Case Studies in Construction Materials 8 (2018) 476–488

Contents lists available at ScienceDirect

Case Studies in Construction Materials

journal homepage: www.elsevier.com/locate/cscm

Case study

Investigating modeling approaches of buckling-restrained braces

T
under cyclic loads
⁎
Rohola Rahnavarda, , Mohammad Naghavib, Maryam Aboudia, Mohamed Suleimanc
a
Structural Engineering, Faculty of Civil Engineering, Jundi-Shapur University of Technology, Dezful, Iran
b
Structural Engineering, Dep. of Civil Engineering, Javid Institute for Higher Education of Jiroft, Iran
c
Lecturer of Structural Engineering, University of Tripoli, Tripoli, Libya

A R T IC LE I N F O ABS TRA CT

Keywords: The Main purpose of this paper is to present a practical method for accurate modeling and
BRB provide a simple model of buckling restrained braces (BRBs). The components of a BRB were
Core-spring introduced. Furthermore, two complete experimental specimens of BRBs were modeled using the
ABAQUS ABAQUS finite element method program. The validity of these models was confirmed after
Hysteresis curves
comparing the results of nonlinear dynamic analysis with the experimental specimen results. An
intended simple model such as steel core and springs was proposed. It was evaluated and ana-
lyzed after modeling, then a method was presented to find the spring stiffness. In this study, a
comparison was made between the experimental and analytical hysteresis curves (the complete
model and the simplified model) and a good agreement was observed between the two results
with 5% difference. Moreover, the two-story moment frame with a BRB in both the complete
details and simple states was modeled to verify the simple method of core-spring. The results
showed a high accuracy of the simple model of core-spring. In addition, the use of the simplified
spring-core model may significantly reduce the time of the analysis compared to the complete
model.

1. Introduction

It is observed that buckling occurs at braces when pressure is exerted on them showing an undesirable behavior and is visible in
most braces. This problem may cause a decrease and energy dissipation capacity of a structure due to the secondary effect of
nonlinear geometric deformation. This plays an important role in cyclic loading like earthquakes due to the nature of the further
reduction in hardness under dynamic seismic loads. The use of braces that have the same behavior in pressure and tension without
buckling, has always been desirable to design structures [1]. The behavior of conventional buckling-restrained braces (BRBs) is
shown in Fig. 1. In fact, the retrofitting corrects inappropriate behavior, such as reducing resistance, stiffness and ductility.
The idea of using submissive steel members for the energy absorption was proposed more than 30 years ago [2,3]. The new idea
was that pressure member was expected to yield before buckling. It should be noted that yield has to be distributed properly that is
not uniformly in local form, so that the dissipated energy reaches its greatest value during a cyclic loading such as earthquakes.
Some research was carried out based on avoiding the buckling of the compressive braces using concrete cover around it. The
bearing compressive load is performed using the steel core of this type of brace and the concrete cover is only assumed to prevent
buckling of the steel core.

⁎
Corresponding author at: Faculty of Civil Engineering, Jundi-Shapur University of Technology, Dezful, P.O.Box 64615-334, Iran.
E-mail addresses: rahnavard1990@gmail.com (R. Rahnavard), mn147013@gmail.com (M. Naghavi), maryamabudi74@gmail.com (M. Aboudi),
suleimmf@mail.cu.edu (M. Suleiman).

https://doi.org/10.1016/j.cscm.2018.04.002
Received 11 December 2017; Received in revised form 29 March 2018; Accepted 5 April 2018
2214-5095/ © 2018 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/BY/4.0/).
R. Rahnavard et al. Case Studies in Construction Materials 8 (2018) 476–488

Fig. 1. Behavior of conventional buckling-restrained braces adopted from [35].

