Conference PaperPDF Available

Strengthening an existing industrial building by optimally designed passive dampers under seismic and service loads

Authors:

Abstract and Figures

Vibration control systems can improve the building performance effectively, not only under earthquake excitation, but also under dynamic service loads. In this paper, it is aimed to evaluate the performance of a two story gable industrial building containing two cranes, both under seismic and cranes load. The structure is designed preliminary based on Iranian national codes. Afterward its responses under seismic and cranes load are evaluated by time history and stochastic analysis method respectively. In order to enhance the responses under both loading, three different optimally designed passive dampers including viscous damper, buckling restrained braced, and friction damper are compared. Building responses show that viscous damper works well in both loadings. However, friction damper only boost the performance of structure under seismic action, and there is no trace of effectiveness for cranes load; exactly the same as what is observed for buckling restrained braced system.
Content may be subject to copyright.
Strengthening an existing industrial building by optimally
designed passive dampers under seismic and service loads
Ali Khansefid1, Ali Bakhshi2
1 Graduate student of Sharif University of Technology, Tehran, Iran
2 Associate professor of Sharif University of Technology, Tehran, Iran
ABSTRACT: Vibration control systems can improve the building performance
effectively, not only under earthquake excitation, but also under dynamic service loads.
In this paper, it is aimed to evaluate the performance of a two story gable industrial
building containing two cranes, both under seismic and cranes load. The structure is
designed preliminary based on Iranian national codes. Afterward its responses under
seismic and cranes load are evaluated by time history and stochastic analysis method
respectively. In order to enhance the responses under both loading, three different
optimally designed passive dampers including viscous damper, buckling restrained
braced, and friction damper are compared. Building responses show that viscous damper
works well in both loadings. However, friction damper only boost the performance of
structure under seismic action, and there is no trace of effectiveness for cranes load;
exactly the same as what is observed for buckling restrained braced system.
Keywords: Industrial building, Stochastic analysis, viscous damper, BRB, Friction
Damper, Crane load.
1 INTRODUCTION
Nowadays beside the development of hi-tech factories, which are highly sensitive to the
vibrations, it is needed to build industrial buildings and factories that suppress the vibration noise
due to different type of loadings. In addition, the buildings shall withstand the natural hazardous
loading like probable earthquake or wind. As an effective method, it is applicable to use the
passive vibration control systems in order to enhance the response of structure. There is different
type of passive devices like viscous dampers, yielding dampers, friction dampers, tuned mass
dampers, and wide variety of base isolations each of which has their own advantages and
disadvantages. In this study, three more practical ones are selected to be studied: viscous damper,
buckling restraint braced (BRB), and friction damper.
There are many theoretical and practical projects conducted to evaluate the performance of these
damping devices separately. However, there are few comparative studies on these systems.
Symans et al. (2008) prepare a comprehensive review on the passive devices from both theoretical
and practical point of view. Fabio et al. (2013) assess the ability of passive energy damping
devices to improve the response of a 6 story reinforced concrete structure. Khansefid and
Ahmadizadeh (2015) also studied the effectiveness of passive and active vibration control devices
in nonlinear building under different seismicity level.
In this research, it is purposed to evaluate the effectiveness of different type of passive vibration
control devices including viscous damper, buckling restrained brace, and friction damper to
improve the response of a two story gable structure with two cranes. Two different loads including
earthquake, as well as handmade noise by the cranes are taken into account. Building responses
to the earthquake excitation are obtained by time history analysis method while the response due
to cranes noise is calculated by using stochastic analysis procedure. Results show that viscous
damper is the best system to reduce the cranes noise while the other two ones do not have any
positive effect. From the other side, viscous damper and friction damper both improve the seismic
behavior of the structure in the same level, far beyond the BRB.
2 BUILDING MODELING
A two story gable structure includes two separate cranes in the first and second story by weight
lifting capacity of 1250 and 320 KN respectively is considered. Building shape and dimensions
are illustrated in Figure 1. All the elements are made of steel with yielding capacity of 360 MPa.
The structure is designed based on the Iranian national seismic code (2013). Floor's dead and live
load are equal to 65 KN/m2 and 100 KN/m2. Obtained beams and columns from the design
procedure are all I-shape sections, their height varying from 240 to 600 mm. Bracing elements
are UNP 110 and 200.
