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

net/publication/321196861

Fiber beam-column model for diagonally reinforced concrete coupling beams

incorporating shear and reinforcement slip effects

Article  in  Engineering Structures · December 2017

DOI: 10.1016/j.engstruct.2017.10.035

CITATION READS

1 455

4 authors, including:

Ran Ding Mu-Xuan Tao

Tsinghua University Tsinghua University
22 PUBLICATIONS   54 CITATIONS    98 PUBLICATIONS   613 CITATIONS   

SEE PROFILE SEE PROFILE

Xin Nie
Tsinghua University
70 PUBLICATIONS   169 CITATIONS   

SEE PROFILE

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

Mechanism and Design Method of Slab Composite Effect in Composite Structural Systems View project

New High-Efficiency Numerical Models for Elaborate Simulation of Structural Systems View project

All content following this page was uploaded by Mu-Xuan Tao on 20 October 2018.

The user has requested enhancement of the downloaded file.

This manuscript was published at: Ding R, Tao M X, Nie X*, Mo Y L. Fiber beam-column model for diagonally reinforced
concrete coupling beams incorporating shear and reinforcement slip effects. Engineering Structures, 2017, 153: 191-204.
The final publication is available at the journal website. The researchers can also privately get the final publication version via
sending Email or ResearchGate message to Prof. Mu-Xuan Tao (taomuxuan@tsinghua.edu.cn).

Fiber Beam-Column Model for Diagonally Reinforced Concrete

Coupling Beams Incorporating Shear and Reinforcement Slip Effects

Ran Ding a, Mu-Xuan Tao b, Xin Nie c* and Y.L. Mo d

Abstract: Due to the improved energy-dissipation and deformation capacity compared to the

conventionally reinforced concrete (RC) coupling beams, diagonally RC coupling beams are

recommended by the ACI 318 code especially for a span-to-depth ratio of less than two and thus

acquire more and more applications in coupled wall and core tube systems for tall buildings. This

paper proposes a sufficiently accurate and efficient displacement-based fiber beam-column model

for the nonlinear seismic analysis of diagonally RC coupling beams with span-to-depth ratios

ranging between one and five. The model is developed on the platform of a general FEA package

MSC.Marc. First, the conventional fiber beam-column element is modified to consider the flexural

contribution of diagonal bars. Then the new section shear force-shear distortion and slip

deformation rules are proposed and incorporated into the modified fiber element, respectively, since

both the shear and reinforcement slip are critical mechanisms influencing the seismic performance

of the beam. The equations for critical model parameters including the cracked shear stiffness and

chord rotation limit are developed and verified based on the results of sixteen test specimens

collected from previous research. The model is utilized to simulate the collected specimens together

a. Postdoctoral researcher, Key Lab. of Civil Engineering Safety and Durability of China Education Ministry, Dept. of Civil
Engineering, Tsinghua University, Beijing, China 100084.
b. Associate Professor, Beijing Engineering Research Center of Steel and Concrete Composite Structures, Tsinghua University,
Beijing 100084, China.
c. Postdoctoral researcher, Key Lab. of Civil Engineering Safety and Durability of China Education Ministry, Dept. of Civil
Engineering, Tsinghua University, Beijing, China 100084. Email: xinnie@mail.tsinghua.edu.cn
d. Professor, Dept. of Civil and Environmental Engineering, University of Houston, Houston, USA 77204.

1
with a coupled wall system and proves to be a powerful tool for the nonlinear seismic analysis of

diagonally RC coupling beams and coupled walls with satisfied accuracy, efficiency and modeling

convenience.

Author Keywords: fiber beam-column model; diagonally reinforced concrete coupling beams;

seismic analysis; hysteretic behavior; nonlinear shear behavior; reinforcement slip

2
 Nomenclature
A area of the coupling beam section s spacing of section confining transverse rebars
Asd sectional area of diagonal reinforcements Vcr shear force of the shear cracking point
Asv sectional area of section confining transverse Vm shear force of unloading starting point from the
reinforcement skeleton curve
b width of the coupling beam Vre shear force of reloading beginning point
d effective depth of the coupling beam Vsd,i shear force of strength deterioration point
Vssd,i
Dd diameter of diagonal bars Vun shear force of unloading starting point from the
other curves
Dl diameter of top or bottom longitudinal bars α inclination angle of diagonal bars
E concrete elastic modulus β the shear stiffness reduction factor
f c' cylinder concrete compressive strength δext slippage of the bar caused by strain
accumulation along the development length
fsd current stress of the diagonal bars δimit slip corresponding to γimit
fyd yield strength of the diagonal bars δm slip deformation of strength deterioration point
fyl yield strength of top or bottom longitudinal bars δre slip of reloading beginning point
G concrete shear modulus = E/2(1+v) δt slip of reloading turning point
h depth of the coupling beam δun slip deformation of unloading starting point
from other curves
i counter tracing successive cycle numbers in one δun,m slip deformation of unloading starting point
direction reaching current maximum strain from the skeleton curve
k1 k2 shape control parameters for monotonic tensile εs current strain of the diagonal bars
and k3 stress-strain curves of rebar
kcr cracked shear stiffness εy yield strain of the diagonal bars
kini initial shear stiffness of the shear force-shear γcr shear strain of the cracking point at the shear
strain skeleton curve skeleton curve
kre reloading reference slope γlimit drift ratio at strength degradation
kt slope before the reloading turning point γm shear strain of the strength deterioration point
l length or span of the coupling beam γre shear strain of reloading beginning point
l/h span-to-depth ratio of the coupling beam γt shear strain of reloading turning point
le length of elastic region for the bar anchorage γun shear strain of unloading starting point from
other curves
lpy length of inelastic region for the bar anchorage γun,m shear strain of unloading starting point from the
skeleton curve
nd number of diagonal bars in each bundle θext angle of the slip-induced crack at the beam ends
nl number of top or bottom horizontal bars ρsv ratio of section confining transverse
reinforcement, ρsv = nsvAsv/bs
nsv number of legs of sectional transverse rebars ρsd ratio of diagonal reinforcements, ρsd = ndAsd/bh
v concrete Poisson’s ratio = 0.17

3
1. Introduction

Nowadays reinforced concrete coupled walls and core tubes have become the most popular

structural system for tall buildings due to their efficient resistance to seismic actions and

compatibility with architectural requirements. In a coupled wall structural system the coupling

beams are key components, which contribute significantly to structural stiffness under frequent

earthquakes and energy dissipation during severe earthquakes. Numerous efforts have been devoted

to investigating the seismic behavior and improving the seismic performance of RC coupling

beams [1-14].

Diagonal confinement

Section stirups Diagonal bar

Longitudinal bar Longitudinal bar

(a) Conventional layout (b) Traditional diagonal layout

1200

800
Sectional confineme nt
400
Force (kN)

α 0

-400
Diagonal layout
Diagonal bar -800 (CB20-1)
Longitudinal bar Conventional layout
(CB20-2)
-1200
-12 -8 -4 0 4 8 12
Drift ratio (%)
(c) Improved diagonal layout (d) Typical hysteretic curves

Figure 1. Comparison of RC coupling beams with different reinforcement layouts

Among all kinds of RC coupling beams, in practice the beams with conventional and diagonal

reinforcement layouts are the most widely used. Compared with the coupling beams constructed

with reinforcement scheme similar to frame beams(Figure 1[a]), it is confirmed by many test

results that the diagonal reinforcement layout (Figure 1[b]) that was first introduced by Paulay and

Binney [3] benefits the coupling beam with significantly a higher deformation and

4
energy-dissipation capacity [3-11], as illustrated by Figure 1(d). In addition, ACI 318-14 [15]

introduced an alternative detailing option, as shown in Figure 1(c), where transverse reinforcement

is placed to confine the entire beam cross section instead of the diagonal bar bundles. Consequently

the construction of a diagonally RC coupling beam is greatly simplified.

