical joint shear reinforcement index (depending mostly on the volumetric joint
strength model in which the contribution of the concrete strut Vch shear reinforcement ratio) and BI is the beam reinforcement
depends on the area and yielding strength of the beam’s long- index (depending mostly on the beam reinforcement ratio rsb ). A
itudinal reinforcement, the concrete compressive strength, column limitation of this model is that column axial load was not taken
axial load and the joint geometry. The contribution of the truss into account as a possible influence parameter for joint shear
mechanism is simply determined as Vsh ¼ Asjh fjhy : A similar strength during development of the model. This is because, in the
contribution of the truss mechanism on joint shear strength was Bayesian parameter estimation model, all parameters should be
proposed by Vollum and Newman (1999), but their model non-zero (Kim and LaFave, 2008) and some experimental tests
additionally accounted for the influence of hb /hc : had zero column axial load.

522
Magazine of Concrete Research A new empirical model for shear strength
Volume 66 Issue 10 of reinforced concrete beam–column
connections
Tran, Hadi and Pham

Based on the strut-and-tie approach, Hwang and Lee (2002) and their diameters are thus examined instead of their total cross-
developed a theoretical joint shear strength model in which the sectional area and yield strength.
joint shear strength was determined as
In the proposed model, the general form of joint shear strength is
5: V jh ¼ K f c9 Astr cos Ł assumed to be a function of influence parameters as illustrated by

!
where K is a factor accounting for the contribution of horizontal X
n
: 0:5
and vertical joint shear reinforcement,  (¼ 3.35/( f c9 )0 5 < 0.52) V jh ¼ Ajh ( f c9 ) ai x i þ c
is a softening coefficient, Astr is the effective area of the diagonal 7: i¼1

strut depending on column axial load and dimensions of the joint

and Ł is the angle of inclination of the diagonal compression
strut with respect to the longitudinal direction of the beam. The
where xi are the input influence parameters, n is the number of
model of Hwang and Lee (2002) can clearly illustrate the strut-
influence parameters, a i are the influence coefficients and c is the
and-tie mechanism in the joint core but the joint shear prediction
intercept. The determination of a i and c was based on regression
is known to be very sensitive to Astr , which is difficult to
analysis of the collected database.
determine accurately.
The proposed model considers eight parameters, as shown in
Based on the average stress approach, Tsonos (2007) developed a
Table 3. Of the eight parameters, the normalised joint shear
theoretical model for the shear strength of beam–column joints.
reinforcement is the sum of the normalised horizontal joint shear
In the model, a biaxial compression–tension failure envelope was
reinforcement and the normalised vertical joint shear reinforce-
adopted to predict the joint shear strength. However, this model
ment. This parameter is taken into consideration to verify whether
exaggerated the role of joint shear reinforcement in confining the
the horizontal, the vertical or their total significantly influence
concrete at the strut, while its influence on the truss mechanism
joint shear strength. Table 4 illustrates the level of significance of
was ignored. Using a similar method, Wang et al. (2012)
the considered parameters on the shear strength of exterior and
developed a theoretical model for the shear strength of beam–
interior joints. The table shows the variation of the average
column joints. However, the limitation of their model is that the
absolute error (AAE) of the model-to-test shear strength with
role of joint shear reinforcement was exaggerated by assuming
variation of the considered parameters. The model joint shear
that the tensile strength of the concrete and the joint shear
strength was calculated using Equation 7 with the number of
reinforcement is reached simultaneously. The model of Wang et
influence parameters ranging from six to one. The AAE of the
al. (2012) is
model to the test joint shear strength indicates the accuracy of the
proposed model and was calculated using
1  (sin2 Æ= f t,n  0:8 cos2 Æ= f c9 ) y
V jh ¼  Ajh
6: (1= f t,n þ 0:8= f c9 ) sin 2Æ
 