Figure 1. Industrial building shape and dimensions, a) 3D view, b) Cross section, c) Longitudinal view.
3 LOADINGS
Industrial buildings located in the high vulnerable seismic area, shall be designed to withstand the
probable earthquake event during their lifetime. In addition to this major loading, in many cases,
movement of the cranes may cause very annoying noise which disturb the application of the other
parts of building. In this research, both foresaid loadings are considered to evaluate the structural
performance.
3.1 Seismic Load
Record selection is one of the most important source of uncertainty in seismic performance
evaluation of any system. To overcome this problem, 10 artificial earthquakes are produced. All
of these records are randomly generated by the SeismoArtif (2016) software and matched with
the Iranian seismic spectrum of soil type 2 for a site with very high seismicity level. The
information summery of artificial records are presented in Table 1, as well as their accelerograms,
and spectrums in Figure 3 and 4.
Table 1. Summary information of randomly generated accelerograms
Earthquake Index PGA(g) PGV(cm/s) PGD(cm) (s)
*
Strong Motion Duration
A01 0.40 39.0 14.0 12.69
A02 0.52 42.9 18.0 18.36
A03 0.47 52.6 24.8 18.98
A04 0.45 54.9 20.8 19.70
A05 0.43 38.7 15.9 19.08
A06 0.47 51.5 18.0 20.01
A07 0.52 48.7 14.0 20.03
A08 0.45 53.6 15.8 18.08
A09 0.42 55.5 27.9 22.27
A10 0.53 54.2 20.5 22.97
mean 0.46 49.1 19.0 19.21
* Strong motion durations are calculated by the method of Trifunac and Brady (1975).
Figure 2. Acceleration time series of randomly generated earthquake.
Figure 3. Acceleration response spectrum of Iranian national seismic code and the artificial accelerograms.
3.2 Non-seismic load (cranes load)
In the industrial building, the cranes may work at different frequencies. Hence, it is a good idea
to consider their loading as a white noise which contains all possible frequencies for the input
vibration imposed on all cranes rail. The white noise power spectral density function (PSD) is a
constant value. Moreover, it is known that the area under the PSD diagram is equal to the variance
of loading as presented by Lutes (2004):
=()

(1)
where S, w, and σ are PSD function, frequency range, and standard deviation of the loading
respectively.
The nominal capacities of the cranes weight lifting are 1250 KN, and 320 KN. To model the
cranes load, it is assumed that the coefficient of variation of their loading is equal to 0.1.
Therefore, the standard deviation of the first and second story cranes will be obtained 125 KN,
and 64 KN. Additionally, the cranes working frequency are presumed to varies from 0 to 200 Hz.
By using Eq. (1) it is found that the PSD of cranes are equal to 20.48 and 78.125 KN2/rad for 320
KN and 1250 KN capacity equipment, respectively.
4 PASSIVE DAMPING DEVICES
There are many different type of passive energy dissipating devices all around the world like
metallic dampers, friction dampers, viscous dampers, viscoelastic dampers, tuned mass dampers
and etc. In this research, among all of them, three more common types are considered including
viscous damper, buckling restrained brace, and friction damper. Viscous dampers are kind of
velocity-dependent devices while the other two ones are displacement-dependent. Each of them
has its own pros and cons from different aspect such as technical capability, installation,
monitoring, maintenance, as well as the price while in this research only the first one is compared.
The most important design parameter of passive dampers is their force-displacement relationship
presenting in Figure 4.
Figure 4. Idealized force-displacement relationship of passive dampers
These devices are used in two different positions. One set is mounted among the truss elements
of first floor deck (Figure 5-a) in order to reduce the floor vertical vibration due to the cranes load.
In each of the transverse frame, two separate devices are used which leads to the total number of
16. Another set of dampers are considered to be installed in two main longitudinal frame of
structure (Figure 5-b) to improve the seismic response of whole structure. As it is seen, there is 8
devices in each of the longitudinal frames. In the other words, in both frames 16 devices are
installed. Finally, 32 viscous dampers are located in the industrial building.
Figure 5. Location of viscous damper. a) Devices used to reduce the cranes vertical vibration, b) Devices
used to reduce the response of building due to earthquake excitation.
