Sunteți pe pagina 1din 12

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

net/publication/260726082

Pushover analysis of reinforced concrete frames considering shear failure


at beam-column joints

Article  in  Earthquake Engineering and Engineering Vibration · September 2013


DOI: 10.1007/s11803-013-0179-8

CITATIONS READS

6 372

4 authors, including:

Yu-Chi Sung Tzu-Kang Lin


National Taipei University of Technology National Chiao Tung University
58 PUBLICATIONS   243 CITATIONS    55 PUBLICATIONS   173 CITATIONS   

SEE PROFILE SEE PROFILE

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

Optimization of after-event evacuation route based on earthquake early warning and structural health monitoring View project

Development and Application of Vibration Isolation System with Controllable Friction Damper Considering Strong Ground Motion Effect
(II)~(III) View project

All content following this page was uploaded by Yu-Chi Sung on 14 January 2015.

The user has requested enhancement of the downloaded file.


Vol.12, No.3 EARTHQUAKE ENGINEERING AND ENGINEERING VIBRATION September, 2013

Earthq Eng & Eng Vib (2013) 12: 373-383 DOI: 10.1007/s11803-013-0179-8

Pushover analysis of reinforced concrete frames considering


shear failure at beam-column joints
Y.C. Sung1†, T.K. Lin2‡, C.C. Hsiao3§ and M.C. Lai1*

1. Bridge Division, Center for Research on Earthquake Engineering, Taipei University of Technology (NTUT), Chinese Taipei
2. Chiao Tung University, Chinese Taipei
3. Taipei University of Technology (NTUT), Chinese Taipei

Abstract: Since most current seismic capacity evaluations of reinforced concrete (RC) frame structures are implemented
by either static pushover analysis (PA) or dynamic time history analysis, with diverse settings of the plastic hinges (PHs) on
such main structural components as columns, beams and walls, the complex behavior of shear failure at beam-column joints
(BCJs) during major earthquakes is commonly neglected. This study proposes new nonlinear PA procedures that consider
shear failure at BCJs and seek to assess the actual damage to RC structures. Based on the specifications of FEMA-356, a
simplified joint model composed of two nonlinear cross struts placed diagonally over the location of the plastic hinge is
established, allowing a sophisticated PA to be performed. To verify the validity of this method, the analytical results for
the capacity curves and the failure mechanism derived from three different full-size RC frames are compared with the
experimental measurements. By considering shear failure at BCJs, the proposed nonlinear analytical procedures can be used
to estimate the structural behavior of RC frames, including seismic capacity and the progressive failure sequence of joints, in
a precise and effective manner.

Keywords: RC frame; beam-column joint; pushover analysis: capacity curve; plastic hinge; performance-based seismic
evaluation

1 Introduction failure mechanism is not always ensured. For example,


the grouting of concrete is often compromised by
Buildings and bridges commonly consist of the complicated arrangement of reinforcements. As
reinforced concrete (RC) structures, and the damage a consequence, failure of BCJs prior to the adjacent
sustained during major earthquakes is chiefly caused by components, which may occur due to insufficient shear
the unexpected failure of these structures. RC structures strength resulting from improper construction, can lead
are currently classified into three categories at the to a catastrophic collapse. It is therefore important to
design stage: beam, column, and beam-column joint precisely evaluate the seismic capacity of existing RC
(BCJ) structures. Now that the strong column-weak structures by considering the possibility of shear failure
beam concept has been enshrined in most seismic codes at BCJs, particularly in the case of designs based on out-
around the world, plastic hinges (PHs) are expected to of-date construction codes.
develop on the beams of moment-resisting frames to BCJs are currently treated as rigid joints by seismic
absorb seismic energy and prevent the collapse of the codes in many countries. ACI 318-11 is considered as
entire structure during a major earthquake (see Fig. 1). the chief reference for most of the seismic codes in
Failure at BCJs is avoided by increasing the seismic the world (ACI Committee 318, 2011), and its basic
capacity of the joint, instead of for other components; regulations concerning BCJs are briefly introduced
i.e., the so called design concept of strong joint- here. Based on the strong column-weak beam design
weak adjacent components. However, because of the concept, PHs should be employed on beam elements in
imperfect conditions that frequently result from the order to dissipate the energy generated by earthquakes.
construction of BCJs in the real world, the expected By properly specifying capacity and ductility, sufficient
shear strength can be provided by BCJs to allow the
Correspondence to: Y.C. Sung, Bridge Division, Center for
development of PHs on beam elements. The strength
Research on Earthquake Engineering, Taipei University of
Technology (NTUT), Chinese Taipei
ratio between beams and columns in the ACI 318-11
Tel: +886-2-27712171-2655; Fax: +886-2-27814518
code is given as:


E-mail: sungyc@ntut.edu.tw
Proessor; ‡Assistant Professor; §Master; *Research Assistant
∑M nc ≥ (6 / 5)∑ M nb (1)
Received July 22, 2012; Accepted January 15, 2013
374 EARTHQUAKE ENGINEERING AND ENGINEERING VIBRATION Vol.12