k V i i 
1X jh,model  V jh,test 
AAE ¼
8: k i¼1 V ijh,test
where  ¼ 1.0 and 0.8 for interior and exterior joints respectively,
Æ is the inclination of the diagonal strut, ft,n is the nominal tensile
strength of concrete with contributions from joint shear reinforce-
ment and y is the column axial stress. where k is the total number of collected experimental results and
V ijh,model and V ijh,test are the predicted and experimental joint shear
The proposed model strengths respectively.
From the above analysis, it can be concluded that there is
inconsistency in the research community about the contribution Table 4 shows that both the horizontal joint shear reinforcement
of strut and truss mechanisms in resisting joint shear force. The and the column intermediate longitudinal reinforcement influence
role of some parameters related to geometry, column axial load, the joint shear strength. The proposed new parameter, the beam
joint shear and beam reinforcement is still under debate. In most bar index b significantly influences the joint shear strength.
of the joint shear strength models, the role of beam reinforcement Consideration of this parameter in the model helps in reducing
was considered in terms of its total cross-sectional area and yield the AAE from 18.2% to 14.3% for exterior joints and from
strength. Joint performance is also known to be influenced by the 13.2% to 12.2% for interior joints. The two parameters with the
reinforcement–concrete bond condition, especially the bond be- least influence include the column depth to beam height ratio and
tween beam bars and the concrete at the joint core. Therefore, a the beam bar diameter to column depth ratio. It is worth
new model is proposed in which the bond condition of beam bars mentioning that the beam bar index was limited to 0.4 in all the
is considered via the new parameter b : The number of beam bars calculations, that is

523
Magazine of Concrete Research A new empirical model for shear strength
Volume 66 Issue 10 of reinforced concrete beam–column
connections
Tran, Hadi and Pham

Parameter Number of considered parameters

6 5 4 3 3 2 2 1

f c9 [ [ [ [ [ [ [ [
j [ [ [ [ — — [ —
jh — — — — [ [ — —
jv — — — — [ — — —
b [ [ [ [ — — — —
N=bc hc f c9 [ [ [ — — — — —
dsb /hc [ [ — — — — — —
hc /hb [ — — — — — — —
AAE, exterior joints 0.102 0.105 0.108 0.143 0.182 0.191 0.182 0.213
AAE, interior joints 0.117 0.117 0.117 0.122 0.132 0.155 0.132 0.162

Table 4. Variation of average absolute error of the model-to-test

joint shear strength with the considered parameters

nb d sb hc
b ¼ < 0: 4 (2012), corresponding to Equations 2, 4, 5 and 6 respectively.
bb hb The ACI model (ACI, 2002) was chosen for the comparison
because it is the simplest model and the only parameter consid-
ered is concrete compressive strength. The model of Kim et al.
As the parameters dsb /hc and hc /hb are the two least important (2009) was chosen because it is of the same level of complexity
factors, these factors were ignored to give a simple formation of as the proposed model and was also developed based on an
the joint shear strength. Moreover, for simplicity, the influence empirical approach. The models of Hwang and Lee (2002) and
coefficients a i and the intercept c for exterior and interior joints Wang et al. (2012) were also chosen for comparison because they
were adjusted to give a unique formulation that can be applied represent theoretical models developed based on the strut-and-tie
for both types of connections. The final equation for joint shear and average stress approaches respectively. The average value of
strength is the predicted-to-test joint shear strength ratio (AVG) and its
coefficient of variation (CoV) are shown in Figure 1 to compare
 