Recently, the nonlinear dynamic analysis has gained increasing attention and application in the

performance-based seismic design of tall buildings, which requires nonlinear models for all

structural members with adequate accuracy, efficiency and numerical stability. Displacement-based

fiber beam-column elements have been successfully used to model the frame beams and columns

due to their simplicity and accuracy [16,17]. However, the diagonally RC coupling beams featured

with complex hysteretic behavior cannot be reasonably simulated by the conventional fiber model.

It is thus necessary to develop an enhanced fiber model for the seismic structural analysis of

coupling beams in tall buildings that is capable of considering the nonlinear shear and

reinforcement slip effects and still keeps the simplicity and convenience of the original fiber model.

To date, several researchers have reported macro models for diagonally RC coupling beams

based on springs, hinges and line elements [18-21]. The prior research has greatly advanced the

modeling of diagonally reinforced coupling beams. However, there remains a significant gap in the

literature. Previous models of coupling beams have not fully considered the three important

deformation modes, which are the shear, flexure and slip effects, with adequately verified model

parameters by a large number of tests reported in literature. Therefore the applicability of the

models remains doubtful. In addition, the models cannot be directly applied into the framework of

fiber models. We have previously proposed a fiber beam-column element for conventionally RC

5
coupling beams [22]. However, the model can consider neither the contribution of diagonal bars nor

the reinforcement slip mechanism and the higher deformation and energy-dissipation capacity of

diagonally RC coupling beams, which are the objects of this study.

Therefore, this study aims to develop a fiber beam-column model for diagonally RC coupling

beams based on a comprehensive representation of the complex deformation mechanism, which

can be directly applied to the seismic analysis of high-rise structures by the general FEA software.

It is worth noting that a total of sixteen test specimens collected from seven different research

groups are summarized to validate adequately the proposed model and calibrate critical model

parameters so that the model can accommodate beams with a broad range of span-to-depth ratios

from one to five. First, the fiber model is modified to consider the flexural contribution of diagonal

bars. Then the modified fiber element is combined with the proposed section shear force versus

shear strain and slip deformation laws in the general FEA software MSC.Marc [23]. The methods

for critical parameters including shear stiffness after cracking and chord rotation limit are also

developed and verified by the collected test results. Finally, the model is used to simulate the

seismic performance of all test specimens and a RC coupled wall specimen, and the predicted

hysteretic behavior is compared to the measured results.

2. Review of diagonally reinforced coupling beam tests

A number of tests have been conducted to assess the seismic behavior of diagonally RC

coupling beams [4-11]. In this study, sixteen test specimens are reviewed and the key parameters

are summarized in Table 1. As shown in Figure 1(b) and (c), two typical reinforcement layouts exist

in the literature. The first is proposed by Paulay and Binney [3] originally and recommended by

6
 ACI 318-05 [24], where the stirrups are placed to confine both the diagonal bars and the horizontal

bars. The second is an improved scheme and specified in ACI 318-14 [15], where stirrups only

surround the horizontal bars, giving rise to a much more convenient construction process.

Comparative tests of the two layouts have been conducted by Naish et al. [4], and it has been

proven that the seismic performance of beams detailed based on ACI 318-14 [15] is equivalent or

slightly better than beams designed as per ACI 318-05 [24].

Table 1. Database of diagonally RC coupling beam specimens

b h l α Dd fyd Dl fyl fc’ Vtest Vmodel Diagonal Section

Reference specimen l/h nd nl
(mm) (mm) (mm) (°) (mm) (MPa) (mm) (MPa) (MPa) (kN) (kN) stirrup (mm) stirrup(mm)

CB-24F 304.8 381 914.4 2.40 15.7 22.2 6 483 9.525 3 483 47.3 769.3 780.1 -1 D9.5@76
CB-24D 304.8 381 914.4 2.40 15.7 22.2 6 483 9.525 3 483 47.3 708.4 780.1 D9.5@63 D6.4@63
Naish [4]
CB-33F 304.8 457.2 1524 3.33 12.3 22.2 6 483 9.525 3 483 47.3 556.2 591.8 - D9.5@76
CB-33D 304.8 457.2 1524 3.33 12.3 22.2 6 483 9.525 3 483 47.3 536.7 591.8 D9.5@63 D6.4@63

CB10 250 500 500 1.00 26.0 25 4 486 10 3 468 34.5 1449.9 1401.3 - D13@100
CB20 300 500 1000 2.00 16.0 29 4 466 13 3 502 52.1 1068.6 1062.9 - D13@100
Lim [5,6]
CB30DA 300 500 1500 3.00 8.8 32 4 465 13 2 441 39.7 701.3 680.9 D10@150 D13@200
CB30DB 300 500 1500 3.00 8.8 32 4 465 10 3 475 38.4 732.9 701.5 - D10@100

Kwan and
CCB11 120 600 600 1.00 31.0 8 6 517 8 2 517 37.8 350.5 363.7 D6@60 D8@140
Zhao [7]

Fortney [8] DCB1 254 356 914 2.57 13.0 25.4 4 431 19 2 418 37.6 615.4 620.9 D9.5@76 D6.4@76

Galano and P07 150 400 600 1.50 22.1 10 4 567 6 2 567 54.0 238.6 251.6 - D6@140
Vignoli [9] P12 150 400 600 1.50 19.7 10 4 567 6 2 567 41.6 242.2 245.9 D6@90 D6@180

CB2A 130 500 500 1.00 26.0 10 4 504 6 3 281 28.5 287.5 300.3 D6@50 D6@120
Tassios [10]
CB2B 130 300 500 1.67 18.0 10 4 504 6 3 281 26.3 172.0 162.0 D6@50 D6@120

9.5 2 487 53.3 46.0

C6 101.6 169.4 423.4 2.50 18.2 5 2 492 18.1 - D4.9@34
12.72 1 408 62.73 57.3
Shiu [11]
9.5 2 569 37.1 34.9
C8 101.6 169.4 845.8 5.00 9.6 5 2 490 23.9 - D4.9@34
12.7 1 433 27.9 24.9
Notes: 1. For specimens following the recommended transverse reinforcement layout of ACI 318-14, diagonal stirrups are removed;
2. The diagonal bars are unsymmetrical for Specimens C6 and C8;
3. The measured maximum shear forces are different in two directions due to the unsymmetrical layout of diagonal bars.

