Вы находитесь на странице: 1из 13

Journal of Sound and Vibration 388 (2017) 141–153

Contents lists available at ScienceDirect

Journal of Sound and Vibration


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

Experimental investigations of tuned liquid damper-structure


interactions in resonance considering multiple parameters
Ali Ashasi-Sorkhabi a,n, Hadi Malekghasemi b, Amirreza Ghaemmaghami b,
Oya Mercan b
a
Department of Civil Engineering, University of Waterloo, 200 University Avenue West, Waterloo, ON, N2L 3G1, Canada.
b
Department of Civil Engineering, University of Toronto, Toronto, ON, Canada

a r t i c l e i n f o abstract

Article history: As structures are constructed more slender and taller, their vibrational response and its
Received 3 February 2016 mitigation become challenging design considerations. Tuned liquid dampers (TLDs) are
Received in revised form cost effective and low maintenance vibration absorbers that can be used to suppress
22 September 2016
structural vibrations. A TLD dissipates energy through liquid boundary layer friction, free
Accepted 20 October 2016
Available online 31 October 2016
surface contamination, and wave breaking. The dynamic characteristics of the TLD and its
interaction with the structure is quite complex. In this paper, using a state-of-the-art
Keywords: experimental testing method, namely real-time hybrid simulation (RTHS), a compre-
Tuned liquid damper hensive parametric study is conducted to investigate the effectiveness of TLDs. During
Real-time hybrid simulation
RTHS the TLD response is obtained experimentally while the structure is modeled in a
Vibration mitigation
computer, thus capturing the TLD-structure interaction in real-time. By keeping the
FVM/FEM method
TLD-structure interaction structure as the analytical model, RTHS offers a unique flexibility in which a wide range of
influential parameters can be investigated without modifying the experimental setup. The
parameters considered in this study with a wide range of variation include TLD/structure
mass ratio, TLD/structure frequency ratio, and structural damping ratio. Additionally, the
accuracy of FVM/FEM method that couples the finite volume and finite element ap-
proaches to model the liquid and solid domains to capture TLD- structure interaction is
assessed experimentally. Results obtained in this study, will not only lead to a better
understanding of TLDs and their interaction with the structures but also, contribute to the
enhanced design of these devices which will in turn result in their wide-spread appli-
cation.
& 2016 Elsevier Ltd. All rights reserved.

1. Introduction

1.1. Tuned liquid dampers (TLDs)

The need to construct flexible and slender buildings (such as skyscrapers) that have relatively low damping character-
istics has attracted engineers’ attention to look for efficient and economical ways to control the vibrational response of the
structures. Several passive energy dissipation devices such as buckling restrained braces (BRBs) [1], metallic dampers [2],
magneto-rheological (MR) fluid dampers [3,4], tuned liquid dampers [5,6], viscoelastic dampers [7], friction dampers [8],

n
Corresponding author.
E-mail address: aliashasi2010@gmail.com (A. Ashasi-Sorkhabi).

http://dx.doi.org/10.1016/j.jsv.2016.10.036
0022-460X/& 2016 Elsevier Ltd. All rights reserved.
142 A. Ashasi-Sorkhabi et al. / Journal of Sound and Vibration 388 (2017) 141–153

and shape memory alloy (SMA)-based reusable hysteretic dampers [9] have been introduced for hazard mitigation of tall
buildings under dynamic loads (e.g. wind or seismic loads).
Owing to their low maintenance requirement, cost effectiveness and ease of installation, tuned liquid dampers attracted
considerable attention [10,11]. Tuned liquid damper (TLD) is a liquid (usually water) filled tank that absorbs energy through
several mechanisms including liquid boundary layer friction, free surface contamination, and wave breaking. Rectangular
TLDs in one-directional motion have been investigated extensively [5,6,12–14]. When a TLD is subjected to motion with
large amplitudes, due to the horizontal component of the liquid velocity related to the wave motion, wave crests descend as
the amplitude increases. This is known as wave breaking. At this point simple linear models can no longer describe the
liquid behavior and wave breaking changes the sloshing frequency of the liquid [14]. Additionally this complicated nonlinear
phenomenon influences the shear force developed at the interface of the TLD with the structure which counteracts the
motion of the structure and it is challenging to model accurately. As another configuration of TLDs, Tuned liquid column
dampers (TLCDs) have been studied both numerically and experimentally by researchers to suppress structural vibrations
due to wind and seismic effects [15–18].
Mass ratio (the ratio of the mass of water to that of the structure) is an important parameter that affects the performance
of TLD-structure system. Mass ratios in the range of 1% [19,20] up to 4% [21] have been suggested in the literature. However,
experimental validation is lacking for higher mass ratios where previous studies on this topic consisted of pure numerical
simulations. With a relatively small mass ratio, without significantly contributing to the overall inertia of the system, TLDs
can provide appreciable structural displacement and acceleration reductions. The liquid sloshing frequency is another
parameter that plays an influential role in the TLD behavior. Previous experimental studies [19,22] have shown that the
effectiveness of the TLD is maximized when the liquid frequency is a value near to the excitation frequency where the liquid
is in resonance with the tank motion; however, the data available from past studies is scattered and limited to only a few
values of the influential parameters.