N : accuracies in predicting the joint shear strength. The figure shows
V jh ¼ ª1 þ þ 1 2b Ajh ( f c9 )0 5
:
bc hc f c9 that the model proposed in this paper gave the best predictions of
joint shear strength.
9: þ ª2 (Asjh f jhy þ Asjv f jvy )
Figures 2 and 3 show comparisons of model accuracy for exterior
and interior joints respectively; the figures indicate that the
in which ª1 ¼ 0.81 and ª2 ¼ 0.14 for interior joints and proposed model can predict the shear strength of both exterior
ª1 ¼ 0.34 and ª2 ¼ 0.22 for exterior joints. and interior connections with a great improvement in accuracy.
The proposed model predicts the shear strength of exterior joints
Model verification with an AAE of 11% and CoV of 15%; these values for the most
The results from tests on exterior and interior joints in the recent model (Wang et al., 2012) are 17% and 22% respectively.
compiled database were used to verify the proposed model. The The ACI 352R-02 model (ACI, 2002) is the least accurate in
test shear strength Vjh,test and the predicted shear strength Vjh,model predicting the shear strength of exterior joints, with AAE ¼ 27%
calculated from Equation 9 are shown in Table 1 for exterior and CoV ¼ 34%. For interior joints (Figure 3), the proposed
joints and Table 2 for interior joints. The tables show that the model predicts the joint shear strength with AAE ¼ 12% and
predicted joint shear strengths are in very close agreement with CoV ¼ 14%; the next most accurate model is ACI 352R-02, with
the experimental values. The proposed model predicts joint shear AAE ¼ 13% and CoV ¼ 19%.
strength with an average model-to-test shear strength ratio of 1.00
and with coefficients of variation (CoV) of 14.7% for exterior Parametric study
joints and 14.0% for interior joints. The proposed model considers four ‘key’ parameters – concrete
compressive strength f c9 , the normalised column axial stress
Figure 1 shows the results from the proposed model and results N =(bc hc f c9 ), the normalised joint shear reinforcement
from the joint shear strength models proposed by ACI (2002), j ¼ jh þ jv and the beam bar index b : These parameters were
Kim et al. (2009), Hwang and Lee (2002) and Wang et al. double checked and compared with existing models to evaluate

524
Magazine of Concrete Research A new empirical model for shear strength
Volume 66 Issue 10 of reinforced concrete beam–column
connections
Tran, Hadi and Pham

2600
2600 Exterior joints:
Exterior joints:
AVG ⫽ 1·00 AVG ⫽ 1·26
CoV ⫽ 0·147 CoV ⫽ 0·336

Interior joints: 1950 Interior joints:

1950 AVG ⫽ 0·96
AVG ⫽ 1·00
CoV ⫽ 0·140 CoV ⫽ 0·193

Vjh,model: kN
Vjh,model: kN

1300 1300

650 650
Proposed ACI (2002)

0 0
0 650 1300 1950 2600 0 650 1300 1950 2600
Vjh,test: kN Vjh,test: kN
(a) (b)
2600 2600
Exterior joints: Exterior joints:
AVG ⫽ 0·80 AVG ⫽ 1·11
CoV ⫽ 0·234 CoV ⫽ 0·183

Interior joints: 1950 Interior joints:

1950 AVG ⫽ 1·03
AVG ⫽ 1·02
CoV ⫽ 0·275 CoV ⫽ 0·201
Vjh,model: kN
Vjh,model: kN

1300 1300

650 650
Hwang & Lee (2002) Kim et al. (2009)

0 0
0 650 1300 1950 2600 0 650 1300 1950 2600
Vjh,test: kN Vjh,test: kN
(c) (d)
2600
Exterior joints:
AVG ⫽ 0·92
CoV ⫽ 0·218

1950 Interior joints:

AVG ⫽ 0·94
CoV ⫽ 0·216
Vjh,model: kN

1300

650
Wang et al. (2012)

0
0 650 1300 1950 2600
Vjh,test: kN
(e)

Figure 1. Performance of the proposed model (a) and the models

of ACI (2002) (b), Hwang and Lee (2002) (c), Kim et al. (2009)
(d) and Wang et al. (2012) (e)

525
Magazine of Concrete Research A new empirical model for shear strength
Volume 66 Issue 10 of reinforced concrete beam–column
connections
Tran, Hadi and Pham

0·4 0·4
0·34 CoV CoV
AAE AAE
0·3 0·27 0·3 0·28
AAE or CoV

AAE or CoV
0·23 0·23 0·22 0·21 0·20 0·22
0·2 0·18 0·2 0·19 0·18
0·16 0·17 0·17
0·15 0·14
0·13
0·11 0·12
0·1 0·1

0 0
9 ) l.

2) al.
)

00 nd

el d
02

9) l.

