Engineering Structures 205 (2020) 110006

Contents lists available at ScienceDirect

Engineering Structures
journal homepage: www.elsevier.com/locate/engstruct

T
Introducing an easy-install precast concrete beam-to-column connection
strengthened by steel box and peripheral plates
Jamshid Esmaeili , Neyram Ahooghalandary
⁎

Department of Civil Engineering, University of Tabriz, 29 Bahman Avenue, Tabriz, Iran

A R T I C LE I N FO A B S T R A C T

Keywords: Precast reinforced concrete frame is one of the industrialized building systems (IBS)s which assembling lim-
Beam-to-column connection itation in its precast beam-to-column connection (PBC) causes many problems and leads to design of pinned
Pseudo dynamic analysis connections inappropriate for moment resisting frames (MRF)s in high seismic zones.
Moment resisting frame Previous PBCs for MRFs in precast reinforced concrete structures (PRCS)s are classified in two “dry” and
Precast connection
“semi-monolithic” connections. Despite of the easy installation procedures, most of dry connections have not
Finite element analysis
showed acceptable performance in ductility and energy dissipation. On the other hand, semi-monolithic con-
nections require long pending time for casting and scaffolding in construction site and lead to reinforcement
congestion in beam-column joint. Also most of the precast beams in the both approaches have not appropriate
bearing surface to create desirable beam-to-slab connection condition.
The current study proposes a simple and efficient precast connection with suitable conditions for beam-to-slab
connection, which could be also used in MRFs, through combining two approaches. To accomplish this, two
external beam-to-column joints on 2/3 scale were designed in precast and in-situ forms. In the following, the
proposed PBC connection was compared using the reverse cyclic test with the in-situ beam-to-column connection
(IBC). The test results demonstrated that the proposed PBC had a higher load-carrying capacity, energy dis-
sipation and ductility than those of the equivalent IBC and met all the seismic requirements of ACI374.1-05
criteria for moment resistant connections. Finally, the connections were simulated with finite element method
(FEM) in ATENA software to develop the test results. At this stage, the finite element analysis predictions for
load-carrying capacity, cracking pattern and strain values indicated good conformance with the test results and
then the effect of parameters such as compressive strength of concrete, axial force of column and beam end
plates length were investigated. An increase in concrete strength and axial force caused to more load-carrying
capacity and strength degradation in the beam-column joints, respectively.

1. Introduction to create structural integrity unlike the in-situ one. Therefore, PBCs play
an effective role in these frames. However, due to assembling and
IBS is an approach for building construction in which structure transportation limitations, providing a PBC that can accommodate a
components are prefabricated partially or generally in off-site and as- capacity level similar to an IBC is difficult. Precast concrete frames
sembled in construction site. Prefabricated steel structures are one of failures during past earthquakes often originated from the unacceptable
the conventional IBSs with lighter mass, higher cost and especial performance of beam-to-column connection [1,2]. In most precast
maintenance during building life time. On the other hand, conventional frames, there are connections with no flexural capacity or semi-rigid
cast in-situ concrete structures require more construction space, long connections that are not solely applicable in MRFs. To overcome this,
pending time and poor quality control. Therefore, PRCSs could be a previous researchers proposed connections which could be classified in
good alternative for in-situ concrete structures in IBS systems. These dry and semi-monolithic types. Dry connection approaches use welded
structures mainly use MRFs as lateral resisting system. Precast concrete plates, bolts and dowels in connection zone. Despite of the easy and
MRFs require a large number of discrete beams and columns assembling quick installation, these connections had less ductility under seismic

Abbreviations: IBS, Industrialized building system; PBC, Precast beam-to-column connection; MRF, Moment resisting frame; PRCS, Precast reinforced concrete
structure; IBC, In-situ beam-to-column connection; FEM, Finite element method
⁎
Corresponding author.
E-mail addresses: [email protected] (J. Esmaeili), [email protected] (N. Ahooghalandary).

https://doi.org/10.1016/j.engstruct.2019.110006

0141-0296/ © 2019 Elsevier Ltd. All rights reserved.

Received 1 July 2019; Received in revised form 25 November 2019; Accepted 26 November 2019
J. Esmaeili and N. Ahooghalandary Engineering Structures 205 (2020) 110006

Nomenclature k4 second stiffness at third cycle of 4% drift ratio, [kN/mm]

Lb clear vertical distance from the column face to the jack
Ag net cross section of column, [mm2] center line, [m]
As total longitudinal bars area, [mm2] Lc length of the column between the two hinge supports, [m]
bc column width, [m] Mb beam flexural strength, [kN·m]
bj effective joint width, [m] Mc connection flexural strength, [kN·m]
c cohesion based on the Mohr-Coulomb theory, [N/mm2] Pmax test maximum resistance of joints, [kN]
d effective depth of the beam cross section, [m] Pn calculated nominal resistance of joints, [kN]
Ec elastic modulus of concrete, [N/mm2] P4 resistance of joint at third cycle of 4% drift ratio, [kN]
EI actual energy dissipation of per cycle, [kN·m] P4max maximum resistance of joint at 4% drift ratio, [kN]
EII elastic-plastic energy dissipation of per cycle, [kN·m] S1 minimum slip of rebars corresponding to peak strength in
Es initial elastic modulus of steel parts, [N/mm2] bond-slip model, [mm]
Esh strain hardening modulus of steel parts, [N/mm2] S2 maximum slip of rebars corresponding to peak strength in
fc’ compressive concrete strength, [N/mm2] bond-slip model, [mm]
fcu’ compressive strength of concrete cube test, [N/mm2] S3 minimum slip of rebars corresponding to failure strength
ft’ tensile strength of concrete, [N/mm2] in bond-slip model, [mm]
fu ultimate strength of steel parts, [N/mm2] Vjmax maximum joint shear, [kN]
fy yield strength of steel parts, [N/mm2] Vn nominal joint shear capacity, [m]
Gf specific fracture energy, [N/mm3] wd critical compressive displacement, [mm]
hc column depth, [m] Δ beam end displacement, [m]
J2 second invariant of stress deviator tensor, [N2/mm4] Δu ultimate displacement, [mm]
jd moment arm length between the centerline of tension and Δy yielding displacement, [mm]
compression areas, [m] ɛf strain in fracture stress of steel parts, [–]
ki initial stiffness of joints, [kN/mm] ϕ coefficient of internal friction based on the Mohr-Coulomb
kp post-yield stiffness of joints, [kN/mm] theory, [–]
ksec second stiffness of joints, [kN/mm] τf failure shear strength in bond-slip model, [N/mm2]
ky yield stiffness of joints, [kN/mm] τmax maximum shear strength in bond-slip model, [N/mm2]

25 25 25 25
300 300
25 300 25

25 300 25
8 125

8 125

8 16 8 16
1 1
1 1
1150mm

1150mm

Section 1-1 Section 1-1

8 100

8 100

2 Connection detail A 2
8 75 8 150 8 75 8 150
2000mm

2000mm
8 100

8 100

25 250 25 25 25
150 2
1150mm

1150mm

2
350 25

25
200 150

1 1 3 20 1 1 2 25
3 12
8 125

8 125

4 10+2 6
(M1) (P1)
25
25

250
Section 2-2 25 25
Section 2-2
2000mm 2000mm
2300mm 2300mm

L60x60x6 In-situ concrete

L=150mm 4
Anchoraged Side plate Steel box 6 320 Seat plates In-situ concrete
bar 300
2PL250x100x12
25 150 25 150

Side plates
3PL276x250x6
15 0

Steel angle
Steel box 2PL300x250x12 Bottom plate
6 250 6 250 Bottom plate
150

Seat plates
Bottom plate Steel box
3 Seat plate 3 Section 3-3 100 70 100
4
50 125 125 50
Connection detail A 50 100 50 Section 4-4
75 100 75

6Stud D14
50 100 50

Longitudinal bar In-situ concrete 4Stud D10 4 10

PL350x250x6
Transverse bar Steel segment PL200x200x6

Precast concrete Headed stud of plates Bottom plate Side plate Bottom bars layout

Headed stud of steel box (for high positive moment condition)

Fig. 1. Details and dimensions of joints.

