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

net/publication/313231676

Cyclic Performance of an Elliptical-Shaped Damper with Shear Diaphragms in

Chevron Braced Steel Frames

Article  in  Journal of Earthquake Engineering · January 2017

DOI: 10.1080/13632469.2016.1277436

CITATIONS READS

3 325

2 authors, including:

Seyed Mehdi Zahrai

University of Tehran
227 PUBLICATIONS   781 CITATIONS   

SEE PROFILE

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

Mitigation of the effects of extreme dynamic loads on timber structures View project

Numerical Analysis of structures View project

All content following this page was uploaded by Seyed Mehdi Zahrai on 04 May 2018.

The user has requested enhancement of the downloaded file.

 Journal of Earthquake Engineering

ISSN: 1363-2469 (Print) 1559-808X (Online) Journal homepage: http://www.tandfonline.com/loi/ueqe20

Cyclic performance of an elliptical shaped damper

with shear diaphragms in chevron braced steel
frames

Seyed Mehdi Zahrai & Mohamad Hosein Mortezagholi

To cite this article: Seyed Mehdi Zahrai & Mohamad Hosein Mortezagholi (2017): Cyclic
performance of an elliptical shaped damper with shear diaphragms in chevron braced steel frames,
Journal of Earthquake Engineering, DOI: 10.1080/13632469.2016.1277436

To link to this article: http://dx.doi.org/10.1080/13632469.2016.1277436

Accepted author version posted online: 31

Jan 2017.

Submit your article to this journal

Article views: 22

View related articles

View Crossmark data

Full Terms & Conditions of access and use can be found at

http://www.tandfonline.com/action/journalInformation?journalCode=ueqe20

Download by: [Northeastern University] Date: 14 March 2017, At: 16:02

 Cyclic performance of an elliptical shaped damper
with shear diaphragms in chevron braced steel frames
Seyed Mehdi Zahrai and Mohamad Hosein Mortezagholi

School of Civil Engineering, The University of Tehran, Tehran, Iran

t
ip
Correspondence to Seyed Mehdi Zahrai, The University of Tehran, School of Civil Engineering,
16th St. Azar, Enghelab Ave, Tehran, Iran. Email: [email protected]

cr
Abstract

us
Several advantages of yielding dampers in controlling seismic energy have attracted attention of

many researchers in designing new buildings and retrofitting existing structures. In recent
an
decades, various shapes and substances of such dampers have been used in engineering

structures and their behavioral features especially including the energy dissipating capacity have
M
been assessed. In this paper, a novel method is presented to obtain design relationship of two

types of yielding elliptical dampers in terms of their selected geometric properties, i.e. distance
d

between the shear diaphragms or virtual diameter and thickness. In addition, the elliptical shaped
te

steel damper equipped with the shear diaphragms in two different manufacturing types are
ep

proposed and modeled using the finite element software ABAQUS and their performance is

investigated. Then, 30 and 25 models respectively on the first and second types are studied using
c

push-over analysis. Designed dampers considering the proposed relationships are used in two
Ac

chevrons braced steel frames placed between the bracing and the beam. Due to their desirable

efficiency in energy dissipation and increase in the equivalent viscous damping of the frame,

better efficiency is achieved in the modified damper with easier fabrication.

1
 Keywords: yielding damper, elliptical shaped damper, shear diaphragms, ABAQUS, energy

dissipation, chevron braced frame.

1. Introduction

t
ip
One main structural and seismic engineering problem which is of utmost importance is to find

cr
solutions for reducing forces exerted upon the main structural members. During an earthquake

event, a huge amount of energy is imposed to the structure. In structures which are designed

us
using conventional methods, the plastic deformations of portions of the structure such as

connections, would cause partial energy dissipation. However when the amount of energy
an
exceeds the structure capacity, it might cause failure of the structure by damaging main
M
members. In case that the induced damages are local, they do not lead to collapse but the high

cost of reconstruction becomes problematic. Hence, providing devices and systems which could
d

prevent damage to the main structural members through absorbing the input energy would be
te

essential. The seismic isolators reduce the input energy to the structure in the event of an

earthquake and utilizing active, semi-active and passive control devices would increase the
ep

damping capacity of structure by dissipating imposed seismic energy.

c

Passive dampers in various types are suitable for application in special types of structures
Ac

according to their functionality. Among passive dampers, steel yielding dampers due to lack of

need for external sources, ease of their installation before and after the earthquake and economic

advantages have been widely considered. Experiencing the plastic behavior, these dampers

concentrate damage in themselves maintaining other members of the structure in the elastic

2
 range reducing at the same time the dynamic response of the structure. Kelly et al. in 1972

proposed application of yielding dampers in the structures. The X-shaped plates (ADAS)

[Whittaker et al., 1991] and the triangular plates (TADAS) [Tsai et al., 1993] comprised of steel

plates between two rigid plates, the Buckling Restrained Braces (BRB) with their significant

t
ductility [Benavent‐ Climent et al., 2011, Hoveidae et al., 2015] and the J-shaped damper

ip
dissipating energy through rolling-bending movement [Kato et al., 2005] are some samples of

cr
steel yielding dampers. The steel rhombic shaped damper (RADAS) made of two connected

us
triangular plates was studied by Ming-hsiang et al., [2004]. The fundamental mechanical features

and energy dissipation capacity of this kind of damper were later examined by Han et al. [2014].

an
Williams and Albermani [2006] introduced yielding shear panel device as the passive damper,

consisting of a thin diaphragm steel plate welded inside a square hollow section. This damper
M
dissipates energy with in-plane shear yielding of its diaphragm plate. Chan et al. [2009] have
d

investigated the yielding shear panel experimentally and found that the slenderness of diaphragm
te

and the surrounding in-plane stiffness of this damper are effective on its efficiency. Hossein et al.

[2011] developed the finite element model for this damper in ANSYS software utilizing material
ep

and geometric nonlinearity. Li et al. [2011] developed technique of Bouc–Wen–Baber–Noori

(BWBN) hysteretic model for considering the pinching and degradation effects in modeling of
c

yielding shear panel device. The hysteretic behavior of the low-yield-strength steel shear panel
Ac

damper (LYSPD) was evaluated by Zhang et al. [2013] through static and dynamic experimental

work.