2) al.
)
00 a

00 nd

el d
od e

02

00 a

od e
m pos
(2 et

01 t
(2 a
20

2)

(2 g e

m pos
(2 et

01 t
(2 a
20

2)

(2 g e
e g

e g
I(

Le an

m
an

I(

Le an

o
Ki

Pr
AC

an
Ki
Hw

Pr
AC
W

Hw

W
Figure 2. Accuracy comparison of the models for shear strength Figure 3. Accuracy comparison of the models for shear strength
prediction of exterior joints prediction of interior joints

the fitness of the influence coefficients as well as the accuracy of demonstrate clearly the dependency of the joint shear strength
the proposed model. Figure 4 illustrates the variation of the ratio on each of the parameters. In general, the figure shows that
predicted-to-test joint shear strength ratio with these four para- the proposed joint shear strength ratio has no clear dependence
meters. Linear regression lines were added to the figure to on these four parameters despite their large range variation.

2·00 2·00
ACI (2002) Kim et al. (2009)
1·75 Proposed 1·75 Proposed

1·50 1·50
Vjh,model /Vjh,test

Vjh,model /Vjh,test

1·25 1·25

1·00 1·00

0·75 0·75

0·50 0·50

0·25 0·25
0 20 40 60 80 100 120 ⫺0·4 ⫺0·2 0 0·2 0·4 0·6
f ⬘c : MPa N/(bc hc f ⬘c )
(a) (b)
2·00 2·00
Wang et al. (2012) Kim et al. (2009)
1·75 Proposed 1·75 Proposed

1·50 1·50
Vjh,model /Vjh,test

Vjh,model /Vjh,test

1·25 1·25

1·00 1·00

0·75 0·75

0·50 0·50

0·25 0·25
0 0·5 1·0 1·5 2·0 2·5 3·0 3·5 4·0 0 0·1 0·2 0·3 0·4 0·5
Normalised joint reinforcement χj Beam bar index χb
(c) (d)

Figure 4. Variation of joint shear strength ratio with each of the

influence parameters

526
Magazine of Concrete Research A new empirical model for shear strength
Volume 66 Issue 10 of reinforced concrete beam–column
connections
Tran, Hadi and Pham

