97 ScissorJack
97 ScissorJack
97 ScissorJack
System
INTRODUCTION
The use of supplemental or damping devices to dissipate seismic energy in building
and bridge structures has gained increasing interest in the past decade. These devices,
commonly known as ‘dampers,’ exhibit either hysteretic behavior (e.g., yielding of met-
als, sliding friction) or viscoelastic/viscous behavior (e.g., fluid viscous dampers, solid
and fluid viscoelastic dampers). The underlying objective of implementing energy dissi-
pation devices in structural systems is to limit or eliminate damage to the structural
frame by dissipating most of the earthquake-induced energy which would otherwise be
absorbed by the load-bearing-system through inelastic deformations. Viscoelastic and
viscous energy dissipation systems are also eminently suitable for wind vibration reduc-
tion. The interested reader is referred to the following for a review of this technology:
Applied Technology Council (ATC) (1997), Soong and Dargush (1997), Constantinou
et al. (1998), and Hanson and Soong (2001).
Today, many countries utilize various types of damping devices as protective systems
to reduce wind and earthquake-induced vibrations in new and retrofit construction. In
Japan, the majority of the applications utilize yielding steel devices and viscoelastic fluid
or solid devices. In the United States, engineers have primarily used fluid viscous de-
vices. In these applications, damping devices have either been installed in-line with di-
agonal bracing or as horizontal elements atop chevron bracing. The popularity of these
a)
Graduate Research Assistant, Department of Civil, Structural and Environmental Engineering, University at
Buffalo, State University of New York, Buffalo, NY 14260
b)
Professor and Chair, Department of Civil, Structural and Environmental Engineering, University at Buffalo,
State University of New York, Buffalo, NY 14260
133
Earthquake Spectra, Volume 19, No. 1, pages 133–158, February 2003; © 2003, Earthquake Engineering Research Institute
134 A. N. ŞIGAHER AND M. C. CONSTANTINOU
configurations is based on the engineers’ familiarity with such bracing systems and the
fact that all experimental research studies have utilized only these two configurations for
energy dissipation systems.
Exception to the rule of use of diagonal or chevron bracing configurations for damp-
ing systems has been the recent construction of two 37-story and one 38-story buildings
in the United States utilizing the toggle-brace configuration. As the name implies, this
configuration is based on the toggle mechanism, which magnifies the damper displace-
ment for a given interstory drift. This amplification results in a reduction in the required
damping force, and reduction in the damper volume, which may lead to reduction of
damper cost. The damper force output is magnified through the same mechanism and
delivered to the framing system. The toggle-brace configuration is suitable for applica-
tions of wind response reduction and of seismic hazard mitigation of stiff structures. A
theoretical treatment of the system’s behavior, along with experimental results confirm-
ing the validity of the concept and the developed theory, and a brief description of ap-
plications can be found in Constantinou et al. (2001).
An additional consideration related to the application of energy dissipation systems
is that in many cases the energy dissipation assemblies occupy entire bays in frames and
often violate architectural requirements such as open space and unobstructed view. With
this intent, the scissor-jack-damper system was developed as a variant of the toggle-
brace-damper system. The system combines the displacement magnification feature with
small size, which is achieved through compactness and near-vertical installation.
This paper presents the concept underlying the scissor-jack-damper configuration
and verifies the theory via experimental results. The experimental study includes earth-
quake simulator testing of a half-scale steel model structure equipped with fluid viscous
dampers. Parts of the work described in this paper have been previously presented or
briefly described by the authors and coworkers (Whittaker and Constantinou 1999a,
1999b, 2000; Constantinou et al. 2000; Constantinou 2000; Constantinou and Şigaher
2000). The scissor-jack system is also briefly described in the recent EERI monograph
of Hanson and Soong (2001). However, this paper represents the first detailed treatment
of the scissor-jack-damper system and the related experimental results, which represent
part of the doctoral dissertation of the first author.