Buckling of the steel core causes lateral distribution of the internal pressure in the concrete cover. To prevent the transmission of
axial compressive force from steel core to concrete cover, a thin layer of a special material is placed at common surface of steel and
concrete. This layer prevents the friction between the core and concrete which causes the compressive force arising from lateral
deformation of steel to transmit the core as extensive transverse load to the concrete. This way of transferring the force among
components has caused this kind of brace to be called Buckling Restrained Brace (BRB).
The BRBs were widely used in the 1980s in Japan [16]. Extensive research on unbounded brace was started by researchers in the
1990s in the United States leading to the use of this type of bracing in different buildings for the seismic rehabilitation. The research is
being followed in other countries including India and Taiwan. The stability analysis of a BRB was investigated by Black et al. [4] and
their study demonstrated that both shear stress and shear strain at the beginning of nonlinear buckling would depend on tangential
shear modulus (Gt) [5], this is due to the existence of single-axis pressure.
In 2002, for the construction of a research center at the University of California, Berkeley, the experimental studies on BRBs were
performed by Lopez et al. [6] and their experimental results showed that connection plates practically cause creation of the rigid
openings and rotation of nodes and should be taken into account in the analysis and design due to high rigidity in connecting plate.
A collaborative study was conducted between National Center for Research on Earthquake Engineering (NCREE) in Taiwan and a
group of researchers at the University of Michigan, USA (2004) [5], the main objective of their study was to provide a convenient
method for obtaining the base shear. Choi and Kim developed a design procedure for the frames with buckling-restrained brace using
energy spectrum hysteresis [7]. In this procedure, it is assumed that beams and columns remain under gravity load in the elastic state
and energy dissipation and its related damage occur at the BRB. Thus, this is necessary and inevitable that any new technology used
in practice requires laboratory and theoretical studies. It has been suggested in the American Institute of Steel Construction (AISC
2016) that two sets of experimental tests should be conducted on braces [8]. The first series of experimental tests was only carried out
on braces. Their axial behavior is carefully observed and the parameters required for the design are obtained. Another result of this
series of experimental test can be the ideal shape of brace. Fahnestock et al. (2006) conducted a research program on BRBs [9]. The
effect of the near-field earthquakes on structures with BRBswas studied by Shmshadyan et al. [10]. The results of the study showed
that such systems had better performance on the energy dissipation [10]. A study entitled “Effect of Design Loads in Buckling
Restrained Braced Frames Performance” was carried out by Shokrgozar and Asgharian. In their study, resistance, ductility and effect
of design loads in buckling-restrained braced frames (BRBFs) were evaluated. For this purpose, some buildings with various stories
and different configurations of braces were studied.
Static pushover, nonlinear incremental dynamic and linear dynamic analyses have been performed using the OpenSees Software.
The effects of some parameters on response of the correction factor including height of a building and the bracing system were
investigated and finally the vibration response of a correction factor for each braced system was separately determined in their study
[11]. A research study was carried out by Almansa and Oller on the cyclical behavior of buckling-restrained brace. In their study,
ductility and the connection resistance based on regulations FEMA-450 and finite element model of the studied sample were modeled,
then plasticity curves and the brace strength were obtained.
In addition, cyclic behavior of a brace member using finite element methods were studied from these curves [12]. A research was
conducted by Chuo and Chi-Yu to analyze and comparison of cyclic performance analysis of BRB screw connections to the beams and
columns. This was performed using the Abaqus software based on changing the horizontal arrangement of connecting screws. The
results obtained from the research were in accordance with the regulations AISC-LRFD for the same beam and column sections
indicating the slenderness ratio for this type of brace was 2. Moreover, the results showed that distance between the screws and the
connection center of the beam to column had a direct relationship to its buckling [13].
Some studies were performed by Ou and Zhao et al. to analyze and compare the cyclic performance of the BRB to the beams and
columns connection as well as investigate the impact of this kind of brace on the lateral buckling. A research was also conducted on

477
R. Rahnavard et al. Case Studies in Construction Materials 8 (2018) 476–488

Fig. 2. Different types of confining sheath adopted from [2].

the shape of metal core and the stiffeners [14]. The periodic performance of a steel moment frame with the BRB and moment frames
without brace was compared by Jia et al. based on several experimental models. Their results showed that the moment frames with
the BRB would bear shearing force up to two times greater than the moment frames without bracing system [15]. Guo et al. in-
vestigated core-separated buckling-restrained brace (CSBRB) using theoretical and experimental specimen. Their findings provide
fundamentals to develop the design method for an energy-dissipation type of SS-BRBs [30–31]. Wei et al. studied local displacement
demands of BRBs in bidirectional end diaphragms [32]. Qu et al. studied on double K-braced reinforced concrete frame subassemblies
with the BRB. Their results showed that the reinforcement concrete frames in the braced specimens were only moderately damaged at
1% inter-story drift when the BRB cores sustained large plastic strain [33]. AlHamaydeh et al. studied on key parameters influencing
performance and failure modes for BRBs. It is found that among the different failure modes, necking of the steel core is potentially the
most serious [34–38]. One of the problems of numerical modeling is that the time of structural analysis is very long. Operations and
mathematical calculations are so numerous in many models that they require a supercomputer to analyze. One of the most important
parameters in numerical modeling is that the numerical model has less complexity. If the complexity is less in the numerical model,
the analysis is simpler and time will be saved more. For modeling of BRBs using ABAQUS software, since concrete and steel are in
contact, the software usually encounters extensive calculations or non-convergence errors. In buckling restrained braces, the concrete
does not play a role in the lateral loading and only prevents the buckling of the steel core. Consequently, concrete can be removed and
replaced by the equivalent spring stiffness. In the present study, BRBs is investigated based on two types of models: complete model
and simplified model, and the results are compared.