Figure 4 also reveals that the proposed model gives the predicted- from the literature. The influence of some ‘key’ parameters on
to-test joint shear strength ratio with less scatter than the other joint shear strength was analysed and four parameters were
models. In particular, the figure shows that the predicted-to-test chosen to generate the proposed model. These four parameters
joint shear strength ratio of the proposed model varies around included a new parameter to reflect the beam bar bond condition
1.00 in a small range from 0.71 to 1.51; this range for the models as well as the possibility of the beam bar transferring a fraction
of ACI (2002), Kim et al. (2009) and Wang et al. (2012) is 0.57– of the joint shear force into the column. The regression analysis
1.77, 0.61–1.83 and 0.51–1.66 respectively. showed that consideration of this parameter helps to significantly
improve the reliability of the model when compared with test
Figure 4(a) compares the dependence of joint shear strength ratio results.
on concrete compressive strength for the proposed model and the
model in ACI 352R-02. In both models, the power term for A parametric study to illustrate the better performance of the
concrete compressive strength is 0.5. However, the joint shear proposed model when compared with other models was also
strength ratio of the ACI model tends to reduce with an increase conducted. This showed that the proposed coefficients in the
in concrete strength while there is no clear dependency in the equation for joint shear strength are justified and that the proposed
proposed model. This indicates that consideration of the remain- equation for joint shear strength is reliable despite the large
ing three parameters effectively improves the reliability of the variation in the parameters of influence. Results from the
proposed model. proposed model were compared with those from four existing
joint shear strength models to demonstrate the effectiveness of the
Figure 4(b) compares the variation of joint shear strength ratio new model. Due to its accuracy, stability and simplicity, applica-
with the normalised column axial stress for the proposed model tion of the proposed model for predicting the shear strength of
and the model of Kim et al. (2009). Both models were developed beam–column connections in practical design is expected.
based on an empirical approach, but the model of Kim et al.
neglects the influence of column axial load and thus its joint REFERENCES
shear strength ratio is significantly influenced by the normalised ACI (American Concrete Institute) (2002) ACI 352R-02:
column axial stress. Recommendations for design of beam–column joints in
monolithic reinforced concrete structures. ACI, Farmington
The dependency of the joint shear strength ratio on the normalised Hills, MI, USA.
joint shear reinforcement in the proposed model and the model of ACI (American Concrete Institute) (2008) ACI 318R: Building
Wang et al. (2012) is shown in Figure 4(c). Once again, it can be code requirements for structural concrete and commentary.
seen that the proposed model has the better performance. The ACI, Farmington Hills, MI, USA.
independence of the proposed joint shear strength ratio on the AIJ (Architectural Institute of Japan) (1999) Design guideline for
normalised joint shear reinforcement indicates that the proposed earthquake resistance reinforced concrete buildings based on
coefficient ª2 in Equation 9 is justified. inelastic displacement concept. AIJ, Tokyo, Japan.
Alva GMS, Debs ALHDC and Debs MKE (2007) An experimental
Figure 4(d) shows a comparison of the dependence of joint shear study on cyclic behaviour of reinforced concrete connections.
strength ratio on the beam bar index. For the proposed model, Canadian Journal of Civil Engineering 34(4): 565–565.
there is very little dependency of the joint shear strength ratio on Antonopoulos CP and Triantafillou TC (2003) Experimental
the beam bar index, while significant dependency is observed for investigation of FRP-strengthened RC beam–column joints.
the model of Kim et al. (2009). Moreover, as there was no Journal of Composites for Construction 7(1): 39–49.
limitation in the variation range of the amount of beam steel bars, Attaalla SA (2004) General analytical model for nominal shear
it is believed that the model of Kim et al. overestimates the joint stress of type 2 normal and high-strength concrete beam–
shear strength when the amount of beam reinforcement is high. It column joints. ACI Structural Journal 101(6): 881–882.
is noted that, in both models, the role of the beam steel bars was Au FTK, Huang K and Pam HJ (2005) Diagonally-reinforced
considered in different manners: Kim et al. considered the beam beam–column joints reinforced under cyclic loading.
bars in terms of their total area together with their yield strength Proceedings of the Institution of Civil Engineers – Structures
while the proposed model considers their contact area with the and Buildings 158(1): 21–40.
surrounding concrete. The better performance of the proposed Bakir PG and Boduroğlu HM (2002) A new design equation for
model compared with that of Kim et al. in this respect indicates predicting the joint shear strength of monotonically loaded
that the reasons for proposing the parameter b to account for the exterior beam–column joints. Engineering Structures 24(8):
influence of beam longitudinal reinforcement are rational. 1105–1117.
BSI (2004) Eurocode 8: Design of structures for earthquake
Conclusions resistance. Part 1: General rules, seismic actions and rules for
A new empirical model for joint shear strength of RC beam– buildings. BSI, London, UK.
column connections has been introduced. The model was devel- Chalioris CE, Favvata MJ and Karayannis CG (2008) Reinforced
oped based on regression analysis using a large database collected concrete beam–column joints with crossed inclined bars

527
Magazine of Concrete Research A new empirical model for shear strength
Volume 66 Issue 10 of reinforced concrete beam–column
connections
Tran, Hadi and Pham