Deng et al. [2015] proposed a novel SPD called Buckling Restrained Shear Panel Damper

3
 (BRSPD). This damper was made of an energy dissipation plate and two external restraining

plates. The tests showed that if restricted plates have sufficient stiffness and strength, the out-of-

plane buckling of the energy dissipation plate can be prevented. The shear-flexural yielding steel

dampers [Sahoo et al., 2015] as a combination of shear link plate and two X-shaped dampers at

t
their end have been recently studied both experimentally and numerically. Zahrai [2015] has

ip
experimentally assessed the vertical link beam as another type of yielding damper. The test

cr
specimens from IPE sections (sections with narrow thickness flanges) showed high ductility and

us
nearly dissipated the whole input energy while other components remained in the elastic zone.

Maleki and Bagheri [2010] have proposed the pipe damper, filled and unfilled with concrete
an
subjected to cyclic shear loading. Their results indicated that the hollow pipes due to appropriate

ductility in comparison with the concrete filled pipes have greater capability of energy
M
dissipation. In order to improve the performance of this kind of damper, the infilled pipe damper
d

(IPD) [Maleki and Mahjoubi, 2014] was presented consisting of two welded pipes in which two
te

smaller pipes were placed and the inner space was filled by Plumb or Zink. In continuation,

Najari [2014] numerically assessed the elliptical shaped damper in the chevron braced frame. It
ep

was found that placing 8 dampers of this type with different diameters and thicknesses and

concentration of damage in this type of damper, the structure elements remain in the elastic zone.
c

Therefore, by considering increase in ductility and also proper energy dissipation in the structure,
Ac

the proposed damper could be utilized for increasing the seismic resistance of steel frames.

Abebe and Choihave [2014] developed and assessed the structural performance of both cases of

stiffened and non-stiffened Circular Shear Panel Damper (CSPD) and Slit Circular Shear Panel

Damper (SCSPD) [Abebe and Choihave, 2014]. The results showed that the cumulative plastic

4
 deformation of SCSP damper is much higher than the CDP damper and for this reason it exhibits

better performance in absorbing the seismic energy.

In this paper, along with developing the yielding steel dampers, the performance of two types of

elliptical shaped dampers with shear diaphragms in the chevron braced steel frames is

t
ip
investigated using the ABAQUS software. The first type of damper is made of a segment with

cr
elliptical cross section and two shear diaphragms placed each on either side, while the second

type has a quasi-elliptical structure and easier fabrication. Since appropriate geometric design of

us
yielding damper would significantly affect its efficiency, a relationship is derived for the damper

design in terms of specific geometric properties. This is conducted after comparing experimental
an
and numerical results and ensuring the validity of modeling using the ABAQUS software.

Geometric properties are: distance between the shear diaphragms and the damper thickness for
M
the first type and virtual diameter and the damper thickness for the second type of damper. It
d

should be noted that in every design relationship, numerous geometric and non-geometric
te

parameters could be involved and considering them all in the relationship makes it complex for

practical application. However, the ultimate goal of design is not to obtain maximum shear force
ep

of a damper with distinct parameters for which there would be a need to involve all parameters in

the relationship. In order to obtain a distinct shear force less than the bracing capacity, it is
c

intended to determine some of the influencing geometric parameters while other parameters are
Ac

considered constant. So, application of two influencing geometric parameters to achieve this

goal is justified. Due to investigating the effect of variation just in two geometric parameters of

the damper, and lacking a known mathematical relationship between the geometric properties

and the ultimate capacity of the dampers also trusting the obtained relationship within a wider

5
 range of variations in the geometric dimensions, there is a need to consider an acceptable number

of dampers. Therefore, in extracting the design relationship for the first and second types of the

damper, respectively 30 and 25 models together with pushover analyses are performed. Finally,

the cyclic performance of the designed dampers is assessed on two different types of frames to

t
investigate the appropriateness of the proposed relationship.

ip
cr
2. Elliptical and Modified Elliptical Shaped Dampers

us
with Shear diaphragms

an
In this study, two types of dampers are investigated and their fabrication and the range of their

dimensions are explained subsequently.

M
2.1. Elliptical shaped damper with shear diaphragms
d
te

Elliptical shaped damper with shear diaphragms is comprised of a horizontal elliptical shaped

profile with two lateral shear diaphragms. Application of an elliptical shaped component instead
ep

of the pipe profile provides more freedom in design of the damper. For example, per each

constant eccentricity we could design a damper with higher energy dissipation capacity just by
c

changing the major diameter of the ellipse, while in case of pipe profile to attain this property
Ac

there would be a need for two profiles with parallel connections. Also in continuation, by placing

an elliptical shaped component with or without shear diaphragms in one of the examined frames,

it could be demonstrated that application of shear diaphragms due to reduced local buckling in

the elliptical shaped component, would result in increased efficiency of the damper. Their

6
 variable geometric parameters are the distance between the shear diaphragms  L  and the

thickness of damper  t  . The thickness of the elliptical shaped segment and the shear diaphragms

are taken equal in each of the models (Fig. 1). The range of distance between the shear

diaphragms of damper is 10-30 cm (with steps of 5 cm) and its thickness is in the range of 0.25-

t
ip
1.50 cm (with steps of 0.25 cm) as presented in Table 1.

cr
Considering this range of variations for geometric properties of the damper, is to achieve a

us
capacity close to the capacity range of the conventional bracings. In this type of damper, the

major external diameter is taken 30cm and the minor internal diameter is taken 15 cm. Naming

an
this damper is in the form of EDLntm, in which ED is abbreviated for Elliptical Damper, L is the

first letter of the word 'Length', n is the distance between the shear diaphragms in cm, t is the first
M
letter of the word 'thickness' and m is the thickness of damper in cm. For example, EDL25t1

introduces an elliptical shaped damper with 25 cm length and 1 cm thickness (Fig. 2).
d
te