It is concluded from the test results that the failure of the coupling beam is mainly attributed to

the crushing of concrete at the ends, and the buckling of diagonal bars under compression as well as

7
fracture under tension. The deformation consists of flexure, shear and reinforcement slip

mechanisms, which are closely related to the span-to-depth ratio l/h. For l/h <2.0, shear

deformation will dominate the overall behavior [5], whereas if l/h >2.0, beam-wall interface

rotation caused by reinforcement slip constitutes most of the total deformation [4,6]. It is worth

noting that the shear-sliding deformation mechanism and shear-tension failure mode both typical in

conventionally RC coupling beams [1,2] are avoided in diagonally RC coupling beams, due to the

contribution of diagonal bars. In addition, typical measured load-displacement curves exhibit much

better fully rounded hysteretic loops without pinching effects and much larger ultimate deformation

(as much as 5%-8% beam chord rotation) than conventional coupling beams [4,5]. The effective

stiffness is reported to be approximately 15%-20% of the initial bending stiffness EcIg generally and

can be even lower than 10% of EcIg [4,25]. Based on the above summary and analysis, the proposed

model should incorporate the three distinct deformation mechanisms and give a precise prediction

to the beam strength, effective stiffness, deformation and energy dissipation capacity.

Compressive force
by concrete strut Tensile force T=fsdndAsd (fsd>fyd )
α

Compressive force C=fydndAsd

Figure 2. Basic mechanism of diagonally reinforced coupling beams

3. Modified fiber model incorporating diagonal reinforcement

In recent years, the authors’ research group [16,17] has developed a conventional

displacement-based fiber beam-column element for RC, steel and steel-concrete composite

structural members in MSC.Marc [23] for the two-node beam element No. 98 with one integration

8
point in the middle. The uniaxial constitutive laws of steel, rebar and concrete have been discussed

in detail [16,17], which will not be repeated in this paper. Based on the traditional fiber element,

this research first extends the fiber model to simulate the flexural contribution of diagonal bars.

l b
A
α 3Al

ltanα

h
2Am
4Ad
Diagonal bar Longitudinal bar A A-A

(a) Actual reinforcement layout

l
A
3Al

ltanα
2Am
2Adcos(α)
Longitudinal bar A A-A
Equivalent diagonal bar
(b) Proposed fiber model

Figure 3. Modified fiber model incorporating diagonal bars (nd=4)

As pointed out by Hwang et al. [5,6] and explained in Figure 2, diagonal bars not only increase

the shear capacity with their vertical components but also improve the bending capacity with their

horizontal components. In addition, it is repeatedly verified by many test results that the flexural

strength of the diagonally RC coupling beams perfectly represents the bearing capacity of the beam,

since the shear strength is sufficient to prevent the shear failure featured with unfavorable wide

diagonal cracks. According to this mechanism, the horizontal component of the diagonal bars is

incorporated into the modified fiber model, as described in Figure 3. Each diagonal bundle is

divided into two equivalent bars, the area of which is 0.5ndAsdcosα. The equivalent bars are located

at the left and right side of the section with the y-coordinate equal to the centroid of the bundle.

Using this modified model, the maximum shear force sustained by the beam and the bending

behavior can be reasonably predicted. Table 1 list the comparison between the simulated (Vmodel)

9
and measured maximum shear force (Vtest) of the sixteen beams. The error ranges between -13.7%

and 10.3% and the average error is only 0.7% which clearly demonstrate that the modified fiber

model is able to predict the beam capacity accurately.

4. Section shear force-shear strain relationship

Shear deformation is a remarkable deformation component for RC coupling beams, especially

for beams with small span-to-depth ratios. In a recent study for conventionally RC coupling beams

[22], the authors have proposed to consider the flexure-shear deformation by incorporating the

nonlinear section shear force (V)-shear strain (γ) relationship into the traditional fiber model. As

illustrated in Figure 4, for the compression-flexure or tension-flexure behavior, the conventional

fiber model calculates the section tangent stiffness matrix as

 n n n

 ∑ E A
tk k ∑ E A y
tk k k − ∑ Etk Ak xk 
=
 k 1 =k 1 =k 1 
 n n n
 (1)
=k 1  ∑ Etk Ak yk ∑ Etk Ak yk2
−∑ Etk Ak xk yk 
=
 k 1 =k 1 =k 1 
 n n n

 −∑ Etk Ak xk −∑ Etk Ak xk yk ∑ 2
Etk Ak xk 
=  k 1 =k 1 =k 1 

where Etk is the fiber tangent stiffness of the kth fiber; Ak is the area of the kth fiber and xk and yk are

the x and y coordinates of the center of the kth fiber in the section local coordination system.

node
My
Y
Vy Mx
X Vx
Z Integration point
N
φy T
node
γy φx
γx
Integration point ε
ω
[N, Mx, My] =k1[ε, φx, φy]
T T

[Vx, Vy, T ]T=k2[γx, γy, ω]T

Figure 4. Generalized sectional stress, strain and stiffness

10
 For the shear-torsion behavior, the section tangent stiffness matrix is

dVx / dγ x 0 0 
 
k2 =  0 dVy / dγ y 0  (2)
 0 0 GI p 


The section force vector D is

T
 n 
D  ∑ σ k Ak , Vx ( γ x ) , Vy ( γ y ) , − ∑ σ k Ak yk , ∑ σ k Ak xk , GI pω 
n n
= (3)
= k 1 = k 1= k 1 

By defining the nonlinear section shear force (V)-shear strain (γ) relationship which is

discussed in the following two sections, the tangent shear stiffness in k2 can be calculated in each

step; thus the nonlinear shear behavior is simulated. This study follows the same principle as that

proposed for conventionally RC coupling beams [22], but modifies the backbone curves and

hysteretic rules so that the shear behavior of diagonally RC coupling beams can be reasonably

simulated. Actually, the shear stiffness after diagonal cracking, the unloading and reloading

stiffness together with the strength deterioration point are all re-investigated and new equations are

proposed.

4.1 Backbone curves

According to the observed damage mode of diagonally reinforced coupling beams, a bilinear

model is proposed as the backbone curve, which considers the significantly reduced shear stiffness

after shear cracking, while neglects the load degradation because no shear failure has ever been

observed and the flexural strength controls the beam capacity. As shown in Figure 5, the key

parameters in the two branches are defined as:

(0.158 f c′ + 34.4 ρsv d / l )bd ≤ 0.29 f c′bd

Vcr = (4)

kini = GA / 1.2 (5)

11
 kcr = β kini (6)

=β 1.2178 ρ sd (l / h)0.5 + 0.0139 (7)

It should be noted that Equation (4) is from ACI 426 [26] and holds for SI units. The

calibration of Equation (7) will be discussed in Section 7.1.

A (γu n,m,Vm) G
kcr
Vsd,1 A1(γu n,Vu n)
Vsd,2 A2
(γcr ,Vcr )
kini
kt F1 F
γm E1 B B1B2
E
(γre,0) γm γ
(γre,Vre) C2
J C1
backbon e u nloading point
D2 Vcr strength deterioration point
C
D1 reloading s tartin g poin t
D reloading turning point

H
K

Figure 5. Shear force-shear strain skeleton curves and hysteretic rules

4.2 Hysteretic rules

The detailed description of the unloading and reloading rules can be found in [22], this section

only emphasizes the difference between the conventionally and diagonally RC coupling beams.