2. Basic principles

Based on Figs. 2 and 3, as mentioned in the introduction, unbounded brace is composed of several parts as follows:

2.1. Enclosed steel core

This is the major member of brace which bears the axial load. Since the basic idea of BRBs yields at a pressure, it is more
appropriate for the type of steel core to be mild with a lower resistance. So the yield in steel must happen in the cyclic loading and
creates more energy dissipation. Different forms of the steel core sections are show in Fig. 2 [16].

2.2. Unconfined metal core

A part of the brace must be inevitably unconfined in order to connect the brace to the building structure. As the buckling must not
happen in the unconfined part, it should have greater cross-sectional area than enclosed zone. Change in cross section should not be
suddenly as this can cause stress concentrations in the area that cross section changed. For this purpose, a transition zone is used.
Cross section is enlarged along the transition zone in a mild way.

2.3. Non-stick material

Non-stick mechanism of buckling-restrained brace (BRB) is provided by this material. Non-stick material eliminates or minimizes
the friction between the steel core and confining concrete. Due to this fact, non-stick material only transfers lateral deformation of
steel core to confining concrete and prevents the transmission of axial force to confining concrete.
The different non-stick materials have been used by some researchers in their experiments. Rubber-like material, polyethylene,
mastic, and Silicone oil are among substances that have been used as non-stick material [17–19]. There should be a proper distance
between steel core and confining concrete. The distance should not be too large that let steel core buckle at the higher modes. In other

478
R. Rahnavard et al. Case Studies in Construction Materials 8 (2018) 476–488

Fig. 3. Buckling Restrained Brace component adopted from [2].

words, the distance between the core and concrete must be smaller than amplitude of bulge caused by buckling. On the other hand,
the distance should not be very small that steel core sticks to confining concrete due to expansion caused by Poisson effect resulting in
the friction between the core and concrete. Because it causes the transfer of axial force from core to confining concrete and con-
sequently the predicted mechanism is eliminated. If the transferring section exists in buckling-restrained brace, an empty interior
space must be considered in order to expand and contract the core according to Fig. 3 [18]. If the space is small or it does not exist in
case of core elongation, it causes the damage in steel core and concrete. Furthermore, it can reduce the bearing capacity and
deformation of the BRB in tension condition. This issue is undesirable in the design. This case can also cause the unbalanced force in
chevron braces, while one of the advantages of using BRB is the removal of these unbalanced forces.

2.4. The joint region

Generally, bracing connection to beams and columns is carried out by screws, rivets and welding. Use of screw is more con-
ventional due to easier possibility of changing.

2.5. Confining sheath

As seen in Fig. 2, different types of confining sheath have been implemented [17]. The most common type of sheath is a square
steel which filled with concrete as shown in Fig. 3 [17]. According to a report by national institute of Taiwan earthquake, other
materials apart from concrete such as wood and sand have been used as filling materials [19].

3. Typical experimental setup

3.1. Geometrical characteristics

Experimental modeling under different loading in the BBRs were carried out by Mir Taheri et al. with different characteristics
[20]. Two specimens of 0.6 and 1 m were used to evaluate the accuracy of the finite element model. Specimen brace has a core with a
rectangular cross-section as shown in Fig. 3. Experimental specimen is also shown in Fig. 4 [20]. ABAQUS software was implemented
for the modeling of the mentioned specimen. Three parts used in the BRB were modeled in three dimensions with actual sizes.
Geometric details in models of 0.6 and 1 ms are summarized in Table 1.

3.2. Material properties

The modulus of elasticity used in all models is 210 GPa. The yield and ultimate stress for the core are 297.5 MPa and 449.8 MPa,

479
R. Rahnavard et al. Case Studies in Construction Materials 8 (2018) 476–488

Fig. 4. Model details a) Experimental [20] and b) numerical.

