Journal of Constructional Steel Research 118 (2016) 105–119

Contents lists available at ScienceDirect

Journal of Constructional Steel Research

Moment-connection between continuous steel beams and reinforced

concrete column under cyclic loading
Seyed Rasoul Mirghaderi, Nasrin Bakhshayesh Eghbali ⁎, Mohammad Mehdi Ahmadi
School of Civil Engineering, University of Tehran, P.O. Box 11365-4563, Tehran, Iran

a r t i c l e i n f o a b s t r a c t

Article history: This paper suggests a new moment connection between steel beams and a reinforced concrete column (RCS). In
Received 7 May 2015 this proposed connection, two parallel beams pass from both sides of the column and are welded to the cover
Received in revised form 28 October 2015 plates surrounding the concrete column in the joint area. This detail provides two main advantages compared
Accepted 7 November 2015
with previous constructions: first, both the beam and column are continuous in the joint area, which provides
Available online 6 December 2015
more reliable performance, and second, the force transfer occurs in such a manner that the cover plates are load-
Keywords:
ed in-plane and stress concentration is prevented in the connection components. Bar shear connectors were
RCS connection installed between the steel and concrete inside the cover plates to restrict sliding. The force transfer mechanisms
Continuous beams and design procedure are described, and the seismic behavior of the proposed connection is studied in two ex-
Experimental test perimental tests under cyclic loading. The test results showed that both specimens sustained 8% story drift
Load transfer mechanism with stable hysteretic loops and that the proposed connection is acceptable as a special moment connection. In
Seismic performance addition, the test results demonstrated that the proposed design relationships were arranged properly such
that the cover plates were maintained in the elastic phase, only slight cracks appeared in the column, and plastic
hinges were formed in the beams in the vicinity of the column. Furthermore, to clarify the behavior and shear ca-
pacity of bar shear connectors embedded in a confined concrete, two push-out specimens were tested under
monotonic loading.
© 2015 Elsevier Ltd. All rights reserved.

1. Introduction
An optimum combination of steel and concrete and efficient
usage of material features renders RCS structures more efficient
and economical than traditional concrete and steel frames. Using
an experimental test on a specimen with improved composite RCS
joint details, Men et al. [1] indicated that the displacement ductility
coefficient of RCS frames is much larger than that of conventional re-
inforced concrete and steel moment frames. Despite these advan-
tages, identification of a beam-to-column connection with proper
seismic performance and an explicit force transfer mechanism is es-
sential. Extensive researches have been conducted since 1987 and
several details have been proposed to effectively transfer induced
forces from steel beam to reinforced concrete column. The investi-
gated details fall into two main types including the beam-through-
type and the column-through-type [2].
To enhance the shear transfer in the panel zone, the following details
were employed in the beam-through-type connections: face bearing
plate welded to the beam at the column face, extended face bearing
plate, steel column attached to beam flanges and shear studs welded
to the beam flanges to mobilize concrete compression field inside and
⁎ Corresponding author.
E-mail address: n.bakhshayesh@ut.ac.ir (N. Bakhshayesh Eghbali).
outside the beam flange [3], steel band plates embedded around the col-
umn to prevent bearing failure [3–5], and cover plates and dowel bars
[4]. These details were suggested to provide new load paths to reduce
stress concentration thereby enhancing the strength and stiffness of
the connection. Beam-through-type connections with continuous
beam eliminate failure concerns in the beam to column connections
where maximum moment exists. By not interrupting the beam at the
point of maximum moment at the column face, the beam-through-
type details avoid the fracture-critical joints that are of concern in con-
ventional structural construction [6]. However, continuous beam inter-
rupts the column reinforcing bars arrangement and concrete continuity
by continuously passing through the joint area.
For seismic design of RCS connections, ASCE 1994 [7] is referenced
by the AISC Seismic Provisions (2002) as the recommended design
guidelines. ASCE 1994 focuses on the beam-through-type connections
and contains a wide variety of joint details. Alizadeh et al. [8,9] tested
four interior beam-through-type connections with standard detailing
and with additional bearing plate, using normal and self-consolidating
concrete. The comparison of the test results showed that all specimens
performed in a ductile manner and they maintained their strength
with little decrease of stiffness. Zhang et al. [10] conducted five experi-
mental tests to evaluate the seismic behavior of confined reinforced
concrete column and composite beam joints with weakening extent of
flanges. Furthermore, some experimental and analytical studies have