1.1.1. Modelling of TLDs and their interaction with structures


To better understand the role of TLDs in mitigating structural vibrations several analytical models with different levels of
complexity have been introduced. With the use of the potential flow theory, Ohyama and Fujji's [23] model could include
nonlinearity; however, computational time was an issue with this model [19]. Kaneko and Ishikawa [24] adopted an in-
tegration scheme to solve the continuity and Navier-Stokes equations without accounting for the wave breaking. Ra-
maswamy et al. [25] solved nonlinear Navier-Stokes equations using Lagrangian description of fluid motion along with finite
element method. There were some shortcomings associated with the model involving sloshing dynamics of inviscid and
viscous fluids. Frandsen [26] introduced a fully nonlinear 2-D s-transformed finite difference model based on inviscid flow
equations in rectangular tanks. In this model, the fluid equations were coupled to a linear elastic base structure. The model
was not able to capture damping effects of liquid and shallow water wave behavior. Sun et al. [19] proposed a model to solve
nonlinear Navier-Stokes and continuity equations that is able to capture the wave breaking as well as the damping effects of
the liquid. In these models the force developed by TLD was calculated fed back to the structural equation of motion as a force
input.
Another approach to model the dynamic behavior of TLD is to replace the TLD by its equivalent Tuned Mass Damper
(TMD) and find the effective TMD properties such as stiffness, damping ratio, and mass. Sunet al. [27] found equivalent TMD
properties based on the nonlinear Navier-Stokes equations and shallow water wave theory. However, the experimental
cases presented in this study are limited. Tait [28] developed an equivalent linear mechanical model that accounts for the
energy dissipated by the damping screens for both sinusoidal and random excitations. Yu [20,29] modeled TLD as a solid
mass damper that can capture nonlinear stiffness and damping of the liquid motion. Halabian and Torki [30] studied the
seismic interaction of tall buildings and TLDs with screens inside by replacing the TLDs with equivalent amplitude de-
pendent TMDs. In the above mentioned models, the equivalent TMD parameters were introduced into the structural
equation of motion assuming the TLD in the form of an additional degree of freedom.
The majority of the previous nonlinear models and experimental studies focused on understanding the behavior of
rectangular or circular TLDs. Love and Tait [31] developed a nonlinear multimodal model to describe the performance of a
liquid inside a flat bottom tank with an arbitrary geometry. In addition, the modal expansion techniques were employed to
conduct nonlinear simulation of the TLDs with damping screens [32]. Recently, Malekghasemi et al. [33] presented a new
analytical model, the Finite Volume/Finite Element (FVM/FEM) method, by using finite volume and finite element ap-
proaches to represent fluid and solid domains, respectively. In the FVM/FEM model the fluid and solid domains are dis-
cretized independently and the interaction between the two domains are provided by the staggered iterations at the in-
terface. The experimental results obtained from the RTHS tests in [54] were employed to calibrate the FEM/FVM method.
Then, the reliability of the proposed numerical method in predicting the TLD-structure interaction was experimentally
verified and the predicted results were compared to three simplified numerical methods from the literature. However, the
verification of the FVM/FEM method was limited to mass ratio of 1% only [33].