Plastic hinge joint, and that the shear strength is determined by the
cross section of the joint. To replicate the RC designs
in the 1970s, six 2/3 scale frame models with three
different types of BCJs, including corner, T-shape,
and inner joints, were tested by Calvi et al. (2002).
Among these tests, the T-shape joint was damaged
due to rebar slippage and shear failure of hooks due to
stress concentration, and suffered serious degradation
of strength and ductility. The study concluded that the
seismic capacity is improved by employing hooks and
anchorages in a joint.
Numerical modeling of joints was performed
concurrently with experimental investigations into
the BCJs of RC structures. A multi-point numerical
model simulating rebar slippage in BCJs under cyclic
Fig. 1 Typical moment-resisting frame with PHs loading was proposed by Elmorsi et al. (2000), and
the results showed that bond-slippage behavior can
be estimated by using the proposed method when
where ∑ M nc is the total nominal flexural strength
numerous parameters are defined. A multi-spring model
and also the minimum flexural strength considering
comprising 12 concrete springs, 12 steel springs, and 12
the axial and lateral forces of columns connected to a
shear springs was established by Youssef and Ghobarah
joint; and ∑ M nb is the total nominal flexural strength
(2001) to simulate a BCJ subjected to shear force with
of beams connected to the joint considering the floor
consideration of the effect of reinforcement slippage. As
reinforcement (ACI Committee 318, 2011).
the reliability of the modeling improved, however, the
The shear capacity of a joint is calculated by
common disadvantage of requiring numerous parameters
considering the repetitive loading on BCJs under
hindered the applicability of this research. Mitra and
earthquakes and energy dissipated by the PHs on beams
Lowes (2004) recommended the use of a simplified two-
near the joint, and the induced shear force Vjh,u is set as
dimensional model with eight bond-slippage springs
the design shear force of the joint Vu. The induced shear
and four shear springs connecting the contact surface
force is expressed as
to describe the behavior of BCJs, in which the loss of
Vjh,u = ( As + AS’ ) f y − Vcol (2) the strength and stiffness of the rebars was simulated
by bond-slippage springs, and the shear deformation
where As and A's are the upper and lower rebar areas of was simulated by shear springs. A new multi-spring
the beam, respectively, αfy is the overstrength of the model using axial springs, inner joints, and rigid bodies
beam rebar, and Vcol is the shear force of the column. The with a global perspective was proposed by Tajiri et al.
A's term can be neglected if the rebar is anchored inside (2006) to analyze the deformation of an elasto-plastic
the joint. plane frame. Another simplified model considering the
The criterion Vn ≥ Vu can finally be examined strength of the BCJ was suggested by Eligehausen et al.
using the calculated Vu and the corresponding nominal (2006). Although improved performance was noted, the
shear strength Vn to ensure sufficient shear strength is estimation of joint strength under various axial loadings
provided by the joint to support the establishment of remained imprecise.
plastic joints on the beams. However, the shear failure Experiments have demonstrated that the global
of the BCJs in practice cannot be illustrated by existing behavior of RC frames is strongly influenced by the
codes. failure of the BCJs, and current efforts to model the
To study the actual failure conditions of BCJs BCJs still face problems such as poor estimation
during earthquakes, many researchers have conducted precision and complexity of application. An effective
experimental investigations and verifications. Hanson simplified method for modeling BCJs is therefore
and Connor (1967) performed a series of tests needed. The basic concept of the method proposed in
concerning the seismic performance of BCJs and this study includes the establishment of PHs on the
became an important reference for subsequent studies. beam and column, and uses an equivalent diagonal
Their test results showed that the shear force of a joint is structural strut model to simulate the BCJ, as briefly
undertaken mostly by the concrete component, and the described in Section 2. The performance of the proposed
confinement of concrete is attributed by the transversal procedure is verified by experimental data for three
stirrups; thus, the concrete does not provide significant specimens with configurations ranging from a one-bay
shear resistance to the joint. Based on experimental one-story specimen to a three-bay three-story specimen
findings, Meinheit and Jirsa (1981) suggested that in Section 3. To further demonstrate the effectiveness
the shear strength of a joint results primarily from the of the methodology, a comparison is made between
diagonal compression strut mechanism formed in the the theoretical analysis and the measured data from the
No.3 Y.C. Sung et al.: Pushover analysis of reinforced concrete frames considering shear failure at beam-column joints 375