5 ANALYSIS PROCEDURE
Since there is two different type of loading on the structure, two separate analysis procedures are
considered in parallel. To calculate the seismic response of building a time history analysis
method is followed whereas the linear stochastic analysis is adopted to model the behavior of
floor deck of building due to cranes vibration.
Dampers design properties may be calculated by two different methods, code based as well as
optimization based. In this work, the second one is selected since there is no specific design code
for using passive energy dissipating device in the industrial building. Each of the three dampers
considered in this study has its own design properties, damping coefficient for viscous damper,
yielding force for BRB, and sliding force for friction damper. To find the optimal value of design
parameter, a sweeping optimization method is used. Accordingly, the best value is selected from
the wide possible range of design parameter. In this regard, for viscous damper 17 different
damping coefficient values including 2, 10, 50, 150, 250, 350, 500, 750, 900, 1000, 1250, 1500,
2500, 3500, 5000, 7500, 10000 KN.s/m; for BRB 12 different yielding forces including 6, 10, 15,
20, 50, 75, 100, 150, 200, 250, 300, and 400 KN; and for friction damper 12 different sliding
forces including 1, 2, 3, 4, 6, 8, 10, 12, 15, 20, 50, 100 are considered. All the dampers used in
the longitudinal frames are in the same type and size whereas the dampers installed in the floor
truss deck are the same as each other. Finally, 410 3-D elaborated finite element analysis are done
by SAP2000 (2016) software under earthquake excitation to find the best values of design
parameter of dampers located in longitudinal frame, while 41 analyses are done to acquire the
best design values for reducing the cranes vibration.
The most important point to reach to the best design value is to define the performance index for
the system. Average of inter-story drift and absolute acceleration (Eq.(2)) is defined as seismic
performance index and the vertical velocity (Eq.(3)) of deck floor is considered for the cranes
response.
 =
|||| 
||||  +
|||| 
|||| 
(2)
 =|||| 
|||| 
(3)
where
A
and
D
are the vectors of maximum absolute acceleration and inter-story drift response
of structure under seismic loading.
V
is the vertical velocity of deck floor due to cranes load
which is monitored at the center of floor deck.
6 RESULTS
In this part, firstly, the optimal value of dampers design parameter will be evaluated, and in the
next section responses of the best designs will be securitized to assess the effectiveness of
dampers on improving the performance of industrial structures under different type of loading.
As it is indicated in Figure 6, for the seismic response, the best damper properties for viscous
damping coefficient, BRB yielding force, and friction damper sliding force are 750 KN.s/m, 250
KN, and 12 KN respectively.
Figure 6. Optimization results of dampers for seismic loading.
For the cranes load, the output of stochastic analysis shows that there are three separate
frequencies at which the vertical vibration of deck floor is intensified. These frequencies are 14,
25, and 43 Hz. Among these three frequencies, 14 Hz is more close to the human comfort zone,
thus it is more critical than other two ones. As it is declared in Figure 7 either BRB or friction
damper, not only can improve the response of structure to cranes load but also in some frequencies
make it worse which is due to inactivation of dampers, i.e. dampers activation force does not
reach at all since the cranes load amplitude is not high enough. On the other hand, viscous damper
enhances the behavior of structure greatly insomuch as it does not need to activation, and works
even in very low level of loads. The best damping coefficient design value is selected equal to
5000 KN.s/m because the higher ones does not improve the performance of system tangibly.
Figure 7. Optimization results of dampers for different frequencies under cranes load.
6.1 Seismic response
In this section, average of seismic responses of industrial building with and without dampers for
all 10 earthquakes are presented. In Figure 8, response of inter-story drift, absolute story
acceleration, and base shear of building are illustrated. It is seen that all the devices improve the
behavior of structure. Among the studied systems viscous damper shows better performance. It
reduces the response of drift, acceleration and base shear equal to 35%, 83%, and 80%
respectively.
Figure 8. Average responses of building to all earthquake excitation.
The next important response of structures is the input seismic energy and its distribution in the
building. Energy distribution in the structure is calculated based on the relationship proposed by
Christopoulos and Filiatrault (2006) and the results are shown in Figure 9. It reveals that by using
dampers, not only the input seismic energy is reduced, but also the total strain and inherent
damping energy, as a representative of structural damage, is decreased significantly. Here again,
viscous damper is the best system among all. It reduces the input energy equal to 57%, and the
total strain and inherent energy up to 59% as well.