Unloading rule

As shown in Figure 5(a), if Vm>Vcr(AB), or Vun>Vcr (A1B1), the unloading slope kun is calculated

as:

β ( k1 − k2 )
kun =
k1 −
120γ cr
(γ un,m − γ cr ) ; k1 =
3β kini ; k2 =
0.2 β kini (8)

Reloading rule

For reloading in the reverse direction, reloading starting from point (γre,0) follows a bilinear

curve before going back to the backbone curves. The curve passes through the reloading turning

12
point (γt, Vt) (B1C1, B2C2, EF and E1F1) and then aims at the strength deterioration reference point

(γm, Vsd) (C1D1, C2D2, F1A2 and FA1). The coordinates of the two reference points are calculated as

=γt (1.15-0.15l /h ) γ re (1 ≤ l /h ≤ 5) (9)

Vsd,i
= =
k t 2.5kre ; kre (10)
γ m − γ re

γ m = γ un,m (11)

γ un,m i
=
Vsd,i Vm (0.95 − ) (12)
1000γ cr

Upon the first unloading from the current maximum shear strain, i is set to be 1. The value of i

is counted every time the load reverses with the maximum strain ranging from 0.8γun,m to

1.2γun,m[22].

5 Section shear force-slip deformation relationship

5.1 Slip-induced beam deformation analysis

For beams with l/h larger than 2.0, it is found that the slip of diagonal reinforcement at the

beam-wall interface contributes significantly to the beam deformations. As shown in Figure 6(a).

the diagonal bars under tension may extend, or slip relative to the wall, due to the accumulated

tensile strain over the embedment length and the slip will cause rigid-body rotation of the beam.

Alsiwat and Saatcioglu [27] proposed an approach to predict reasonably the reinforcement

anchorage slip and Naish et al. [18] utilized the method with a few modifications to calculate the

slip-related rotations corresponding to the yield of diagonal bars.

13
 A s fs
reba r

uf 0.5 fc'

Stepped bond stress ub

ue

Figure 6. Diagonal reinforcement slip model in the proposed fiber model

This study aims to simulate the whole-process hysteretic response of slip deformations on the

basis of the method proposed by Naish et al. [18]. If adequate embedment of the diagonal

reinforcement is provided as is the case for all test specimens, the slip at the unloaded end can be

neglected. As shown in Figure 6(b), the reinforcement slip is calculated using a stepped-bond stress

distribution where an elastic uniform bond stress ue is assumed for elastic steel stresses and a

frictional uniform bond stress uf is assumed for stresses exceeding the yield strength. The elastic

bond stress ue can be calculated according to ACI Committee 408 [28] and equals to 0.87 f c′ MPa

for most cases. The inelastic bond stress uf is calculated as 0.5 f c′ MPa for simplicity as suggested

by Sezen et al. [29] and Pan et al. [30]. Thus the reinforcement slip can be calculated by integrating

the strains over the development length that equals the area below the strain diagram in Figure 6(b).

l f D
δ ext= 1.25 × ε s × e ; le= sd d ; ue= 0.87 f c′ (ε s ≤ ε y ) (13)
2 4ue

14
 l l +l l f D 2 1
δ ext= 1.25 × ε y × e + ∫ ε ( x)dx= 1.25 × ε y × e + ε y yd d (k3 − 1)( k1 + k2 );
e py

2 l e 2 4uf 3 3
(14)
( fsd − f yd ) Dd
=lpy = ; uf 0.5 f c′ (ε s > ε y )
4uf
δ ext
θ ext = (15)
d−x
where δext is the bar slippage caused by strain accumulation along the development length; le and lpy

represent the elastic and inelastic region lengths; θext represents angle of the slip-induced crack at

the beam ends; fsd, εs and εy are the current stress, strain and yield strain of the diagonal bars. The

factor 1.25 in Equation (13) is suggested by Naish et al. [18] to account for approximately the

severe beam end damage due to the shear effects in coupling beams with low span-to-depth ratio

and hysteretic loading influences. k1, k2 and k3 are the shape control parameters for the monotonic

tensile stress-strain curves of reinforcement [16,17], as shown in Figure 6(c); x is the neutral axis

depth and d is the effective depth with respect to the outmost diagonal bar.

lslip
Slip element at the end

Spread pl asticity fi ber element for flexure

node node
nonlinear shear hinge incorporated in the fiber model
Internal fiber element considering
nonlinear shear effects

Figure 7. Proposed coupling beam fiber model considering shear and slip deformation

Based on the above method, the slip-related rotation at the beam end can be calculated and then

the corresponding additional displacement of the beam is obtained as lθext. In the proposed coupling

beam model, as shown in Figure 7, the shear and flexural deformation are simulated by the internal

shear element which has been developed in the above section. On the other hand, the slip

15
deformation is simulated by two additional slip elements at the ends which are similar to the

internal shear element but modify the shear element by setting the shear stiffness according to the

shear force (V) versus slip deformation (δext) relationship. The slip deformation, regarded as the

shear strain of the slip element, is lθext/(2lslip), where lslip is the length of the slip element at the

beam ends. It is worth noting that despite the fact that the slip-related additional deformation is

actually the rigid-body rotation of the beam, the model assumes an equivalent deformation that is

the local shear deformation at beam ends so that the shear element can be conveniently converted to

the slip element in order to simulate the slip deformation with only a few modifications. Therefore,

the proposed coupling beam fiber model with shear elements in the middle and slip elements at the

ends in the series can reasonably reflect the deformation modes of diagonally RC coupling beams.

The V - δext relationship is illustrated in detail in Sections 5.2 and 5.3.

5.2 Backbone curves

(δs,3,Vs,3)

(δs,2,Vs,2)
Shear force V

(δs,1,Vs,1)
Calculated curve
Simplified trilinear curve

Slip deformation δ

Figure 8. Proposed simplified trilinear shear force-slip deformation model

The flexural analysis with the modified traditional fiber model considering diagonal

reinforcement is conducted first to obtain the monotonic relationship between the shear force and

the tensile stress of the diagonal bar. Then the slip deformation (δext) at each step can be calculated

16
by Equations (13)-(15), and the V - δext monotonic curve is thus obtained. However, the curve

cannot be explicitly expressed, which brings great difficulty to the development and numerical

stability of the element. To overcome this barrier, as shown in Figure 8, a simplified trilinear curve

is suggested to represent the calculated V - δext curve with Equations (13)-(15). Therefore the V

- δext backbone curves can be defined before analysis using an explicit expression, as shown in

Figure 9.

V
δlimit
Vs,3 G
Vs,2
Vssd,1 F R Z
Vssd,2 (δu n,Vu n) A1
A V
Descending
Backbone curve
branch Vs,2

H S
Y (δt,Vt) Q Vs,1 B1
kt E C1 k3
Vs,1 k2
δm δpY D δpE δpU I
X P δpQ O B T δm δ Vs,1 k
1 k4
(δre,0) Vs,1
Vs,1 C
U backbon e u nloading point
J δ
strength deterioration point
W N L(δre,Vre) reloading s tartin g poin t
reloading turning point
K

V (a) Definition of hysteretic rules (b) Definition of critical stiffnesses

Vssd,1
-δlimit M (δu n,m,Vm)

Figure 9. Proposed hysteretic shear force-slip deformation model

5.3 Hysteretic rules

The detailed unloading and reloading rules are similar to those of the shear force-shear sliding

deformation in conventionally RC coupling beams, which can be found in [22], thus this section

only emphasizes the difference between the two beams with different reinforcement layouts.

Unloading rule

1. If Vm < Vs,1, and δun,m < δs,2, or Vun < Vs,1 and δun < δs,2 (CD), the unloading slope equals to k1.

2. If Vm>Vs,1 and δun,m<δs,2, or Vun>Vs,1 and δun<δs,2, unloading will follow a linear curve. The

17
 stiffness is given by Equation (16) (AB and KL).

k −k Vs,2 + Vs,1
k1 − 1 2 (δ un,m − δ s,1 ) ; k1 =
kun = ks,1 ; k2 = (16)
δ s,2 − δ s,1 δ s,2 + δ s,1

where k1 and k2 are defined in Figure 9(b).