http://dx.doi.org/10.1016/j.jcsr.2015.11.002
0143-974X/© 2015 Elsevier Ltd. All rights reserved.
106 S.R. Mirghaderi et al. / Journal of Constructional Steel Research 118 (2016) 105–119
been carried out on the behavior of RCS frames with beam-through type
connections [11–14].
Column-through-type connections, using diaphragms [15] or cover
plates and horizontal stiffeners [16], may simplify construction process;
however, additional research is required to improve the strength and
ductility of the connection. According to the review by Li et al. [17],
the research conducted by Iizuka et al. revealed that the hysteretic
loops of the column-through-type connections were more plentiful
than the beam-through-type connections; however the through-
column type joint was designed stiffer than the through-beam type
joint. Therefore, column-through-type connection showed better seis-
mic performance. Li et al. [18] tested three full scale bolted endplate
connections to steel beam and reinforced concrete column. Experimen-
tal results showed that specimens without slab exhibited good ductility
and energy-dissipating capacity; however the specimen with floor slab
fractured at the weld between the bottom flange of the beam and the
endplate.
Furthermore, various details for other types of composite connec-
tions have been studied as well. Li et al. [19] conducted finite element
analyses of rebar-penetrated connection between gangue concrete
filled steel tubular column and reinforced gangue concrete beam to in-
vestigate the failure modes and force transferring mechanism. The seis-
mic behavior of bolted endplate connections for concrete filled circular
steel tube (CFT) column to steel beam was investigated by Li et al. [20]
and Sheet et al. [21].
In order to improve the performance of steel beam to concrete
column connections, this paper proposes a seismic moment connection
with proper features of both beam-through-type and column-through-
type connections. In the proposed connection, two parallel beams are
placed outside the column and are welded to the steel cover plates
that surround the concrete column in the joint area. Passing the
beams outside the column provides the advantages of a continuous
beam and column connections without disturbing the column in the
joint area. Two relatively similar specimens of the proposed connection
were experimentally studied. In the first specimen, as depicted in Fig. 1,
the flanges of the channel-section beams are longitudinally welded to
the parallel cover plates (referred to as the side plates), and the beam
web is directly welded to the edges of the side plate. In the second spec-
imen, two continuous beams are located on the protruding sections of
the cover plates (referred to as the front plates) as vertical supports lo-
cated perpendicular to the beam web (Fig. 2). To provide sufficient con-
nection, the top and bottom flanges of the I-section beams are welded to
the beveled edges of the protruding sections of the front plates. In addi-
tion, the beam web is connected to the front plate vertical edges via con-
tinuity plates. In both specimens, bar shear connectors are used inside
the cover plates to minimize slippage at the steel–concrete interfaces.
Studies conducted by MacRae et al. [22] showed that the connection
Fig. 1. The first specimen (SPU): (a) Connection configuration, (b) connection assembly.
between a steel plate and the concrete column with ribs (with similar
behavior to bar shear connectors) were more effective in transferring
force than the other types of the studied connection tools. Also it was re-
vealed that slip deformations between the steel and concrete were in-
sufficient to fully activate flexible shear studs in the joint area.
However, the possibility of the proposed connection using bolts also
exists in the form depicted in Fig. 3, and this structure can be accurately
evaluated through further research. In the connection method that uses
bolts, a rectangular connecting plate on one side is connected to the pro-
truding section of the front plates by penetration groove welds, and is
bolted to the beam flange on the other side.
According to the configuration of the suggested connection, some
advantages are expected including: more resistance against gravity
and seismic loads because of continuity of the beams, simultaneous
transfer of the shear force and bending moment to the column, in-
plane loading of the connection components, more plastic rotation ca-
pacity by using two beams with less depth, increase in the capacity of
the column in the connection region, possibility of repair in the case of
shear yielding, more reliable performance by locating the panel-zone
out of the column [23] and appropriate arrangement of reinforcement
bars and continuous access inside the column.
In this research program, two interior RCS connections were tested
under cyclic loading. The objective of this research is to describe the
load transfer mechanisms, design method and seismic behavior of the
proposed connection. Furthermore, two push-out specimens were test-
ed under monotonic loading to estimate the shear capacity of a bar
shear connector embedded in a confined concrete block.
2. Force transfer mechanisms
2.1. First specimen (SPU)
The moment (Mpb) and shear force (Vb) that result from formation of
plastic hinges are transferred to the side plates, which are located paral-
lel to the beam web via the weld connection of the beam flanges and
web (Fig. 4). These induced forces tend to cause in-plane rotation in
the side plates that is resisted by three mechanisms, including the tor-
sional resistance of the side plates and the in-plane and out-of-plane re-
sistance of the front plates. The torsional resistance of the side plates is
provided by shear connectors that are welded on the side plates and
mobilize the concrete through torsional deformation.
As depicted in Fig. 5, the in-plane mechanism of the front plate pro-
duces two sets of in-plane forces along the front plate edges, and the
out-of-plane mechanism creates a force distribution perpendicular to
the column face. The in-plane resistance of the front plates is provided
by the shear resistance of the shear connectors that are welded on the
front plates inside the column (Fig. 6). The out-of-plane forces apply
 S.R. Mirghaderi et al. / Journal of Constructional Steel Research 118 (2016) 105–119
Fig. 2. The second specimen (SPI): (a) Connection configuration, (b) connection assembly.
compressive stresses on the concrete surface, and their intensity de-
creases in the direction moving toward the middle of the column
width. These compressive stresses lead to formation of a concrete diag-
onal strut via the contribution of the compressive stresses from the col-
umn. These resisting mechanisms represent three parallel paths that
reliably transfer the moment and shear forces of the beam's plastic
hinges to the reinforced concrete column.
2.2. Second specimen (SPI)
The moment and shear forces of the beam's plastic hinges are trans-
ferred simultaneously to the front plates through the horizontal and
vertical weld connections of the beam flanges and continuity plates, re-
spectively. In this specimen, the protruding sections of the front plates
behave as vertical supports for the beams in the connection zone, as
shown in Fig. 7.
As illustrated in Fig. 8, these induced forces tend to move the front
plates parallel to the column faces that encounter three resisting mech-
anisms, including the torsional resistance of the side plates, and the in-
plane and out-of-plane resistances of the front plates, similar to the pre-
vious specimen (Fig. 6).
As previously described, although the load transfer mechanisms act
such that the beam forces are initially transferred to the side plates and
front plates in the first and second connections, respectively, the force
resisting mechanisms are identical in both specimens.
Fig. 3. Proposed details of the connection using bolts.
107
3. Design procedure for the proposed connection
The design of the specimens was based on a strong-column weak-
beam criterion to concentrate plasticity in the beams. Therefore, the
connection components were proportioned according to the beam mo-
ment and shear induced in the plastic hinges. The AISC seismic provi-
sions [24] were applied in the design process of the steel components.
3.1. First specimen (SPU)
3.1.1. Design of the connection between the beam and side plate
The moments of the beam plastic hinges are transferred to the side
plates via the beam flanges. The beam flanges are connected to the
side plate and the beam web in the panel zone such that the distribution
of the moment between the side plate (M1) and the beam web (M2) in
this region depends on their height and thickness, and these moments
are described as follows,
t s db
M1 ¼ M ð1Þ
ts db þ tw ðdb  2t f Þ
tw ðdb  2t f Þ
M2 ¼ M ð2Þ
ts db þ tw ðdb  2t f Þ
Where M = 2ME + Vbdc, ME = Mpb + VbLa is the moment at the col-
umn face, La is the distance between the location of the plastic hinge
and the column face, and Mpb = 1.1RyZbFy is the plastic hinge moment
of the two beams. Based on AISC, Ry for A36 steel hot-rolled structural
shapes is set to 1.5 [24], Zb is the plastic modulus of the two beams at
the plastic location, Fy is the minimum specified yield strength of the
Fig. 4. The efforts resulted from formation of plastic hinges in the first specimen (SPU).
108 S.R. Mirghaderi et al. / Journal of Constructional Steel Research 118 (2016) 105–119
Fig. 5. In-plane and out-of-plane resisting forces for the first specimen (SPU).
Mpb
steel material, V b ¼ Lp is the beam shear force at the plastic hinge, Lp is
the distance from the beam end to the plastic hinge, dc is the column
width, db is the beam depth, ts is the side plate thickness, and tw and tf
are the web and flange thicknesses of the beam, respectively.
As depicted in Fig. 4, the reaction forces of the weld connections of
the beam flanges and the web on the side plate include horizontal (F1)
and vertical (F2) components that are calculated as follows,
M1
F1 ¼
db
M2
F2 ¼ Vb þ
dc
The plastic hinge forces induce parallel to the axis shear stresses on
the weld connections. The beam flanges are welded to the side plate
via partial joint penetration groove welds, and fillet welds are used to
connect the beam web and the side plate.
3.1.2. Design of the panel zone
As described previously, the beam web and flanges are welded to the
side plate, leading to the participation of the side plate in the panel zone
shear stresses. Therefore, each beam has its own separate panel zone,
including a section of the beam web and side plates between the conti-
nuity plates and beam flanges.
Participation of the side plate in the panel zone usually provides suf-
ficient shear resistance that reinforcement of the beam web with dou-
bler plates is not necessary. The shear capacity of the side plate (Vp1)
Fig. 6. Assumed stress transferring mechanisms: (a) Column section, (b) concrete diagonal strut due to the out of plane resistance of the front plates, (c) in-plane resistance of the front
plates, and (d) torsional resistance of the side plates.
Fig. 7. Moment and shear force transfer mechanism from beam to column through front
plates in the second specimen (SPI) [9].
and the beam web (Vp2) can be calculated as follows,
!
3b f t2f
Vp1;2 ¼ 0:577 Fy dc ts;w 1 þ ð5Þ
db dc ts;w
The shear capacity (Vp1,2) of each component of the panel zone (the
beam web or the side plate) must be greater than or equal to the design
shear force(Vpz1,2), which can be computed as follows,
M1
Vpz1 ¼ ð6Þ
db
M2
Vpz2 ¼ ð7Þ
ð3Þ db  2t f
3.1.3. Shear connection at the steel and concrete interface
ð4Þ According to the load transfer mechanisms, the total moment
(M) induced from the moment and shear forces of the beam plastic
hinge are transferred to the side plate via the weld connections of the
beam flanges and web. A portion of the total moment (m1) is transferred
to the concrete column via shear connectors on the side plates (first
mechanism), and the remaining portion (m2) is transferred to the
front plate through the connection of the side plate to the front plate,
as shown in Fig. 9. As described previously, the induced forces in the
front plate are countered with two resisting mechanisms, including
in-plane and out-of-plane resistances. In the design process, the portion
of the out-of-plane resistance that leads to formation of the concrete di-
agonal strut in the load transfer is conservatively ignored because it
transfers a small portion of the total moment relative to the other mech-
anisms, especially in the absence of the axial forces of the roof level col-
umns. The portions of the shear connectors on the side plate (m1) and
front plate (m2) of the total moment are calculated as follows,
m1 ¼ f 1 d1 ð8Þ
 S.R. Mirghaderi et al. / Journal of Constructional Steel Research 118 (2016) 105–119 109