Table 1
Geometric details in models of 1 and 0.6 m [20].
Model Core Section area (mm2 ) Core length (mm) Tube section (mm)

BRB100 640 1000 1200*1200*30

BRB60 640 600 1200*1200*30

while the mentioned values for confining steel are equal to 370 and 403.4 MPa, respectively. Bilinear kinematic hardening was used
to model the elasto-plastic behavior of a steel (Fig. 5a). Details of the mechanical properties of steels material are shown in Table 2.
The concrete damage plasticity model was used for the modeling of concrete. In order to define the concrete plastic damage model,
the compressive and tensile properties of concrete are separately described in the two graphs. Fig. 5b shows the cyclic behavior of
concrete properties [21–28]. Table 3 represents the details of the mechanical properties of concrete

3.3. Interaction

There are two types of concrete and steel contacts for the modeling this type of brace including contact between the brace core
and the concrete as well as between the concrete and enclosing steel (Steel cover plate). Surface to surface contact was implemented
to represent the interaction between the concrete and the steel. Since a thin layer of non-stick material is applied between the steel
core and the concrete in the BRBs, coefficient of friction is reduced to 0.2. Moreover, a thin layer of non-stick material in the gap
between the steel core and concrete allows for de-bonding effects.
The friction coefficient between the concrete and enclosing steel is 0.8 [21–29]. Also hard contact is used to define the vertical

480
R. Rahnavard et al. Case Studies in Construction Materials 8 (2018) 476–488

Fig. 5. Cyclic behavior a) steel b) concrete.

Table 2
Mechanical property of steel.
Section Young modulus (MPa) Yield Stress (MPa) Yield Strain Ultimate Stress (MPa) Ultimate Strain

Core 210000 297.5 0.0022 449.8 0.21

Tube 210000 370 0.0025 403.4 0.33

Table 3
Mechanical property of concrete.
Material Young modulus (MPa) Dilation Angle Eccentrically Viscosity parameter Tensile Stress (MPa) Compressive Stress (MPa)

Concrete 21000 31 0.1 0.001 2.9 30

481
R. Rahnavard et al. Case Studies in Construction Materials 8 (2018) 476–488

Fig. 6. a) Finite element meshing b) Convergence test curve.

properties of the interaction between steel and concrete. More information about interaction between steel and concrete presented
herein [21–29].

3.4. Boundary conditions and mesh

Example of the numerical mesh with loading and boundary conditions is shown in Fig. 6. The cyclic loading located on the end of
the blade and the inner surface of the sheath (beginning of the brace) is considered as support. C3D8R element is used to mesh steel
and concrete parts. The element is defined by eight nodes having three degrees of freedom at each node. Reduced integration is used
to perform its calculations. Also, to achieve the optimal size mesh, six mesh sizes including 100 mm (348 elements), 80 mm (466
elements), 60 mm (575 elements), 40 mm (923 elements), 20 mm (3869 elements) and 10 mm (8276 elements) were considered for
the models. Fig. 6b shows the mesh sensitivity analysis test (Convergence study) for a model with a length of 1 m. In this test, the
relationship between mesh size and the percent error of the experimental-numerical evaluation of the specimens is shown. According
to Fig. 6b, the lowest error is related to the model with a mesh size of 10 mm. Therefore, all specimens were molded with a mesh of
10 mm.

3.5. Numerical results

Axial force-displacement curves derived from experimental specimen and numerical modeling for the models of 0.6 and 1 m are
shown in Fig. 7a and b, respectively. As can be seen in the Figures, loops of behavioral hysteresis curve are very close to each other in
both finite element analysis results and the experimental results, therefore they are well-matched.
The greater curvature of loops obtained from experimental work in the region of start of yield is due to the low loading rate in the
experimental model compared to the finite element analysis model. Stiffness of the BRB in both experimental and finite element
analysis states in the different cycles are very close to each other and are almost equal. It is shown that resistance of braces is not
much different in the tensile state. The maximum tensile strength in the experimental specimen is about 4% more than the numerical
model, but the maximum compressive strength of experimental specimen is about 10% higher than the maximum compressive
strength of the finite element analysis models.
This difference is due to the lack of consideration of friction between the steel core and surrounding concrete. In the actual case,
there is friction effect of non-stick material, although it may be little effect. The friction may cause interaction between steel core and
confining concrete in the compressive state. Finally it leads to an increase in the compressive strength as shown in the charts.

482
R. Rahnavard et al. Case Studies in Construction Materials 8 (2018) 476–488

Fig. 7. Comparison of experimental and numerical results a) BRB100 b) BRB60.