2
J. Esmaeili and N. Ahooghalandary Engineering Structures 205 (2020) 110006

efforts [3,4] and usually failed at welded parts and steel billets [5]. way that dry parts remained elastic and plastic hinge was located in the
Modularized hybrid system (MHS) is one of the recent dry con- precast concrete beam to achieve higher ductility under cyclic loads. It
nections using steel profiles as beam and column cores. Hong et al. [6] is similar to the approaches used in reduced steel beam section (RBS)
indicated that utilization of steel components confined by concrete in- [16,17], accordion-web reduced beam section (AW-RBS) [18] and
crease probability of shear failures in large displacements at contacts. double reduced beam section (DRBS) connections [19] in prefabricated
Because steel cores have less dimensions in comparison of beam or steel and wooden structures [20]. Furthermore, the embedded bearing
column cross section, these connections usually utilize reinforcing bars surfaces for locating the precast slabs on the prefabricated part of beam
beside steel cores to reach enough capacity [7]. Applying reinforcing provide a condition for casting the semi-monolithic part without con-
bars and steel cores together in connection zone make installation crete formworks.
process difficult.
Another type of PBCs is based on the use of in-situ concrete. Unlike
2. Introducing a new PBC
the relatively more desirable performance of these connections in
ductility and energy dissipation under cyclic loads, long pending time
In order to provide a new PBC and ensure its performance, this
for scaffolding and installation of concrete formworks are required and
connection was compared with an equivalent IBC in a MRF. To achieve
result in reinforcement congestion in joint [8,9]. Also, using different
this, a symmetrical four stories structure was considered with a story
concrete ages in a connection cause to cold joints and poor reinforcing
height of 3 m and four bays with a length of 6 m. The structure was
bars bond-slip condition in places with thin concrete cover [10].
analyzed based on the seismic requirements of intermediate MRFs that
Some recent studies categorized semi-monolithic connections as low
are conventional for ordinary residential buildings in high seismic
to medium ductility and proposed new stirrups arrangement [11] and
zones. To evaluate the connection performance, an exterior joint of an
ultra-high performance fiber reinforced concrete (UHPFRC) [12] for
interior frame was selected on the third floor of the structure.
more ductility and less pinching in hysteresis curves under cyclic loads.
Depending on the applied efforts, the selected joint was designed in the
Hybrid connection is another type of PBCs used in pre-stressed
precast and the in-situ form based on the regulations of ACI318-14 [21]
structures. In these connections, post-tension cables and reinforcing
and PCI handbook [22] with 2/3 scale. Fig. 1 shows the dimensions of
bars pass together through beam-column joint [13]. Utilization of
the in-situ (M1) and precast joint (P1).
tendons usually is associated with more limited cracking, higher load-
In the precast column, instead of using a corbel in the joint area, a
carrying capacity and less energy dissipation in comparison with con-
steel box is used. The box is welded to the stiffener confined by concrete
ventional PBCs. Externally attached energy dissipaters [14] or replic-
to avoid out of plan buckling. The peripheral plates of the steel box are
able mild steel bars [15] were proposed to increase the amount of en-
welded by groove weld with full penetration. In a condition with high
ergy dissipation for these connections especially in large drifts.
tension, headed studs could be used in the steel box on opposite side of
According to the previous studies more important challenges could
the seat plate to reduce out of plan deformation.
be summarized for PBCs as follows:
The proposed beam consists of two parts. The lower prefabricated
part and the upper in-situ part that is casted in construction site after
1. Lack of an appropriate bedding surface on precast beam for beam-
placement of hollow core slabs. As shown in Fig. 1, the prefabricated
to-slab connection
part of the beam has the side and bottom plates. These plates are an-
2. Requirement for scaffolding and concrete formworks in construction
chored to the concrete of the beam end through the headed studs.
site, especially in semi-monolithic connections
At the installation stage, at first, the seat plates and the web angles
3. Reinforcement congestion in beam-column joints especially in semi-
are welded to the steel box and then the beam prefabricated part is
monolithic connections
placed on the seat plates and through welding of them to the bottom
4. Premature failure and lower ductility especially in dry connections.
plate and the angles to the side plates, the connection of this part is
completed. After installation of the beam prefabricated part, hollow
In this research, a connection is proposed in the dry type for the
core slabs could be placed on it, and the longitudinal bars of the upper
lower part and a semi-monolithic type for its upper part. Utilization of
part pass through the beam U bars as well as the column holes, and are
dry connection in the lower part of beam provides a suitable condition
mechanically anchored on its opposite side. Then casting of the beam
for easy installation procedure without needing scaffolding thereby
upper part is carried out without requiring a formwork, since the lower
saving time and lowering costs. The connection was designed in such a
face is confined by the prefabricated part, and the side faces could be

Fig. 2. Different parts of joint P1.

3
J. Esmaeili and N. Ahooghalandary Engineering Structures 205 (2020) 110006

confined by hollow core slabs. Fig. 2 shows the different parts of joint 4. Loading pattern
P1.
In order to create a rigid diaphragm on story levels and identical Fig. 7 shows the test setup. As can be seen, at two ends of the
lateral displacement of frames under seismic loads, hollow core slabs column, the hinged support was used to provide rotational freedom at
and the precast beams, should have enough integrity through an ap- the turning points (the points in which column bending moment is zero
propriate connection to each other. To achieve this, after installation of under lateral load).
hollow core slabs, the anchorage bars could be located at the specified Two hydraulic jacks were simultaneously used for the tests. The first
intervals in the holes of these slabs and casted at the same time with the jack was used to apply axial force on the column by force control
upper part of the beam (Fig. 3). In another way, to maintain better method and its value was considered 0.1fc’Ag. In this regard, fc’ is the
integrity, hollow core slabs could be used with end U bars. In this compressive concrete strength and Ag is the column net cross section.
method, firstly, the hollow core slabs are placed on the prefabricated The jack centerline was such that it did not have eccentricity with the
part of the beam, and then the upper longitudinal bars of the beam pass column axis and did not cause bending efforts. On the opposite side of
simultaneously through the beam and slab U bars. This part is then the jack, the hinged support was placed as a bearing of the axial force.
casted according to Fig. 4. The second jack was used to apply cyclic forces at the end of the
Following list consists of the advantages of proposed connection in beam and perpendicular to it. The applied load was displacement
precast MRFs: control with a loading pattern shown in Fig. 7. This jack was fixed to
the end of the beam using threaded dowels and a stiffened plate. To
1. No need for cast-in-place concrete formworks increase accuracy and avoid the effect of jack weight on the test results,
2. Appropriate dimensions for transportation and installation condi- the jack was connected to the strong frame with a chain. Figs. 8 and 9
tions show the jacks and their controller systems.
3. Easy assembling due to utilize the seat plates The selected hysteresis loading pattern was similar to the proposed
5. No need for tight tolerances one in ACI374.1-05 [25]. According to this code, the initial drift ratio
6. Establishing a suitable bearing surface on the both sides of the beam should be within the elastic range and the subsequent drift ratios to the
prefabricated part to place hollow core precast slabs previous one are suggested to be about 1.25–1.5 and the test should
7. No need for column corbel and an increase in ceiling height of continue up to minimum 3.5% drift ratio. As shown in Fig. 7, the se-
stories lected experimental loading pattern was severe in comparison with
ACI374.1-05 [25] criteria because the maximum of mentioned ratios
was selected 2 and the test was continued up to minimum 5% drift ratio
3. Materials properties and preparation process: for observation of failure mechanisms. As can be seen, each drift ratio
included three load cycles and in the absence of 3.5% drift ratio in
Concrete, reinforcing bars and steel parts of the experimental joints loading pattern, the performance of the test specimens was evaluated
(specimens) were tested under compression and tensile tests respec- conservatively not less than 4% drift ratio.
tively. The maximum concrete aggregate size used in the specimens was
20 mm. Additionally, grade 60 longitudinal bars and grade 50 trans-
verse bars in accordance with ASTM A615 [23] and steel parts of the 5. Cyclic test results
type S235JRG2 in accordance with European Standard EN10025 [24]
were used in the specimens. Table 1 presents the characteristics of the Fig. 10 shows the hysteresis curves of specimens M1 and P1. The
materials and the beams cross section specifications. drift ratio in the graphs was obtained from the following equation:

θ = ∆ (Lb + 0.5hc )
For convenience, the joints were casted and assembled in a hor-
(1)
izontal plane. Fig. 5a and b show specimen M1 during and after casting.
In this case, unlike specimen P1, the beam and column were casted
In which, Δ is beam end displacement, Lb is clear vertical distance
simultaneously. from the column face to the jack center line and hc is column depth.
In specimen P1, before casting the column, the steel box, the end
The parameters of a hysteresis curves shown in Table 2 consist of
plate and the filler were placed in the formwork. After the concrete three parts: load-carrying capacity, deformation capacity and stiffness.
hardening, the filler was pulled out to create the needed hole. So, the
According to this table, the maximum resistance (Pmax) for specimens
column is full prefabricated. For casting of the lower part of the precast M1 and P1 in push and pull direction (negative and positive moment)
beam, the side and bottom plates were placed in the formwork at first.
occurred at 4% and 5% drift ratio and at 3% and 2% drift ratio, re-
Following the concrete hardening, the narrower upper part of the beam spectively. The nominal resistance (Pn) was obtained from the calcu-
was casted. This part practically will be casted without concrete
lated flexural strength divided by the vertical distance from the column
formwork in construction site. Finally, after the beams and columns face to the jack center line. The results show that maximum resistance is
were prepared, the longitudinal top bars of the beam were passed
0.99–1.51 times the nominal resistance value. The obtained ratio of
through the created holes in the column and were mechanically an- 1.51 for specimen P1 under positive moment is due to the minimum
chored in the opposite face of it; then the holes were filled with a non-
20 cm distance between the beam plastic hinge and column face,
shrink grout. Fig. 6a and b shows the beam and column of specimen P1 thereby shortening of the lever arm.
in the formwork before casting. After preparing the specimens, their
The maximum joint shear (Vjmax) values in Table 2 were obtained
surface area was meshed with a width of 10 cm. (Figs. 5c and 6c).
from the following equation:

Fig. 3. Connection detail of precast beam to

hollow core slabs with anchorage bars.

4
J. Esmaeili and N. Ahooghalandary Engineering Structures 205 (2020) 110006

Fig. 4. Connection detail of precast beam to

hollow core slabs with U bars.

Table 1 ⎧ (bb + bc ) 2
bj = bb + mhc 2
⎨
Material properties and beam cross section specifications.
Material properties: For both specimens ⎩ bc (4)

Steel parts properties: Longitudinal bars Transverse bars Steel plates where bc is the column width and in absence of any eccentricity be-
Yield strength fy (MPa) 505 351 301 tween the beam and the column center line, coefficient of m is equal to
Ultimate strength fu (MPa) 623 524 423
0.5. The results show that in all the specimens, the ratio of the max-
Strain in fracture stress ɛf 15.58 18.15 24.90
(%)
imum joint shear to the nominal joint shear capacity is less than unit.
Elastic modulus Es (MPa) 2.03 × 105 1.71 × 105 1.73 × 105 The deformation capacity parameters of the joints include the yield
Strain hardening Esh (MPa) 1032 1225 656 (Δy) and ultimate displacement (Δu) which were obtained from the
Concrete property: For specimen M1 For specimen P1 envelope curve of the test specimens, are shown in Fig. 11. To obtain
Compressive strength of 49.55 47.40 the deformation parameters, the idealized curve of Fig. 11 was used. In
cube fcu’ (MPa) the post-peak region, according to ACI374.1-05 [25], a displacement
Beam cross section Bottom Top Bottom Top
corresponding to 75% maximum resistance was considered as the ul-
specifications:
Total longitudinal bars 339 942 314 982 timate displacement and in the pre-peak region the displacement cor-
area As (mm2) responding to the yield resistance was obtained based on the equal
Beam flexural strength* Mb 53.26 134.17 47.27 137.78 energy principle through idealizing the actual curve with the bilinear
(kN·m) curve by identical enclosed area. Ratio of the ultimate displacement to
Connection flexural – – 226.84 154.05
strength** Mc (kN·m)
the yield displacement is known as the connection ductility. For idea-
lization of the actual curve after the post-peak region, the linear curve
* Beam flexural strength was calculated based on Whitney rectangular stress was used with equivalent enclosed area. The first line slope of the
distribution theory at beam cross section. idealized curve in the pre-peak region and the line slope of the idealized
** Connection flexural strength was calculated based on passed top bars and curve in the post-peak region were considered as the yield stiffness (ky)
seat plates capacity at column face. and post-yield stiffness (kp), respectively. The kp/ky ratio is shown in

Vj max = Pmax × (Lb jd − (Lb + 0.5hc ) Lc )

Table 2.
(2) Fig. 12a and b show the energy dissipation and cumulative energy
dissipation in terms of drift ratio, respectively. As shown in Fig. 12a, the
where jd is the moment arm length between the centerline of tension maximum energy dissipation in specimens M1 and P1 was achieved at
and compression areas in beam cross section and Lc is length of the 5% drift ratio. At this drift ratio, according to Fig. 12b, the total energy
column between the two hinge supports. Also, the nominal joint shear dissipation in specimen P1 is 1.08 time of that in specimen M1. This
capacity (Vn) was obtained from the following equation (ACI352R-02 ratio is approximately 0.92 at 5% drift ratio of specimen P1 to 6% drift
[26]): ratio in specimen M1. Consequently, specimen P1 has more energy

Vn = γ f ′c bj hc
dissipation than specimen M1 until 5% drift. Utilization of the bottom
(3) and side plates and the steel box resulted in more concrete confinement
in the precast joint and led to better performance in this part.
In which, γ is the coefficient of joint confinement and has unit value The second stiffness (ksec) in term of drift ratio (as shown in
for an exterior joint without slabs and bj is effective joint width, which Fig. 13b) was obtained from the slope of the line passing through the
was obtained from the minimum value of the following equation: peak points at per drift ratio in the first cycle under positive and ne-
gative moments (as shown in Fig. 13a). As can be seen, the most

(a) Beam and column in casting (b) Beam and column after casting (c) Specimen with surface mesh
Fig. 5. Preparation of specimen M1.

5
J. Esmaeili and N. Ahooghalandary Engineering Structures 205 (2020) 110006

(a) Column ready to cast (b) Beam ready to cast (c) Specimen with surface mesh
Fig. 6. Preparation of specimen P1.

stiffness degradation is occurred at the initial drift ratios of the tests. 5.1. Comparison of test results with acceptance criteria in ACI374.1-05
The second stiffness at 6% and 5% drift ratios to initial stiffness in [25]
specimen M1 is 6% and 11%, respectively, and at 5% drift ratio in
specimen P1 is 12%. As Fig. 13b shows, the results indicate more In order to evaluate the cyclic performance of the connections,
second stiffness for specimen P1. Application of dry components in the different parameters of the hysteresis curves of the specimens were
beam prefabricated part and the steel box in the beam-column joint compared with ACI374.1-05 [25] acceptance criteria. These parameters
partially limited the deformations in the connection zone. including strength degradation, energy dissipation and stiffness de-
To check the strength degradation at each cycle, Fig. 14a and b gradation at the third cycle are evaluated at a drift ratio not less than
show the strength ratios of specimens M1 and P1, respectively. The 3.5%. In this study, ACI374.1-05 [25] performance criteria were ap-
strength ratios in the figures were obtained by dividing the strength plied at 4% drift ratio. According to ACI374.1-05 [25], the strength
value at the second and the third cycles of each drift ratio by that of the degradation of the third cycle to first cycle should not be less than 25%
first cycle. The minimum strength ratio under negative and the positive (in same load direction). Also the energy dissipation ratio should not be
moments occurred with a value of 0.71 and 0.27 at 5% and 6% drift less than 0.125. As shown in Fig. 15, this ratio was obtained from the
ratios, respectively, for specimen M1 and 0.87 and 0.47 at 5% drift division of the enclosed area of the hysteresis curve at the third cycle to
ratio, respectively, for specimen P1 corresponding to the third cycles. the enclosed area of the dotted line. Also, the second stiffness at the
The results indicate less strength degradation in specimen P1 compared third cycle should not be less than 0.05 of the initial stiffness at the first
to specimen M1 at their last drift ratios. The considerable strength cycle of the elastic drift ratio. The second stiffness was obtained from
degradations at these drift ratios are directly related to the fracture of the slope of the passing line from −0.10 to +0.10 maximum drift ratio
reinforcing longitudinal bars in the lower part of the beams under po- of the third cycle.
sitive moments. Comparison of the cyclic performance of the specimens with
ACI374.1-05 [25] acceptance criteria is presented in Table 3. As can be
seen, specimen P1, like specimen M1, is able to meet all criteria in this
code.