Fig. 8. In-plane and out-of-plane resisting forces for the second specimen (SPI).

Fig. 10. Parameters of bar shear connectors EC4 (1994) [16].

m2 ¼ f 2 dc ð9Þ

area. Furthermore, the thicknesses of the side plates should be deter-

where f1 is the total shear force of the shear connectors on the side
mined such that they can resist the shear forces of the panel zone to-
plates acting in the same direction, f2 is the total shear force of the
gether with the beam web.
shear connectors on the front plate, and d1 is the distance between the
centerline of the shear connectors on the side plate.
According to Fig. 9, based on the principle of compatibility of dis-
placements, the relationship of the shear forces distributed between
the shear connectors on the side plates and front plate can be calculated
as follows,

dc
f2 ¼ f1 ð10Þ
d1

According to Eq. (10), locating the shear connectors near the edges
of the side plates provides optimum participation of the side plate in
the load transfer via the shear connectors. The strength estimation of
the bar shear connector is described in Section 3.3.

3.1.4. Design of the cover plates

The cover plate height (including front plates and side plates) is de-
termined based on the need to provide the required surface for placing
shear connectors that will transfer the existing shear forces between the
steel and concrete.
As described in Section 3.1.3, the weld connection between the
side plate and front plate should be designed for the portion of the
total moment that is transferred to the front plate (m2). According
to Fig. 5, the shear distribution on the front plates occurs in such a
manner that the forces are primarily transferred over a length
equal to the beam height. Therefore, the welds and the thickness of
the cover plates are compared against the shear stresses in this

Fig. 9. Load distribution between shear connectors on the front plate and side plate. Fig. 11. The geometry and details of the push-out specimens (dimensions are in mm).
110 S.R. Mirghaderi et al. / Journal of Constructional Steel Research 118 (2016) 105–119
Fig. 12. Load–relative displacement curves for two specimens.
3.2. Second specimen (SPI)
3.2.1. Design of the connection between the beam and front plate
As depicted in Fig. 7, the total demands of the beam plastic hinges
are transferred to the column through the reactions induced in the pro-
truding sections of the front plate. The vertical reaction forces (F) are
calculated as follows,
F ¼ M=2dc
The reaction forces (F) should be resisted by the horizontal and ver-
tical weld connections of the beam flanges and the web to the front
plates. The beam flanges are connected to the front plate by complete
joint penetration groove welds, and the connection of the beam web
to the front plate is provided via the weld connection of the continuity
plate to the front plate.
Fig. 13. Specifications and dimensions of the test specimens (dimensions are in mm).
3.2.2. Design of the panel zone
In this specimen, the panel zone is composed of the beam web be-
tween the continuity plates, and the doubler plates which are welded
to the beam flanges and continuity plates. Depending on the beam sec-
tion, a doubler plate can be used on one side or both sides of the beam
web for channel section or I-section beams, respectively. The shear ca-
pacity of the panel zone and the design shear force can be calculated
using Eqs. 5 and 7 by substituting ts,w = tdp + tw , where tdp is the total
thickness of the doubler plates and M2 = M.
3.2.3. Shear connection at the steel and concrete interface
According to the load transfer mechanisms, the total moment
(M) induced from the moment and shear forces of the beam plastic
hinge are transferred to the front plate through the reactions induced
in the supports. A portion of the total moment (m2) is transferred to
the concrete column via the shear connectors on the front plate, and
the remaining portion (m1)is transferred to the side plate via the weld
connection of the front plate to the side plate, as shown in Fig. 9. Addi-
tionally, the out-of-plane stresses are applied to the front plate because
of the shear deformation of the side plates, leading to the formation of
the concrete diagonal strut. Because the out-of-plane resistance in-
cludes a small portion of the total moment relative to the other mecha-
ð11Þ nisms, especially in the absence of the axial forces of the columns at the
roof level, a portion of the out-of-plane resistance in load transfer is con-
servatively ignored in the design process.
Similar to the first specimen, the portions of the shear connec-
tors on the side plate (m1) and front plate (m2) of the total moment
are calculated from Eqs. 8 and 9. The evaluation of the bar shear
connector strength as a shear transfer component is explained in
Section 3.3.
 S.R. Mirghaderi et al. / Journal of Constructional Steel Research 118 (2016) 105–119 111