2.2. Modified Elliptical Shaped Damper with shear diaphragms

ep

Modified Elliptical Shaped Damper with shear diaphragms is made of a quasi-elliptical segment

with two shear diaphragms (Fig. 3). As it would be observed from the results of elliptical shaped
c

damper with the shear diaphragms, the distance between shear diaphragms  L  in design
Ac

relationship of the damper is not an effective factor upon design process. So in the modified

model, in addition to ease of damper fabrication at the site, the distance between shear

diaphragms is replaced by virtual diameter parameter for the purpose of enhancing effectiveness

(Fig. 4). In this case the distance between shear diaphragms is taken a constant value equal to 10

7
 cm. The small distance taken between the shear diaphragms for type 2 damper is due to the

capability of being installed on a narrow beam. The geometric parameters include the virtual

diameter of ellipse with a range of 10-30 cm (with steps of 5 cm) and the thickness of damper

with a range of 0.5-1.5 cm (with steps of 0.25 cm) as presented in Table 2.

t
ip
The obtained results for type 1 damper reveals its reduced performance due to the premature

cr
local buckling in models with small thicknesses. Therefore, with respect to type 2 damper, those

models with thicknesses of 0.25 cm have been removed. The reason for application of virtual

us
diameter term is due to the quasi ellipse shape of the damper and impossibility to define the

major and minor diameters like a real ellipse shaped section. Naming various models of quasi
an
elliptical damper is also indicated in the form of MEDDntm, in which MED is abbreviated for

'Modified Elliptical Damper', D is the first word of 'Diameter' , n is the virtual minor diameter in
M
cm, while t and m have a similar definition as explained in the previous section. For example, a
d

modified elliptical damper with a virtual minor diameter of 25 cm and thickness of 1 cm, is
te

named as MEDL25t1.
ep

3. Modeling Details
c

Solid elements in 3D space are used for modeling. The material behavior curve used by
Ac

considering the effects of hardening steel is elasto-plastic. Assuming that the loading is gradual

and slow, the effects of velocity and acceleration are negligible and therefore the type of analysis

is general, Static. Also the secondary effects due to the large deformations and buckling of the

members are considered in the analysis process. The interaction between all structural members

8
 is so that all degrees of freedom are translated to the connection levels. The elements used in this

study are 8-node brick elements with reduced integration, in which the hourglass effects are

controlled (C3D8R). For validation of the numerical method, the number of used meshes in the

analysis problem is increased in a few stages (the meshing size is halved) and distribution of one

t
of the variables is taken into account. The proper size of meshes is obtained when the change in

ip
size, would lead to less than 10% difference, corresponding to reasonable accuracy and

cr
processing cost.

us
4. Numerical Model Validation
an
Since the experimental results of the chevron braced steel frame with elliptical shaped cross-
M
section link beam are not available, an experimental work of I-shaped vertical link beam in a

chevron braced steel frame is considered. The finite element software ABAQUS is used for
d

numerical modeling of SPS2 and SPS4 test specimens [Zahrai, 2015] as shown in Fig. 5.
te

Comparison of the numerical and experimental results assured the accuracy of modeling the

main components of chevron braced frames, the connections, boundary conditions and the
ep

applied loads. Also due to the appropriate compliance observed between the results, one could

conclude the accuracy of dimension choice and shape of the applied meshing. The difference
c
Ac

between the numerical and experimental models was caused by ignoring details of the welded

and bolted connections in the numerical model. These details appear to be negligible due to the

small difference in the performance curves, as shown in Fig. 6.

9
 ST37 steel was used for the main members of the frames. The yield stress ( Fy ) and ultimate

stress ( Fu ) of the vertical link beam (VLB) were 3372 and 4822 for SPS2, respectively. These

values for SPS4 are 3641 and 5175, respectively. In addition, the geometric properties of the

frames and vertical link beams were presented in Table 3. In the following, cyclic curves

t
ip
obtained from the numerical and experimental results are compared to each other. It should be

noted that due to the gradual loading of the experimental model according to the AISC2010

cr
code, the effects of acceleration and velocity are negligible. Therefore, in the numerical model

us
the static analysis is incorporated.

5. Proposing Design Relationship an

M
As one of the main objectives of damper design is to control level of forces induced in the

braces, in order to derive the maximum producible shear force by the dampers, all of the 30
d

elliptical shaped dampers and 25 modified dampers with appropriate boundary conditions are
te

analyzed using the pushover analysis method. As restraining braces against buckling is one of
ep

the goals for accurate design of dampers with a good approximation, it is assumed that the

damper lower plate is anchored as fixed. Also due to the high axial stiffness of the column, the
c

beam connected to the damper prevents downward movement of the upper plate and finally the
Ac

lateral movement of the beam causes displacement in the damper upper plate. Having known all

these, the equalized boundary conditions for pushover analysis of the damper only, is shown in

Fig. 7. To draw the pushover curve, a displacement about 4% of the conventional frames height

(3 m) is applied to the dampers. The pushover curves of the whole models are presented in the

11
 appendix. The maximum shear force induced in each pushover curve is an appropriate

approximation of this controlling force. After determining the necessary curves for models, the

maximum induced shear force is determined. For example, using the pushover curves of

EDL10t1.5 and EDL20t1 models, maximum shear force of dampers are obtained as 1676.42 kN

t
and 1034.56 kN, respectively (Fig. 8). This process is conducted for all pushover curves and the

ip
results are shown in Tables 4 and 5 for each type of damper.

cr
In Fig. 9 the yielded zones of the damper due to applying final displacement are shown where the

us
active yielding output of the software for two cases: type1 and type2 dampers indicate the

yielded elements, so that each element after yielding is shown with number 1 and red color
an
(bright parts) and in case of not yielding is shown with number zero and blue color (dark parts).