perspective of both the capacity curve and the behavioral between the transformed moment Mv and rotation θ, as
sequence of the PHs. A summary and conclusions are shown in Fig. 2(b), can be obtained. By superimposing
provided in Section 4. the diagrams of ( M b −  ) and ( M v −  ) , three different
types of failure modes (shear failure, flexure to shear
2 Proposed method failure, and flexure failure) can be illustrated. The PH
characteristics indicated by points A through E in Fig. 4,
A new PA method was used to accurately describe expressed by the relationship between moment and
the process in which structural damage propagates from flexural rotation, are therefore definable.
slight damage to the limiting state of ultimate strength of
components of a RC frame under a monotonic loading. 2.2 PH settings for BCJs
The behavior of the entire structure can be derived by 2.2.1 PH characteristics of the BCJs of RC structures
properly defining the PH of the structural components. Since more details concerning BCJs are now
The PH setting of columns and beams in the frame are provided by FEMA-356 (2000) rather than by the
introduced, and the structural nonlinearity of the BCJs existing code, these guidelines are therefore used as
is simulated through the use of an equivalent diagonal the main reference for defining the PH characteristic
structural strut method. The details of each aspect are of beam-column-joints. According to FEMA-356, the
provided below. nominal shear strength of BCJs can be calculated as
follows:
2.1 PH settings of beams and columns
Vn =  f c’ Aj (3)
The PH settings of the beams and columns of the
frame were established using the method developed
where λ is the coefficient of the concrete, and is set as 1
and demonstrated by Sung et al. (2005) for determining
for regular concrete and 0.75 for lightweight concrete;
the PH characteristics of the frame columns. For a
specific RC component, the relationship between the γ is a constant depending on the volumetric ratio of the
moment and curvature ( M −  ) can be obtained using horizontal confinement reinforcement in the joint and
the strip section method, which is based on the detailed the classification of the BCJ. Specific values of γ can be
configuration of the component. By integrating the found in Table 1, where f 'c is the strength of concrete and
curvature along the RC component, the relationship Aj is the effective cross sectional area of the joint.
between the bending moment and flexural rotation A schematic diagram of the PH of a BCJ is
( M b −  ) can be established when considering the shown in Fig. 3, where the parameters were based
flexural capacity of the component, as shown in Fig. 2(a). on the specifications of FEMA-356, in which the
Note that the condition where the shear capacity of building performance levels are classified as Collapse
the RC component decreases as inelastic deformation Prevention (CP), Life Safety (LS), and Immediate
proceeds is also included in this approach. As a result, Occupancy (IO). In Fig. 4, points A through E illustrate
the shear capacity, which consists of the relationship the behavior of the PH. Point A is the origin, B is the yield

M M M
Mb Mv
Mv
Mv Mb Mb

θy θu θy θu θy θu
(a) Shear failure (b) Flexural-shear failure (c) Flexural failure
Fig. 2 Failure modes of a column or beam and their PH characteristics (Sung et al., 2005)

Table 1 Values of the constant γ specified in FEMA 356


Interior joint External joint
Without Without
ρ" * With transverse With transverse Knee joint
transverse transverse
beams beams
beams beams
< 0.003 12 10 8 6 4
≥ 0.003 20 15 15 12 8
*
ρ" = volumetric ratio of the horizontal confinement reinforcement in the joint.
376 EARTHQUAKE ENGINEERING AND ENGINEERING VIBRATION Vol.12

point, C is the end point of the plastic range, D represents


Axial force Pstrut
the residual strength, and E is the failure state. The
parameters of C, D, and E are dependent upon the ratio
Bj
of P/Ag f 'c, the level of conformity or nonconformity of Cj
Pn
the horizontal confinement reinforcement, and the ratio
of V/Vn, as shown in Fig. 3, where P is axial force, Ag is Dj
0.6 Pn
the gross cross sectional area of the joint, and V is the 0.4EcAg
shear force of the joint. The parameters a and b refer to Ej
the portions of the deformation that occur after yielding; 0.2 Pn
that is, the plastic deformation. The parameter c is Aj δy a 0.5(a+b) b
the reduced resistance after the sudden reduction Axial displacement δstrut
from C to D. Fig. 4 Proposed behavior of the PH of a BCJ
I.O.