3.6. Provision of core-spring model

Buckling-restrained brace includes steel core, concrete filler and confining steel which can be so difficult for both the modeling of
all parts and the interaction among them. Moreover, it faces the non-convergence error.
In addition, the analysis of BRB frames using the described model in the previous section cannot be justified in terms of economic
issues and spending much time. This is due to the large number of elements and existence of three-dimensional elements.
Therefore, a simpler style was suggested for the modeling of the BRBs as described below. The presented model, which is im-
plemented in this article, is provided for rectangular core (Fig. 2).
It is shown that performing the following steps is accessible for other different forms. In this simple model, steel core is made using
the four-node shell element (S4R)
The following nonlinear spring element is used instead of the modeling the non-stick material, concrete and confining steel. This
element has a large deformation capacity, as a result, it allows to investigate the possible buckling of the steel core. It should be noted
that spring is modeled by a linear property in this case. In this type of modeling, a spring is placed outside of the plate at each node of
confining steel core. A spring in the transverse direction of a rectangular plate is also located on the present nodes on the upper and
lower sides of the rectangular core. The model which is made by division of core element to four sections in the longitudinal direction
and one section in the transverse is shown in Figs. 8 and 10c.
It should be noted that a number of meshes and the used springs have a direct impact on finding spring stiffness. Model made by

Fig. 8. Finite element core-spring model.

483
R. Rahnavard et al. Case Studies in Construction Materials 8 (2018) 476–488

Fig. 9. Comparison of Experimental and numerical results for the model with core-spring a)BRB100 b)BRB60.

spring is placed under cyclic loading to find stiffness of the spring. It can be achieved to the minimum rate of hardness by changing
the stiffness of the spring in order to obtain hysteresis curves similar to the real model without the occurrence of buckling. If the
stiffness is exceeded this value, the same hysteresis curves are again observed. Axial force-displacement curve derived from the
simplified model along with actual prototype is shown Fig. 9. Very good and acceptable compliance is indicated in both the initial
and secondary stiffness regions and also resistance level of the two models. If the steel core division is too small that the buckling load
becomes greater than core yield stress, placement of spring can be ignored in some nodes. Buckling force of a core with a rectangular
profile in its weak direction is written as follows:

π 2Ei bt 3
Pcr =
12 L2 (1)

Where, Ei , b, t and L are elasticity modulus of steel core, height of section, cross-section and length of an element, respectively. The
yield strength of steel core is also obtained from the yield stress σy by the following equation:
Fy = btσy (2)

When Eq. (1) is palcedgreater than the Eq. (2), the minimum length of the mesh in which spring is required for all the nodes, can
be calculated.
πt
L≤
12 σy
Ei (3)

It is shown that if meshing dimensions are less than the obtained L from Eq. (3), the springs shall be placed in the nodes so that
their distance from each other should not be more than the value of L.
Therefore, it is recommended that dimensions of elements in the modeling are selected equivalent to the value obtained from Eq.
(3), if possible. A transverse loading is exerted on them in the weak direction of brace to find a unique number for stiffness of springs.
In the different types of test available, the buckling in an enclosed area has not been reported and failure occurred in the connection
areas with very high rate of rotation.
Since the buckling of axial member of brace occurs as an out of plane displacement, a transverse loading is exerted on the steel
core in the direction that the buckling is likely to happen. Due to the loading, transverse displacement should be the same in both
cases of the actual model with one spring so that the required energy for buckling in both cases would be the same. Considering these
matters, for models built by 4 parts of meshing and 20 springs, the axial stiffness of springs is obtained 10,000 Nm/mm. The above
mentioned method can be used to find equivalent stiffness of models with different types of mesh.
For example, the axial stiffness of the springs for a model having 36 springs was obtained 5000 N/mm by dividing the length
height of core into 8 parts and 1 section respectively. The axial stiffness of the springs was obtained 2500 N/mm for steel core with its
division into 16 parts of length and 1 part of the height. Due to these explanations, it can be concluded that the spring constant

484
R. Rahnavard et al. Case Studies in Construction Materials 8 (2018) 476–488

Fig. 10. Model details a) Experimental [15] b) The full FE model and c)Simplified FE model with core-spring.

485
R. Rahnavard et al. Case Studies in Construction Materials 8 (2018) 476–488

Table 4
Experimental model details from [15].
Story Height Span Column Section Tube Section Beam section of middle beam (mm) Beam section of top and bottom beams
(mm) D × t (mm) D × t (mm) (mm)

1500 2000 219 × 4 219 × 4 H194 × 150 × 6.5 × 9 H300 × 150 × 6.5 × 9

multiplied by the number of springs according to series or parallel connection of springs has a constant value for a specified section of
the BRB.