SCISSOR-JACK-DAMPER THEORY
The scissor-jack-damper system is best explained by first reviewing the conventional
diagonal and chevron brace configurations, in which the displacement of the energy dis-
sipation devices is either less than (case of diagonal brace) or equal to (case of chevron
brace) the drift of the story at which the devices are installed. If u and uD denote the
interstory drift and the damper relative displacement, respectively, then
uD!f•u (1)
where f!magnification factor. For the chevron brace configuration, f!1.0; for the di-
agonal configuration f!cos !, where !!angle of inclination of the damper with respect
to the horizontal axis. The force FD along the damper axis is similarly related to F, the
horizontal component of the damper force exerted on the frame, through
SCISSOR-JACK-DAMPER ENERGY DISSIPATION SYSTEM 135
F!f•FD (2)
Figure 1 illustrates a single-story structure with diagonal and chevron brace configu-
rations. Also shown in the figure are the force, F, and the interstory drift, u. Consider
that this single-story structure has an effective weight, W, and a fundamental period un-
der elastic conditions, T, and that it is equipped with a linear fluid viscous damper for
which
FD!Co•u̇D (3)
136 A. N. ŞIGAHER AND M. C. CONSTANTINOU
where Co!damping coefficient, and u̇D!relative velocity between the ends of the
damper along its axis. The damping force F, exerted on the frame by the damper assem-
bly is given by
Co•f 2•g•T
4•#•W
"! (5)
It is essential to realize the effect of the magnification factor on the damping ratio.
As Equation 5 suggests, the damping ratio varies proportionally with the square of the
magnification factor. In the two conventional configurations in Figure 1, a damper de-
signed to provide a damping ratio of 5 percent of critical when installed horizontally
(chevron brace), will provide a damping ratio of 3.2 percent of critical in the diagonal
configuration.
In contrast to the familiar diagonal and chevron brace configurations, the scissor-jack
configuration can achieve magnification factors substantially greater than unity. This is
also true for the toggle-brace-damper systems. Figure 1 also illustrates the scissor-jack-
damper and toggle-brace-damper systems as implemented in a single-story frame. These
systems make use of shallow trusses that amplify the effect of the interstory drift on the
damper displacement and also amplify the small damper force and deliver it to the struc-
tural frame. The expression for the magnification factor, f, under the assumption of
small rotations, and its value for a typical geometry are also given in Figure 1. A sub-
stantial increase in the damping ratio with respect to that provided by conventional
damper configurations demonstrates the efficacy of these systems.
The presence of the magnifying mechanism in the scissor-jack system extends the
utility of fluid viscous devices to cases of small interstory drifts and velocities, which are
typical of stiff structural systems under seismic excitation and structures subjected to
wind load. Fluid viscous devices require special detailing when operating at small
stroke. This results in an increased volume of the device and, accordingly, cost. In ad-
dition, in stiff structural systems, the required damping forces are large, leading to a fur-
ther increase in the cost of the energy dissipation system. These damper configurations,
therefore, may lead to cost savings, provided that the cost of the support framing is not
substantially greater than the cost of the framing that would be required to support the
dampers in conventional configurations. Moreover, the scissor-jack system may be con-
figured to allow for open space, minimal obstruction of view and slender configuration,
features that are often desired by architects.
The scissor-jack damping system may be installed in a variety of configurations as
shown in Figure 2.
SCISSOR-JACK-DAMPER ENERGY DISSIPATION SYSTEM 137
uD !A!B!"AB!
f! ! (6)
u u
where AB and A!B! denote the initial and the deformed lengths of the damper, respec-
tively.
It should be noted that the deformed configuration of Figure 3 does not take into
account any deformations in the frame (only rigid body motion is considered) and any
reduction in height due to column rotation. The latter has negligible effect on the mag-
nification factor for typical values of interstory drift and for low-rise structures. The
frame deformations, on the other hand, may have notable effect on the magnification fac-
tor. As an example, consider that the beam in Figure 3 is simply connected to the column
on the left and rigidly connected to the column on the right. Upon an interstory drift
138 A. N. ŞIGAHER AND M. C. CONSTANTINOU
towards the right, the beam will deflect upwards, causing a decrease in the damper de-
formation and thus, in the magnification factor. The opposite will occur when the beam-
to-column connections are reversed. This type of amplification/deamplification of the
magnification factor will depend on the relative stiffnesses of the beam and the column,
and the position of the point of connection of the scissor-jack on the beam. Additionally,
the displacements due to the forces in the damper and in the scissor-braces (i.e., dis-
placements due to finite stiffness of the scissor-braces) will reduce the magnification
factor, regardless of the structural system configuration. A treatment of the effect of the
deformations in the frame and in the damper assembly, on the behavior of the frame will
be presented in sections that follow.