L.S. of cross struts in the diagonal direction when resisting


horizontal loading, as illustrated in Fig. 5(b). The
Q C.P.
surrounding components of the BCJ are simulated by
Qy
a rigid bar with a hinge connection on the end point,
b where the height of the model is the depth of the beam,
a
and the width equals the effective width of the column.
The complex behavior of the BCJ is subsequently
1.0 C simulated by a cross-strut model with an equivalent
B two-force component. With help from the proposed
model, based on the characteristics of the BCJ suggested
D E c in the previous section, the complex force-deformation
A
relationship of the beam-column connection zone can
θ or Δ
Fig. 3 Behavior of the PH of a BCJ be modeled by setting the axial PH in the SAP 2000
software and performing PA. As shown in Fig. 6, the
relationship between the horizontal shear force V and
2.2.2 Simulation of a BCJ in a RC frame displacement δ is transformed into the direction of the
Since the intersection of a beam and column in a RC strut, and is derived as
frame is often treated as a rigid connection, with the basic
assumption that failure of the joint can be simplified and Pstrut = V / 2 cos  (4)
neglected in numerical analysis, the influence of shear
failure on BCJs may not be fully reflected in the PA
process.
Based on FEMA-356, the values used to define the P
PH characteristics of a BCJ are calculated as shown in
Fig. 4, where Aj is the initial point and Bj represents the
yielding. Since no definition of the initial stiffness of
the PH between Aj and Bj is suggested in FEMA-356, the
proposed stiffness is taken from the values for 15 different
BCJ specimens derived by Sezen and Alendar (2007), Column, Beam plastic hinge
where the initial recommended stiffness of 0.4 EcAg Column, Beam plastic hinge Axial plastic hinge
best fits practical experimental data. Since shear failure P
P
is a common cause of failure of a BCJ, the strength
at point Cj, the final point of the nonlinear stage, is
conservatively set as the same value as at Bj. Point Dj
is defined to represent the residual strength, and the
strength and axial displacement can be estimated as the Column plastic hinge
mean values at points Cj and Ej, where the strength at Ej Rigid bar Hinge
is 0.2 Pn. The axial displacement at Cj and Ej are obtained Two plastic hinge Rigid bar
Beam plastic hinge
from FEMA-356. Beam
Beam plastic hinge
Beam
With the help of the proposed PH curve, the Rigid point Equivalent structural
diagonal strut
commonly seen beam-column model shown in Fig. 5(a) Column
Column plastic hinge
Rigid point
can be modified as an equivalent numerical model Column
by adopting the cross-strut model for BCJs proposed (a) Beam column model (b) Beam column joints model
by Youssef and Ghobarah (2001). The force and
deformation of the BCJ is simulated by utilizing a pair Fig. 5 Cross-strut model for BCJ simulation
No.3 Y.C. Sung et al.: Pushover analysis of reinforced concrete frames considering shear failure at beam-column joints 377

Horizontal shear V Axial force Pstrut

Pstrut = V/2cosθ
Bj Cj Bj Cj
Vn Pn δstrut = δ×cosθ

Dj Dj

Ej Ej
Aj Aj

Horizontal displacement δ Axial displacement δstrut

(a) Plastic hinge of beam-column joints (b) Plastic hinge of diagonal struts
Fig. 6 Relationship between the PHs of a BCJ and diagonal struts

 strut =  × cos  (5) spacing of the stirrups was set as 8 cm in the BMDF
specimen, which constitutes a ductile design, the beams
where Pstrut is the equivalent axial force on the strut; V and columns were governed by flexural failure. The
is the equivalent horizontal shear force on the strut; δstrut specimen was tested in a typical pushover experiment,
is the equivalent axial displacement; δ is the equivalent and with the degradation of lateral force, shear cracks
horizontal displacement; and θ is the angle of the strut occurred on the BCJs when the drift ratio of the roof
from horizontal. floor reached 0.75%. Then, shear failure was observed at
The global behavior of RC frames can be easily the same location with a ± 3.0% drift ratio, as shown in
estimated analytically by setting the PHs on the beam, Table 2, while PHs were investigated on the two column
column, and BCJ using PA. joints with a ± 2.0% drift ratio. To accurately evaluate
the effectiveness of the procedures in the experiment,
two sets of settings for the PH of the BCJs were used
3 Experimental verification
in numerical analysis. The first settings used the values
suggested by the FEMA-356 guidelines. Unlike other
In order to verify the applicability of the proposed
analytical procedures where the BCJ is treated as a rigid
pushover procedures, the behavior of specimens with
joint, all the joints were simulated using the cross-strut
three different scales were examined under cyclic
loading. These specimens consisted of one-bay one-
story, two-bay two-story, and three-bay three-story
300 mm

500 mm

frames with different quantities of stirrups and different 300 mm


spacing. The failure mechanisms of the beams and 2500 mm 300 mm
columns were first determined to be flexural, shear-
500 mm
flexural and shear failure according to the design
drawings. The seismic capacity curves of the specimens 25-D13@100 D19
were then analyzed by considering the PHs on beams
1700 mm

and columns with the corresponding parameters shown


in Fig. 2 and possible shear failure at joints through D13@80 D19

numerical modeling, and the findings were compared


700 mm
with the experimental results.
Beam section
3.1 Benchmark ductile frame (BMDF) specimen Column section
300 mm

As shown in Fig. 7, the BMDF specimen tested at


300 mm

10-D19
40 mm

the Center for Research on Earthquake Engineering


500 mm

D13
8-D19
(NCREE) consisted of a 3 m × 5 m in plane and 3/5 500 mm
40 mm
scaled RC frame with one-bay and one-story, and was D13
designed to simulate a practical low-rise moment-
resisting frame structure (Feng et al., 2011). Since the Fig. 7 Detailed layout of the BMDF specimen