3.2.4. Design of the cover plates Table 1.

The thickness of the front plate is determined based on the reaction Mechanical properties of the tested steel coupons.

forces (F) in the plate. Therefore, the thickness of the front plate is equal Test Member Yield Tensile Elongation
to the maximum value resulting from Eqs. (12) and (13). In these equa- spec. strength strength (%)
tions, the effect of the weld between the beam web and the front plate is (MPa) (MPa)

ignored to obtain safer and more conservative results [23]. 1 Beam flange (UNP220) 321 463 27.6
1 Beam web (UNP220) 339 471 30.1
2 Beam flange (IPE220) 345 520 27.6
F
t1 ¼ ð12Þ 2 Beam web (IPE220) 379 505 21.5
Fy L1 1,2 Cover plate (thickness = 15 mm) 329.6 474.5 30.2
1,2 Bending reinforcement φ18 bar 544.4 687.4 16
1,2 Stirrup φ10 bar 426.8 660.4 19
F
t2 ¼ ð13Þ
0:55Fy L2

3.3.2. Test program for the shear connector

L1and L2 are illustrated in Fig. 8. Required surface for placing shear
connectors for transfer of shear forces at the steel–concrete interface,
and the dimensions of the beam section are determining parameters
on the cover plates' height (including front plates and side plates).
Furthermore, the weld connection between the side plate and front
plate should be assessed based on the portion of the total moment that
is transferred to the side plate (m1), thus inducing parallel shear stresses
on the weld connection, as described in Section 3.1.4. It should be noted
that distribution of the shear stresses induced in the side plates occurs in
such a manner that they are primarily transferred over a length equal to
the beam height. Therefore, the thickness of the cover plates and welds
should be compared against the shear stresses in this area.
3.3. Strength estimation of the bar shear connector
Various types of shear connectors are used to transfer the shear
forces on the contact surface of the steel and concrete in many compos-
ite components (including composite beams, bridges, etc.). Two fea-
tures of bar shear connectors make them more appropriate than other
types of shear connectors for application in the panel zone of a joint
composed of steel and concrete. First, because of the large amount of
shear forces in the steel–concrete interface of a composite joint, bar
shear connectors with high strength and rigid behavior can prevent
slippage at the steel–concrete interface, which is essential for a rigid
moment connection. Second, the small dimensions of these shear con-
nectors make them a feasible option for use in the limited conditions
of the panel zone with minimum interruption to the reinforcement ar-
rangement of the column. Therefore, bar shear connectors are used on
the cover plates inside the column to transfer the induced shear forces
to the concrete column in both specimens.
3.3.1. Provisions of the design code
Shear strength of a bar shear connector embedded in a solid concrete
slab is determined according to EC4 formula ENV 1994-1-1 [25] Clause
6.3.4 as follows,
To estimate the shear connector's strength for the purpose of design-
ing the proposed connection, two push-out specimens with plain
concrete and reinforced concrete (RC) were tested under monotonic
loading, as shown in Fig. 11(a). The other features of the two specimens
were identical. The geometric properties of the push-out specimens are
shown in Fig. 11(b). Bar shear connectors with dimensions of 30 ×
20 × 50 mm were welded on two opposite sides of a built up box
with dimensions of 300 × 300 × 450 mm that was filled with concrete.
To avoid dominant weld failure, the surroundings of the bar shear con-
nector were welded to the steel plate by fillet welds. The specimens
were built such that the steel box and concrete block could move rela-
tive to each other up to 50 mm under applied load. The steel box was
built from ST37 steel sheets. The concrete was normal concrete with
compression strength of 32 MPa. In the specimen reinforced with steel
bars, as illustrated in Fig. 11(b), six longitudinal bars with a 14 mm di-
ameter were closed with two hoops of steel bars with a 10 mm diame-
ter. The middle reinforcing bars were placed exactly in front of the shear
connectors to investigate the effect of the reinforcing bars on the shear
connector's strength and the failure mode.
The applied load versus relative displacement between the steel box
and concrete is presented in Fig. 12 for both specimens. The results indi-
cated that the load–displacement curves are relatively identical for
specimens, especially in the linear region. According to the results,
each shear connector experienced a relative displacement of approxi-
mately 1 mm at a load of 170 kN; therefore, a bar shear connector
with small dimensions can provide high strength under the small rela-
tive displacement. In both specimens, the failure mode was concrete
crushing–splitting.
A 1 mm relative displacement is considered to be the maximum rel-
ative displacement, and the load corresponding to this displacement is
considered as the strength of the shear connector in the design process.
According to Eqs. (8)–(10), the total shear forces mobilized across
the steel and concrete interface on the front plate (f2) were 1078 and
1204 kN in the first (SPU) and second (SPI) specimen, respectively. If

Prd ¼ ηA f1 f ck =γc ð14Þ

sffiffiffiffiffiffiffiffi
A f2
η¼ ≤2:5 ð15Þ
A f1

Where:
fck is the compressive strength of concrete, Af1 and Af2 are the area of
the front surface and the area of the front surface enlarged at a slope of
1:5 to the rear surface of the adjacent connector, as shown in Fig. 10, and
γc is Partial safety factor for concrete [25].
Because the EC4 formula gives the strength of each shear connector
in an unconfined condition and the shear connectors are employed in
the confined condition of the joint with cover plates in the test speci-
mens of this paper, the strength obtained from EC4 formula is too con-
servative for this application. Fig. 14. Uniaxial tension tests of steel beam's web and flange.
112 S.R. Mirghaderi et al. / Journal of Constructional Steel Research 118 (2016) 105–119

Fig. 15. A photo of the second experimental specimen and its test setup at the end.

the front plate portion of the total shear force is distributed uniformly
between the shear connectors on the front plate, each shear connector
transfers approximately 135 and 150 kN to the concrete in the first
(SPU) and second (SPI) specimen, respectively. According to the push-
out tests results, we expect that these induced forces cause a relative
displacement of less than 1 mm. Because the shear connectors in the
joint area of the proposed connection are applied under slightly differ-
ent conditions (concrete with higher strength, shear connector with
greater length, small distance between shear connectors in a group, cy-
clic loading), thus affecting the test result by making changes to the

Fig. 16. Test setup configuration (dimensions are in mm).