3. If δun,m>δs,2 and Vm>Vs,1, or Vun>Vs,1 and δun>δs,2, unloading will follow a bilinear curve. The

stiffness when the shear force is larger than Vs,1 is given by Equation (17) (GH, RS, MN, VW

and A1B1) and less than Vs,1 is given by Equation (18) (HI, ST, NP, WX and B1C1).

2δ k
kun = s,2 2 ≥ k3
δ un,m
(V > V )
s,1 (17)

4δ k
kun =s,2 2 ≥ k4
5δ un,m
(V ≤ V )
s,1 (18)

where k3 and k4 are defined in Figure 9(b).

4. If Vm<Vs,1 and δun,m>δs,2 (UV), or Vun<Vs,1 and δun>δs,2, the unloading slope can be calculated

with Equation (18).

Reloading rule

For reloading in the reverse direction, reloading starting from point (δre, 0) follows a bilinear

curve before reaching the backbone curve. The curve passes through the reloading turning point (δt,

Vt) (DE, PQ, TU and XY) and then aims at the strength deterioration reference point (δm, Vssd) (EF,

QR, UV and YZ). The coordinates of the two reference points are calculated as:

=δt (1.15-0.15l /h ) δ re (1 ≤ l /h ≤ 5) (19)

Vssd,i
= =
k t 2.5kre ; kre (20)
δ m − δ re

δ m = δ un,m (21)

18
  δ δ 
=
Vssd,i Vm exp  −0.1 un,m × i − 0.015 i  un,m  (22)
 δ limit  δ limit  

It may be noted that although Equation (22) seems a little complex, it has already been

programed and can be calculated automatically. Actually, the basic form of Equation (22) was first

proposed by Ozcebe and Saatcioglu [31] to calculate the strength deterioration reference point in

hysteretic shear models for columns, and also adopted by some other researchers such as Xu and

Zhang [32] and Ding et al. [22]. In this study, the equation is found to be able to simulate

reasonably the strength deterioration in the diagonally RC coupling beams after setting the two

coefficients to be 0.1 and 0.015.

6 Fiber beam-column element considering shear and slip deformation

As shown in Figure 7, the shear element considering nonlinear shear effect and the slip element

considering the nonlinear slip mechanism are developed by incorporating the traditional fiber

element with section shear force-shear strain and slip deformation relations, respectively. When the

two elements are combined together to model the coupling beam, the shear force in all the elements

are always kept the same.

The failure of the coupling beam is mainly caused by the crushing of concrete at the end, the

buckling of diagonal bars under compression and fracture under tension, which are all closely

related to the large lateral displacement of the beam. Based on test data analysis, the following

formula is proposed to calculate the drift ratio at strength degradation γlimit:

γ limit =
0.0239 + 4.3209 ρsv l / h ( ρsv l / h ≤ 0.0125)
=γ limit 0.07791 ( ρsv l / h > 0.0125)

Asv
ρsv = (without diagonal stirrups) (23)
bs

19
 1.2Asv
ρsv = (with diagonal stirrups)
bs

where ρsv is the transverse reinforcement ratio and the factor 1.2 is an approximation to consider

the contribution of stirrups surrounding diagonal bars.

When the beam drift ratio reaches the drift limit, the load-deformation curve will go into the

descending branch. As shown in Figure 9, this phenomenon is realized by the end slip elements

which enter the descending stage at δlimit corresponding to the beam drift limit γlimit, while the

internal shear element will unload. The slope of the descending branch of the shear force-drift ratio

curve is set to be 0.1 times the initial stiffness, according to the test data analysis. Equation (23)

will be discussed in detail in Section 7.2.

7 Calibration of critical model parameters

7.1 Shear stiffness after diagonal cracking

In the proposed model, the nonlinear flexural deformation has been properly simulated by the

modified fiber model incorporating diagonal bars in Section 3, and the reinforcement slip

deformation can be reasonably considered based on the method proposed in Section 5.1. Therefore,

the remaining problem is the calculation of nonlinear shear deformation, which mainly depends on

the shear stiffness after diagonal cracking. By assuming that the flexure and slip deformation are

properly calculated, the actual cracked shear stiffness ratio β can be acquired by matching the

proposed model with measured shear force-total deformation curves. According to the test

observations, cracked shear stiffness is closely related to the diagonal reinforcement ratio ρsd and

span-to-depth ratio l/h. To further reveal their relationship, a key parameter ρsd(l/h)0.5 reflecting the

integrated influences of ρsd and l/h is proposed, and the correlation between ρsd(l/h)0.5 and α is

20
 plotted in Figure 10. A satisfying positive correlation can be found and the regression formula

Equation (7) is thus proposed.

0.08 10

Chord rotation at strength degradation

γlimit =7.791%
Shear stiffness reduction factor

8
0.06

6
β=kcr/kini

γlimit (%)
0.04
y=4.3209x+0.0239
y=1.2178x+0.0139 4
R2=0.6632
R2=0.7917
0.02
2
ρsv(l/h)0.5=1.25%

0.00 0
0.00 0.01 0.02 0.03 0.04 0.05 0.000 0.005 0.010 0.015 0.020 0.025
Key parameter ρsd(l/h)0.5 Key parameter ρsv(l/h)0.5

Figure 10. Regression analysis of cracked shear stiffness ratio Figure 11. Regression analysis of drift ratio limit γlimit

7.2 Drift ratio at strength degradation

Based on the large number of test results, it can be found that the drift ratio where load begins

to drop (γlimit) is significantly affected by the span-to-depth ratio and transverse reinforcement ratio

including the stirrups surrounding the diagonal bars and horizontal bars. Figure 11 plots the

relationship between the measured value of γlimit and the unified parameter ρsv(l/h)0.5. A clear

positive correlation can be found before 1.25%; thus Equation (23) is proposed by regression of

data points.

8 Model application in diagonally RC coupling beam and coupled wall analysis

The above proposed and calibrated model is now applied to all the test specimens listed in

Table 1, to further verify the accuracy of the model. In addition, a RC coupled wall specimen is

simulated to show the feasibility of the proposed model in the seismic analysis of a coupled wall

structural system.

8.1 Beam tests by Lim et al. [5,6]

21
 Lim et al. [5,6] reported the test results of four diagonally reinforced beams with different

span-to-depth ratios--CB10, CB20, CB30DA and CB30DB--and the detailed information can be

found in Table 1. Specimens CB30DA and CB30DB have different transverse reinforcement

schemes, which follows the ACI318-05[23] and ACI318-14[15] specifications respectively. It is

concluded that Specimen CB10 with l/h=1.0 shows significant shear behavior and fails in the

shear-flexure mode. The other three specimens show obvious concrete crushing and rotation at the

beam ends and all fail in the flexure mode. The behavior of Specimens CB30DA and CB30DB are

basically the same before the 7.7% rotation, while after that Specimen CB30DA failed and

Specimen CB30DB continues to sustain considerable shear forces until a drift ratio of 9.9%. This

might be attributable to the different transverse reinforcement detailing.