Fig. 7. Test setup.

6
J. Esmaeili and N. Ahooghalandary Engineering Structures 205 (2020) 110006

(a) Controller system (b) Jack with force control programming

Fig. 8. First hydraulic jack.

(a) Controller system (b) Jack with displacement control programming

Fig. 9. Second hydraulic jack.

5.2. Cracking pattern and failure mode ratio.

• Crack formation at 8 cm distance from the column face under po-
The crack distribution and crushing of the reinforced concrete of the sitive moment and creating diagonal shear cracks in the beam-
specimens as well as the steel components performance until the last column joint at 3% drift ratio.
drift ratio were considered. • First concrete spalling in the bottom concrete cover of the beam near
the column face at the third cycle of 4% drift ratio (Fig. 16a).
5.2.1. Specimen M1 • Concrete crushing within 20 cm length from the column at the first
cycle of 5% drift ratio (Fig. 16b). Collapse of the crushed concrete at
• Cracking initiation in the bottom concrete cover of the beam from the second cycle of 5% drift ratio. Fracture of a bottom longitudinal
the column face and at 30 cm intervals under positive moment at reinforcing bar under positive moment at the third cycle of 5% drift
0.5% drift ratio. ratio (Fig. 16c).
• More crack propagation in the column face at 0.75% and 1% drift • Fracture of second and third bottom longitudinal reinforcing bars
ratios. under positive moment at the third cycle of 6% drift ratio (Fig. 16d).
• Penetration of existing cracks in the column face to the beam-
column joint at 1.5% drift ratio. 5.2.2. Specimen P1
• Increasing of the crack width in the column face and at a distance of
15 cm from it in the beam and in the beam-column joint at 2% drift • Cracking initiation of the concrete cover of the beam in the column

(a) Specimen M1 (b) Specimen P1

Fig. 10. Hysteresis curves.

7
J. Esmaeili and N. Ahooghalandary Engineering Structures 205 (2020) 110006

Table 2
Parameters of hysteresis curves.
Connection Load-carrying capacity Deformation capacity Stiffness

Load Maximum Nominal Pmax/Pn Maximum Vjmax/Vn Yield Ultimate Ductility ky kp (kN/ kp/ky
direction* resistance Pmax resistance Pn joint shear displacement Δy displacement Δu Δu/ Δy (kN/ mm) (%)
(kN) (kN) Vjmax (kN) (mm) (mm) mm)

M1 – 72.10 72.52 0.99 398.39 0.74 36.40(1.82) 108.60(5.43) 2.98 1.98 −0.19 10
+ 29.43 28.79 1.02 153.58 0.29 23.60(1.18) 108.60(5.43) 4.60 1.25 −0.86 69

P1 – 77.70 74.47 1.04 383.76 0.90 37.40(1.87) 93.20(4.66) 2.49 2.08 −0.13 6
+ 38.46 25.55 1.51 171.11 0.33 20.20(1.01) 93.20(4.66) 4.61 1.90 −0.14 7

* − Related to negative moment; + Related to positive moment.

drift ratio. Growth of the collapsed concrete zone up to 47 cm from

the column face at the third cycle of 4% drift ratio (Fig. 17b).
• Fracture of two bottom side longitudinal reinforcing bars at a dis-
tance of 30 cm from the column face at the first cycle of 5% drift
ratio (Fig. 17c). Growth of the collapsed concrete zone to a distance
of 60 cm from the column face at the second cycle of 5% drift ratio
(Fig. 17c). Fracture of third and fourth bottom longitudinal re-
inforcing bars under positive moment at the third cycle of 5% drift
ratio (Fig. 17d).
• No damage observed in the steel beam-to-column connection com-
ponents.

As can be seen from Figs. 16 and 17, failure mechanism of the both
specimens was associated with the fracture of longitudinal bars in the
Fig. 11. Deformation parameters definition based on equal energy principle. beams. However, the concrete plastic hinge in specimen P1, unlike
specimen M1, located away from the column face and out of the con-
face under negative moment and at the end of the bottom plate nection zone. Reduced distance between the plastic hinge zone and the
under positive moment at the 0.25% drift ratio. jack centerline in the precast beam caused earlier cracking initiation
• Formation of a concrete cracked zone at a 25–40 cm distance from and more applied deflection demand in the precast beam. As a result,
concrete spalling and collapse in specimen P1 occurred at lower drift
the column face under positive moment and a crack at 35 cm dis-
tance from the column face under negative moment at 0.75% drift ratio respect to specimen M1. Despite that, more concrete confinement
ratio. created through the beam end plates, postponed the fracture of re-
• Crack formation at a distance of 50 cm from the column face under inforcing bars in specimen P1.
positive moment and joining of the existing cracks at 28 and 35 cm
distance under negative and positive moments respectively, at 1.5% 5.3. Strain of reinforcing bars and steel plates
drift ratio.
• Discrete cracks propagation in the narrower top part of beam from To evaluate the yielding procedure in different parts of specimens
M1 and P1, the strain was measured at their specified points. These
the column face to 60 cm distance of it at 10 cm intervals.
Penetration of the existing cracks in the column face to the beam- points for specimens M1 and P1 are shown in Fig. 18. The strains in
column joint at 2% drift ratio. specimen M1, were measured on longitudinal reinforcing bars and in
• Formation of a concrete cracked zone at a distance of 20–40 cm specimen P1, in addition to them, were measured on the steel plates.
The measured strains in specimen M1 are shown in Fig. 19. The
from the column face and diagonal shear cracks in the beam-column
joint at 3% drift ratio. First concrete spalling near the end of the positive and negative values in the horizontal axis are related to the
bottom plate at the first cycle of 3% drift ratio (Fig. 17a). compressive and tensile strains and in vertical axis denote to the push
• Concrete crushing at 25–40 cm from the column face at 4% drift and pull directions, respectively. The results show that at first, the beam
bottom longitudinal reinforcing bars started to yield in the column face
ratio. Collapse of the crushed concrete at the second cycle of 4%
and at the distance d (effective depth of beam cross section) from it

(a) Energy dissipation in per drift ratio (b) Cumulative energy dissipation
Fig. 12. Energy dissipation.

8
J. Esmaeili and N. Ahooghalandary Engineering Structures 205 (2020) 110006

(a) Second stiffness definition (b) Second stiffness per drift ratio
Fig. 13. Second stiffness degradation.