4. Numerical model of moment frame with buckling-restrained braces (additional research)

In this section, the modeling of moment frames using BRB was investigated for the development, evaluation and comparison of
the actual modeling and core spring method. Experimental model of Mingming et al. [15] was chosen to reach this goal. Geometric
dimensions of the experimental specimen are depicted in Fig. 10a and Table 4. Modeling techniques were used as described above. To
define the mechanical properties of steel, modulus of elasticity is 210 GPa and the yield and ultimate stress for the core are 263 and
379 MPa, respectively. Moreover, these values are 261 and 413 MPa for beam and 298 and 366 MPa are for the confining steel and
column, respectively. The actual numerical model and model with core-spring are shown in Fig. 10b and c.
Shear force-displacement curves obtained from the experimental specimen and numerical modeling for models with and without
springs (real modeling mode) are demonstrated in Fig. 11a and b. As can be seen in the Fig. 11a, the maximum shear forces for both
the experimental and numerical models (Full modeling) are closed. However, there are noticeable differences in the post-yield
stiffness, hardening, and width of hysteretic cycles. The reason for this kind of error is that in the initial cycles, part of the concrete
enters the plastic area and it cannot prevent buckling in the final cycles. As can be seen in the Fig. 11b, loops of behavioral hysteresis
curve in both finite element analysis (core-spring) and the experimental results are very similar so that the results have a good
agreement with each other. However, it is observed in the real modeling that hysteresis graph is converged slightly to stiffness.

5. Time consuming

One of the problems of numerical modeling is that the time of structural analysis is very long. Operations and mathematical
calculations are so numerous in many models that they require a supercomputer to analyze. One of the most important parameters in

Fig. 11. Shear force-displacement a) the full FE model and b) The simplified FE model with core-spring.

486
R. Rahnavard et al. Case Studies in Construction Materials 8 (2018) 476–488

Table 5
Time of the analysis for all models.
Modeling Techniques Time (h)

BRB100 BRB60 BRB Frame System

Full Modeling 43 39 65
Core-Spring Modeling 6 4 11

numerical modeling is that the numerical model has less complexity. If complexity is less in the numerical model, the analysis will be
simpler and time will be saved more. For the modeling of BRBs using ABAQUS software, since concrete and steel are in contact, the
software usually encounters extensive calculations or non-convergence errors. In buckling restrained braces, the concrete does not
play a role in the lateral loading and only prevents the buckling of the steel core. Consequently, concrete can be removed and
replaced with equivalent hardness springs. The time taken to analyze the models is shown in Table 5. Analysis of all models is
performed using the same computer. As shown in Table 5, the analysis time are reduced from 43 to 6 h, 39 to 4 h, and 65 to 11 h for
models of BRB100, BRB60, and BRB Frame System, respectively.

6. Conclusion

Modeling of Buckling Restrained Braces (BRB) could increase analysis time and of output data volume. In order to avoid these
disadvantages, in this paper a simple model for steel of core-spring is presented. According to performing dynamic nonlinear analysis
on the real-full model and simplified model, it can be concluded that:

1 Full analytical model was presented for buckling-restrained brace. The above mentioned modeling had a very good match with
the experimental results.
2 Creating spring in both transverse directions of section could prevent the buckling of steel core and yield under the effect of
pressure at the steel core occurred like actual specimen.
3 Spring must be used in both directions in all parts of the meshing along the steel core. Lack of implementing springs in only one
node could cause buckling of the steel core. In case the core is meshed so small that the buckling load in proportion to the length of
each element is more than force of core yield pressure, inserting spring in all nodes is not necessary.
4 A primary relationship to select the dimensions of the elements in simple core-spring model was recommended.
5 The spring constant number has an inverse relation with steel core therefore, if the value of core division is doubled, the spring
constant would be halved.
6 The simple model of core-spring can be extended to model the steel frame with buckling-restrained brace.
7 The use of the simplified spring-core model may significantly reduce the analysis time compared to the full model.

Conflicts of interest

The authors have no conflicts of interest to declare.