Table 2 Observed drift ratio values at different states of the BMDF specimen
Joint number Shear crack Shear failure PH forming
1J1 0.75% ± 3% ----
1J2 0.75% ± 3% ----
1C1-B ---- ---- ± 2%
1C2-B ---- ---- ± 2%
378 EARTHQUAKE ENGINEERING AND ENGINEERING VIBRATION Vol.12

model. The second settings used different values for the The pushover curve of the specimen cannot be clearly
beam and column joints instead of those recommended observed, since the initial stiffness cannot be successfully
by FEMA-356. The global behavior of the RC frame estimated before the yielding point. In comparison with
can be precisely described by the combination of these the curve predicted by FEMA-356, where a significant
settings. difference occurred throughout the entire process, an
As shown in Fig. 8, the differences in the initial acceptable conservative result can be obtained by the
stiffness, strength, and ductility of the structure were proposed method.
investigated when all the joints were analyzed while Along with the experimental failure sequence and
not considering the BCJ effect and using the default final mechanism of the specimen illustrated in Fig. 9,
settings in FEMA-356. The green line illustrating the the behavior of a PH obtained by the experiments and
model without considering the BCJ effect expresses the numerical analysis was also compared to evaluate the
initial stiffness correctly; however, the ultimate strength feasibility of the proposed method. As shown in Table
is overestimated throughout the whole nonlinear region, 3, the initial degradation of stiffness was predicted
which reflects the main drawback of traditional methods from the drift ratio of ± 0.87%. This fit well with the
that cannot describe the degradation phenomena. experimental results of 0.75%, displayed in Fig. 8,
This method was therefore neglected in the case of where the PHs on the BCJs were observed. The
the following two specimens. Furthermore, the initial estimated strength decreased rapidly, with shear failure
stiffness was overestimated, while the stiffness after the on BCJs occurring when the drift ratio reached ± 1.64%,
yielding point was underestimated using the guidelines and remained at the same level during the remaining
in FEMA-356 regarding PHs on the beams and columns. stages. Shear failure can therefore be expected in BCJs
when the drift ratio is ± 2.92 %, which was observed as
± 3.0% in the experiment.
Drift ratio (%)
-5 -4 -3 -2 -1 0 1 2 3 4 5
In general, the strength and initial stiffness derived
700 from the experiments can be effectively modeled with
600 the proposed method. An error was only observed when
500
400
300 Axial hinges
Lateral force (kN)

200
100
0
-100
BMDF frame
-200
Experimental
-300 Proposed method
FEMA
-400 Not consider beam-column joints
Plastic hinges
-500
-600
-700
-75 -50 -25 0 25 50 75
Displacement (mm) Failure 1 2
Fig. 8 Comparison of pushover curves for BMDF specimen Fig. 9 Failure sequence and final mechanism of BMDF

Deformed shape (PUSHEQX)-step 39


200
1J1 1J2 200
150
Axial force (kN)

150
Axial force (kN)

100

100
50 Input P plastic hinge
SAP hinge state step 39
50 Input P plastic hinge
0
0 2.5 5.0 7.5
Axial hinges SAP hinge state step 39

Displacement (mm)
(E) 0
0 2.5 5.0 7.5
Displacement (mm)

400 400

300
1C1-B Plastic hinges 1C2-B 300
(103kN.mm)

(103kN.mm)

200 (D) 200

Input M3 plastic hinge Input M3 plastic hinge


SAP hinge state step 39 SAP hinge state step 39
Marr

Marr

100 100

0 00 0.03 0.06 0.09 0.12 0.15


0 0.03 0.06 0.09 0.12 0.15
Rotation (Rad.) Rotation (rad.)

Fig. 10 Failure sequence of BMDF by using the proposed method (at Step 39)
No.3 Y.C. Sung et al.: Pushover analysis of reinforced concrete frames considering shear failure at beam-column joints 379

Table 3 The analytical sequence of the PH of the BMDF mechanism and behavioral sequence of a PH in the NF
specimen are shown in Figs. 13–15 and Tables 4 and
Drift ratio PH
State step 5. Shear cracks were first observed at a roof drift ratio
(%) 1J1 1J2 1C1-B 1C2-B of 2%, and significant failure was then observed at the
1 0.02 B-C central BCJ on the second floor at a drift ratio of ± 2.5%.
2 0.02 B-C
Meanwhile, the phenomenon of concrete spalling was
investigated for the right column of the first story at a
13 0.87 Bj-Cj Bj-Cj drift ratio of ± 3.0%. Although a slight difference was
14 1.03 C-D C-D observed between the results of the numerical analysis
15 1.08 D-E D-E and experimental observation, the failure sequence was
captured well by the proposed method.
22 1.64 Dj- Ej With regard to the capacity curve of the NF
39 2.92 Ej Ej specimen, the data derived from numerical analysis
was compared with the experimental observations as
shown in Fig. 14. Again, the experimental pushover
the curve reached a level of negative stiffness. The curve results were effectively predicted by the proposed
estimated behavioral sequence of the PH and the failure method. The initial stiffness and the curve before a drift
mechanism of the BMDF (see Fig. 10 and Table 3) also ratio of 2% were estimated correctly. Shear failure was
matched the experimental results. Flexural failure was also observed at the central BCJ of the first story when
first found at the column bottoms at levels D-E, and the drift ratio of the roof exceeded ± 1.46%. The lateral
shear failure at Level Ej then developed. This was highly
consistent with the experimental failure mechanism 400 mm
600 mm