(a) Specimen M1 (b) Specimen P1

Fig. 14. Strength degradation.

under positive moment at 0.75% and 1% drift ratios, respectively. Then

the beam top longitudinal reinforcing bars in the column face and the
column longitudinal reinforcing bars in the top face of the beam
showed inelastic behavior at 1.5% drift ratio simultaneously. Finally,
with enlarging plastic hinge zone, the beam top longitudinal reinforcing
bars yielded at the distance d from column face at 2% drift ratio. But,
they remained elastic at beam-column joint until last drift.
The measured strains indicate that the plastic hinge was formed in
the beam near the column face and in a d length from it. Higher plastic
strains in the bottom longitudinal reinforcing bars are accommodating
with more crack propagation in the lower part of the beam.
The results of the strain values in specimen P1 are shown in Fig. 20.
As can be seen, similar to specimen M1, the beam top longitudinal re-
inforcing bars started to yield under negative moment in the column
face and within the distance d from it at 1.5% drift ratio. Also, the beam
Fig. 15. ACI374.1-05 [25] seismic performance criteria. bottom longitudinal reinforcing bars yielded at the distance d from
column face at 3% drift ratio. With more extension of the plastic hinge

Table 3
Comparison of the test results with ACI 374.1–05 [25] acceptance criteria.
Connection Load-carrying capacity* Energy dissipation*+ Stiffness*−

Load Maximum Resistance at third Resistance Elastic-plastic Actual energy EI/ EII Initial Second stiffness at k4/ki
direction# resistance at 4% cycle of 4% drift ratio P4/ P4max energy dissipation dissipation EI stiffness ki third cycle of 4%
drift ratio P4max ratio P4 (kN) EII (kN·m) (kN·m) (kN/mm) drift ratio k4 (kN/
(kN) mm)

M1 – 72.10 67.79 0.94 11.36 3.87 0.34 4.28 0.20 0.05

+ 28.25 27.57 0.98 4.49 0.23 0.05

P1 – 75.64 67.98 0.90 11.89 4.15 0.35 3.65 0.19 0.05

+ 35.22 33.84 0.96 4.30 0.30 0.07

#
− Related to negative moment; + Related to positive moment.
* Maximum allowable degradation in load-carrying capacity at 4% drift ratio is 25% according to ACI 374.1-05 [25].
*+ Minimum allowable energy dissipation ratio at 4% drift ratio is 12.5% according to ACI 374.1-05 [25].
*− Minimum allowable second stiffness at 4% drift ratio to initial stiffness is 5% according to ACI 374.1-05 [25].

9
J. Esmaeili and N. Ahooghalandary Engineering Structures 205 (2020) 110006

(a) 4% drift ratio (b) 5% drift ratio (cycle1) (c) 5% drift ratio (cycle3) (d) 6% drift ratio
Fig. 16. Cracking pattern and failure mechanism of specimen M1.

zone at 4% drift ratio, the top reinforcing bars in the beam-column joint moments, respectively. By more inelastic deformation demands in the
entered to inelastic behavior zone. However, the beam end plates and plastic hinge zone at larger drifts, the bending deformation contribution
the steel box remained elastic until last drift. of the beam and column were reduced. As can be seen, the shear de-
Comparison of measured strains indicate that application of the formation of the beam-column joint had approximately 10% contribu-
bottom and side plates and the steel box restrict the deformations in the tion at all drift ratios. Measurement of the hinge support deformation
precast connection zone such that yielding of reinforcing bars move indicated that the rigid body rotation had less than 1% contribution and
away from the column face. Plastic hinge formation in a place away was not considered in the calculations.
from the beam-column joint is in accordance with weak beam-strong The contribution to the total drift in specimen P1 under negative
column design philosophy in MRFs. and positive moments from 1% to 5% drift ratios is shown in Fig. 23. As
its obvious, contribution of the plastic hinge rotation under positive and
5.4. Drift contribution negative moments between 1% and 5% drift ratio with a value of about
26% and 18% was increased to 66% and 50%, respectively. However,
To measure the deformation of different parts of the specimens, the shear deformation of the beam-column joint contribution was de-
including the bending deformation of the beam and column, the shear creased from about 6% to 1% at the end of the experiment, which
deformation of the beam-column joint and the plastic hinge rotation, 10 alongside with the bending deformation of the column, showed a
LVDT were used according to Fig. 21. downtrend. Also, the bending deformation contribution of the beam
Two vertical LVDTs on the hinge supports determined the rigid was decreased from 3% drift ratio, which is simultaneous with the
body rotation of the specimens. Two other vertical LVDTs on the crushing of the concrete in the plastic hinge region.
column and adjacent to the beam faces measured the assembled rota- Comparison of the results shows that in specimens M1 and P1, the
tion due to the column curvature and the beam-column joint de- plastic hinge rotation contribution under positive moment is more than
formation and Two diagonal LVDTs measured the shear deformation of its contribution under negative moment. It is due to more inelastic
the beam-column joint. Also, four LVDTs were located on the beam in deformation and damage concentration in the lower part of the beam at
the vicinity of the column face and at distance of 300 mm and 600 mm the plastic hinge zone. Due to the widespread crack propagation over
from it in order to calculate the deformation of the plastic hinge based the precast beam length, the plastic hinge rotation contribution in
on the differential curvature of beam between these points. With dif- specimen P1 is less than that of specimen M1. Also, due to more con-
ference of the assembled resultant deformation and the applied de- crete confinement by the peripheral steel box, amount of the beam-
formation of the hydraulic jack (100% deformation), the bending de- column joint shear deformation contribution to the overall deformation
formation of the beam was calculated. and its damages in specimen P1 is less than specimen M1.
The contribution to total drift in specimen M1 under negative and
positive moments from 1% to 6% drift ratio is shown in Fig. 22. The 6. Finite element simulation
plastic hinge rotation contribution was increased from zero at 1% drift
ratio to 71% and 62% at 6% drift ratio under positive and negative In order to develop the test results for using in precast structures,

(a) 3% drift ratio (b) 4% drift ratio (c) 5% drift ratio (cycle2) (d) 5% drift ratio (cycle3)
Fig. 17. Cracking pattern and failure mechanism of specimen P1.

10
J. Esmaeili and N. Ahooghalandary Engineering Structures 205 (2020) 110006

Fig. 18. Strain gage layout.

Fig. 19. Strains measured from specimen M1.

11
J. Esmaeili and N. Ahooghalandary Engineering Structures 205 (2020) 110006

Fig. 20. Strain measured from specimen P1.

the specimens were simulated through finite element models and the To determine concrete stress-strain curve, the suggested models of
effect of concrete compressive strength, column axial forces, beam CEB-FIP [27] was used. This curve consists of four parts:
concrete confinement, side plates and bottom plate length were in-
vestigated as important parameters. 1. Tension before cracking: Behavior of concrete in this region is
considered linear.
2. Tension after cracking: In this region, the exponential crack opening
6.1. Materials modeling: function provided by Hordijk [28], which was obtained experi-
mentally, is used.
Evaluating the connection performance by FEM requires the defi- 3. Compression before peak stress: The proposed relation for this re-
nition of the material properties in an idealized form. Therefore, it was gion is in conformance with the proposed CEB-FIP [27] models and
tried to simulate the behavior of the connections by using existing be- can cover a wide range of diagrams from linear to curvy.
havioral models and the materials properties.

12
J. Esmaeili and N. Ahooghalandary Engineering Structures 205 (2020) 110006

(a) LVDTs layout in tests setup (b) Drift contribution of different parts
Fig. 21. Drift contribution measurement.

4. Compression after peak stress: In this region, the fictitious com- interactions depend on normal and tangent stresses and the adjacent
pression plane model is used. This theory was confirmed by ex- material properties. Failure criterion in this theory is in accordance
periments conducted by Mier [29]. with the followings equations:
τ ⩽ c − σ . ϕ, σ ⩽ ft (9)
Table 4 shows the concrete parameters of models M1 and P1. The
biaxial stress failure criterion of concrete was selected based on the τ = 0, σ > ft (10)
relations of Kupfer and Gerstle [30].
A “discrete model” was used to model the reinforcing bars. In this Where c is cohesion and ϕ is coefficient of internal friction. After stress
model, the bars are defined separately and in contact with concrete. In is increased in comparison with existing strength on a surface, the in-
order to determine the stress-strain curve of the reinforcing bars and the terface will be failed and the residual strength will be only due to
steel plates, the three-component curve with respect to the initial elastic friction at the residual interface. This behavioral model was used for the
and strain hardening modulus was used. The necessary parameters for interface between the new concrete of the beam top part and the har-
the longitudinal and the transverse reinforcing bars and the steel plates dened concrete in the lower part of it, as well as the weld of the seat
were derived from Table 1. plate to the bottom plate and the weld of the angles to the side plates
The CEB-FIP [27] behavioral model was used to consider reinfor- and the steel box. The equivalent shear capacity of the headed studs
cing bars bonding to concrete. This behavioral model consists of four was also modeled in the same way. Table 6 shows the parameters of the
parts: interfaces.