1.2. Objectives

This paper presents a comprehensive experimental study that is conducted to evaluate the dynamic interaction of TLD-
structure systems. In this study, using the state-of-the-art experimental testing technique (i.e., real-time hybrid simulation,
A. Ashasi-Sorkhabi et al. / Journal of Sound and Vibration 388 (2017) 141–153 143

RTHS), a wide range of influential parameters (namely, TLD/structure mass ratio, TLD/structure frequency ratio, and
structural damping ratio) are considered to explore the efficiency of the TLDs in suppressing structural vibrations under
different operational conditions. The experimental validation data available in the literature comes from different studies
with varying test specifications that only investigated a few values for some of the parameters. As such they don’t provide an
overarching database. The use of RTHS in this study addresses this need. It should also be pointed out that without RTHS,
this experimental investigation of TLD-structure interaction would have been very expensive and time consuming as several
different structural models with the TLD would need to be physically constructed and tested in the laboratory.
Additionally, in this study the reliability of the FVM/FEM model is experimentally validated for TLD/structure cases with
mass ratios greater than 1%.

1.3. Finite volume/finite element method (FVM/FEM)

Modeling the sloshing response of the TLD requires robust and accurate computational tools due to highly nonlinear
fluid motions especially under large-amplitude dynamic loading. Simplified computational approaches such as potential
flow methods are inadequate in simulating general sloshing cases. There are two main methods available in the literature
which are able to handle complex free-surface behavior: 1) Traditional grid-based methods such as finite difference
methods (FDM), finite volume methods (FVM) and the finite element method (FEM) which have been widely used in various
areas of computational fluid dynamics (CFD) and computational solid mechanics (CSM) [34,35] 2) Mesh-free methods which
are able to provide accurate numerical solutions using a set of arbitrary distributed nodes or particles. Smoothed particle
hydrodynamics (SPH) method is a mesh-free particle method which has been successfully applied to sloshing problems
with very steep deformations of free-surface including breaking waves [36–39] and [40].
This study employs a grid-based FVM/FEM method is for the numerical simulation purpose. The main features of the
FVM/FEM method are briefly introduced in this section. FVM/FEM method captures the free surface fluid motion inside the
TLD as well as the interaction between the fluid and the structure. The fluid motion is governed by the Navier-Stokes
equations. The solution domain is regarded as having two phases: air and water with negligible surface tension. The physical
domain is treated as homogeneous where both fluid phases share the same velocity field and separation never occurs.
In this study, an element-based finite volume method (FVM) using CFX modulus (ANSYS 2006) was employed. The
fundamentals of FVM can be found in the works done by Versteeg and Malalasekara [41] and are not repeated here. The
solution procedure involves the discretization of the solution domain into a number of control volumes over which the
governing equations are numerically integrated. This results in a set of linear algebraic equations which are then solved
using an appropriate solver.
In modeling the TLD-structure systems using FVM/FEM method, two different codes separately solve the fluid and solid
domains, and the two domains are coupled through the exchange of information at the interface. These solutions are
synchronized through numerical iterations within a numerical loop called as stagger loop and each numerical iteration is
referred to as staggered iteration. To increase the accuracy of the numerical results, in this study, a second order backward
Euler algorithm was selected as the transient scheme, and a high-resolution scheme was employed to treat the advection
terms. To reduce the numerical error, finer mesh spacing was used in the direction of rapid solution variation at the free

Fig. 1. Mesh grid for the structure and fluid domains. (a) Structure. (b) Fluid.
144 A. Ashasi-Sorkhabi et al. / Journal of Sound and Vibration 388 (2017) 141–153

surface as illustrated in Fig. 1.


The geometry and dynamic properties of the numerical model are set the same as the tank in the RTHS setup. The
numerical model includes 76,813 fluid elements and 1488 structural elements. The structural model is treated as a rigid
body attached to a spring and the total mass is assigned to the base slab as shown in Fig. 1.