1500 1500
CB10 CB20
1000 1000
Lateral load (kN)

Lateral load (kN)

500 500

0 0

-500 -500

-1000 Measured results -750 Measured results

Predicted res ults Predicted res ults
-1500 -1000
-12 -8 -4 0 4 8 12 -12 -8 -4 0 4 8 12
Beam chord rotation (%) Beam chord rotation (%)
(a) CB10 (b) CB20
800 800
600 CB30DA 600 CB30DB

400 400
Lateral load (kN)
Lateral load (kN)

200 200

0 0
-200 -200
-400 -400
Measured results Measured results
-600 -600
Predicted res ults Predicted res ults
-800 -800
-12 -8 -4 0 4 8 12 -12 -8 -4 0 4 8 12
Beam chord rotation (%) Beam chord rotation (%)
(c) CB30DA (d) CB30DB

Figure 12. Comparison of predicted and measured lateral load versus beam chord rotation curves

As shown in Figure 12, the predicted shear force versus chord rotation hysteretic relations

correlate well with the experimental results. The maximum load, peak rotation, loading and

unloading stiffness and the overall shape of loops are all accurately simulated. The drop of load

22
caused by the crushing of concrete and buckling of bars is generally so sudden that it is quite

difficult to capture this behavior accurately.

1500 1500 1500
CB10 CB20 CB30DB
1000 1000 1000

Lateral load (kN)

Lateral load (kN)

Lateral load (kN)

500 500 500

0 0 0

-500 -500 -500

Slip response Slip response Slip response
-750 Shear res ponse -750 Shear res ponse -750 Shear res ponse
Total response Total response Total response
-1000 -1000 -1000
-12 -8 -4 0 4 8 12 -12 -8 -4 0 4 8 12 -12 -8 -4 0 4 8 12
Beam chord rotation (%) Beam chord rotation (%) Beam chord rotation (%)
(a) CB10 (l/h=1.0) (b) CB20 (l/h=2.0) (c) CB30DB (l/h=3.0)

Figure 13. Analysis of deformation components for specimens with different span-to-depth ratios

The comparison of different response components are then analyzed based on the proposed

model, as shown in Figure 13. Generally, the shear and slip deformation take up most of the total

responses; thus the flexure deformation is not plotted for clarity. It is indicated that for specimens

with small l/h such as CB10, the shear deformation contributes a large proportion to the total

response; for specimens with moderate l/h such as CB20, the shear and slip deformation are almost

the same; and for specimens with relatively large l/h such as CB30DB, slip rotation greatly

overshadows the shear deformation and contributes most of the deformation. Figure 14 further

summarizes the ratio of slip rotation to total deformation at flexural strength. It can be seen that the

slip deformation ratio tends to increase with the span-to-depth ratio and can reach as large as 50%,

which is also emphasized by Naish et al. [4,18].

The cumulative energy consumed by the specimens before the sudden failure are also

calculated with the predicted hysteretic curves and compared to the measured results, as illustrated

in Figure 15. It can be demonstrated that the proposed model is able to give a satisfactory

prediction of the cumulative energy of diagonally reinforced coupling beams, which is of concern

23
 for structural seismic analysis.

60 1500
Slip deformation/Total deformation(%)

Numerical results
50 1200

Cumulative energy (kNm)

Test results
40
900
30
600
20
300
10

0 0
0 1 2 3 4 CB10 CB20 CB30DA CB30DB
Span-to-depth ratio l/h Speicemen

Figure 14. Trends of slip rotation to total deformation ratio Figure 15. Comparison of cumulative energy

8.2 Beam tests by other researchers [4,7-11]

All the other twelve specimens in the database in Table 1, tested by Naish et al. [4], Kwan and

Zhao [7], Fortney et al. [8], Galano and Vignoli [9], Tassios et al. [10] and Shiu et al. [11], are also

simulated by the proposed model. Figure 16 plots both the numerical and experimental results,

which extensively demonstrate the accuracy of the proposed model.

It is worth noting that Naish [33] attempted to apply his model to coupling beams with a small

span-to-depth ratio, as shown in Figure 16(g). The specimen CCB11 was tested by Kwan and Zhao

[7] and its span-to-depth ratio was only 1.17. The predicted shear force-beam chord rotation curve

is found to overestimate the stiffness significantly and underestimate the shear capacity, thus

demonstrating that the model is not able to predict the behavior of coupling beams with small

span-to-depth ratios. However, the proposed model in this paper successfully predicts the hysteretic

behavior of this specimen due to the reasonable consideration of shear effects and adequately

calibrated parameters.

24
 1000 1000 1000
750 CB24F 750 CB24D 750 CB33F

500 500 500

Lateral load (kN)

Lateral load (kN)

Lateral load (kN)

250 250 250
0 0 0

J
-250 -250 -250
-500 -500 -500
Measured results Measured results Measured results
-750 -750 -750
Predicted res ults Predicted res ults Predicted res ults
-1000 -1000 -1000
-12 -8 -4 0 4 8 12 -12 -8 -4 0 4 8 12 -12 -8 -4 0 4 8 12
Beam chord rotation (%) Beam chord rotation (%) Beam chord rotation (%)
(a) CB24F (b) CB24D (b) CB33F
1000 400 200

750 CB33D 300 CB2A 150 CB2B

500 200 100

Lateral load (kN)

Lateral load (kN)

Lateral load (kN)

250 100 50

0 0 0

-250 -100 -50

-500 -200 -100

Measured results Measured results Measured results
-750 -300 -150
Predicted res ults Predicted res ults Predicted res ults
-1000 -400 -200
-12 -8 -4 0 4 8 12 -12 -8 -4 0 4 8 12 -12 -8 -4 0 4 8 12
Beam chord rotation (%) Beam chord rotation (%) Beam chord rotation (%)
(d) CB33D (e) CB2A (f) CB2B
400 800 300
300 CCB11 600 DCB1 225 P07

200 400 150

Lateral load (kN)
Lateral load (kN)

Lateral load (kN)

100 200 75
0 0 0
-100 -200 -75
-200 Naish et al. 2010 -400 -150
Measured results Measured results Measured results
-300 -600 -225
Proposed model Predicted res ults Predicted res ults
-400 -800 -300
-12 -8 -4 0 4 8 12 -12 -8 -4 0 4 8 12 -12 -8 -4 0 4 8 12
Beam chord rotation (%) Beam chord rotation (%) Beam chord rotation (%)
(g) CCB11 (h) DCB1 (i) P07
300 80 40
225 P12 60 C6 30 C8

150 40 20
Lateral load (kN)

Lateral load (kN)

Lateral load (kN)

75 20 10
0 0 0
-75 -20 -10
-150 -40 -20
Measured results Measured results Measured results
-225 -60 -30
Predicted res ults Predicted res ults Predicted res ults
-300 -80 -40
-12 -8 -4 0 4 8 12 -12 -8 -4 0 4 8 12 -12 -8 -4 0 4 8 12
Beam chord rotation (%) Beam chord rotation (%) Beam chord rotation (%)
(j) P12 (k) C6 (l) C8

Figure 16. Comparison of predicted and measured lateral load versus beam chord rotation curves

8.3 Coupled wall tests by Cheng et al. [34]

Cheng et al. [34] conducted quasi-static tests on coupled walls with diagonal coupling beams.

The specimen CW-RC, as shown in Figure 17, is chosen to be simulated by the proposed coupling

beam model. The traditional fiber model without shear and slip effects is also applied to the system

to further demonstrate the accuracy of the proposed coupling beam model. The finite element

25
modeling concept for the coupled wall system is briefly introduced first.