under cyclic deformations. Earthquake Engineering and Hwang SJ, Lee HJ and Wang KC (2004) Seismic design and
Structural Dynamics 37(6): 881–897. detailing of exterior reinforced concrete beam–column joints.
Choi H and Kim J (2011) Progressive collapse-resisting capacity Proceedings of 13th World Conference on Earthquake
of RC beam–column sub-assemblage. Magazine of Concrete Engineering, Vancouver BC, Canada, paper 0397.
Research 63(4): 297–310. Hwang SJ, Lee HJ, Liao TF, Wang KC and Tsai HH (2005) Role of
Chun SC, Oh B, Lee SH and Naito CJ (2009) Anchorage strength hoops on shear strength of reinforced concrete beam–column
and behavior of headed bars in exterior beam–column joints. joints. ACI Structural Journal 102(3): 445–453.
ACI Structural Journal 106(5): 579–590. Ishibashi K (1993) Experimental Study on Interior Beam-to-
Chutarat N and Aboutaha RS (2003) Cyclic response of exterior column Joints of Reinforced Concrete Frames. PhD thesis
reinforced concrete beam–column joints reinforced with Tokyo University, Tokyo, Japan (in Japanese).
headed bars–experimental investigation. ACI Structural Joyklad P, Pimanmas A and Dhakal RP (2012) Cyclic
Journal 100(2): 259–264. performance of beam–column joints with extended column
Clyde C, Pantelides CP and Reaveley LD (2000) Performance-based fixed at base. Part I: experimental investigation. Magazine of
evaluation of reinforced concrete building exterior joints for Concrete Research 64(9): 807–807.
seismic excitation. Earthquake Spectra 18(3): 449–480. Kaku T and Asakusa H (1991) Ductility estimation of exterior
Dhakal RP, Pan TC, Irawan P et al. (2005) Experimental study on beam–column subassemblages in reinforced concrete frames.
the dynamic response of gravity-designed reinforced concrete In Design of Beam-column Joints for Seismic Resistance.
connections. Engineering Structures 27(1): 75–87. American Concrete Institute, Detroit, MI, USA, ACI SP-123,
Durrani A and Wight J (1985) Behaviour of interior beam-to- pp. 167–185.
column connections under earthquake-type loading. ACI Karayannis CG and Sirkelis GM (2005) Response of column and
Journal 82(3): 343–349. joints with spiral shear reinforcement. In Computational
Ehsani MR and Alameddine F (1991) Design recommendations Methods and Experimental Measurements, XII (Brebbia CA
for type 2 high-strength reinforced concrete connections. ACI and Carlomagno GM (eds)). WIT Press, Southampton, UK,
Structural Journal 88(3): 277–291. vol. 41.
Ehsani MR and Wight JK (1985) Exterior reinforced concrete Karayannis CG and Sirkelis GM (2008) Strengthening and
beam-to-column connections subjected to earthquake-type rehabilitation of RC beam–column joints using carbon-FRP
loading. ACI Structural Journal 82(4): 492–499. jacketing and epoxy resin injection. Earthquake Engineering
Ehsani MR, Moussa AE and Vallenilla CR (1987) Comparison of and Structural Dynamics 37(5): 769–790.
inelastic behavior of reinforced ordinary and high-strength Karayannis CG, Chalioris CE and Sirkelis GM (2008) Local retrofit
concrete frames. ACI Structural Journal 84(2): 161–169. of exterior RC beam–column joints using thin RC jackets –
El-Amoury T and Ghobarah A (2002) Seismic rehabilitation of an experimental study. Earthquake Engineering and
beam–column joint using GFRP sheets. Engineering Structural Dynamics 37(5): 727–746.
Structures 24(11): 1397–1407. Kim J and LaFave JM (2007) Key influence parameters for the
Fisher MJ and Sezen H (2011) Behavior of exterior reinforced joint shear behaviour of reinforced concrete (RC) beam–
concrete beam–column joints including a new reinforcement. column connections. Engineering Structures 29(10): 2523–
Structural Engineering and Mechanics 40(6): 867–883. 2539.
Ghobarah A and El-Amoury T (2005) Seismic rehabilitation of Kim J and LaFave JM (2008) Probabilistic joint shear strength
deficient exterior concrete frame joints. Journal of models for design of RC beam–column connections. ACI
Composites for Construction 9(5): 408–408. Structural Journal 105(6): 770–780.
Ghobarah A and Said A (2001) Seismic rehabilitation of beam– Kim J and Yu J (2012) Analysis of reinforced concrete frames
column joints using FRP laminates. Earthquake Engineering subjected to column loss. Magazine of Concrete Research
5(1): 113–129. 64(1): 21–33.
Ghobarah A and Said A (2002) Shear strengthening of beam– Kim J, LaFave JM and Song J (2009) Joint shear behaviour of
column joints. Engineering Structures 24(7): 881–888. reinforced concrete beam–column connections. Magazine of
Hanson NW and Connor HW (1967) Seismic resistance of Concrete Research 61(2): 119–132.
reinforced concrete beam–column joints. Journal of Kitayama K, Otani S and Aoyama H (1991) Development of
Structural Division ASCE 93(5): 533–559. design criteria for RC interior beam–column joints. Design of
Hegger J, Sherif A and Roeser W (2003) Nonseismic design of Beam–Column Joints for Seismic Resistance. American
beam–column joints. ACI Structural Journal 100(5): 654–664. Concrete Institute, Detroit, MI, USA, ACI SP-123, pp. 97–
Hwang J and Lee HJ (1999) Analytical model for predicting shear 123.
strengths of exterior reinforced concrete beam–column joints Kuang JS and Wong HF (2006) Effects of beam bar anchorage on
for seismic resistance. ACI Structural Journal 96(5): 846–857. beam–column joint behaviour. Proceedings of the Institution
Hwang SJ and Lee HJ (2002) Strength prediction for discontinuity of Civil Engineers – Structures and Buildings 159(2): 115–
regions by softened strut-and-tie model. Journal of Structural 124.
Engineering ASCE 128(12): 1519–1526. Le-Trung K, Lee J, Lee DH, Lee K and Woo S (2010) Experimental