400 mm
400 mm

600 mm
depicted in Fig. 9. 3600 mm 3600 mm

3.2 Non-ductile frame (NF) specimen


2150 mm

Following the satisfactory results from the one-bay D10@300


one-story RC frame, the proposed method was applied 75 mm
75 mm
50 mm
75 mm
75 mm
50 mm

to the NF specimen. This specimen was designed to 400 mm


represent a non-ductile two-bay two-story moment- 1000 mm
D10@100
1200 mm 1000 mm
D10@150 D10@100
1000 mm
D10@100
1200 mm
D10@150
1000 mm
D10@100
2150 mm

resisting frame with no wall (Wang et al., 2006), since


10-D19
the spacing of the stirrup in the columns was set as 30 cm,
and the columns were controlled by flexural-shear or
600 mm
shear failures. The height of the specimen was 21.5 m
and the width of each bay was 36 m. D10 rebar was Column section Beam section
used as the main reinforcement, and also comprised
the stirrup in both of the beam and column sections. In
400 mm

contrast with the layout of reinforcement in the columns 10-D19 10-D19

and BCJs, the beam elements were designed with high


D10 D10
ductility (see Fig. 11). A simplified model of the RC
400 mm 300 mm
frame with six BCJs is illustrated in Fig. 12. 40 mm 40 mm
A full comparison of the results of numerical
analysis and the experiment concerning the failure Fig. 11 Detailed layout of NF specimen

Column Beam-column joints Column, Beam plastic hinge Axial plastic hinge
Beam
P
P

Fig. 12 Simplified model of RC frame of BCJs with shear failure


380 EARTHQUAKE ENGINEERING AND ENGINEERING VIBRATION Vol.12

force began to degrade along with the flexural failure of


the top of the central column at the second floor. Details
concerning how the PHs progressed are illustrated in
Plastic hinges Fig. 15 and Table 5. The PH of the column bottom at
the first story progressed to the yielding level C-D with
slight stiffness degradation when the drift ratio of the
roof reached ± 0.92%. The lateral force was absorbed
primarily by the BCJ until the drift ratio achieved
Axial hinges ± 1.87%, at which point the stiffness decreased further,
Failure 1 2 3 leading to the occurrence of the shear failure Ej of the
central BCJ of the first floor and the flexural failure E
of the central column top at the second story. The global
behavior of the two-bay two-story specimen can thus be
Fig. 13 Behavioral sequence of failure mechanism and final accurately predicted by the proposed method.
condition of the NF Specimen

Drift ratio (%) 3.3 Three-bay three-story frame (TTF) specimen


-4 -3 -2 -1 0 1 2 3 4
500 The pushover procedures were applied to moment-
400 resisting frames to demonstrate their feasibility in
300
analyzing real structures. The TTF specimen consisted
of three stories and three bays, and sought to replicate
200 a moment-resisting frame structure used in Italy in the
Lateral force (kN)

100 mid-1970s (see Fig. 16) (Calvi et al., 2002). Each story
0
of this specimen, which consisted of three beam sections
and one column section, had a height of 2 m. The width
-100 of each bay was 3 m.
-200 NF Frame The behavioral sequence and failure mechanism
-300 Experimental obtained from experimental results and via numerical
Proposed method
modeling using pushover procedures was compared
-400 (see Fig. 17 and Tables 6 and 7). According to the
-500 experimental observations, shear cracks were first found
-196 -147 -98 -49 0 49 98 147 196
on the outside BCJs of the second story when the roof
Displacement (mm)
drift ratio was 0.6%. Shear failure was then found on the
Fig. 14 Comparison of analytical and experimental same location when the drift ratio was ± 1.0%. As shown
pushover curves

Deformed shape (PUSHEQX)-step 47

500
250
400
Moment (103kN.mm)

200
2C2-T
Moment (103kN.mm)

300

200 Axial hinge (E) Plastic hinge (D) 150

100
Input M3 plastic hinge
100 SAP hinge state step 47 Input M3 plastic hinge
50 SAP hinge state step 47
00 0.01 0.02 0.03 0.04
Rotation (rad)
1J2 00 0.01 0.02
Rotation (rad)
0.03 0.04

Plastic hinge (C)

1C1-B 1C2-B 1C3-B

2.5E+05 2.5E+05
250
Moment (103kN.mm)

2.0E+05
Moment (103kN.mm)

2.0E+05
Moment (103kN.mm)

200
1.5E+05 1.5E+05

150
1.0E+05 1.0E+05
Input M3 plastic hinge Input M3 plastic hinge
Input M3 plastic hinge SAP hinge state step 47
SAP hinge state step 47 100 SAP hinge state step 47
0.0E+00 0.0E+00
0 0.01 0.02 0.03 0.04 0.05 0 0.01 0.02 0.03 0.04 0.05
Ratice (rad) 50 Ratice (rad)
0 0.015 0.03 0.045
Rotation (rad)