Based on rigid body kinematics, the damper displacement may be expressed as
uD!!A!B!"AB!!#2•!1•$sin%!#&!'"sin !( (7)
where &!!angle of rotation of the scissor-braces. Preservation of lengths between
points C and D requires that
SCISSOR-JACK-DAMPER ENERGY DISSIPATION SYSTEM 139
" # $
uD!#2•!1• sin cos"1 cos !$
cos )
2•!1 %& '
•u "sin ! (9)
where in Equations 7 to 9, positive signs hold for drift towards the right (u and &! as
shown in Figure 3) and for damper extension (uD%0). For drift towards the left, these
equations are valid with negative signs.
Equation 9 may be used to calculate the damper displacement given a value of drift,
provided the latter is small. However, the equation cannot be solved for the ratio of two
displacements, which is of much practical value. Realizing that for most applications &!
is very small, Equations 7 and 8 may be significantly simplified to easily yield the mag-
nification factor
cos )
f! (10)
tan !
It can be shown that Equations 1 and 10 provide a very good approximation to the exact
damper deformation (Equation 9) for &!*0.2!. Moreover, Equation 10 provides insight
into the major factors affecting the performance of the scissor-jack configuration.
The dependence of the magnification factor, f, on angles ! and ) is illustrated in
Figure 4. As the figure suggests, the magnification factor assumes very large values as !
approaches 0"; but this has no meaning since the scissors tend to act as a single brace
inclined at angle ). Rather, when designing such systems, emphasis should be placed on
the fact that the magnification factor should have minimal sensitivity to small changes
140 A. N. ŞIGAHER AND M. C. CONSTANTINOU
1
T! •F (11)
2•sin ! D
where T and FD denote the forces in the brace and damper, respectively. Note that the
forces T are greater than the force FD by a factor of 1/2•sin ! due to the shallow truss
configuration of the scissor-braces. The resultant of the horizontal component of forces
T equals force F, that is,
F!2•T•cos !•cos ) (12)
Equation 12 together with Equation 11, result in Equation 2. That is, the magnification
factor can be written as
uD F
f! ! (13)
u FD
which proves the accuracy of the analysis presented.
and , is the frequency of free vibration of the damped structure. It should be noted that
for infinitely stiff bracing (+ j!0), the energy dissipation system behaves as a pure vis-
cous system and Equations 14 to 16 reduce to Equation 4.
The approximate energy method starts with the assumption that the frequencies and
mode shapes of the nonclassically damped structure are identical to those of the un-
damped structure with the added effect of storage stiffness, but not of damping, from the
energy dissipation assembly. Thus, the frequencies and mode shapes can be determined
from standard eigenvalue analysis. Assuming that the structure undergoes vibration in
the kth mode with period Tk (or with frequency of vibration in the kth mode equal to ,k),
the damping ratio at the jth story may be expressed (Constantinou and Symans 1992,
ATC 1997, FEMA 2000) as
142 A. N. ŞIGAHER AND M. C. CONSTANTINOU
where T!k is the kth mode period of the undamped structure and K j is the horizontal stiff-
ness of story j. Moreover,
where parameter + was assumed to be the same for all stories of the structure, or + is an
average representative value for all stories. Equation 20 is implicit in the calculation of
period Tk , requiring an iterative procedure to solve. However, given the approximate na-
ture of the calculation, the following equation, representing the first iteration result, may
be used.