τ = τmax (S S1)α , 0 ⩽ S ⩽ S1 (5) 6.2. Finite element models

τ = τmax , S1 ⩽ S ⩽ S2 (6)
The models were analyzed with FEM through ATENA 3D software.
τ = τmax − (τmax − τf ) × (S − S1) (S3 − S2) , S2 ⩽ S ⩽ S3
This software is able to consider nonlinear behavior of material prop-
(7)
erties, exhibit stresses and strains in concrete and steel members and
τ = τf , S3 ⩽ S (8)
also concrete cracking. The FEM in joints analysis requires the defini-
tion of 3D truss elements for one-dimensional members, such as re-
where S1, S2 and S3 are parameters which are defined based on the inforcing bars. These isotropic elements have two integration points in
compressive strength of concrete, the reinforcing bar diameter and its 3-node elements. Also, for 3D members, tetrahedral elements with 10
type. Table 5 shows the bond-slip parameters of the reinforcing bars. nodes and brick elements with 20 nodes simultaneously were used with
In order to define a failure criterion in the steel plates, the von Mises 5 cm mesh size. The finite element models of the joints are shown in
plasticity model [32] was used, which is also called the plasticity model Fig. 24. In the models two steel plates at the beginning and end of the
J2. column as supporting members and two other plates above the beam as
In the interfaces, a behavioral model was used to simulate the load bearing members were used to prevent stress concentration caused
contacts based on the Mohr-Coulomb theory [32]. In this theory, by the point loads. Tetrahedral elements were used in the supporting

Fig. 22. Drift contribution in specimen M1.

13
J. Esmaeili and N. Ahooghalandary Engineering Structures 205 (2020) 110006

Fig. 23. Drift contribution in specimen P1.

and the load bearing members and brick elements were used for the 3D bars is about 2000 μs.
joint members. The hinge supports were simulated by fixing the points As is obvious, in model M1, the maximum concrete strains occurred
located on the horizontal neutral axis of the supporting members. in the column face and in the length of 25 cm from it in beam, which is
in accordance with the cracking pattern in the test. In model P1, the
6.3. Comparison of test and finite element analysis results maximum concrete strains occurred under negative moment at a dis-
tance of 20 cm from the column face and were extended in beam length.
6.3.1. Load-carrying capacity curves Under positive moment, the most strains occurred 35 cm away from the
Comparison of the finite element pushover analysis results for column face and cover an area with about 35 cm length from the side
models M1 and P1 with the envelope curve of cyclic tests for their plates. Base on the results, the steel parts remained elastic until the last
specimens is shown in Fig. 25a and b, respectively. drift which is coincided with their strain values (measured by strain
Comparison of predicted strength in model M1 at 4% and 5% drift gauges 6–9) in the test.
ratios (according to maximum resistance in the test) under negative and The yield strain location of the longitudinal reinforcing bars in the
positive moments, shows a 1% and 15% difference, respectively. This model M1 was predicted at the beam from the column face and in 2d
comparison for model P1 depicts 2% and 1% difference at 3% and 2% length from it. This has good conformity with the test results, because
drift ratios, respectively. According to the experimental and numerical the strain gauges 1, 2, 5 and 6 measured post-yield strains from the
curves, it is obvious that the finite element pushover analysis predicted longitudinal bars in this zone. The predicted post-yield strains for the
the experimental results in term of load carrying-capacity, properly. model P1 under negative moment had more extension accordance with
The initial stiffness of numerical curve for model M1 to the initial values measured by the strain gauges 1, 2 and 4 and covered the top
stiffness of the experimental envelope curve shows ratio of 1.13 and longitudinal bars in the beam-column joint. Also, under positive mo-
1.11 under negative and positive moments, respectively. For model P1, ment, the maximum strains occurred at 20 cm distance from the column
this ratio is 1.10 and 1.07, respectively, which seems have good con- face and in 2d length from it. It can be seen, the comparison of ex-
formity. perimental and numerical strain values shows good agreement.

6.3.2. Cracking pattern 6.4. Effective parameters

The predicted numerical analysis results for the cracked region of
the models and its comparison with the cracking pattern in the speci- 6.4.1. Compressive strength of concrete
mens are indicated in Figs. 26 and 27. The cracks shown in the models The proposed PBC should have capability to use in precast MRFs
have more than 0.5 mm width. These progressive cracks in model M1 with different concrete strengths. Hence, the effect of this parameter
grew on the column face and for model P1 were extended at the end of was evaluated in the connections. For this purpose, the connection
the side plates. As can be seen, the predicted progressive cracks loca- models with 30, 40, 50 and 60 MPa concrete strengths were analyzed.
tions in the numerical models coincide with the test specimens. Other concrete parameters like tensile strength, elastic modulus and
specification fracture energy were calculated based on Table 4. The
6.3.3. Strain values analysis results in term of load-carrying capacity curve for M1 and P1
The strains values of concrete, steel parts and reinforcing bars at the are indicated in Fig. 30a and b. As indicated, concrete strength had
last drift under negative and positive moments in models M1 and P1 are same effect on both connections behavior. The maximum and minimum
shown in Figs. 28 and 29. The strain values shown for concrete and load-carrying capacity were associated to models with 60 and 30 MPa
reinforcing bars are corresponding to the principle strain. The fracture with peak strength ratios equal to 1.11 and 1.10 under negative mo-
strain of concrete is about 5000 μs and the yield strain of reinforcing ment and 1.15 and 1.11 under positive moment for M1 and P1,

Table 4
Concrete parameters.
Parameter Definition Relation Value (M1,P1) Unit

fc′ Compressive strength fc′ = 0.85fcu

′ (42.11,40.29) MPa
ft′ Tensile strength ft′ = 0.24(fcu
′ )2 3 (3.24,3.14) MPa
Ec Elastic modulus ′ ) × (fcu
Ec = (6000 − 15.5fcu ′ )1 2 (3.68E + 4, 3.62E + 4) MPa
ν Poisson’s ratio – (0.20,0.20) –
wd Critical compressive displacement1 – (0.50,0.50) mm
Gf Specific fracture energy2 Gf = 2.5 × 10−5ft′ (8.10E − 5, 7.85E − 5) kN/mm

1
Critical compressive displacement is the end point in descending part of stress-displacement diagram. In this way, the required energy for generation of failure
plane with unit area is indirectly defined [29].
2
Specific fracture energy is the required energy to create a unit area of stress-free crack [31].

14
J. Esmaeili and N. Ahooghalandary Engineering Structures 205 (2020) 110006

Table 5
Bond-slip parameters.
Parameter Definition Relation or explanation Value (M1,P1) Unit