 S.R. Mirghaderi et al. / Journal of Constructional Steel Research 118 (2016) 105–119 113

Fig. 17. Instrumentation of the test specimens (dimensions are in mm).

strength and stiffness of each shear connector [26–29], the results of the beam flanges and continuity plate along the doubler plates on both
push-out tests are used for primary estimation. Moreover, further tests sides of the beam web.
are required to achieve an accurate formula.

4. Experimental program 4.2. Material properties

4.1. Test specimens The mean compressive strengths of the concrete were 38 MPa and
35 MPa for the first and second specimens, respectively. The mechanical
Two interior RCS connections with 3/4 scales were constructed for properties of the steel materials used in construction of the specimens
the experimental program. As shown in Fig. 13(a) and (b), each speci- are listed in Table 1. The stress–strain relationships resulting from uni-
men consisted of concrete column with dimensions of 400 × 400 mm axial tension tests on four coupons of the steel materials are shown in
reinforced with twelve Φ18 steel bars. Φ10 bars were used for column Fig. 14.
stirrups. Due to the confining effect of the cover plates, there was no
need to stirrup in the joint region. Normal concrete was employed in
both specimens. The column height was 3000 mm and the beam span
was 4000 mm with the cross sections of 2UNP 220 and 2IPE220 in the
first and second specimens, respectively.
The side plate was welded to the front plate using complete joint
penetration groove welds. In the first specimen (SPU), the channel
beams at the conjunction of the web and flange were trimmed, and
the beam flanges were directly welded to the cover plates through lon-
gitudinal penetration groove welds. Additionally, the beam web was
welded to the side plate edges with fillet welds. In the second specimen,
a steel plate with a special shape was used as the front plate, as present-
ed in Fig. 13(b). The beams were placed between the protruding sec-
tions of the front plates, and four weld lines between the beam flanges
and beveled edges of the front plates provided a load path for transfer
of the beam demands to the column. Furthermore, the weld connection
of the continuity plates to the front plate provided an extra path to con-
servatively transfer the shear forces.
In both specimens, continuity plates were placed along the axis of
the front plates. In the first specimen, the side plate and the beam web
between the continuity plates together form the panel zone. In the sec-
ond specimen, the panel zone consists of the doubler plates and the
beam web between two continuity plates.
To avoid deformation of each beam separately under lateral loading,
vertical constraints were used to restrict the beams' lateral deformation
relative to each other. Vertical constraints were welded to the beam
webs at the location of the plastic hinges and in the middle of the
beam length on both side of the column.
To locate the plastic hinges at an appropriate distance from the col-
umn face and to avoid damage to the welds in the connection region,
the side plates were extended 25 mm from each side of the panel
zone in the first specimen. For the same purpose, in the second speci-
men, two vertical plates with a length of 30 mm were welded to the Fig. 18. Hysteretic curves for the specimens.
114 S.R. Mirghaderi et al. / Journal of Constructional Steel Research 118 (2016) 105–119
4.3. Test setup and loading pattern
The test setup, as shown in Fig. 15, was provided according to the
boundary conditions of an interior beam to column connection of a mo-
ment frame under cyclic lateral loading. The column was supported at
the base by a real hinge; roller supports were used at the beam ends
to provide free movement in the direction of the lateral loading. Two
lateral supports were placed between two parallel beams on each side
of the column to restrict out-of-plane displacement. The lateral supports
were set 0.7 m and 1.25 m from the column centerline, which are less
than the maximum spacing of the beam lateral support. In addition, a
lateral support was used at the top of the column to restrain it in the pri-
mary plane. The geometry of the test specimens and the configuration
of the test setup are depicted in Fig. 16.
Cyclic lateral displacement was applied using two 500 kN compres-
sion actuators at each side of the column tip. The cyclic displacements
were repeated twice at 0.25%, 0.375%, 0.5%, 0.75%, 1%, 1.5%, 2%, 3%, 4%,
5%, 6%, 7% and 8% total story drift angles. Two load cells were located be-
tween the actuators and the column tip on each side of the column to
record the applied loads. Two LVDTs were installed to measure the lat-
eral displacement at the same point.
To measure the shear deformations of the panel zone, two diagonal
LVDTs (Nos. 1 and 2) were installed as shown in Fig. 17(a) and (b).
Moreover, to collect the information required to calculate the connec-
tion rigidity, two horizontal LVDTs (Nos. 3 and 4) were placed at the
Fig. 19. The first specimen at 8% drift. (a) Concrete crack propagation & beam flanges and web local buckling, (b) beam lateral torsional buckling from a top view, (c) weld connection of the
beam to the side plate, (d) white wash in the steel–concrete interface.
top and bottom of the cover plate parallel to the beams that were con-
nected to the center of the column by rigid steel bars. LVDT No. 5 was
installed to measure the relative displacement between the steel cover
plates and concrete. Rosette and linear strain gauges were used at spe-
cial locations of the test specimens, as depicted in Fig. 17(a) and (b). Fi-
nally, the specimens were coated with a thin layer of lime to provide
better visibility of the progression of yielding and concrete cracking dur-
ing the test. To consider the effect of the gravity loads, a constant axial
force was applied on the column using four pre-stressed bars out of
the column. The axial force applied to the column was equal to
500 kN, as measured by a load cell located at the top of the column.
4.4. Test results
4.4.1. General test observations
In both specimens, as expected, yielding occurred at the beam ends
at a small distance from the column because the specimens were de-
signed according to strong column and weak beam criteria. The lateral
loads versus story drift angle of the first and second specimens are pre-
sented in Fig. 18(a) and (b). Both specimens were loaded up to 0.08 rad
story drift angle, and the hysteretic loops showed stable behavior and
good energy dissipation capacity. The specimens did not exhibit any
strength degradation until 7% drift, which exceeded the AISC seismic
provisions [24] requirement for qualifying a connection for special
moment-resisting frames. No yield or fracture was observed in the
 S.R. Mirghaderi et al. / Journal of Constructional Steel Research 118 (2016) 105–119
cover plates and welds during the test. General test observations for the
first and second specimens are described in the following subsections.
4.4.1.1. First specimen (SPU). This specimen began to yield at 1% story
drift angle in the beam flanges near the edges of the side plates. During