EXPERIMENTAL PROGRAM
The scissor-jack-damper system was first tested in a frame connected to a strong
floor and subjected to an imposed displacement history, and then in a model structure on
the earthquake simulator at the University at Buffalo. The model was the half-scale
steel-framed structure, which was previously utilized for testing of the toggle-brace-
damper system (Constantinou et al. 2001). The model structure consisted of two identi-
cal frames that could be tested individually on the strong floor or together on the earth-
quake simulator with an added mass on the top of the frames. Figure 5 illustrates one of
the two tested frames with the scissor-jack-damper system. A view of the structure on
the earthquake simulator is presented in Figure 6. The model features the following char-
acteristics:
1. Beam-to-column connections of the model frames were easily convertible from
simple to rigid, by bolting stiffened angles to the flanges of the beam and col-
umn (see Figure 5). This enabled testing with one rigid and one simple connec-
tion per frame (referred to as rigid-simple or simple-rigid configurations) and
with two rigid configurations per frame (rigid-rigid configuration). The rigid-
simple, simple-rigid, and rigid-rigid configurations were tested in the strong
floor experiments, whereas the earthquake simulator testing included only rigid-
simple and simple-rigid configurations.
2. The scissor-braces were connected to the frame (scissors-to-beam and scissors-
to-column connections) utilizing plates, which were designed to undergo mainly
rotation. As shown in the detail of Figure 5, these plates were designed with
sufficient length to prevent inelastic action. The damper-to-brace connections
were designed as true pins to avoid transfer of bending forces to the damper.
However, this was not fully accomplished because of tight pin configuration that
exhibited considerable friction.
3. The concrete weight used for earthquake simulator testing comprised of two
blocks weighing a total of 142.3 kN, and was secured atop the columns by way
of simple connections. The center of mass of these blocks was 1,113 mm above
the centerline of the beam.
4. A total of three viscous fluid dampers were utilized for the floor and earthquake
simulator testing. Shown in Figure 7 are the peak force-peak velocity charac-
teristics of the dampers (shown by triangular symbols), extracted from harmonic
testing. It follows that for damper 1 (used in strong floor testing), the behavior
is practically linear, which can be described by Equation 3 with Co
!25.8 N-s/mm for velocities up to 500 mm/s. Dampers 2 and 4 (used in earth-
quake simulator testing) exhibit linear behavior for velocities to 250 mm/s (with
Co!40.0 N-s/mm in Equation 3), after which the behavior deviates towards
nonlinearity. For these dampers, the overall behavior can be described via FD
!CNo•u̇D0 , where CNo!137.3 N-(s/mm)0 and 0!0.76. All dampers were of the
run-through rod type construction, and had a length of 273 mm, a diameter of
44.5 mm, and a stroke of #28.6 mm.
144 A. N. ŞIGAHER AND M. C. CONSTANTINOU
Figure 6. Model with scissor-jack damping system on the Buffalo earthquake simulator.
Figure 7. Peak force versus peak velocity relation of tested fluid viscous dampers.
mations under the action of forces in the scissor-jack system. Although the
damper forces are low (force FD in Figure 3), the resultant force on the frame
(resultant of forces T in Figure 3 equals FD /tan !) is large (for the tested con-
figuration, !!9" so that FD /tan !!6.3 FD) resulting in deflection of the beam.
2. Under quasi-static conditions, the magnification factor is higher than predicted
by theory. For the tested frame, uD!f !•u, where f !/f•(h/hs), and h/hs is a factor
accounting for the geometry in which the vertical projection, h, of the scissor-
jack is less than the story height, hs , where drift, u, occurs (see Figure 5). In this
case, h/hs!0.838 and f/1.8; whereas testing under quasi-static conditions re-
vealed f/2.9 for negative drift (damper undergoes compression) and f/2.5
otherwise. The difference between the predicted and the observed magnification
factors can be explained by the fact that with the beam-to-column connections
configured as rigid-simple, part of the damper deformation is caused by vertical
deflection of the beam (for either direction of drift)—a factor not accounted for
in theoretical predictions. In addition, under negative drift (towards the left), as
SCISSOR-JACK-DAMPER ENERGY DISSIPATION SYSTEM 147
the scissor-braces close, the angle ! decreases, causing an increase in the in-
stantaneous magnification factor, which results in a larger value of f (see Figure
4).
3. Under dynamic conditions, for which considerable forces develop in the scissor-
jack assembly, the magnification factor attains values of about 1.6 to 1.9, de-
pending on whether the angle ! increases or decreases, respectively. The sub-
stantial reduction of the magnification factor from the values attained under
quasi-static conditions is due to deformations of the energy dissipation assem-
bly (primarily beam deflections) caused by the damping forces.