As shown in Figure 18(a), a coupled wall system consists of two individual wall piers,

boundary confined columns and coupling beams. The individual walls can be modeled with a thick

shell element provided by the MSC.MARC [23]. Because the shear walls are reinforced with

horizontal and vertical distributed rebars, the layered material model can be applied as the material

constitutive relationship [35,36]. The detailed parameters for the material models have been

suggested by Nie et al. [35]. The confined boundary column consists of longitudinal and transverse

reinforcement, which is mainly subjected to axial force. Therefore, it can be simulated with the

traditional fiber beam-column element. The coupling beam is modeled with the coupling beam

element proposed in this paper.

122.5kN 122.5kN 122.5kN 122.5kN

300

P 4#5 Diagonal rebars 180

300

A 4#5
3#3 Horizontal rebars
1500

19°
#3 300

#3@75mm transverse rebars 2#3

A
A-A
1500

Coupling beam

P/2 8#7 Longitudinal reinforcement

260
#3@150mm web reinforcement both directions
1500

200

#3@60mm transverse reinforcement

confined boundary element
Wall pier
1500

#3 rebar fy: 454MPa; fu: 684MPa fc': 30MPa (base block)

#5 rebar fy: 475MPa; fu: 691MPa 35MPa (lower 2 floors)
825

#7 rebar fy: 455MPa; fu: 656MPa 37MPa (upper 2 floors)

(unit:mm)
1080 1300 450 1300 1080

Figure 17. Details of the coupled wall specimen

26
 Boundary confined column Coupling beam y Local coordinate system
Tied node x
Retained node

Coupling beam dx,i=0

element dy,i=0

Shear element
Slip element

Shear wall
（Multilayer shell
element）

Fiber beam-
column element

Offset of fiber beam-

column element

Node sharing between

beam-column and
multilayer shell element

(a) scheme of coupled wall system (b) modeling concept of coupled wall system

Figure 18. Scheme and modeling concept of coupled wall system

The above three types of elements are then combined to model a coupled wall system as

illustrated in Figure 18(b). The fiber beam-column elements modeling the confined boundary

column are linked to the multilayer shell elements modeling the shear wall through the share-node

approach to make sure they can work together. An offset equal to half of the column width is given

to the beam-column element to reflect its exact position.

To assure the internal forces in the coupling beam can be reliably transferred to the wall,

constraint equations of degrees-of-freedoms need to be defined between the end nodes of coupling

beams and the wall nodes in the range of beam height, as illustrated in Figure 18(b). A local

coordinate system is first attached to the end node of the coupling beam element, which is defined

as the retained node. The local x-axis is fixed in the longitudinal direction of the coupling beam and

co-rotates according to the rotation of end nodes. The y-axis is kept perpendicular to the x-axis. The

constraint conditions as Equations (24) and (25) are applied to the corresponding wall nodes which

are defined as the tied nodes here. Equation (24) can be used to define the deformation

27
compatibility in the local x-direction for transferring the axial force and bending moment and

Equation (25) can realize the deformation compatibility in the local y-direction for transferring the

shear force.

d x,i = 0 (24)

d y,i = 0 (25)

where dx,i and dy,i are the displacements of the ith tied node in the x-direction and y-direction of the

local coordinate system, respectively.

Based on the proposed coupled wall model, the seismic behavior of Specimen CW-RC is

simulated with different coupling beam models and the predicted base-moment versus roof drift

ratio relationships are compared to the experimental results. The specimen is loaded horizontally

with two hydraulic actuators. The force applied by the roof-floor actuator is kept twice that of the

third-floor actuator. In addition, the 245kN additional vertical load is applied to the top of each wall

pier using four hydraulic jacks before applying lateral displacement.

800 10000 10000
Traditional fiber model Traditional fiber model
600 8000 8000
Proposed model Proposed model
6000 6000
400
Base moment (kN·m)

Base moment (kN·m)

4000 4000
Lateral load (kN)

200 2000 2000

0 0 0

-200 -2000 -2000

-4000 -4000
-400
-6000 -6000
Tes t results
-600 -8000 -8000
Proposed model
-800 -10000 -10000
-10 -7.5 -5 -2.5 0 2.5 5 7.5 10 -0.04 -0.02 0.00 0.02 0.04 -0.04 -0.02 0.00 0.02 0.04
Beam chord rotation (%) Roof drift ratio Roof drift ratio
(a) Coupling beam (b) Coupled wall system

Figure 19. Comparison of predicted and measured results for beams and coupled walls of specimen CW-RC

It can be seen from Figure 19(b) that the predicted results with the proposed coupling beam

model are well correlated with the test results in terms of the initial stiffness, shear capacity and the

overall unloading and reloading loops. However, compared to the proposed coupling beam model,

28
the predicted results with the traditional fiber model yield obviously larger initial unloading and

reloading stiffness of the coupled wall system, which is attributed to the different predicted

behavior of coupling beams as plotted in Figure 19(a). It can be found that the traditional fiber

model greatly overestimates the initial unloading and reloading stiffness of the coupling beam

because the significant shear and slip effects are neglected.

9 Conclusions

Based on the summary and analysis of the experimental results, this paper proposes an accurate,

efficient and practical fiber model for diagonally RC coupling beams on the platform of general

FEA software MSC.Marc. It can be concluded from the research that :

1. The proposed model can accommodate diagonally RC coupling beams with a wide range of

span-to-depth ratios from one to five since all the deformation components including flexure,

shear and slip are considered and the critical parameters are calibrated by a total of sixteen test

specimens collected from seven different research groups.

2. The modified fiber model incorporating the horizontal component of the diagonal bars is able to

consider the flexural contribution of diagonal bars reasonably and give precise prediction of

beam strength.

3. The proposed section shear force-shear strain and slip displacement model can reasonably

represent the complex hysteretic behavior including unloading and reloading stiffness, strength

and stiffness deterioration.

4. The formulas for cracked shear stiffness and chord rotation at strength degradation are proposed

and verified.

29
5. The hysteretic behavior of sixteen test specimens are simulated by the proposed model and

good correlation between numerical and experiment results is demonstrated in terms of the

overall hysteretic loops and energy consumed. The different deformation components are

analyzed and the slip rotation ratio is found to increase with the span-to-depth ratio and can be

as large as 50%, whereas the shear deformation is dominant when the beam span-to-depth ratio

is less than 2.0.

6. The accuracy, efficiency and applicability of the proposed model is further shown by means of

the simulation of a RC coupled shear wall specimen.

Acknowledgments

The writers gratefully acknowledge the financial support provided by the National Science Fund of

China (Grant No. 51708328 and No. 51722808) and the China Postdoctoral Science Foundation

(Grant No. 2016M601039, 2017T100083).

References

[1] Paulay T. Coupling beams of reinforced concrete shear walls. ASCE J Struct Div 1971; 97(3): 843-862.

[2] Breña SF, Ihtiyar O. Performance of conventionally reinforced coupling beams subjected to cyclic loading. J

Struct Eng 2011; 137(6): 665-676.

[3] Paulay T, Binney JR. Diagonally Reinforced Coupling Beams of Shear Walls. ACI Spec. Publ. 1974; 42: 579-598.