the first cycle of 3% story drift, slight flexural cracks appeared in the con-
crete column at the bottom and top of the cover plates and propagated
slightly at 4% story drift. In addition, during the second cycle of 3% story
drift, slight sliding was observed via cracks that occurred in the white-
wash on the steel cover plates and concrete interface. The yielding ex-
tended approximately 270 mm along the beam flanges at 4% story drift.
The beam flanges and web displayed local buckling at 5% and 6%
story drift, respectively. Slight lateral torsional buckling of the beams
was initiated at 7% story drift. The test was terminated due to restricted
laboratory conditions after 8% story drift. Fig. 19(a)–(d) illustrate the
first specimen at 8% story drift.
4.4.1.2. Second specimen (SPI). In the second specimen, the beam flanges
initiated yield at 1% story drift in the vicinity of the cover plates imme-
diately after the reinforcing web plates, and slight cracks appeared in
the concrete column at the bottom and top of the cover plates at 2%
story drift and propagated slightly up to 4% story drift. At 4% story
drift, slight sliding was observed at the cover plates and concrete inter-
face. Local buckling at the beam flanges and web began at 6% and 7%
story drift, respectively. At 8% story drift, the local buckling increased
Fig. 20. The second specimen at 8% drift (a) Concrete crack propagation & beam flanges and web local buckling, (b) beam lateral torsional buckling from a top view, (c) weld connection of
the beam to the front plate, (d) white wash in the steel–concrete interface.
115
significantly, and slight lateral torsional buckling of the beams occurred.
Finally, the test was terminated due to fracture of the beam compres-
sion flange at 8% story drift during unloading. Fig. 20 (a)–(d) illustrate
the second specimen at 8% story drift.
4.4.2. Behavior of the plastic hinge
As demonstrated in Figs. 19(a) and 20(a), in both specimens, plastic
hinges were concentrated in the beams in the vicinity of the column
faces, and the cover plates, welds and panel zones remained elastic
throughout the test. Strength degradation occurred in the hysteretic di-
agram at 8% and 7% story drift angle (Fig. 18(a) and (b)) due to the ex-
pansion of the beam local buckling that was initiated at 5% and 6% story
drift in the first and second specimens, respectively. Because two beams
with less depth were used in the proposed connections instead of one
equal beam, the level of plastic strain in the beam flanges is lower,
and consequently, beam local buckling and the subsequent strength
degradation occur at higher drifts. In the early stages of the beam local
buckling, the strength degradation was compensated by a strength in-
crease due to strain hardening, and no strength reduction was observed
before 7% story drift.
The magnitudes of the moments in the plastic hinges at the maxi-
mum lateral loads were 231 and 258 kN·m for the first and second spec-
imens, respectively, as presented in Table 2. These amounts are,
respectively, 1.36 and 1.3 times greater than the theoretical beam plas-
tic moment obtained according to the yield stress of the steel beam
116 S.R. Mirghaderi et al. / Journal of Constructional Steel Research 118 (2016) 105–119
Table 2
The strength and stiffness of the specimens based on the experimental results.
Specimens Experimental results
Maximum lateral load (kN)
First: SPU 180
Second: SPI 201
material. This increase in the plastic hinge moment occurred due to the
increase of the strain beyond the yield point and subsequently led to
strain hardening behavior in the steel.
Normalized strains in the location of the strain gauges on the beam
flanges and web versus story drift angle are shown in Fig. 21(a) and
(b) for both specimens. Strains are normalized with the yield strain of
the steel beam material. These figures show that yielding began at 1%
story drift for both specimens in the beam flanges at a small distance
from the column, and plasticity developed in the beam web at 3% drift
angle, as measured using a pasted strain gauge located slightly higher
than the neutral axis of the beam section. In addition, a strain reduction
for strain gauges Top2 and Bot2 was observed at 7% and 8% story drift
angles for the first and second specimen, respectively, due to lateral tor-
sional buckling.
4.4.3. Behavior of the cover plates
To examine the behavior of the cover plates, two strain gauges were
installed horizontally at a small distance from the weld line of the beam
flange to the side plate in specimen SPU, and a strain gauge was installed
vertically on the front plate in the vicinity of the beam flange connection
in specimen SPI, as depicted in Fig. 17(a) and (b). The variations in
strains are presented in Fig. 22(a). The strain values are normalized
against the steel nominal yield strain. In the mentioned region in
Fig. 21. Envelope of the normalized strains along the beam length versus total story drifts.
(a) The first specimen, (b) the second specimen (see Fig. 15 for numbering of the strain
gauges).
Theoretical results
Beam moment at the maximum load (kN·m) Beam nominal moment (kN·m)
231 170
258 197
specimen SPI, the amount of strain shows that the front plate was opti-
mally designed based on the plastic moment value of the beam and that
the strains are limited to the yield point of the steel material of the cover
plates. In specimen SPU, due to neutralization of the tensile and com-
pressive forces in the middle of the weld line, the strain gauges show
small values, as expected.
In both specimens, a strain gauge was installed in the vertical direc-
tion on the front plate, slightly higher than the middle shear connector,
as depicted in Fig. 17(a) and (b). These strain gauges measured the
strain mobilized in the front cover plate due to the shear connectors.
As shown in Fig. 22(b), strain values are limited to 0.42 and 0.28 for
the first and second specimens, respectively. The cover plates act as con-
finement for the concrete in the connection zone; moreover, yielding of
these plates leads to total failure of the connection region. Therefore, to
avoid excessive sliding in the steel and concrete interface and also to
provide reliable confinement for the concrete in the connection region,
the cover plates were preserved as elastic.
4.4.4. Behavior of the panel zone
According to Fig. 17, two diagonal LVDTs were installed in the panel
zone of the specimens to record the shear deformations. Fig. 23(a) and
(b) present diagrams of the total story drift that resulted from the
shear deformation of the panel zone for both specimens. The maximum
story drifts due to the panel zone deformation were 0.47% and 0.52% at
8% story drift for specimens SPU and SPI, respectively.
Fig. 22. Normalized strains of the cover plates versus total story drifts (a) on the front plate
behind shear connectors, (b) near the connection of the beam flange to the column (see
Fig. 15 for numbering of the strain gauges).
 S.R. Mirghaderi et al. / Journal of Constructional Steel Research 118 (2016) 105–119 117

Fig. 25. Relative displacement on the steel and concrete interface versus total story drifts
for both specimens.