As observed, the major dissipated energy in these dampers is resulted from yielding of the shear
M
diaphragms (Fig. 9).
d

The maximum shear force is a function of geometric parameters of dampers (distance between
te

shear diaphragms and thickness for the ED and virtual diameter for the MED). In the first step, to
ep

obtain the intended design relationship among two available geometric parameters for each

damper, the thickness is taken constant and the maximum shear force is shown in terms of the
c

other parameter (Fig. 10).

Ac

Considering the location of points in the coordinate system, the best fitted curves for maximum

shear force in terms of the distance between shear diaphragms and the virtual diameter for the

ED and MED, respectively and corresponding to a constant thickness value is in the form of a

line. Table 4 shows the equation of these lines and the coefficients of determination. The

11
 ‘determination coefficient’ is shown as and defined as follows:

SSTO  SSE
R2  (1)
SSTO

t
ip
where, SSTO and SSE are abbreviation of Sum of Squared Total and Sum of Squared Error,

respectively. Comparing the slope of the lines in Table 6, especially in small thicknesses, the

cr
increase in the sensitivity of the design relationship for the geometric parameter of virtual

us
diameter of modified elliptical damper is well understandable with respect to the distance

between shear diaphragms of the elliptical damper.

an
The slope and intercept for each linear equation in the form Fmax  a(L or D )  b vary
M
corresponding to different thicknesses. Therefore, it could be concluded that the parameters a

and b are both functions of the thickness. In continuation for considering the effect of second
d

geometric parameter (thickness) in the design relationship, the best curve fitting corresponding to
te

the slope and intercept of above lines are depicted with respect to the thickness. Fig. 11 shows
ep

the best fitted curves for these parameters. Having obtained the equations corresponding to the

parameters a and b from the curve fitting and replacing them in the initial linear equation, the
c

basic design relationships of dampers are obtained as follows:

Ac

12
 a  19.184 t 2  17.478 t  4.577, SSR  0.9872
  ED
Fmax  aL  b
b  1026.24 t  94.911, SSR  0.9981
(2-a)

 Fmax
ED
 19.184 t 2  17.478 t  4.577  L 1026.24 t  94.911

t
ip
a  29.315 t  0.888, SSR  0.9962
  MED
Fmax  aD  b

cr
b  532.73 t  46.616, SSR  0.9980
(2-b)

us
 Fmax
MED
  29.315 t  0.888 D   532.73 t  46.616 

an
Constant coefficients a and b of studied dampers are function of the geometric and non-

geometric parameters which have not been incorporated in the design relationship. For example
M
by changing the type of steel used in damper, the constant coefficients also have different values.

In design related issues, it is essential to identify the influential parameters and to subsequently
d

investigate and assess them. The behavioral curve of the constituent material of the damper is
te

among the effective factors in the design relationship. However, for each arbitrary behavioral
ep

curve one can obtain the design relationship of the elliptical shaped damper based on a similar

process. In this particular case, these coefficients are function of the behavioral characteristics of
c

the material used in damper. Also as shown in Fig. 12, depicting the maximum shear force in
Ac

terms of thickness of the elliptical shaped damper in various lengths, it could be concluded that

due to extreme similarity of the curves to each other, the distance between shear diaphragms is

not an effective parameter and the related shear force especially in small thicknesses has

negligible variation. Therefore, choosing the distance between shear diaphragms of damper is not

13
 a good factor in design here.

Lack of sensitivity of maximum shear force with respect to the distance between shear

diaphragms of the first type damper especially in small thicknesses and increase in the sensitivity

of second type damper with respect to the virtual diameter is well observable in the 3D diagram

t
ip
of Fig. 13 (a and b). Albeit in larger thicknesses, these variations become more prominent. Also,

cr
the maximum shear force corresponding to the specified range would vary for geometric

parameters of the first type damper in the range of 200-2000 kN and for geometric parameters of

us
the second type damper in the range of 400-2000 kN. Note that due to formation of linear terms

with respect to the existing geometric parameters in the design relationships, reducing the
an
number of models is possible such that even with 4 models, one can achieve these design

relationships with good accuracy.

M
6. Assessment of Proposed Design Relationship
d
te

6.1. ED Damper
ep

In order to install the elliptical shaped damper in SPS4 test frame, with respect to selected
c

diameters, it is necessary to change the length of braces. Lack of ability in changing the in-plane
Ac

dimensions of this type of damper is among its basic design shortcomings. Hence in this section,

the length of SPS4 frame braces is changed and the cyclic performance of the elliptical shaped

damper is examined. This modified frame is named SPS4′. The capacities of chevron bracings

for SPS4′ and SPS4 frames in case of their negligible yield and lack of buckling are about 625

14
 kN and 255 kN, respectively. To determine the maximum base shear at the braces, one could put

the upper gusset plate under a controlled displacement and obtain the maximum shear capacity

before buckling, but because of the distance between beam and gusset plate, the set of braces is

not in fact subjected to a pure shear force and an additional moment is also applied on the setting

t
which results in reduced bracings capacity.

ip
cr
Therefore another method is used to determine real capacity of the set of braces. In this case as

shown in Fig. 14, a reference point on the lower flange of the beam in the frames without

us
dampers (SPS4 and SPS4') is defined and 3 translational degrees of freedom at this point are

transferred to the upper gusset plate. It is noteworthy that due to axial stiffness of the columns,
an
the beam connected to the damper prevents movement of the upper gusset plate downward or in

other words restrains it in the vertical direction. Hence, to simulate the above conditions, the
M
translational degree of freedom of reference point at the vertical direction is translated to the
d

beam. Next, the reference point would be subjected to a controlled displacement. This
te

displacement continues as long as the compressive brace has not buckled, meanwhile a

negligible yield occurs in the set of braces and the gusset plates. Finally, the maximum base
ep

shear of the bracing before buckling and yielding would be selected as the capacity of the

bracing. It should be noted that the aforementioned method is the idea of the authors while not in
c

conflict with other methods to determine the bracing capacity. But in this method using the
Ac

computer facilities, other details like effects of the axial stiffness of columns and the extra

moment due to eccentricity of the bracing are considered in the accurate determination of the

bracing capacity.