4. There is considerable increase in the effective stiffness of the frame when sig-
nificant damping forces develop, as it is evident in the hysteresis loops of the
top graph in Figure 8. These loops show a 60 percent increase in effective stiff-
ness, which corresponds to about 25 percent increase in frequency. This is con-
sistent with the predictions of Equation 21 and experimental results to be pre-
sented later in this paper.
148 A. N. ŞIGAHER AND M. C. CONSTANTINOU
Figure 9. Amplitude of transfer function of model structure with rigid-simple and simple-rigid
connections.
Fundamental
Beam-to-Column Frequency Damping
Connections Configuration (Hz) Ratio
No Dampers 3.2 0.035
Rigid-Simple Scissor-Jack-Damper 4.0 0.130
No Dampers 3.2 0.031
Simple-Rigid Scissor-Jack-Damper 3.5 0.055
tems. Accordingly, the frequency and damping ratio on the basis of the assumption of
linear elastic and linear viscous behavior may be easily determined. They are presented
in Table 1 for each of the four tested configurations.
An observation to be made in the results of Table 1 and Figure 9 is the significant
difference in the added damping in the two configurations of frames, of which the origin
has been previously explained. Another important observation is the significant stiffen-
ing of the structure, marked by the increase in frequency. For the rigid-simple configu-
ration, the increase in frequency from 3.2 to 4.0 Hz is significant and consistent with the
approximately 60 percent increase in stiffness of the frame observed in testing under
imposed displacement. This increase in frequency is the result of viscoelastic behavior
caused by frame and energy dissipation assembly deformations under the action of the
damping forces.
Peak earthquake
simulator motion
Damper Damper
Displ. Veloc. Accel. Drift Accel. displ. force Scissor Magnification
Excitation (mm) (mm/s) (g) (mm) (g) (mm) (kN) damper factor
earthquake simulator displacement, velocity and acceleration, and peak frame response
in terms of drift (displacement of the beam-to-column joint with respect to the simula-
tor), beam-to-column joint acceleration, damper displacement, and damper force. The
peak frame output values are the average of the two quantities measured at east and west
frames. The two values differ slightly due to accidental asymmetry caused by slight
variations in stiffness of the two frames, and in damper properties.
The table also provides information on the measured range of values of the magni-
fication factor for positive and negative directions of drift. As noted earlier, the magni-
fication factor is dependent on the direction of movement due to changes in the geom-
etry of the scissor-jack system. Measured values of the magnification factor lie in the
range of 1.5 to 2.0 and are dependent on the excitation type. By comparison, the theo-
retical value is 1.8, whereas values measured in the testing of the frame under imposed
harmonic displacement varied between 1.6 and 1.9.
SCISSOR-JACK-DAMPER ENERGY DISSIPATION SYSTEM 151
Figure 10. Peak response of model structure as a function of peak earthquake simulator accel-
eration: peak drift ratio (top) and normalized peak structural acceleration (bottom).
The results of Table 2 clearly demonstrate that the scissor-jack system operates as an
effective damping system. Drift is substantially reduced. For example, observe the re-
sults in the El Centro motion. Without the scissor-jack system, the structure (which is
essentially elastic with damping ratio of about 0.04) undergoes drift of 14.3 mm. The
damped structure undergoes less than half that drift when excited by the El Centro mo-
tion scaled up by a significant amount. Interestingly in this case of elastic response, the
peak-recorded acceleration is also smaller in the damped structure despite the stronger
input. Similarly, the reader may want to compare the responses of the undamped and
damped structures in the Mexico City motion. For the same input, the damped structure
undergoes 50 percent lesser drift and 40 percent lesser acceleration.