4.4.5. Steel-concrete relative displacement in the connection zone

Sliding in the concrete and steel interface was measured using a
LVDT placed vertically at the top of the cover plate and connected to
the concrete column immediately after the cover plate for both speci-
mens. Fig. 25 shows that the relative displacement between steel and
concrete increased gradually with increasing applied displacement up
to 7% story drift. Subsequently, due to a decrease in connection forces,
the relative displacement slightly decreased. At 3% story drift, when
the beam full plastic moment was developed, the relative displacement
was approximately 0.5 mm for both specimens. This amount of small
sliding did not cause any pinching in the hysteresis diagram.

Fig. 23. Total story drift developed by shear deformation of the panel zone. (a) First spec-
imen, (b) second specimen.

The results showed that the connection components, including the

panel zone, cover plates and welds, remained relatively elastic during
the test and that the plasticity was concentrated only in the beams.
In both specimens, a rosette strain gauge was used on the beam web
in the panel zone to determine the von Mises strain (Fig. 17). Fig. 24 il-
lustrates the normalized von Mises strain versus story drift angles of the
specimens. As shown in Fig. 24, the panel zone of specimen SPU
remained elastic up to 8% story drift, whereas specimen SPI experienced
slightly inelastic behavior in the panel zone. In specimen SPU, the panel
zone includes the beam web and the side plate, which is involved with
the concrete via shear connectors. In fact, certain regions of the concrete
in the vicinity of the shear connectors participate in the panel zone
shear forces, which were ignored in the calculation of the panel zone
strength.

Fig. 26. (a) Normalized strains of the column reinforcements versus total story drifts for
specimen SPI. (b) The location of the strain gauges on the corner and middle reinforce-
Fig. 24. Strain response in the panel-zone for both specimens. ment in specimen SPI.
118 S.R. Mirghaderi et al. / Journal of Constructional Steel Research 118 (2016) 105–119

4.4.6. Behavior of the column reinforcements 5. Conclusion

Fig. 26(a) represents the normalized measured strains of the rein-
forcements on the front side of the column and 12 cm inside the cover
plates, as shown in Fig. 26(b). According to Fig. 26(a), the column rein-
forcements experienced strains up to 85% yield strain. The results veri-
fied that the column was optimally designed based on the beam
plastic moment.
4.5. Evaluation of the connection rigidity
The rigidity of the proposed connection evaluated based on AISC
Specification [30]. The secant stiffness of the connection (KS) at the ser-
vice loads is taken as follows,
In this research program, a new moment connection for steel beam
to reinforced concrete column (RCS) is proposed. The seismic behavior
of the proposed connection was studied through two experimental
tests in 3/4 scale under cyclic loading. In the proposed connection,
two parallel beams pass from both sides of the column and are welded
to the cover plates surrounding the concrete column in the joint area.
Bar shear connectors are used on the contact surface of the steel and
concrete to restrict sliding. To clarify the bar shear connector behavior
embedded in the confined concrete by steel cover plates, two push-
out specimens were tested under monotonic loading. The results of
this study are described as follows:

MS • Because the steel beams pass outside the column, both beams and col-
KS ¼ ð16Þ
θS umn are continuous in the joint area to provide an appropriate ar-
rangement of the reinforcement bars and continuous access inside
8 the column.
>
> L
>
> KS≥20 fully restrained connection • The stress transfer mechanisms described in the paper are composed
>
< EI
L of three mechanisms including torsional resistance of the side plates,
2bKS b 20 partially restrained connection ð17Þ
>
> EI in-plane and out-of-plane resistances of the front plates. The torsional
>
> L
>
: KS ≤2 simple connection resistance of the side plates is provided by shear connectors which are
EI
welded on the side plates and mobilize the concrete through torsional
deformation. The in-plane resistance of the front plates is provided by
Where MS is the moment at service loads level and θS is the connec- the shear resistance of the shear connectors that are welded on the
tion rotation at service loads. L and EI are the length and bending rigid- front plates. The out-of-plane forces apply compressive stresses to
ity, respectively, of the beam [30]. the concrete leading to formation of a concrete diagonal strut via the
In order to determine the classification of the proposed connection, contribution of the compressive stresses from the column. The resul-
the moment-rotation diagrams of the specimens were employed as tant of these resisting mechanisms reliably transfers the moment
shown in Fig. 27(a) and (b). Two LVDTs (Nos. 3 and 4) measured the ro- and shear forces of the beam's plastic hinges to the reinforced con-
tation of the connection relative to the center of the column. The initial crete column.
values of the secant stiffness of the specimens revealed that the pro- • The proposed design method was arranged properly such that plastic
posed detail can be considered as a fully restrained connection. hinges were formed in the beams in the vicinity of the column, and
the panel zone and cover plates remained elastic during the test; only
negligible damage (i.e., slight cracks) was observed in the column.
• According to the test results, both specimens sustained an 8% story drift
with stable hysteretic loops and high energy dissipation capacity. The
specimens did not exhibit any strength degradation until 7% story
drift angle, which exceeded the AISC seismic provision requirements
for special moment-resisting frames.
• It is notable that in both connections, all of the forces resulting from the
beam plastic hinges occur in the cover plate planes, and all welds are
loaded parallel to the axis; consequently, no yield or fracture occurred
in the cover plates and welds throughout the test.
• Evaluation of the connection rigidity revealed that the proposed con-
nection with high initial stiffness could be classified as a fully restrained
connection between steel beams and reinforced concrete columns.
• According to the push-out test results, the bar shear connector exhibit-
ed high strength with small relative displacement. The bar shear con-
nector with high initial stiffness is a proper component for restricting
relative displacement in the steel–concrete interface. Additionally,
their small dimensions make bar shear connectors a feasible option
for use in the joint region of the proposed RCS connection. Despite the
acceptable tests results, additional research on the behavior of the bar
shear connector in a confined condition is required to propose an accu-
rate design formula.
• The test results on the RCS connection revealed that bar shear connec-
tors employed in the joint region inside the cover plates could effective-
ly restrict the relative displacement between steel and concrete to
0.9 mm at 8% story drift angle.