15
 In the following by considering the SPS4′ frame, chevron bracing capacity and the obtained

design relationship, the geometric dimensions of the elliptical shaped damper are determined.

Considering the width dimension of the beam in the SPS4′ frame (9.1 cm), the geometric

parameter of distance between shear diaphragms of damper is taken 9 cm.

t
ip
Fmax  t , L   19.184 t 2  17.478 t  4.577  L 1026.24 t  94.911
(3)

cr
Fmax  t ,9   625  t 0.68 cm

us
After determining the geometric dimensions of damper, its cyclic performance within SPS4′

an
frame is assessed. Considering the bi-linearization of force-displacement diagram, the equivalent

yield displacement of the frame is estimated to be 0.6 cm (Fig. 15).

M
In which, V is the base shear, Vy is the yield shear force, and Drift is the ratio of relative lateral

displacement to the frame height. The following figure shows the time history of cyclic loading
d

and the cyclic hysteretic curves (Fig. 16).

te

Finally the energy dissipation due to plastic behavior is shown in Fig. 17. From the dissipated
ep

energy diagram for ED damper with respect to the whole frame it could be concluded that:
c

A: for 6 initial cycles, all structure members remained in the elastic zone and the input
Ac

energy to the structure is stored like an elastic spring.

B: for 3 next cycles, almost all energy dissipation takes place in the damper and main

members like the beam, braces and columns remain within elastic range.

16
 C: Afterwards, the main members of the structure together with the damper participate in

the dissipation of input energy to the structure. In this case 76% of energy is dissipated through

yielding of the damper.

t
ip
PDE for Damper 206.7340

cr
Efficiency up to Cycle 23   0.76 (4)
PDE for Frame 272.0750

us
where, PDE is the abbreviation for the Plastic Dissipated Energy.

an
Design of the ductile member is implemented correctly when not only dissipates main part of the

energy imposed to its structure, but also as a fuse prevents from damage to the main structure
M
members such as braces. The results obtained from both introduced dampers indicate the proper
d

performance of the ductile members and meeting above–mentioned- conditions.

te

Using the elliptical shaped profile without shear diaphragms in chevron bracing does not induce
ep

acceptable energy dissipation in the system due to instability and premature buckling and it could

be stated that the set of damper and chevron bracings does not affect the seismic performance of
c

structure, while utilizing shear diaphragms in the elliptical shaped damper could significantly
Ac

increase the efficiency. For example, in the diagram of Fig. 18 which demonstrates the energy

dissipation by a damper having no shear diaphragm compared to the whole frame, no force is

induced in the set of damper and bracing due to the buckling of the elliptical shaped plates and as

a result there would be no significant energy dissipation.

17
 But addition of shear diaphragms to the set, not only prevents from buckling of the elliptical

shaped segment but also the major portion of this energy is dissipated through yielding of these

shear diaphragms. It is noteworthy that shear diaphragms are actually accounted as the main

yielding members in the ductile element and the elliptical shaped component has the role of

t
constraining these lateral plates. As a result, the effective role of these lateral plates becomes

ip
more significant as shown in Fig. 19.

cr
6.2. MED Damper

us
In order to determine the geometric dimensions of the modified elliptical shaped damper with
an
respect to the amount of eccentricity and the capacity obtained from chevron bracings in both

SPS4 and SPS4' frames considering expression (2-b), appropriate dimensions of the damper are
M
derived in each case.
d

Fmax  t , D    29.315 t  0.888 D   532.73 t  46.616 

te

(5-a)
Fmax  t ,30   255  t 0.2 c
ep

Fmax  t , D    29.315 t  0.888 D   532.73 t  46.616 

(5-b)
Fmax  t ,15  625
c

 t 0.67 cm
Ac

In SPS4 frame, the longitudinal dimensions of damper are very large (virtual major diameter 60

cm, virtual minor diameter 30 cm) and its thickness is very small (0.20 cm). Note that 0.20 cm

thickness is obtained by extrapolating the design relationship with respect to the range defined

18
 for geometric parameters of damper. So in this model, not only the extrapolation method is

investigated in determination of geometric dimensions of the damper but also performance of the

shear diaphragms with large dimensions and small thickness is examined. By placing the damper

within SPS4 and SPS4' frames and performing pushover analysis, the force-displacement curve

t
as shown in Fig. 20 determines the yield displacements for both frames as 0.60 and 0.70 cm,

ip
respectively.

cr
Considering the yield displacement, the time history of the loading protocol with a domain of 7

us
times of the yield displacement is shown in Fig. 21.

an
Figs. 22 and 23 show the cyclic curves and the rate of energy dissipation for the damper with

respect to the whole frame, respectively, each for SPS4 and SPS4' frames.
M
Considering the energy dissipation diagram for the damper and the whole frame, the following

results are derived:

d
te

A: For 9 initial cycles, all structure members remained in the elastic zone which means

zero energy dissipation in the whole frame.

ep

B: For next 3 cycles in both models, due to negligible difference between the two curves,
c

all energy dissipation is due to inelastic behavior by the damper.

Ac

C: In the remaining cycles, due to imposing the main members to the plastic range, the

performance of damper is reduced so that its efficiency at the end of loading is the same for both

models:

19
 SPS4 frame

PDE for Damper 140.468

Efficiency up to Cycle 23   0.94 (6)
PDE for Frame 155.560

t
ip
SPS4′ frame

cr
PDE for Damper 248.999
Efficiency up to Cycle 23   0.69 (7)

us
PDE for Frame 359.772

an
Similar to the elliptical shaped damper, addition of shear diaphragms to the quasi elliptical

segment significantly increases its efficiency (Figs. 24 and 25).

M
7. Equivalent Viscous Damping
d
te

The inherent damping in the structures under factors such as the friction between the members,

looseness of the connections and material degradation for the steel structures ranges between 2-
ep

5%. Application of energy dissipation devices in these structures would result in significant

increase in damping ratio. This damping is non-linear and induces complexities in the process of
c
Ac

analysis. In order to diminish these complexities, the non-linear damping is replaced by the

equivalent viscous damping concept. By setting the dissipated energy by non-viscous damping

equal to the damping of a viscous material, the amount of equivalent viscous damping is

obtained.