528
Magazine of Concrete Research A new empirical model for shear strength
Volume 66 Issue 10 of reinforced concrete beam–column
connections
Tran, Hadi and Pham

study of RC beam–column joints strengthened using CFRP Reinforced Concrete Beam-column Joints, Monterey,
composites. Composites Part B 41(1): 76–85. California, USA.
Lee WT, Chiou YJ and Shih MH (2010) Reinforced concrete Pantazopoulou S and Bonacci J (1992) Consideration of
beam–column joint strengthened with carbon fiber reinforced questions about beam–column joints. ACI Structural Journal
polymer. Composite Structures 92(1): 48–60. 89(1): 27–36.
Leon RT (1990) Shear-strength and hysteretic behavior of interior Pantelides CP, Hansen J, Nadauld J and Reaveley LD (2002)
beam–column joints. ACI Structural Journal 87(1): 3–11. Assessment of RC Building Exterior Joints with Substandard
Lima C, Martinelli E and Faella C (2012) Capacity models for Details. University of California, Berkeley, CA, USA, PEER
shear strength of exterior joints in RC frames: state-of-the-art Report 2002/18.
and synoptic examination. Bulletin of Earthquake Park R and Paulay T (1974) Behaviour of reinforced concrete
Engineering 10(3): 967–983. external beam–column joints under cyclic loading.
Liu C (2006) Seismic Behaviour of Beam–Column Joint Proceedings of 5th World Conference on Earthquake
Subassemblies Reinforced with Steel Fibres. Masters thesis, Engineering, Rome, Italy, pp. 772–781.
University of Canterbury, Christchurch, New Zealand. Park S and Mosalam KM (2012) Parameters for shear strength
Lu X, Urukap TH, Li S and Lin F (2012) Seismic behavior of prediction of exterior beam–column joints without transverse
interior RC beam–column joints with additional bars under reinforcement. Engineering Structures 36: 198–209.
cyclic loading. Earthquakes and Structures 3(1): 37–57. Parker DE and Bullman PJM (1997) Shear strength within
Marques JLG and Jirsa JO (1975) A study of hooked bar reinforced concrete beam–column joints. Structural Engineer
anchorages in beam–column joints. ACI Structural Journal 75(4): 53–57.
72(5): 198–209. Patel PA, Desai AK and Desai JA (2013) Evaluation of RC and
Megget LM (1974) Cyclic behaviour of exterior reinforced SFRC exterior beam–column joint under cyclic loading for
concrete beam–column joints. Bulletin of The New Zealand reduction in lateral reinforcement of the joint region.
National Society for Earthquake Engineering 7(1): 27–47. Magazine of Concrete Research 65(7): 405–414.
Meinheit DF and Jirsa JO (1977) The Shear Strength of Paulay T and Priestley MJN (1992) Seismic Design of Reinforced
Reinforced Concrete Beam–Column Joints. University of Concrete and Masonry Buildings. Wiley, New York, NY,
Texas, Austin, TX, USA, CESRL Report 77-1. USA.
Morita S, Kitayama K, Koyama A and Hosono T (1999) Effects of SANZ (Standards Association of New Zealand) (1995) NZS-3101:
beam bar bond and column axial load on shear strength in Concrete structures standard. Part 1: the design of concrete
reinforced concrete interior beam–column joints. Transactions structures. SANZ, Wellington, New Zealand.
of the Japan Concrete Institute 21: 453–460. Sarsam KF and Phipps ME (1985) The shear design of in situ
Murty CVR, Rai DC, Bajpai KK and Jain SK (2003) Effectiveness reinforced concrete beam–column joints subjected to
of reinforcement details in exterior reinforced concrete monotonic loading. Magazine of Concrete Research 37(130):
beam–column joints for earthquake resistance. ACI Structural 16–28.
Journal 100(2): 149–156. Shrestha R, Smith ST and Samali B (2009) Strengthening RC
Najafian HA, Vollum RL and Fang L (2013) Comparative beam–column connections with FRP strips. Proceedings of
assessment of finite-element and strut and tie based design the Institution of Civil Engineers – Structures and Buildings
methods for deep beams. Magazine of Concrete Research 162(5): 323–334.
65(16): 970–970. Supaviriyakit T and Pimanmas A (2008) Comparative
Noguchi H and Kashiwazaki T (1992) Experimental studies on performance of sub-standard interior reinforced concrete
shear performances of RC interior column–beam joints with beam–column connection with various joint reinforcing
high-strength materials. Proceedings of 10th World details. Materials and Structures 41(3): 543–557.
Conference on Earthquake Engineering, Balkema, Rotterdam, Tsonos AG (2007) Cyclic load behavior of reinforced concrete
the Netherlands, pp. 3163–3168. beam–column subassemblages of modern structures. ACI
Noguchi H and Kurusu K (1988) Correlation of bond and shear in Structural Journal 104(4): 468–478.
RC beam–column connections subjected to seismic forces. Tsonos AG, Tegos IA and Penelis GG (1992) Seismic resistance of
Proceeding of 9th World Conference on Earthquake type 2 exterior beam–column joints reinforced with inclined
Engineering, Tokyo-Kyoto, Japan, pp. 597–602. bars. ACI Structural Journal 89(1): 3–12.
Oka K and Shiohara H (1992) Test of high-strength concrete Vatani-Oskouei A (2010) Repairing of seismically damaged RC
interior beam-to-column joint subassemblages. Proceedings exterior beam–column connection using CFRP. Journal of
of 10th World Conference on Earthquake Engineering. Reinforced Plastics and Composites 29(21): 3257–3274.
Balkema, Rotterdam, the Netherlands, pp. 3211–3217. Vollum RL and Newman JB (1999) Strut and tie models for
Otani S, Kobayashi Y and Aoyama H (1984) Reinforced concrete analysis/design of external beam–column joints. Magazine of
beam–column joints under simulated earthquake loading. In Concrete Research 51(6): 415–425.
Proceedings of 1st US–NZ–Japan Seminar on Design of Wang GL, Dai JG and Teng JG (2012) Shear strength model for