A comparison of the performances of the structure without and with the damping
system is presented in Figure 10 (using circular and triangular symbols, respectively),
where the recorded peak drift ratio (peak drift—from Table 2—divided by the height of
the tested structure) and normalized peak structural acceleration (peak acceleration of
beam-to-column joint divided by peak earthquake simulator acceleration—both from
Table 2) are plotted against the peak earthquake simulator acceleration. The latter may
be regarded as representative of the intensity of the seismic excitation given the low pe-
riod of the tested model. The benefits offered by the damping system are clearly evident
in this figure: lower drift and lower acceleration for a given intensity of seismic excita-
152 A. N. ŞIGAHER AND M. C. CONSTANTINOU
tion. These benefits are typical of what damping systems may offer—the intent in show-
ing these graphs being to demonstrate the equivalence of the scissor-jack-damper con-
figuration to more conventional configurations.
Of interest is to discuss the effect of the stiffening of the structure on the reduction
of response. The reduction in displacement response is certainly the combined result of
stiffening of the structure and of increased damping. However, the reduction of accel-
eration response is primarily the result of increased damping. It should be noted that the
model structure is stiff with a fundamental period that falls within the acceleration-
sensitive region of the spectrum for all of the earthquake motions used in the testing, for
which, reductions in period (stiffening) do not result in reduction of acceleration.
SIMPLIFIED ANALYSIS
Response history analysis represents the best means of calculating the seismic re-
sponse of a structure with the scissor-jack system. Illustrated in Figure 12a is the com-
plete structural representation of the tested model that was used in the response history
analysis reported in the previous section with the exception that the damper is linear vis-
cous as described by Equation 3. This representation may be simplified for ease in the
dynamic analysis by replacing the scissor-jack assembly by an equivalent spring and
dashpot system as shown in Figure 12b. Note that the quantity Ka represents the stiffness
SCISSOR-JACK-DAMPER ENERGY DISSIPATION SYSTEM 153
Figure 11. Comparison of analytical (SAP2000) and experimental response of model structure
with rigid-simple beam-to-column connections for: 1940 El Centro earthquake, PGA!0.34 g
(left), and 1994 Northridge earthquake, Sylmar record, component 90, PGA!0.60 g (right).
of the assembly only, which is typically very large. It is determined by the procedure
illustrated in Figure 12b. Note that in the calculation of stiffness, the damper is consid-
ered ‘‘locked’’ so that it acts as a spring with stiffness equal to that of the oil column in
the damper. The stiffness is given by Ar2•B/V, where Ar is the piston rod area, V is the
effective volume of fluid, and B is its bulk modulus. In the case of the damper in the
tested model, the stiffness of the locked damper was represented by a steel element hav-
ing a diameter of 10 mm and length equal to that of the damper.
Simplified analysis is based on the premise that a linear elastic and proportional lin-
ear viscous representation of the structural system produces estimates of the seismic re-
sponse that are of acceptable accuracy. A discussion on the subject may be found in
Hanson and Soong (2001); several examples of application of simplified methods of
analysis and evaluation of the accuracy of the methods may be found in Ramirez et al.
(2000). Herein, we concentrate on the prediction of the fundamental period and associ-
ated damping ratio of the tested model using simplified methods of analysis described by
Equations 17 and 20.
154 A. N. ŞIGAHER AND M. C. CONSTANTINOU
energy dissipation system, and KS and CS are the stiffness and damping constant, respec-
tively, contributed by the scissor-jack-damper system (two scissor-jacks) inclusive of the
effect of its interaction with the frame. These parameters are given by
KS!2•Kb•cos2 )• # &
.r1 2
.1
/2•Kb•cos2 )•
h
H #& 2
(22)
CS!2•
tan2 !
•
.1 # &
Co•cos2 ) .r1 2
/2•
Co•cos2 ) h
tan2 !
•
H #& 2
(23)
where .r1 is the relative modal horizontal displacement of the two ends of the scissor-
braces and .1 is the modal displacement of the center of mass of the concrete block. For
the model structure, .r1 /.1/h/H, where heights h and H are identified in Figures 4 and
12a.
Analysis of the single-degree-of-freedom representation of Figure 12d is itself com-
plicated, given the viscoelastic nature of the system. For example, calculation of the pe-
riod and damping ratio requires complex eigenvalue analysis (Constantinou and Symans
1992). Simplified analysis would require a further step of replacing the Maxwell element
in Figure 12d by an equivalent Kelvin element of stiffness k! and damping constant c!