Fig. 15 Behavioral sequence of PH under PA


No.3 Y.C. Sung et al.: Pushover analysis of reinforced concrete frames considering shear failure at beam-column joints 381

in Fig. 17, the first PHs suffering shear failure on level Table 4 Experimental observation of the NF specimens
Ej were predicted to be on both sides of the external BCJ Concrete
on the first story, and were followed by flexural failure of Joint number Shear crack Shear failure
spoiling
levels D-E and C-D at the top and bottom of the columns 1J2 0.375% ± 2.5%
on the first floor. The numerical analysis results shown 2C2-T
in Fig. 18 prove that the behavioral sequence can be 1C1-B
estimated correctly. 1C2-B
The estimated capacity curves of the TTF specimen 1C3-B 3%
are illustrated in Fig. 19. The stiffness slightly decreased

Table 5 PH distribution at different state steps observed from PA for NF

Drift ratio PH
State step
(%) 1J2 2C2-T 1C1-B 1C2-B 1C3-B
1 0.04 B-C
2 0.09 B-C B-C B-C
14 0.92 C-D C-D
15 1.03 D-E C-D
23 1.46 Bj-Cj
30 1.87 Dj-Ej
31 1.87 Ej
33 1.97 E
48 2.73 E
50 2.76 E

Table 6 Experimental observation of specimen TF


Joint number Shear crack Shear failure
1J1 0.6% ± 1.0%
at the PHs of level C-D at the column bottoms of the first 1J4 0.6% ± 1.0%
floor when the drift ratio was ± 0.47%. The lateral force
was successively absorbed mostly by the BCJs on both
sides of the first floor and the PH of the column in the story when the drift ratio was ± 1.289%. The fact that the
second story before the drift ratio reached ± 0.9%. The proposed method predicted the global behavior of the
lateral force started to degrade significantly when the three-bay three-story RC frame most precisely among
drift ratio reached ± 1.0%, at which point shear failure the three different specimens used in this study verified
of the BCJs was observed. Shear failure occurred on the feasibility of this method.
level Ej on both sides of the external BCJ of the first
382 EARTHQUAKE ENGINEERING AND ENGINEERING VIBRATION Vol.12

Table 7 PH distribution at different State steps observed from PA for specimen TF


Drift ratio PH
State step
(%) 1J1 1J4 1C2-T 1C3-T 1C1-B 1C2-B 1C3-B 1C4-B
1 0.007 B-C B-C
2 0.030 B-C
3 0.048 B-C B-C
4 0.069 B-C
19 0.363 D-E
20 0.363 C-D
21 0.386 C-D
22 0.403 C-D
25 0.474 C-D
35 0.660 Bj-Cj
37 0.687 D-E
38 0.687 Bj-Cj
67 1.181 Dj-Ej
68 1.181 Ej
75 1.289 Dj-Ej
76 1.289 Ej

Deformed shape (PUSHEQX)-step 76

75 75
Axial force (kN)

50 50

Axial force (kN)


25 25
Input P plastic hinge Input P plastic hinge
SAP hinge state step 76 SAP hinge state step 76
0
0 0.5 1.0 1.5 2.0 2.5
Displacement (mm)
3.0 3.5
Axial hinge (E) 00 0.5 1.0
Rotation (rad)
1.5 2.0

20 17.5

15
1J1 1J4
15.0

1C2-T
Moment (103kN.mm)

1C3-T
12.5
Axial force (kN)

Input M3 plastic hinge 10.0


10 SAP hinge state step 76

Plastic hinge 7.5

5 (C&D) 5.0
Input M3 plastic hinge
1C1-B 2.5 SAP hinge state step 76

0
0 0.02 0.04 0.06 1C2-B 1C3-B 1C4-B 0
0 0.03 0.06 0.09 0.12 0.15 0.18
Rotation (rad) Rotation (rad)

12.5 12.5 17.5 17.5

15.0 15.0
10.0 10.0
Moment (103kN.mm)

Moment (103kN.mm)

Moment (103kN.mm)
Moment (103kN.mm)

12.5 12.5

7.5 7.5 Input M3 plastic hinge 10.0 10.0


SAP hinge state step 76
7.5 7.5
5.0 5.0
Input M3 plastic hinge 5.0 5.0 Input M3 plastic hinge
SAP hinge state step 76 Input M3 plastic hinge SAP hinge state step 76
2.5 2.5 SAP hinge state step 76
2.5 2.5

0 0 0
0 0 0.02 0.04 0.06 0 0.03 0.06 0.09 0.12 0.15 0.18 0 0.03 0.06 0.09 0.12 0.15 0.18
0 0.02 0.04 0.06
Rotation (rad) Rotation (rad) Rotation (rad) Rotation (rad)

Fig. 18 Behavioral sequence of PH under PA

4 Summary and conclusions are considered: shear failure, flexural-shear failure, and
flexural failure. The characteristics of the PHs of BCJs
This study proposed a novel pushover analysis (PA) are set on the basis of FEMA-356, and a cross-strut
method for reinforced concrete (RC) frame structures model is used to simulate the detailed behavior of the
that takes shear failure at beam-column joints (BCJs) joint in the PA. By using simplified simulation methods
into consideration by utilizing a cross-strut model to for the PHs of beams, columns, and BCJs, the nonlinear
simulate the joint behavior. In contrast to other methods behavior of concrete frame structures can be derived.
that follow the guidelines of common specifications To evaluate the performance of the proposed
such as FEMA-356, in the proposed method the plastic method, three different specimens were studied under
hinges (PHs) of the beams and columns are set while pushover testing, and the possible shear failure at joints
three different failure modes of the beam and column was compared with the experimental results. Nonlinear
No.3 Y.C. Sung et al.: Pushover analysis of reinforced concrete frames considering shear failure at beam-column joints 383