References

[1] P. Clark, I. Aiken, K. Kasai, E. Ko, I. Kimura, Design procedures for buildings incorporating hysteretic damping devices, Proc. 69th Annual Convention of SEAOC,
Sacramento, CA, 1999.
[2] J.M. Kelly, R.I. Skinner, A.J. Heine, Mechanism of energy absorption in special devices for use in earthquake resistant structures, Bull. N. Z. Soc. Earthq. Eng. 5
(3) (1972) 63–88.
[3] R.I. Skinner, M.J. Kelly, A.J. Heine, Hysteretic dampers for earthquake-resistant structures, Earthq. Eng. Struct. Dyn. 3 (1975) 287–296.
[4] C.J. Black, N. Markis, I. Aiken, "Component Testing, Stability Analysis and Characterization of BuckLing-Restrained Unbonded Braces", Technical Rep. No. PEER
Report 2002/8, Pacific Earthquake Engineering Research Center, Univ. of California at Berkeley, Berkeley, Calif, 2002.
[5] N. Makris, Plastic torsional buckling of cruciform compression members, J. Eng. Mech. 129 (6) (2003).
[6] W.A. Lopez, D.S. Gwie, C.M. Saunders, T.W. Lauck, Lessons learned from largescale tests of unbonded braced frame subassemblages, Proceedings 71st Annual
Convention, Structural Engineers Association of California, Sacramento, California, 2002.
[7] H. Choi, J. Kim, Energy-based seismic design of buckling-restrained braced frames using hysteretic energy spectrum, Eng. Struct. 28 (2006) (2006) 304–311.
[8] AISC, Seismic Provisions for Structural Steel Buildings, Chicago (2005).
[9] L.A. Fahnestock, R. Sause, J.M. Ricles, Seismic Response and Performance of Buckling-Restrained Braced Frames, J. Struct. Eng. ASCE (September) (2006)
1195–1204.
[10] M.A. Shmshadyan, D. Vafaee, S.M. Zahraei, Effects of forward directivity and fling step in assessing control parameters of the buckling restrained braced frames,
The Fifth National Congress of Civil Engineering, Ferdowsi University of Mashhad, Iran, 2010.
[11] Behrouz Asgarian, Masoud Hamed Rahman Shokrgozar, Abitorabi, Effect of design loads in buckling restrained braced frames performance, The 14th World
Conference on Earthquake Engineering October 12–17, Beijing, China, 2008.
[12] F.L. pez-Almansa, J.C. Castro-Medina, S. Oller, A numerical model of the structural behavior of buckling-restrained braces, Eng. Struct. 41 (2012) 108–117.
[13] Chung-Che Chou, Gin-Show Liou, Jiun-Chi Yu, Compressive behavior of dual-gusset-plate connections for buckling-restrained braced frames, J. Constr. Steel Res.
76 (2012) 54–67.
[14] Junxian Zhao, Bin Wu, Jinping Ou, Effect of brace end rotation on the global buckling behavior of pin-connected buckling-restrained braces with end collars,
Eng. Struct. 40 (2012) 240–253.
[15] Mingming Jia, Dagang Lu, Lanhui Guo, Lin Sun, Experimental research and cyclic behavior of buckling-restrained braced composite frame, J. Constr. Steel Res.