[4] Naish D, Fry A, Klemencic R, Wallace J. Reinforced concrete coupling beams-Part I: testing. ACI Struct J 2013;

110(6): 1057-1066

[5] Lim E, Hwang SJ, Wang TW, Chang YH. An investigation on the seismic behavior of deep reinforced concrete

coupling beams. ACI Struct J 2016; 113(2): 217-226.

[6] Lim E, Hwang SJ, Cheng CH, Lin PY. Cyclic tests of reinforced concrete coupling beam with intermediate

span-depth ratio. ACI Struct J 2016; 113(3): 515-524.

[7] Kwan AKH, Zhao ZZ. Cyclic behaviour of deep reinforced concrete coupling beams. Proc. ICE, Struct & Build,

2002; 152(3): 283-293.

30
[8] Fortney PJ, Rassati GA, Shahrooz BM. Investigation on Effect of Transverse Reinforcement on Performance of

Diagonally Reinforced Coupling Beams. ACI Struct J 2008; 105(6): 781-788.

[9] Galano L, Vignoli A. Seismic behavior of short coupling beams with different reinforcement layouts. ACI Struct J

2000; 97(6): 876-885.

[10] Tassios TP, Moretti M, Bezas A. On the behavior and ductility of reinforced concrete coupling beams of shear

walls. ACI Struct J 1996; 93(6): 711-720.

[11] Shiu KN, Barney GB, Fiorato AE, Corley WG. Reversing load tests of reinforced concrete coupling beams. Proc.

Central American Conf. on Earthquake Engineering, Central America, 1978.

[12] Park WD, Yun HD. Seismic behaviour of coupling beams in a hybrid coupled shear walls. J Constr Steel Res 2005;

61(11): 1492-1524.

[13] Gong BN, Shahrooz BM. Concrete-steel composite coupling beams. I: Component testing. J Struct Eng 2001;

127(6): 625-631.

[14] Nie JG, Hu HS, Eatherton MR. Concrete filled steel plate composite coupling beams: Experimental study. J Constr

Steel Res 2014; 94(1): 49-63.

[15] ACI 318-14. Building code requirements for structural concrete (ACI 318-14) and commentary/reported by ACI

Committee 318. Farmington Hills, Mich.: American Concrete Institute; 2014.

[16] Tao MX, Nie JG. Fiber beam-column model considering slab spatial composite effect for nonlinear analysis of

composite frame systems. J Struct Eng. 2014; 140(1): 04013039.

[17] Tao MX, Nie JG. Element mesh, section discretization and material hysteretic laws for fiber beam-column

elements of composite structural members. Mater and Struct 2015; 48(8): 2521-2544.

[18] Naish D, Fry A, Klemencic R,Wallace J. Reinforced concrete coupling beams-part II: modeling. ACI Struct J 2013;

110(6):1067–1076

[19] Lu X, Chen Y. Modeling of coupled shear walls and its experimental verification. J Struct Eng 2005;

131(1):75–84

[20] Barbachyn S, Kurama Y, Novak LC. Analytical evaluation of diagonally reinforced concrete coupling beams

under lateral loads. ACI Struct J 2012; 109(4):497–508.

[21] Toprak AE, Bal IE, Gülay FG. Review on the macro-modeling alternatives and a proposal for modeling coupling

beams in tall buildings. Bull Earthquake Eng 2015; 13: 2309-2326

[22] Ding R, Tao MX, Nie JG, Mo YL. Shear deformation and sliding-based fiber beam-column model for seismic

analysis of reinforced concrete coupling beams. J Struct Eng, 2016; 142(7): 04016032.

31
[23] MSC. Marc Version 2007r1 [Computer software]. MSC. Software Corp., Santa Ana, CA.

[24] ACI 318-05. Building code requirements for structural concrete (ACI 318-05) and commentary/reported by ACI

Committee 318. Farmington Hills, Mich.: American Concrete Institute; 2005.

[25] Vu NS, Li B, Beyer K. Effective stiffness of reinforced concrete coupling beams. Eng Struct 2014; 76: 371-382.

[26] ASCE-ACI Joint Task Committee 426. Shear strength of reinforced concrete members. ASCE J Struct Div 1973;

99(6): 1091-1187．

[27] Alsiwat JM, Saatcioglu M. Reinforcement anchorage slip under monotonic loading. J Struct Eng 1992; 118(9):

2421-2438.

[28] ACI Committee 408. Suggested development, splice, and standard hook provisions for deformed bars in tension

(ACI 408.1R-79). American Concrete Institute, Farmington Hills, MI, 1979.

[29] Sezen H, Setzler EJ. Reinforcement Slip in Reinforced Concrete Columns. ACI Struct J 2008; 105(3):280-289.

[30] Pan WH, Tao MX, Nie JG. Fiber beam-column element model considering reinforcement anchorage slip in the

footing. Bull Earthquake Eng 2017; 15(3): 991-1018.

[31] Ozcebe G, Saatcioglu M. Hysteretic Shear Model for Reinforced Concrete Members. J Struct Eng 1989; 115(1):

132-148.

[32] Xu SY, Zhang J. Hysteretic shear-flexure interaction model of reinforced concrete columns for seismic response

assessment of bridges. Earthq Eng Struct D, 2011; 40(3):315-337.

[33] Naish DAB. Testing and modeling of reinforced concrete coupling beams. Ph.D. Ann Arbor. University of

California, Los Angeles; 2010.

[34] Cheng MY, Fikri R, Chen CC. Experimental study of reinforced concrete and hybrid coupled shear wall systems.

Eng Struct 2015; 82: 214-225.

[35] Nie JG, Tao MX, Cai CS, Chen G. Modeling and investigation of elasto-plastic behavior of steel-concrete

composite frame systems. J Constr Steel Res 2011; 67(12): 1973-1984.

[36] Ding R, Tao MX, Zhou M, Nie JG. Seismic behavior of RC structures with absence of floor slab constraints and

large mass turbine as a non-conventional TMD: a case study. Bull Earthquake Eng 2015; 13(11): 3401-3422.

32
LIST OF FIGURES

Fig. 1: Comparison of RC coupling beams with different reinforcement layouts

Fig. 2: Basic mechanism of diagonally reinforced coupling beams

Fig. 3: Modified fiber model incorporating diagonal bars (nd=4)

Fig. 4: Generalized sectional stress, strain and stiffness

Fig. 5: Shear force-shear strain skeleton curves and hysteretic rules

Fig. 6: Diagonal reinforcement slip model in the proposed fiber model

Fig. 7: Proposed coupling beam fiber model considering shear and slip deformation

Fig. 8: Proposed simplified trilinear shear force-slip deformation model

Fig. 9: Proposed hysteretic shear force-slip deformation model

Fig. 10: Regression analysis of cracked shear stiffness ratio

Fig. 11: Regression analysis of drift ratio limit γlimit

Fig. 12: Comparison of predicted and measured lateral load versus beam chord rotation curves

Fig. 13: Analysis of deformation components for specimens with different span-to-depth ratios

Fig. 14: Trends of slip rotation to total deformation ratio

Fig. 15: Comparison of cumulative energy

Fig. 16: Comparison of predicted and measured lateral load versus beam chord rotation curves

Fig. 17: Details of the coupled wall specimen

Fig. 18: Scheme and modeling concept of coupled wall system

Fig. 19: Comparison of predicted and measured results for beams and coupled walls of specimen CW-RC

33
 LIST OF TABLES

Table 1: Database of diagonally RC coupling beam specimens

34

View publication stats