21
 Considering Fig. 26, the equivalent viscous damping (  eq ) would be [Priestly et al. 1996]:

Ah
eq  (8)
2 Vm  m

t
ip
In which Ah is the area under the curve,  m and Vm are the mean maximum displacement and

cr
shear force in a complete cycle of the cyclic curve, respectively. The equivalent viscous damping

us
for the last cycle of the cyclic loading curve for the 3 tested models are given in Table 7.

As presented in Table 7, the last cycle damping for all models is in the range of 28-32% which
an
indicates effective performance of the elliptical dampers.
M
8. Conclusion
d

To propose new yielding dampers, in this paper two types of dampers with elliptical cross
te

section were studied. By modeling 30 elliptical dampers and 25 modified elliptical dampers and
ep

with variable geometric parameters and performing pushover analysis, the design relationships

with simple and accurate formulation were obtained for the proposed yielding dampers.
c

By designing the proposed dampers based on the presented relationships and placing them into
Ac

the chevron braced steel frames and also examining the cyclic performance of the set of frame

and damper in dissipating energy and increasing the damping ratio, the assessment of the design

relationships was accomplished. The results indicate the appropriateness of the design

relationships because in addition to significant energy dissipation share by the elliptical damper,

21
 such a device acts like a fuse and prevents the braces from buckling or yielding by controlling

the force level. Investigating the cyclic performance of the elliptical shaped component placed in

the chevron braced frame with or without shear diaphragms, reveals the effective role of such

diaphragms in improving performance of the ductile member. It should be mentioned that after

t
assuring of the appropriate performance of the elliptical shaped component with the shear

ip
diaphragms, the second generation of this damper was recommended for practical purposes.

cr
As a primary step for any complementary studies and experimental work, design relations for the

us
proposed dampers were developed in this paper. In addition, described regression analysis

method is not limited to the proposed energy dissipation system and can be used for defining

design relations of other dampers.

an
M
References

ABAQUS.[2010]. “Analysis user's manual,” version 6.10, Dassault Systèmes, Simulia Corp.
d
te

Abebe, D. Y., & Choi, J. [2014]. “Analytical Evaluation on Hysteresis Performance of Circular

Shear Panel Damper.” World Academy of Science, Engineering and Technology 8(6), 744-750.
ep

Abebe, D. Y., & Choi, J. [2014]. “Development and Structural Performance Evaluation on Slit
c

Circular Shear Panel Damper,” World Academy of Science, Engineering and Technology 8(7),
Ac

803-808.

Benavent‐Climent, A., Morillas, L., & Vico, J. M. [2011]. “A study on using wide‐flange section

web under out‐of‐plane flexure for passive energy dissipation,” Earthquake Engineering &

Structural Dynamics 40(5), 473-490.

22
 Chan, R. W., Albermani, F., & Williams, M. S. [2009]. “Evaluation of yielding shear panel

device for passive energy dissipation,” Journal of Constructional Steel Research 65(2), 260-268.

Deng, K., Pan, P., Li, W. and Xue, Y. [2015]. “Development of a buckling restrained shear panel

damper.” Journal of Constructional Steel Research, 106, 311-321.

t
ip
Han, Q., Jia, J., Xu, Z., Bai, Y., & Song, N. [2014]. “Experimental Evaluation of Hysteretic

cr
Behavior of Rhombic Steel Plate Dampers,” Advances in Mechanical Engineering 6, 185629.

Hossain, M. R., Ashraf, M., & Albermani, F.[2011]. “Numerical modelling of yielding shear

us
panel device for passive energy dissipation.” Thin-Walled Structures, 49(8), 1032-1044.

an
Hoveidae, N., Tremblay, R., Rafezy, B., & Davaran, A. [2015]. “Numerical investigation of

seismic behavior of short-core all-steel buckling restrained braces,” Journal of Constructional

M
Steel Research 114, 89-99.

Kato, S., Kim, Y. B., Nakazawa, S., & Ohya, T. [2005]. “Simulation of the cyclic behavior of J-
d

shaped steel hysteresis devices and study on the efficiency for reducing earthquake responses of
te

space structures,” Journal of Constructional Steel Research 61(10), 1457-1473.

ep

Kelly, J. M., Skinner, R. I., & Heine, A. J. [1972]. “Mechanisms of energy absorption in special

devices for use in earthquake resistant structures,” Bulletin of NZ Society for Earthquake
c

Engineering 5(3), 63-88.

Ac

Li, Z., Albermani, F., Chan, R.W. and Kitipornchai, S.[2011]. “ Pinching hysteretic response of

yielding shear panel device.” Engineering Structures, 33(3), 993-1000.

Maleki, S., & Bagheri, S. [2010]. “Pipe damper, Part I: Experimental and analytical study,”

23
 Journal of Constructional Steel Research, 66(8), 1088-1095.

Maleki, S., & Mahjoubi, S. [2014]. “Infilled-pipe damper,” Journal of Constructional Steel

Research 98, 45-58.

Ming-hsiang, S. H. I. H., Wen-pei, S. U. N. G., & Cheer-germ, G. O. [2004]. “Investigation of

t
ip
newly developed added damping and stiffness device with low yield strength steel,” Journal of

cr
Zhejiang University Science 5(3), 326-334.

Najari Varzaneh. M., Hosseini. M., and Akbarpoor., A. [2014]. “The Study of EADAS Elliptical

us
Steel Damper Function in Seismic Resisting of Steel Frames”. International Research Journal of

Applied and Basic Sciences, 8, 1569-1573.

an
Priestly, M.J.N., Seible, F., and Clavi, G.M. [1996]. “Seismic Design and Retrofit of Bridges”,
M
John Wiley and Sons, New York.

Sahoo, D. R., Singhal, T., Taraithia, S. S., & Saini, A. [2015]. “Cyclic behavior of shear-and-
d

flexural yielding metallic dampers,” Journal of Constructional Steel Research 114, 247-257.
te

Tsai, K. C., Chen, H. W., Hong, C. P., & Su, Y. F. [1993]. “Design of steel triangular plate
ep

energy absorbers for seismic-resistant construction,” Earthquake spectra 9(3), 505-528.