487
R. Rahnavard et al. Case Studies in Construction Materials 8 (2018) 476–488

95 (2014) 90–105.
[16] K. Kimura, Y. Takeda, K. Yoshida, N. Furuya, Y. Takemoto, An experimental study on braces encased in steel tube and mortar, Annual Meeting of the
Architectural Institute of Japan (in Japanese), (1976).
[17] M. Iwata, T. Kato, A. Wada, Buckling-restrained braces as hysteretic dampers, Proceedings of Third International Conference on Behavior Steel Structures in
Seismic Areas (STESSA 2000), Montreal, Canada, 33-38. Rotterdam; and Brookfield, Vt.: A.A. Balkema, 2000.
[18] R. Tremblay, G. Degrange, J. Blouin, Seismic rehabilitation of a four-story bulding with a stiffened bracing system, Proc. 8th Canadian Conference on Earthquake
Engineering, Vancouver, Canada, 1999, pp. 549–554.
[19] C.C. Chen, Recent Advances of Seismic Design of Steel Building in Taiwan, International Training Programs for Seismic Design of Building Structures Hosted by
National Center for Research on Earthquake Engineering Sponsored by Department of International Programs, National Science Council, 1999.
[20] Masoud Mirtaheri, Ali Gheidi, Amir Peyman Zandi, Pejman Alanjari, Hamid Rahmani Samani, Experimental optimization studies on steel core lengths in
buckling restrained braces, J. Constr. Steel Res. 67 (2011) 1244–1253.
[21] Rohola Rahnavard, Akbar Hassanipour, Ali Mounesi, Numerical study on important parameters of composite steel-concrete shear walls, J. Constr. Steel Res. 121
(2016) 441–456.
[22] Rohola Rahnavard, Akbar Hassanipour, Mohamed Suleiman, Ali Mokhtari, Evaluation on eccentrically braced frame with single and double shear panels, J.
Build. Eng. 10 (2017) 13–25.
[23] Rohola Rahnavard, Akbar Hassanipour, Steel Structure Analysis Using ABAQUS, ACECR Kerman Branch, 978-600-6789-44-6, 2016, pp. 400–540 (in Persian).
[24] R. Rahnavard, A. Hassanipour, N. Siahpolo, Analytical study on new types of reduced beam section moment connections affecting cyclic behavior, Case Stud.
Struct. Eng. 3 (2015) 33–51.
[25] Rohola Rahnavard, Navid Siahpolo, Mohammad Naghavi, Akbar Hassanipour, Analytical study of common rigid steel connections under the effect of heat, Adv.
Civ. Eng. 2014 (2014) 10, http://dx.doi.org/10.1155/2014/692323 Article ID 692323.
[26] Rohola Rahnavard, Navid Siahpolo, Function comparison between moment frame and moment frame with centrically braces in high-rise steel structure under
the effect of progressive collapse, J. Struct. Constr. Eng. 4 (4) (2018) 42–57, http://dx.doi.org/10.22065/jsce.2017.77865.1084 Serial Number 14, (in Persian).
[27] Rohola Rahnavard, Maziyar Taghi Khaje, Akbar Hassanipour, Navid Siahpolo, Parametric study of seismic performance of steel bridges pier rehabilitated with
composite connection, J. Struct. Constr. Eng. (2018), http://dx.doi.org/10.22065/JSCE.2017.92128.1259 (in Persian).
[28] Rohola Rahnavard, Faramarz Fathi Zadeh Fard, Ali Hosseini, Mohamed Suleiman, Nonlinear analysis on progressive collapse of tall steel composite buildings,
Case Stud. Constr. Mater. 8 (2018) 359–379.
[29] Akbar Hassanipour, Rohola Rahnavard, Ali Mokhtari, Najaf Rahnavard, Numerical investigation on reduces web beam section moment connections under the
effectson cyclic loading, J. Multidiscip. Eng. Sci. Technol. (JMEST) 2 (8) (2015) ISSN:3159-0040.
[30] Yan-Lin Guo, Jing-Shen Zhu, Peng Zhou, Bo-Li Zhu, A new shuttle-shaped buckling-restrained brace, Theor. Study Buckling Behav. Load Resist. 147 (2017)
223–241, http://dx.doi.org/10.1016/j.engstruct.2017.05.033.
[31] Yan-Lin Guo, Jing-Zhong Tong, Bo-Hao Zhang, Bo-Li Zhu, Yong-Lin Pi, Theoretical and experimental investigation of core-separated buckling-restrained braces,
J. Constr. Steel Res. 135 (2017) 137–149.
[32] Xiaone Wei, Michel Bruneau, Analytical investigation of buckling restrained braces’ applications in bidirectional ductile end diaphragms for seismic performance
of slab-on-girder bridge, Eng. Struct. 141 (2017) 634–650.
[33] Zhe Qu, Jinzhen Xie, Tao Wang, Shoichi Kishiki, Cyclic loading test of double k-braced reinforced concrete frame subassemblies with buckling restrained braces,
Eng. Struct. 139 (2017) 1–14.
[34] K. Assaleh, M. AlHamaydeh, I. Choudhary, Modeling nonlinear behavior of buckling-restrained braces via different artificial intelligence methods, Appl. Soft
Comput. 37 (December) (2015) 923–938.
[35] M. AlHamaydeh, F. Abed, A. Mustapha, Key parameters influencing performance and failure modes for BRBs using nonlinear FEA, J. Constr. Steel Res. 116
(January) (2016) 1–18.
[36] M. AlHamaydeh, I. Choudhary, K. Assaleh, Virtual testing of buckling-restrained braces via nonlinear autoregressive exogenous neural networks, J. Comput.Civ.
Eng. ASCE 27 (November (6)) (2013) 755–768.
[37] M. AlHamaydeh, F. Abed, A. Mostafa, Finite element simulation of different failure modes for buckling-restrained braces under cyclic loading April, The 9th
International Conference on Composite Science and Technology (ICCST-9) (2013) 24–26.
[38] I. Choudhary, K. Assaleh, M. AlHamaydeh, Nonlinear AutoRegressive eXogenous artificial neural networks for predicting buckling restrained braces force, The
2012 International Symposium on Mechatronics and Its Applications (ISMA 2012) 10 (2012) April.

488