(Constantinou et al. 1998), as shown in Figure 12e, where
CS•+•,2
1&+2•,2
k!! (24)
CS
1&+2•,2
c !! (25)
CS Co
KS tan !•Kb
+! ! 2 (26)
Parameter ,!2•#/T1 is the frequency of vibration of the damped structure. Note that
parameter + is the relaxation time, which also appears in Equations 20 and 21.
On the basis of the representation shown in Figure 12e, the period, T1 , and damping
ratio, "1 , of the damped structure are
"1!
c!•T1 Co•cos2 )•T1
2 •
4•#•m 2•#•m•%1&+ •, '•tan ! H
! 2 2
h
#& 2
(28)
where T1! is the period of vibration of the structure exclusive of the energy dissipation
system,
156 A. N. ŞIGAHER AND M. C. CONSTANTINOU
T1!!2•#•
m
"'
KF
1/2
(29)
Equations 27 and 28 are identical to Equations 20 and 17, respectively, for the model
structure, with the exception that .r1 /.1 appears instead of h/H.
For the purpose of performing calculations Co!40.0 N-s/mm was used, a value
which represents well the behavior of the dampers for velocities less than about 250
mm/s (see Figure 7). Equation 26 results in +!0.046 s. Period T1!!0.31 s based on the
model identification (frequency of 3.2 Hz). Moreover, .r1 /.1/h/H!0.68 and Equations
27 and 28 are iteratively solved to result in T1!0.27 s and "1!0.12. Note that the period
of 0.27 s corresponds to a frequency of 3.7 Hz, which is less than the identified value of
4.0 Hz (see Table 1) but sufficiently close for practical purposes. Also, the calculated
damping ratio of 0.12 represents the added value supplied by the damping system. Since
the inherent damping was of the order of 0.04, the total damping ratio is about 0.16.
Identification of the structure showed a total damping ratio of 0.13 (see Table 1). Sim-
plified analysis predicts well the dynamic properties of the model.
CONCLUSIONS
Scissor-jack-damper energy dissipation systems offer a new opportunity to apply
damping systems. As the toggle-brace-damper system that preceded its development, the
scissor-jack-damper system utilizes shallow trusses to magnify the damper displacement
for a given interstory drift, and to magnify the damper force output delivered to the
structural frame. The system thus extends the applicability of damping devices to cases
of small interstory drifts, such as stiff structures under seismic loading and structures
subjected to wind loading. Moreover, the scissor-jack damping system can be configured
to allow for open space through its compactness and near-vertical installation, a feature
that is often desired for architectural purposes.
This paper presented a theoretical treatment of the scissor-jack-damper system and
experimental results that demonstrated its effectiveness. Testing of a half-scale steel
model structure on the earthquake simulator indicated a significant increase in the damp-
ing ratio, accompanied with reduction in drift and acceleration responses. The scissor-
jack-damper system also caused stiffening of the structure, marked by the increase in
frequency. This viscoelastic behavior occurred as a result of frame and energy dissipa-
tion assembly deformations under the action of damping forces.
The response of the model structure was reproduced analytically with acceptable ac-
curacy by response history analysis. The analytical model satisfactorily captured signifi-
cant characteristics of the model such as the stiffening effect, and peak values of drift
and acceleration. In addition, the application of simplified analysis methods for predict-
ing period and damping ratio of the model structure was presented. Simplified analysis
requires a proper representation of the increase in stiffness of the structure due to the
damping system. Results of this analysis were in close agreement with those of the ex-
periments.
SCISSOR-JACK-DAMPER ENERGY DISSIPATION SYSTEM 157
ACKNOWLEDGMENTS
Financial support for this project was provided by the Multidisciplinary Center for
Earthquake Engineering Research (MCEER), tasks on Rehabilitation Strategies for
Buildings and Experimental Facilities Network, and by Taylor Devices, Inc., N.
Tonawanda, New York. Taylor Devices also manufactured the damping devices. The au-
thors thank Professor Andrew S. Whittaker for his constructive comments.