Whittaker, A. S., Bertero, V. V., Thompson, C. L., & Alonso, L. J. [1991]. “ Seismic testing of
c
Ac

steel plate energy dissipation devices,” Earthquake Spectra 7(4), 563-604.

Williams MS., Albermani F. [2006]. “Monotonic and cyclic tests on shear diaphragm dissipaters

for steel frames.” Advanced Steel Construction, 2(1), 1–21.

Zahrai, S. M. [2015]. “Cyclic testing of chevron braced steel frames with IPE shear panels.”

24
 Steel and Composite Structures 19(5), 1167-1184.

Zhang, C., Aoki, T., Zhang, Q., & Wu, M. [2013]. “Experimental investigation on the low-yield-

strength steel shear panel damper under different loading.” Journal of Constructional Steel

Research, 84, 105-113.

t
ip
cr
us
an
M
d
te
c ep
Ac

25
 Appendix

Pushover curves of dampers type I and type II with respect to geometric parameters are shown in

Fig. A1 and A2, respectively.

t
Fig. A1 First Damper ; a: L=10, b: L=15, c: L=20, d: L=25 e: L=30

ip
cr
us
an
M
d
te
c ep
Ac

26
 Fig. A2. Second Damper ; a: D=10, b: D=15c: D=20, d: D=25, e: D=30

t
ip
cr
us
an
M
d
te
c ep
Ac

27
 Fig. 1 Elliptical shaped damper with shear diaphragms.

t
ip
cr
us
an
M
d
te
c ep
Ac

28
 Fig. 2 Elliptical damper; a: Schematic view; b: Longitudinal section; c: Cross section.

t
ip
cr
us
an
M
d
te
c ep
Ac

29
 Fig. 3 Modified elliptical shaped damper with shear diaphragms.

t
ip
cr
us
an
M
d
te
c ep
Ac

31
 Fig. 4 Modified elliptical damper; a: Schematic view; b:Longitudinal section; c: Cross section.

t
ip
cr
us
an
M
d
te
c ep
Ac

31
 Fig. 5 Result verification; a: experimental specimen [SPS4 in Zahrai, 2015]; b: Numerical model.

t
ip
cr
us
an
M
d
te
c ep
Ac

32
 Fig. 6 Cyclic curves based on numerical and experimental [Zahrai, 2015] results of test frames:
SPS2, (b) SPS4.

t
ip
cr
us
an
M
d
te
c ep
Ac

33
 Fig. 7 Equalized boundary conditions; a: the lower plate of damper; b: the upper plate of damper.

t
ip
cr
us
an
M
d
te
c ep
Ac

34
 Fig. 8 The maximum shear capacity extracted from the pushover curves for models EDL10t1.5
and EDL20t1.

t
ip
cr
us
an
M
d
te
c ep
Ac

35
 Fig. 9 Yielded zones at maximum applied displacement, a: EDL10t1; b: MEDD20t1.

t
ip
cr
us
an
M
d
te
c ep
Ac

36
 Fig. 10 Maximum base shear force for different damper thicknesses. a: ED in terms of distance between
shear diaphragms  L  based on 30 models; b: MED in terms of virtual diameter  D  based on 25
models.

t
ip
cr
us
an
M
d
te
c ep
Ac

37
 Fig. 11 Curve of the slope of line and curve of the intercept of line. a and b: ED; c and d: MED.

t
ip
cr
us
an
M
d
te
c ep
Ac

38
 Fig. 12 Maximum base shear of ED damper in terms of thickness for various distances between
shear diaphragms.

t
ip
cr
us
an
M
d
te
c ep
Ac

39
 Fig. 13 3D diagram of design relationship, a: ED; b: MED.

t
ip
cr
us
an
M
d
te
c ep
Ac

41
 Fig. 14 The location of reference point. a:schematic shape; b: software model.

t
ip
cr
us
an
M
d
te
c ep
Ac

41
 Fig. 15 Force-displacement curve of SPS4′ frame equipped with ED.

t
ip
cr
us
an
M
d
te
c ep
Ac

42
 Fig. 16 a: Loading protocol; b: cyclic hysteretic curves of SPS4' frame equipped with ED.

t
ip
cr
us
an
M
d
te
c ep
Ac

43
 Fig. 17 Comparison of dissipated energy by the damper with respect to that of the whole SPS4'
frame.

t
ip
cr
us
an
M
d
te
c ep
Ac

44
 Fig. 18 Energy dissipation in damper with no shear diaphragms compared to that of the whole
SPS4' frame.

t
ip
cr
us
an
M
d
te
c ep
Ac

45
 Fig. 19: a: Yielded zones for elliptical damper with no shear diaphragms; b: yielded zones for
damper with shear diaphragms.

t
ip
cr
us
an
M
d
te
c ep
Ac

46
 Fig. 20 Normalized force-displacement relationships for the frame equipped with MED damper;
a: SPS4 ; b: SPS4'.

t
ip
cr
us
an
M
d
te
c ep
Ac

47
 Fig. 21: Lateral loading protocol applied to the frame equipped with MED damper; a: SPS4 ;
b:SPS4'.

t
ip
cr
us
an
M
d
te
c ep
Ac

48
 Fig. 22: Cyclic curves of the frame equipped with MED; a; SPS4 ;b: SPS4'.

t
ip
cr
us
an
M
d
te
c ep
Ac

49
 Fig. 23: Comparison of energy dissipation by the damper with respect to the whole frame; a:
SPS4 ; b: SPS4'.

t
ip
cr
us
an
M
d
te
c ep
Ac

51
 Fig. 24: Energy dissipation rate for the damper with no shear diaphragms with respect to the
whole SPS4 frame.

t
ip
cr
us
an
M
d
te
c ep
Ac

51
 Fig. 25: a: Yielded zones for the modified elliptical damper with no shear diaphragms; b:yielded
zones for the damper with shear diaphragms.