Drift ratio (%)


-2.0 -1.5 -1.0 -0.5 0 0.5 1.0 1.5 2.0 “Experimental Test on a Three Story RC Frame
70 Designed for Gravity Only,” 12th European Conference
60 on Earthquake Engineering, London.
50
Eligehausen R, Ožbolt J, Genesio G, Hoehler MS and
40
Pampanin S (2006), “Three-dimensional Modelling of
30
Poorly Detailed RC Frame Joints,” New Zealand Society
Lateral force (kN)

20
10
of Earthquake Engineering (NZSEE), New Zealand.
0 Elmorsi M, Kianoush MR and Tso WK (2000),
-10 “Modeling Bond-slip Deformations in Reinforced
-20 Concrete Beam-column Joints,” Canadian Journal of
-30 Test frame Civil Engineering, 27: 490–505.
-40 Experimental
Federal Emergency Management Agency (2000),
Proposed method
-50 Prestandard and Commentary for the Seismic
-60
Rehabilitation of Buildings, FEMA-356, Washington
-70
-125 -100 -75 -50 -25 0 25 50 75 100 125 D.C., Building Seismic Safety Council.
Displacement (mm) Feng Lin Chou, Yaw Shen Tu, Jainn Cheng Chen
Fig. 19 Comparison of analytical and experimental pushover and Shih Yi Kuo (2011), “Simplified Assessment of
curves of specimen TF Ultimate Strength and Deflection of RC Squat Wall with
Opening,” Journal of Architecture, 78: 19–37.
pushover procedures were implemented using finite
element analysis and the proposed PH settings. The Hanson NW and Conner HW (1967), “Seismic
relationship between the base shear and roof displacement Resistance of Reinforced Concrete Beam-column
of the RC frames when shear failure of the BCJs occurred Joints,” Journal of Structural Engineering, ASCE,
was obtained and compared with pushover curve data 93(ST5): 553–560.
from the experiments. The behavioral sequence and Meinheit DF, and Jirsa JO (1981), “Shear Strength
failure mechanisms of the frames obtained via numerical of Reinforced Concrete Beam-column Connections,”
modeling and experiments were also verified to check Journal of Structural Engineering, ASCE, 107(ST11):
their consistency. The results of this comparison 2227–2244.
demonstrated that the proposed method is able to yield Mitra N and Lowes L (2004), “Evaluation and
improved predictions and estimate the initial stiffness of Advancement of a Reinforced Concrete Beam-
the pushover curve well, while the method suggested by column Joint,” 13th World Conference on Earthquake
FEMA-356 results in a wide margin of error. Acceptable Engineering, Vancouver, B.C., Canada, Paper No.
results after yielding can also be obtained by using the 1001.
proposed procedures. As demonstrated by the three
Sezen H and Alemdar F (2007), “Evaluation of
specimens of different scales and types examined in this
FEMA-356 Models for Reinforced Concrete Columns
study, the behavior of the BCJs and adjacent elements,
and Beam-column Joints,” Journal of the Structural
as well as the failure sequence of PHs of the frame as a
Division, ASCE.
whole, can be precisely and effectively modeled using
the proposed method. Sung YC, Liu KY, Su CK, Tsai IC and Chang KC
The three main contributions of the new method, (2005), “A Study on Pushover Analyses of Reinforced
which consists of a new definition of beam and column Concrete Columns,” Journal of Structural Engineering
failure modes, the proposed characteristics of the PHs of and Mechanic, 21(1): 35–52.
BCJs, and the use of a novel cross-strut model to simulate Tajiri S, Shiohara H and Kusuhara F (2006), “A
detailed joint behavior in PA, render PA simpler and New Macro Element of Reinforced Concrete Beam-
more effective. In comparison with existing methods, column Joint for Elasto-plastic Plane Frame Analysis,”
this approach can provide an alternative method for use Proceedings of the 8th U.S. National Conference on
by engineers. Earthquake Engineering, San Francisco, California,
USA, Paper No. 674
Wang YC, Huang JS, and Chen BC (2006), “Testing
References on 2-story and 2-bay Reinforced Concrete Frame with
Substandard Reinforcing Details,” Third International
ACI Committee 318 (2011), Building Code Requirements Conference on Urban Earthquake Engineering, Tokyo,
for Structural Concrete (ACI 318-11) and Commentary, Japan.
Farmington Hills, Michigan, U.S.A., American Concrete Youssef M and Ghobarah A (2001), “Modeling of RC
Institute. Beam-column Joints and Structural Wall,” Journal of
Calvi GM, Magenes G and Pampanin S (2002), Earthquake Engineering, 5(1): 93–111.

View publication stats

S-ar putea să vă placă și