Figure 9. Input seismic energy and its distribution in the building for different type of dampers.
6.2 Response of floor deck
Here, the velocity response of floor deck is assessed owing to the cranes operation load. Figure
10, shows the maximum vertical velocity response of floor deck regarding to the crane operation
frequency. This is clearly seen that viscous damper reduces the peak velocity at all resonance
frequencies (14.0, 25.0, and 43.0 Hz) very well, almost up to 50 percent while the other two
systems do not.
In accordance to the AISC design guide No.11 (2003) for floor vibration, the allowable velocity
amplitude of floor vibration for buildings containing computer systems is equal to 200 μm. This
criterion shall be satisfied for all dynamic activities with frequencies ranging between 8 to 80 Hz.
By comparing the results of Figure 10, it is indicated that by using the viscous damper, velocity
response of floor deck drops down below the allowable threshold.
Figure 10. Velocity response frequency spectrum of floor deck with and without damper
7 CONCLUDING REMARKS
In this paper, the effectiveness of optimally designed passive dampers including viscous damper,
BRB, and friction damper on the behavior of industrial building with cranes are evaluated. In this
regard, two of the most important type of loading during the life of structure is taken into account
containing seismic load and cranes noise vibration, and imposed on both with and without damper
structure. By implementing stochastic analysis, it is observed that the operating cranes vibration
noise or in the other words the vertical velocity induced by the cranes is remarkably reduced
(about 50 percent) when viscous damper is used as a supplemental damping device. The other
two dampers do not improve the structural performance at all which is due to their inactivation in
low intensity loading. Moreover, viscous damper shows the best performance in seismic loading
as well. It reduces the building responses up to 80%. These results reveal that viscous damper can
greatly improve the structural performance under loading either with low intensity level like
service loads or high intensity ones like earthquake excitation. However, BRB and friction damper
cannot work under low level loads.
8 REFERRENCE
Building and Housing Research Center, 2013, Iranian code of practice for seismic resistance building
design, Tehran, Iran.
Christopoulos, C., A. Filiatrault, 2006, Principles of Passive Supplemental Damping and Seismic Isolation,
Italy: IUSS Press.
American Institute of Steel Construction, 2003, Floor Vibrations Due to Human Activity, Chicago, USA:
AISC.
Computers and Structures Inc., 2016, SAP2000 Integrated Software for Structural Analysis and Design,
Berkeley, California, USA.
Fabio, M., V. Alfonso, M. Mirko, and M. Giuseppe, 2013, Modeling and Nonlinear Seismic Analysis of
Framed Structures equipped with damped braces. 7th WSEAS International Conference on Computer
Engineering and Application, Milan, Italy.
Khansefid, A., M. Ahmadizadeh, 2015, An investigation of the effects of structural nonlinearity on the
seismic performance degradation of active and passive control systems used for supplemental energy
dissipation, Journal of Vibration and Control, 22(16): 3544-3554.
Seismosoft Ltd, 2016, SeismoArtif V.2.1.: Software applications for generating artificial earthquake
accelerograms, Pavia, Italy.
Lutes, L.D., S. Sarkani, 2004, Analysis of Structural and Mechanical Systems, Burlington, USA: Elsevier
Butterworth–Heineman.
Symans, M.D., A.S. Charney, M.C. Constantinou, M.W. Johnson, and R.J. McNamara, 2008, Energy
dissipation systems for seismic application: current practice and recent development. Journal of
Structural Engineering, 134(1): 3-21.
Trifunac, M.D., A.G. Brady, 1975, A study on the duration of strong earthquake ground motion, Bulletin
of the Seismological Society of America, 65(3): 581-626.
... The next important parameter is the ground motion duration of the accelerograms evaluated based on the Trifunac and Brady method (Trifunac and Brady 1975). This is one of the key parameters in the selection of accelerograms for the nonlinear dynamic analysis of structures (ASCE7-16, 2016, Khansefid andBakhshi 2017). As it is observed in Fig. 16, the larger the site-to-source distance, the higher the strong motion duration is. ...