S1 – Good bond condition (1,1) mm

S2 – Good bond condition (3,3) mm
S3 Clear rib spacing Good bond condition According to bar size mm
α – Ribbed bar (0.40,0.40) –
τmax Maximum shear strength 2.5 fc′ (16.22,15.86) MPa
τf Failure shear strength 0.4τmax (6.49,6.34) MPa

respectively. Increasing in load-carrying capacity is proportional to before casting of the narrower top part of it in construction site.
more flexural strength in beam concrete cross section at plastic hinge Therefore, the concrete of this part will be confined by precast slabs,
zone. As shown, the other parameters such as strength degradation and practically. This confinement effect in model P1 was evaluated with
initial stiffness were not affected, significantly. considering nonlinear springs placed perpendicular to the side faces of
beam upper part. These springs had stiffness equal to concrete elastic
6.4.2. Column axial force modulus (Ec) in compression and zero in tension.
The proposed precast connection should be applicable in frames The comparison of results indicated in Fig. 32a and b and Fig. 27b
with different column axial force levels. This parameter could be ef- and c depicts that the concrete cracking in the plastic hinge zone is
fective on beam-column joint deformation. For evaluating this para- decreased specially in the upper part of beam under positive moment.
meter, the models with 0.1, 0.3, 0.5 and 0.7fc’Ag were analyzed. The Less damage in compressive region of the concrete cross section in the
load-carrying capacity curves of models are depicted in Fig. 31a and b plastic hinge zone cause more load-carrying capacity with a value of 7%
for M1 and P1, respectively. According to these figures, increasing in in the peak strength (Fig. 32c). However, this confinement had not any
axial force of column caused more strength degradation in post-peak effect on the cracking pattern and the precast joint behavior under
region, especially in model M1. It seems that higher axial forces restrict negative moment.
deformation of the column and joint and lead to more deflection de-
mand in the beam at the same drift ratio especially after the plastic 6.4.4. Length of side plates
hinge formation. The maximum strength degradation ratio in model P1 For easy casting of the proposed beam in manufacture, the side
to M1 with 0.7fc’Ag was 0.74 and 0.54 under negative and positive plates were designed with the same height of the beam prefabricated
moments at 5% drift ratio, respectively. Less strength degradation in part and their thickness was chosen based on the maximum calculated
model P1 is due to more limited deformation and cracking in the beam- values required for shear strength in the connection zone and their fillet
column joint and larger distance between the plastic hinge zone and and stud welds. In addition, the length of side plates was designed
column face. As indicated, different axial force levels of column had not based on the minimum allowable distance of their headed studs.
effect on load carrying capacity in pre-peak region and initial stiffness. As it is obvious in Fig. 27b and c the results indicated that the side
plates not only could be applicable as shear resistance elements but also
6.4.3. Beam concrete confinement have confining effect on the concrete in a way that crushing of the
As be mentioned before, the hollow core slabs in the precast joint concrete occurred out of them. Therefore, increasing in length of these
are placed on the prefabricated part of beam at installation process plates with a maximum value equal to the bottom plate length could be

Table 6
Interfaces parameters for model P1.
Parameter Definition Explanation Value Unit

Interface between in-situ and precast concrete

c Cohesion Equivalent resistance based on concrete shear strength 1.27* MPa
ϕ Coefficient of internal friction For concrete to hardened concrete 0.6+ –

Interface between side plate and concrete

c Cohesion Equivalent resistance based on concrete breakout strength 0.38** MPa
ϕ Coefficient of internal friction For headed studs embedded in concrete 0.7++ –

Interface between bottom plate and concrete

c Cohesion Equivalent resistance based on concrete breakout strength 0.22*+ MPa
ϕ Coefficient of internal friction For headed studs embedded in concrete 0.7 –

Interface between seat plate and bottom plate

c Cohesion Resistance based on weld metal strength divided to contact surface area 12.11 MPa
ϕ Coefficient of internal friction Shear strength after weld fracture 0 –

Interface between angle and side plate

c Cohesion Resistance based on weld metal strength divided to contact surface area 30.28 MPa
ϕ Coefficient of internal friction Shear strength after weld fracture 0 –

Interface between angle and steel box

c Cohesion Resistance based on weld metal strength divided to contact surface area 25.23 MPa
ϕ Coefficient of internal friction Shear strength after weld fracture 0 –

* Strength based on shear friction theory from ACI318-14 [21] divided to concrete contact surface area.
+
Coefficient of friction based on ACI318-14 [21] for concrete placed against hardened concrete not intentionally roughened.
** Concrete breakout strength of group anchors based on PCI handbook [22] divided to side plate surface.
++
Coefficient of friction based on ACI318-14 [21] for concrete anchored to as-rolled structural steel, headed studs or reinforcing bars.
*+ Concrete breakout strength of group anchors based on PCI handbook [22] divided to bottom plate surface.

15
J. Esmaeili and N. Ahooghalandary Engineering Structures 205 (2020) 110006

(a) Model M1 (b) Model P1

Fig. 24. Finite element modeling in software.

(a) Model M1 (b) Model P1

Fig. 25. Pushover analysis results in comparison with envelope curve of tests.

effective on the plastic hinge location. For evaluating this, a model of 6.4.5. Length of bottom plate
the precast joint with equal length of the side and bottom plates was In order to provide a suitable surface on the bottom plate for its
analyzed and its results in term of cracking pattern and load-carrying connections to the group headed studs and the seat plates, the bottom
capacity curve were considered. Comparing the results of current model plate was designed with a same width of the beam prefabricated part
(Fig. 33a and b) with the first one (Fig. 27b and c) indicate that the and its thickness was calculated based on required bending strength in
concrete cracked zone is moved away to beam end plates, completely the connection zone. Also the length of bottom plate was determined
and load-carrying capacity is increased maximum 9% and 10% under with considering the following parameters:
Negative and positive moments, respectively (Fig. 33c). These incre-
ments have reverse proportionality with beam plastic hinge to jack • Minimum allowable distance of group headed studs in shear
center line distance ratio of models (1.65 m/1.50 m). Therefore, the • Concrete breakout strength of anchors in shear
moment capacity has about equal values in both models. • Required weld length between the seat and bottom plates
• Minimum development length of the bottom longitudinal bars with

(a) Concrete cracked zone in (b) Cracking pattern under positive (c) Cracking pattern under
specimen moment in model negative moment in model
Fig. 26. Cracking pattern in model and specimen M1.

16
J. Esmaeili and N. Ahooghalandary Engineering Structures 205 (2020) 110006

(a) Concrete cracked zone in (b) Cracking pattern under positive (c) Cracking pattern under
specimen moment in model negative moment in model
Fig. 27. Cracking pattern in model and specimen P1.

(a) Strain contour in negative moment (b) Strain contour in positive moment
Fig. 28. Predicted strain values for model M1.

(a) Strain contour in negative moment (b) Strain contour in positive moment
Fig. 29. Predicted strain values for model P1.

standard hook positive moment. Besides that, the analysis result in previous part
shown that application of the bottom and side plates with the equal
Due to this limitation the length of bottom plate was considered length moved the plastic hinge zone location to the end of bottom plate,
350 mm in specimen P1. As an approximate criterion the length of this completely. Hence, the length of bottom plate in this situation could
plate could be predicted about equal to the beam height. have important role in beam mechanism under negative and positive
As can be seen in Fig. 17 the first concrete spalling and the bottom moments. For evaluating this, two models with 450 and 550 mm end
longitudinal bars fracture occurred at the end of bottom plate under plate length analyzed. Fig. 34 shows the cracking patterns in each

17
J. Esmaeili and N. Ahooghalandary Engineering Structures 205 (2020) 110006

(a) Model M1 (b) Model P1

Fig. 30. Pushover analysis results with different compressive strengths of concrete.

(a) Model M1 (b) Model P1

Fig. 31. Pushover analysis results with different column axial forces.

(a) Cracking pattern under (b) Cracking pattern under

(c) Load-carrying capacity curve
positive moment negative moment
Fig. 32. Analysis result of model with top confined concrete.

(a) Cracking pattern under (b) Cracking pattern under

(c) Load-carrying capacity curve
positive moment negative moment
Fig. 33. Analysis result of model with 350 mm side plate length.