t
ip
cr
us
an
M
d
te
c ep
Ac

52
 Fig. 26: Determination of equivalent damping coefficient in terms of dissipated energy in each
cycle.

t
ip
cr
us
an
M
d
te
c ep
Ac

53
 Table1. Characteristics of damper type 1

Geometric parameters
Behavioral

characteristics of steel
constant variable

t
ip
Yield Parameter Parameter Parameter Range Step

cr
240 MPa
stress name value name changes(cm) changes(cm)

us
distance

Tensile
Small
an between

shear
370 MPa diameter 15 10-30 5
M
strength
(cm) diaphragms

 L
d
te

large
thickness
ep

Elongation
0.2 diameter 30 0.25-1.5 0.25
rate t 
(cm)
c
Ac

54
 Table 2. Characteristics of damper type 2

Behavioral Geometric parameters

characteristics of

steel constant variable

t
ip
Yield 240 Parameter Parameter Parameter Range Step

cr
stress MPa name value name changes(cm) changes(cm)

us
distance

Tensile 370
between
an virtual

shear 10 diameter 10-30 5

M
strength MPa
diaphragms  D
 L  (cm)
d
te

virtual Large
ep

Elongation diameter(cm)/ thickness

0.2 2 0.5-1.5 0.25
rate virtual small t 
c

diameter(cm)
Ac

55
 Table 3. Geometric properties of the frames and their vertical link beams

Brace Beam Column Member Frame

2UNP80 IPB140 IPB120 Section SPS2

t
345 420 300 Length(cm)

ip
2UNP80 IPE180 2IPE140 Section SPS4

cr
337 420 300 Length(cm)

us
Stiffener Shear panel
an VLB
M
1 Thickness IPE140 Section SPS2

10 Distance 20 Length(cm)
d
te

1 Thickness IPE140 Section SPS4

ep

10 Distance 30 Length(cm)
c
Ac

56
 Table 4. Maximum shear forces for elliptical dampers

Base Shear Base Shear Base Shear

Model Model Model
(kN) (kN) (kN)

t
ip
EDL10 t 0.25 195.682 EDL15 t 1.25 1384.45 EDL25 t 0.75 726.978

cr
us
EDL10 t 0. 5 413.48 EDL15 t 1. 5 1791.16 EDL25 t 1 1069.69

EDL10 t 0.75 679.243 an

EDL20 t 0.25 204.382 EDL25 t 1.25 1494.12
M
EDL10 t 1 982.17 EDL20 t 0.5 429.297 EDL25 t 1. 5 1987.84
d

EDL30 t 0.
EDL10 t 1.25 1320.66 EDL20 t 0.75 715.823 209.127
te

25
ep

EDL10 t 1. 5 1676.42 EDL20 t 1 1034.56 EDL30 t 0. 5 446.373

c

EDL15 t 0.25 200.503 EDL20 t 1.25 1437.86 EDL30 t 0.75 737.632

Ac

EDL15 t 0.5 423.558 EDL20 t 1.5 1871.69 EDL30 t 1 1095.34

57
 EDL15 t 0.75 699.046 EDL25 t 0.25 206.104 EDL30 t 1.25 1553.88

EDL15 t 1 1008.51 EDL25 t 0. 5 441.172 EDL30 t 1.5 2136.41

t
ip
cr
us
an
M
d
te
c ep
Ac

58
 Table 5. Maximum shear forces for MEDs

Base Shear Base Shear Base Shear

Model Model Model
(kN) (kN) (kN)

t
ip
MEDD10 t MEDD15 t MEDD25 t
367.323 1402.08 1565.81
0.5 1.5 1.25

cr
us
MEDD10 t MEDD20 t 0. MEDD25 t
571.759 542.072 1874.47
0.7 5 5 1.5

MEDD20
an t MEDD30 t
M
MEDD10 t 1 787.111 700.687
799.556 0.75 0.5
d

MEDD10 t MEDD30 t
te

1017.50 MEDD20 t 1 1078.55 1027.26

1.25 0.75
ep

MEDD10 t MEDD20 t
1224.78 1350.52 MEDD30 t 1 1404.91
c

1.5 1.25
Ac

MEDD15 t 0. MEDD20 t MEDD30 t

471.478 1633.36 1743.49
5 1.5 1.25

59
 MEDD15 t MEDD25 t MEDD30 t
706.522 615.383 2124.79
0.75 0.5 1.5

MEDD25 t

t
MEDD15 t 1 929.783 897.139

ip
0.75

cr
MEDD15 t
1175.35 MEDD25 t 1 1242.73

us
1.25

an
M
d
te
c ep
Ac

61
 Table 6. Equations of the fitted lines

Thicknes
Equation R2
s (cm)

t
ip
0.25 ED
Fmax  0.650L  190.16 0.9708

cr
0. 5 Fmax  1.668L  397.42 0.9869

us
Fmax  2.894L  653.86
0.75
an 0.9802
M
1 Fmax  5.750L  923.05 0.9971
d

1.25 Fmax  11.522L  1207.75 0.9993

Elliptical Damper

te
ep

1. 5 Fmax  22.333L  1446.04 0.9898

Elliptical
c

Thicknes
R2
Ac

Equation
s (cm)

Fmax  16.213D  215.14

Modified

Damper

0. 5 0.9954

61
 0.75 Fmax  20.032D  357.31 0.9942

1 Fmax  30.473D  481.64 0.9979

t
ip
1.25 Fmax  36.849D  633.56 0.9974

cr
Fmax  45.448D  742.93

us
1.5 0.9962

an
M
d
te
ep
c
Ac

62
 Table 7. Equivalent viscous damping ratios for the frame samples.

Model Ah (kN.m) Vm (kN ) m (m) %eq

t
SPS4′-ED 30.052 449.505 0.038 28.00

ip
cr
SPS4-MED 24.684 297.898 0.041 32.17

us
SPS4′- 49.572 524.261 0.045 33.44

MED an
M
d
te
c ep
Ac

63

View publication stats