Acknowledgments

The authors would like to appreciate Padena Hoor Company for pro-
Fig. 27. Determination of the connection rigidity (a) first specimen, (b) second specimen. viding materials and construction costs.
 S.R. Mirghaderi et al. / Journal of Constructional Steel Research 118 (2016) 105–119
References
[1] J. Men, Y. Zhang, Z. Guo, Q. Shi, Experimental research on seismic behavior of a com-
posite RCS frame, Steel Compos. Struct. 18 (4) (2015) 971–983.
[2] C.T. Cheng, C.C. Chen, Seismic behavior of steel beam and reinforced concrete col-
umn connections, J. Constr. Steel Res. 61 (2005) 587–606.
[3] T.M. Sheikh, G.G. Deierlein, J.A. Yura, J.O. Jirsa, Beam–column moment connections
for composite frames: part 1, J. Struct. Eng. ASCE 115 (11) (1989) 2858–2876.
[4] G.J. Parra-Montesinos, J.K. Wight, Seismic response of exterior RC column-to-steel
beam connections, J. Struct. Eng. ASCE 126 (10) (2000) 1113–1121.
[5] M.N. Bugeja, Seismic Behavior of Composite Moment Resisting Frame Systems.(PhD
dissertation) Texas A&M University, USA, 1999.
[6] G.G. Deierlein, H. Noguchi, Overview of U.S.–Japan research on the seismic design of
composite reinforced concrete and steel moment frame structures, J. Struct. Eng.
ASCE 130 (2) (2004) 361–367.
[7] ASCE Task Committee on Design Criteria for Composite Structures in Steel and
Concrete, Guidelines for design of joints between steel beams and reinforced con-
crete columns, J. Struct. Eng. ASCE 120 (8) (1994) 2330–2357.
[8] S. Alizadeh, N.K.A. Attari, M.T. Kazem, The seismic performance of new detailing for
RCS connections, J. Constr. Steel Res. 91 (2013) 76–88.
[9] S. Alizadeh, N.K.A. Attari, M.T. Kazem, Experimental investigation of RCS connections
performance using self-consolidated concrete, J. Constr. Steel Res. 114 (2015)
204–216.
[10] Q. Zhang, Q. Yu, J. Guan, Z. Wu, Seismic behavior of confined RC column-composite
beam joints, Trans. Tianjin Univ. 20 (3) (2014) 174–181.
[11] W. Li, Q.N. Li, W.S. Jiang, Parameter study on composite frames consisting of steel
beams and reinforced concrete columns, J. Constr. Steel Res. 77 (2012) 145–162.
[12] B. Farahmand Azar, H. Ghaffarzadeh, N. Talebian, Seismic performance of composite
RCS special moment frames, KSCE J. Civ. Eng. 17 (2) (2013) 450–457.
[13] C.C. Chou, J.H. Chen, G. MacRae, Tests and analyses of a full-scale post-tensioned RCS
frame subassembly, J. Constr. Steel Res. 66 (2010) 1354–1365.
[14] H. Noguchi, K. Uchida, Finite element method analysis of hybrid structural frames
with reinforced concrete columns and steel beams, J. Struct. Eng. 130 (2) (2004)
328–335.
[15] I. Nishiyama, H. Kuramoto, H. Noguchi, Guidelines: seismic design of composite re-
inforced concrete and steel buildings, J. Struct. Eng. ASCE 130 (2) (2004) 336–342.
[16] H. Kuramoto, I. Nishiyama, Seismic performance and stress transferring mechanism
of through-column-type joints for composite reinforced concrete and steel frames, J.
Struct. Eng. ASCE 130 (2) (2004) 352–360.
119
[17] W. Li, Q. Li, W. Jiang, L. Jiang, Seismic performance of composite reinforced concrete
and steel moment frame structures — state-of-the-art, Compos. Part B 42 (2011)
190–206.
[18] X. Li, Y. Xiao, Experimental evaluation of steel beam bolted to reinforced concrete
column connection, Key Eng. Mater. 400–402 (2008) 311–319.
[19] G. Li, C. Fang, X. Zhao, Y. An, Y. Liu, Cyclic behavior of rebar-penetrated connection
between gangue concrete filed steel tubular column and reinforced gangue concrete
beam, Adv. Steel Constr. 11 (1) (2015) 54–72.
[20] X. Li, Y. Xiao, Y.T. Wu, Seismic behavior of exterior connections with steel beams
bolted to CFT columns, J. Constr. Steel Res. 65 (2009) 1438–1446.
[21] I.S. Sheet, U. Gunasekaran, G. MacRae, Experimental investigation of CFT column to
steel beam connections under cyclic loading, J. Constr. Steel Res. 86 (2013) 167–182.
[22] G. MacRae, C. Roeder, C. Gunderson, Y. Kimura, Brace–beam–column connections
for concentrically braced frames with concrete filled tube columns, J. Struct. Eng.
130 (2004) 233–243.
[23] S.R. Mirghaderi, R.M. Dehghani, The rigid seismic connection of continuous beams to
column, J. Constr. Steel Res. 64 (2008) 1516–1529.
[24] AISC/ANSI 341-10, Seismic Provisions for Structural Steel Buildings, American Insti-
tute of Steel Construction, Inc., Chicago (IL), 2010.
[25] ENV1994-1-1 Eurocode4, Design of composite steel and concrete structures. Part
1.1: General rules and rules for buildings, European Committee for Standardization,
Bruxelles, 1994.
[26] J. Qureshi, D. Lam, V. Jianqiao, The influence of profiled sheeting thickness and shear
connector's position on strength and ductility of headed shear connector, J. Constr.
Steel Res. 33 (2011) 1643–1656.
[27] J. Qureshi, D. Lam, V. Jianqiao, Effect of shear connector spacing and layout on the
shear connector capacity in composite beams, J. Constr. Steel Res. 67 (2011)
706–719.
[28] E. Baran, C. Topkaya, An experimental study on channel type shear connectors, J.
Constr. Steel Res. 74 (2012) 108–117.
[29] M. Shariati, N.H.R. Sulong, M.M. Arabnejad khanouki, Experimental assessment of
channel shear connectors under monotonic and fully reversed cyclic loading in
high strength concrete, J. Constr. Steel Res. 34 (2011) 325–331.
[30] AISC ANSI 360-05, Commentary on the Specification for Structural Steel Buildings,
American Institute of Steel Construction, Inc., Chicago(IL), 2005.