529
Magazine of Concrete Research A new empirical model for shear strength
Volume 66 Issue 10 of reinforced concrete beam–column
connections
Tran, Hadi and Pham

RC beam–column joints under seismic loading. Engineering ratio on joint seismic behaviour. Proceeding of the Institution
Structures 40: 350–360. of Civil Engineers – Structures and Buildings 161(2):
Wang YC and Hsu K (2009) Shear strength of RC jacketed 91–101.
interior beam–column joints without horizontal shear Zhang L and Jirsa JO (1982) A Study of Shear Behavior of
reinforcement. ACI Structural Journal 106(2): 222–232. Reinforced Concrete Beam–Column Joints. University of
Wong HF and Kuang JS (2008) Effects of beam–column depth Texas, Austin, TX, USA, PMFSEL Report 82–1.

WHAT DO YOU THINK?

To discuss this paper, please submit up to 500 words to
the editor at journals@ice.org.uk. Your contribution will
be forwarded to the author(s) for a reply and, if
considered appropriate by the editorial panel, will be
published as a discussion in a future issue of the journal.

530
Copyright of Magazine of Concrete Research is the property of Thomas Telford Ltd and its
content may not be copied or emailed to multiple sites or posted to a listserv without the
copyright holder's express written permission. However, users may print, download, or email
articles for individual use.