18
J. Esmaeili and N. Ahooghalandary Engineering Structures 205 (2020) 110006

(a) Cracking pattern (with 450mm end plates length) (b) Load-carrying capacity curve

(c) Cracking pattern (with 550mm end plates length) (d) Load-carrying capacity curve
Fig. 34. Analysis result of model with 450 mm and 550 mm end plates length.

model and their load-carrying capacity curves, respectively. The results prefabricated steel structures leading to weak beam-strong column
indicate, similar to the previous model, the plastic hinge locations are in precast MRFs.
depending on the plates length and their strength ratios are about 1.07 5. Owing to the confining effect of the peripheral steel box on the
and 1.17 in models with end plates length equal to 450 and 550 mm to proposed precast joint, the concrete damage of this part was less
350 mm which are reverse proportional to their moment lever arm than that of the in-situ one.
ratios equal to 1.5 m/1.4 m and 1.5 m/1.3 m, respectively. As the lo- 6. The load-carrying capacity parameters as well as the plastic hinge
cation of plastic hinge getting farther from the column face and closer location and the cracking pattern were well predicted by the finite
to the loading point, the models experienced a significant strength loss element pushover analysis.
at last drifts and the ductility in the joint connection was reduced. This 7. Increasing in compressive strength of concrete and axial force of
could be seen in the model with the end plates length equal to 450 mm column, respectively, lead to more load-carrying capacity and
at 4.5% drift ratio under positive moment and the model with the end strength degradation in the both evaluated connections. But, the
plates length equal to 550 mm at 4.3 and 4.8% drift ratio under ne- precast proposed connection had less strength degradation in com-
gative and positive moments, respectively. Increasing in deformation parison with the in-situ one at the same drift ratio.
demand at the hinges located closer to the loading point play a crucial 8. Confining effect of precast slabs on the top part of proposed precast
role in the reduction of ductility. beam increased the load-carrying capacity of PBC model about 7%
under positive moment.
7. Conclusion 9. The side and bottom plates had confining effect on the concrete of
beam end in a way that the plastic hinge formed after them in PBC
The results of this study can be summarized as follows: models. But increasing of end plates length could lead to less duc-
tility of the precast joint.
1. The proposed PBC according to the ACI374.1-05 acceptance criteria
can be used as a moment resistant connection in PRCS. Funding
2. Even with dry components, the proposed PBC under negative and
positive moments had 1.08 and 1.30 strength ratio, respectively, This research did not receive any specific grant funding agencies
regarding to the equivalent IBC and comparable ductility and energy public, commercial, or not-for-profit sectors.
dissipation. In this regard, the proposed connection can be com-
pared with previous successful semi-monolithic connections.
3. The precast proposed beam has a simple assembling process similar CRediT authorship contribution statement
to other dry conventional PBCs. Furthermore, it provides the ne-
cessary bearing surface to install the precast hollow core slabs Jamshid Esmaeili: Methodology, Data curation, Writing - original
without requiring scaffolding and casting with formworks. draft, Writing - review & editing, Project administration. Neyram
4. The plastic hinge formation in the proposed PBC, unlike the Ahooghalandary: Conceptualization, Software, Validation,
equivalent IBC, was not located on the beam-column joint and had a Investigation, Resources.
minimum distance of 20 cm. Furthermore, all steel components of
the precast connection remained elastic during the test. This result is Declaration of Competing Interest
similar to the approach used in connections with RBS in
None.

19
J. Esmaeili and N. Ahooghalandary Engineering Structures 205 (2020) 110006

References column connections. Comput Concr 2008;5(6):525–44.

[16] Shin M, Kim SP, Halterman A, Aschheim M. Seismic toughness and failure me-
chanisms of reduced web-section beams: phase 2 tests. Eng Struct
[1] Savoia M, Buratti N, Vincenzi L. Damage and collapses in industrial precast build- 2017;141:607–23.
ings after the 2012 Emilia earthquake. Eng Struct 2017;137:162–80. [17] Li R, Samali B, Tao Z, Kamrul Hassan M. Cyclic behaviour of composite joints with
[2] Marzo A, Marghella G, Indirli M. The Emilia-Romagna earthquake: damages to reduced beam sections. Eng Struct 2017;136:329–44.
precast/prestressed reinforced concrete factories. Ing Sismica [18] Imanpour A, Torabian S, Mirghaderi SR. Seismic design of the double-cell ac-
2012;29(2–3):132–45. cordion-web reduced beam section connection. Eng Struct 2019;191:23–38.
[3] Rodríguez ME, Torres-Matos M. Seismic behavior of a type of welded precast con- [19] Morshedi MA, Dolatshahi KM, Maleki S. Double reduced beam section connection. J
crete beam-column connection. PCI J 2014;58(3):81–94. Constr Steel Res 2017;138:283–97.
[4] Fan L, Lu X. Investigation on seismic behavior of jointed precast concrete frame [20] Montuori R, Sagarese V. The use of steel rbs to increase ductility of wooden beams.
structures. 14th world conference on earthquake engineering (14WCEE). 2008. Eng Struct 2018;169:154–61.
[5] Elliott KS, Davies G, Ferreira M, Gorgun H, Mahdi AA. Can precast concrete [21] ACI Committee 318. Building Code Requirements for Structural Concrete (ACI 318-
structures be designed as semi-rigid frames? Part 1 - The experimental evidence. 14) and Commentary. Farmington Hills, MI, USA, American Concrete Institute;
Struct Eng 2003;81(16):14–27. 2014.
[6] Hong W-K, et al. Composite beam composed of steel and precast concrete (mod- [22] Wilden H, Precast/Prestressed Concrete Institute. PCI design handbook : precast
ularized hybrid system). Part III: application for a 19-storey building. Struct Des Tall and prestressed concrete. Precast/Prestressed Concrete Institute; 2010.
Spec Build 2010;19(6):679–706. [23] ASTM. Standard Specification for Deformed and Plain Carbon-Steel Bars for
[7] Choi HK, Choi YC, Choi CS. Development and testing of precast concrete beam-to- Concrete Reinforcement. West Conshohocken, PA, USA: ASTM A615/A615M,
column connections. Eng Struct 2013;56:1820–35. American Society for Testing and Materials; 2012.
[8] Yan Q, Chen T, Xie Z. Seismic experimental study on a precast concrete beam- [24] Bringas JE. Handbook of comparative world steel standards. ASTM International;
column connection with grout sleeves. Eng Struct 2018;155:330–44. 2016.
[9] Lu C, Dong B, Pan J, Shan Q, Hanif A, Yin W. An investigation on the behavior of a [25] ACI Committee 374. Acceptance Criteria for Moment Frames Based on Structural
new connection for precast structures under reverse cyclic loading. Eng Struct Testing (ACI 374.1-05) and Commentary. Farmington Hills, MI, USA: American
2018;169:131–40. Concrete Institute; 2005.
[10] Im HJ, Park HG, Eom TS. Cyclic loading test for reinforced-concrete-emulated [26] ACI Committee 352. Recommendations for Design of Beam‐Column Connections in
beam-column connection of precast concrete moment frame. ACI Struct J Monolithic Reinforced Concrete Structures (ACI 352R‐02). Farmington Hills, MI,
2013;110(1):115–26. USA: American Concrete Institute; 2002.
[11] Yuksel E, Karadogan HF, Bal IE, Ilki A, Bal A, Inci P. Seismic behavior of two ex- [27] Committee Euro-International du Beton. CEB-FIP Model Code, Bulletin information
terior beam-column connections made of normal-strength concrete developed for No. 195. Lausanne, Switzerland: CEB-FIP; 1990.
precast construction. Eng Struct 2015;99:157–72. [28] Hordijk DA. Local approach to fatique of concrete. Netherland: Delft University of
[12] Maya LF, Zanuy C, Albajar L, Lopez C, Portabella J. Experimental assessment of Technology; 1991.
connections for precast concrete frames using ultra high performance fibre re- [29] Mier V. Multi-axial strain-softening of concrete, Part I: fracture, materials and
inforced concrete. Constr Build Mater 2013;48:173–86. structures. RILEM 1986;19:111.
[13] Guan D, Guo Z, Xiao Q, Zheng Y. Experimental study of a new beam-to-column [30] Kupfer H, Gerstle KH. Behavior of concrete under biaxial stresses. ACI J Proc
connection for precast concrete frames under reversal cyclic loading. Adv Struct 1973;66(8).
Eng 2016;19(3):529–45. [31] Vos E. Influence of loading rate and radial pressure on bond in reinforced concrete
[14] Bradley BA, Dhakal RP, Mander JB, Li L. Experimental multi-level seismic perfor- Dissertation Netherland: Delft University; 1983.
mance assessment of 3D RC frame designed for damage avoidance. Earthq Eng [32] Chen WF, Saleeb AF. Constitutive equations for engineering materials. Volume 1,
Struct Dyn 2008;37(1):1–20. Elasticity and modeling. Elsevier; 2014.
[15] Cheng CT. Seismic behavior of post-tensioned precast reinforced concrete beam-to-

20