Article
Full-text available
In this paper, a comprehensive accelerogram database of the Iranian plateau containing 3585 data with all three components is gathered. The raw data are processed by the wavelet-based denoising method, and results are compared with the contaminated data. All the data are classified into mainshock and aftershock categories using the time and spatial window method. Afterward, the data are categorized into the pulse-like and non-pulse-like events based on the detection of velocity pulse in any of horizontal and/or vertical directions. Eventually, among 3585 data, the ones with an average shear wave velocity of top 30 m of subsurface soil profile are selected and their important ground motion parameters such as peak ground acceleration, velocity, and displacement, Arias intensity, ground motion duration, and acceleration response spectra are studied. It is observed that the intensity of vertical component of the accelerograms is almost at the same level as the strong horizontal direction or even more, in some cases. Near-field events cause more severe responses than far-field earthquakes. Generally, the mainshocks responses are slightly tougher than the aftershocks. By comparing the acceleration response spectra of the accelerograms with the ones obtained from the building design codes, the necessity of revising the existing standard spectra is revealed for considering the effects of mainshocks, aftershocks, near-field effects, as well as separate vertical component of the ground motions directly. Finally, the presented database, as a powerful tool, will give an interesting opportunity to the scientists and engineers who are enthusiastic to work on the Iranian plateau accelerograms.
Article
Full-text available
It is generally accepted that active control systems provide better structural performance when compared to their passive counterparts. On the other hand, the design of active control systems based on linear control theory is highly dependent on the structural properties. For this reason, their performance is expected to be affected more severely by variations in structural properties compared to those of passive systems. These variations can occur due to nonlinear structural behavior, or even before that due to uncertainties in the estimation of these properties and in numerical modeling. The present work is an investigation of the dependency of various control systems used for supplemental energy dissipation to the changes in structural properties. For this purpose, the performance degradations of most common control systems are studied for structures entering their nonlinear response range. The considered control systems include active control using linear quadratic regulator algorithm and passive control using viscous fluid dampers and yielding devices. These control systems are optimized separately for linear and nonlinear structures by minimizing performance indices based on inter-story drift and absolute acceleration response. It is shown that among these control systems, viscous dampers are affected the least by the alteration of structural properties, which may further support their utilization contrary to the more costly active control systems. It is also shown that the amount of performance degradation depends on the selected structural performance index, and more importantly, the severity of the earthquake excitation.
Article
This paper presents a summary of current practice and recent developments in the application of passive energy dissipation systems for seismic protection of structures. The emphasis is on the application of passive energy dissipation systems within the framing of building structures. Major topics that are presented include basic principles of energy dissipation systems, descriptions of the mechanical behavior and mathematical modeling of selected passive energy dissipation devices, advantages and disadvantages of these devices, development of guidelines and design philosophy for analysis and design of structures employing energy dissipation devices, and design considerations that are unique to structures with energy dissipation devices. A selection of recent applications of passive energy dissipation systems is also presented.
Article
A simple definition of the duration of strong earthquake ground motion based on the mean-square integral of motion has been presented. It is closely related to that part of the strong motion which contributes significantly to the seismic energy as recorded at a point and to the related spectral amplitudes. Correlations have been established between the duration of strong-motion acceleration, velocity, and displacement and Modified Mercalli intensity, earthquake magnitude, the type of recording site geology, and epicentral distance. Simple relations have been presented that predict the average trend of the duration and other related param- eters as a function of Modified Mercalli intensity, earthquake magnitude, site geology and epicentral distance.
Iranian code of practice for seismic resistance building design
  • Housing Research Building
  • Center
Building and Housing Research Center, 2013, Iranian code of practice for seismic resistance building design, Tehran, Iran.
Modeling and Nonlinear Seismic Analysis of Framed Structures equipped with damped braces
  • M Fabio
  • V Alfonso
  • M Mirko
  • M Giuseppe
Fabio, M., V. Alfonso, M. Mirko, and M. Giuseppe, 2013, Modeling and Nonlinear Seismic Analysis of Framed Structures equipped with damped braces. 7th WSEAS International Conference on Computer Engineering and Application, Milan, Italy.
SeismoArtif V.2.1.: Software applications for generating artificial earthquake accelerograms
  • Seismosoft Ltd
Seismosoft Ltd, 2016, SeismoArtif V.2.1.: Software applications for generating artificial earthquake accelerograms, Pavia, Italy.
  • L D Lutes
  • S Sarkani
Lutes, L.D., S. Sarkani, 2004, Analysis of Structural and Mechanical Systems, Burlington, USA: Elsevier Butterworth-Heineman.