REFERENCES
Applied Technology Council (ATC), 1997. NEHRP Guidelines for the Seismic Rehabilitation of
Buildings, (FEMA 273 Guidelines and FEMA 274 Commentary), prepared for the Building
Seismic Safety Council, published by the Federal Emergency Management Agency, Wash-
ington, DC.
Computers and Structures Inc., 2000. SAP2000: Static and Dynamic Finite Element Analysis of
Structures (Nonlinear Version 7.40), Computers and Structures Inc., Berkeley, CA.
Constantinou, M. C., 2000. Damping systems for structural design and retrofit, Presentation,
EERI 52nd Annual Meeting, St. Louis, MO.
Constantinou, M. C., and Şigaher, N., 2000. Energy dissipation system configurations for im-
proved performance, Proceedings of the 2000 Structures Congress & Exposition, ASCE,
Philadelphia, PA.
Constantinou, M. C., Soong, T. T., and Dargush, G. F., 1998. Passive energy dissipation systems
for structural design and retrofit, Monograph No. 1, Multidisciplinary Center for Earthquake
Engineering Research, Buffalo, NY.
Constantinou, M. C., and Symans, M. D., 1992. Experimental and Analytical Investigation of
Seismic Response of Structures with Supplemental Fluid Viscous Dampers, Technical Report
NCEER-92-0032, National Center for Earthquake Engineering Research, Buffalo, NY.
Constantinou, M. C., Tsopelas, P., and Hammel, W., 1997. Testing and modeling of an improved
damper configuration for stiff structural systems, Center for Industrial Effectiveness, Uni-
versity at Buffalo, SUNY, Buffalo, NY, 1http://civil.eng.buffalo.edu/users–ntwk/2.
Constantinou, M. C., Tsopelas, P., Hammel, W., and Şigaher, A. N., 2000. New configurations
of fluid viscous dampers for improved performance: Proceedings of the Passive Structural
Control Symposium 2000, Tokyo Institute of Technology, Yokohama, Japan.
Constantinou, M. C., Tsopelas, P., Hammel, W., and Şigaher, A. N., 2001. Toggle-brace-damper
seismic energy dissipation systems, J. Struct. Eng. 127 (2), 105–112.
Federal Emergency Management Agency (FEMA), 2000. NEHRP Recommended Provisions for
Seismic Regulations for New Buildings and Other Structures, (FEMA 368), Washington, DC.
Hanson, R. D., and Soong, T. T., 2001. Seismic Design with Supplemental Energy Dissipation
Devices, Monograph No. 8, Earthquake Engineering Research Institute, Oakland, CA.
Ramirez, O. M., Constantinou, M. C., Kircher, C. A., Whittaker, A. S., Johnson, M. W., Gomez,
J. D., and Chrysostomou, C. Z., 2000. Development and Evaluation of Simplified Procedures
for Analysis and Design of Buildings with Passive Energy Dissipation Systems, Technical Re-
port MCEER-00-0010, Revision 1, Multidisciplinary Center for Earthquake Engineering Re-
search, Buffalo, NY.
Soong, T. T., and Dargush, G. F., 1997. Passive Energy Dissipation Systems in Structural En-
gineering, John Wiley & Sons Ltd., London (UK) and New York (USA).
158 A. N. ŞIGAHER AND M. C. CONSTANTINOU
Veletsos, A. S., and Ventura, C. E., 1986. Modal analysis of non-proportionally damped linear
systems, Earthquake Eng. Struct. Dyn. 14, 217–243.
Whittaker, A. S., and Constantinou, M. C., 1999a. Advances in supplemental damping systems
for civil infrastructure, Proceedings of the VIIth Mexican Conference on Earthquake Engi-
neering, Morelia, Mexico.
Whittaker, A. S., and Constantinou, M. C., 1999b. Supplemental damping for new and retrofit
construction, Revista de Ingenieria Sismica, Vol. 61, Sociedad Mexicana de Ingenieria Sis-
mica A. C., Mexico City.
Whittaker, A. S., and Constantinou, M. C., 2000. Fluid viscous dampers for building construc-
tion, Proceedings of the Passive Structural Control Symposium 2000, Tokyo Institute of
Technology, Yokohama, Japan.
(Received 26 December 2001; accepted 26 September 2002)