1.4. Real-time hybrid simulation (RTHS)

In a pseudodynamic (PSD) test the dynamic equation of motion of the structural system to be tested is expressed as
follows:

Mx¨ ( t )+Cẋ ( t )+R ( x,ẋ, x¨ ,t )=F ( t ) (1)

where M is the mass matrix, C is the damping matrix (to model the inherent viscous damping), R is the restoring force
vector, F is the vector of externally applied or effective forces, and x ̇ and ẍ are the velocity and acceleration vectors, re-
spectively. When the structure is subjected to ground acceleration, ẍg the external force vector becomes F ( t )= − M Γx¨ g ( t ),
where Γ is the mass influence vector.
In traditional PSD test to capture the R ( x,x ̇, x¨ ,t ) term in Eq. (1) the physical model of the entire test structure is built and
( )
tested in the laboratory whereas, the inertia Mx¨ ( t ) and damping (Cx ̇ ( t ) ) terms are kept analytical. During the PSD test Eq.
(1) is solved using a step by step time integration algorithm. The command displacements generated by the integration
algorithm are imposed on the test structure by displacement controlled hydraulic actuators. The corresponding restoring
forces and displacements developed by the deformed structure are measured and fed back to the integration algorithm for
generation of the next step command positions [42].
Hybrid simulation is an extension of PSD testing method [43]. In a hybrid simulation, unlike PSD testing, only a part of
the test structure (i.e. the component(s) for which a reliable model does not exist; also known as the critical components) is
built and tested in the laboratory and the remainder of the test structure is modeled analytically in a computer. The resulting
experimentally measured restoring forces and analytically calculated restoring forces together constitute the R ( x,x ̇, x¨ ,t ) term
in Eq. (1).
To capture the realistic vibrational characteristics of load-rate dependent structures the hybrid simulation needs to be
conducted dynamically in real-time [44,45]. This requires an efficient and robust computational platform with fast and
synchronized data communication [46]. Moreover, for accuracy and stability of the RTHS, special attention needs to be paid
to the accurate control of the experimental substructure and synchronization of the command and feedback signals during
the simulation [47–50].
The RTHS provides a means to investigate the realistic performance of complicated structural systems such as the ones
that employ passive and semi-active control devices (e.g. base isolators, tuned liquid dampers, tuned mass dampers and MR
dampers, etc.) [50–53]. In RTHS, as only the critical components of a structural system need to be tested physically and the
remaining parts are modeled analytically, a wide range of influential parameters and loading cases can be considered rapidly
and also in a cost-effective manner [54].
As an application example, in Fig. 2, the steps needed for running the RTHS experiments of a single degree of freedom

Fig. 2. Schematic view of real-time hybrid simulation.


A. Ashasi-Sorkhabi et al. / Journal of Sound and Vibration 388 (2017) 141–153 145

Fig. 3. Real-time hybrid simulation setup.

structure (SDOF) with a TLD on the roof are described. First the TLD-structure assembly is divided into two parts (i.e.,
substructuring): the SDOF structure which is modelled numerically as a mass-spring-dashpot oscillator in MATLAB/Simulink
(i.e., the analytical substructure) and the TLD which is built and tested physically on a shake table (i.e., the experimental
substructure). At each time step of the time integration, the displacement at the interface DOF between the two sub-
structures (i.e., the roof displacement) is computed. This displacement is imposed on the experimental substructure using a
shake table and the resulting interface force is measured by a load cell and fed back to the Simulink code for computation of
the next step command displacement.

2. Test setup

The real-time hybrid simulation is implemented using a Quanser earthquake simulator (Quanser shake table II) con-
trolled by a digital controller, displacement (i.e., position encoder), force (i.e., load cell) and acceleration (i.e., accelerometer)
measurement devices, and a Pentium class computer with a board for real-time execution (to solve the equation of motion
and generate the target displacements).The system also has digital to analog (D/A) and analog to digital (A/D) converters for
data acquisition purposes. The communication with the hardware is carried out through a digital signal processing board
called MultiQ-2E. Physical setup of the RTHS experiments is shown in Fig. 3. The board communicates with the host PC
using a proprietary software package called WinCon which is fully compatible with MATLAB and Simulink. WinCon is a real-
time application that is capable of running the generated Simulink codes in real time on the same PC (also known as local
PC) or on a remote computer [55]. Prior to a test, the numerical substructure model is compiled in Simulink and down-
loaded to the board. The schematic of the test setup and associated signal routings are shown in Fig. 4. The details of the
hardware and the software developed for the introduced RTHS setup, steps taken to improve the controller and the accuracy
of the test results along with the verification tests can be found in the previous work published by the authors [54].

Fig. 4. Schematic diagram of test setup.


146 A. Ashasi-Sorkhabi et al. / Journal of Sound and Vibration 388 (2017) 141–153

Fig. 5. Outer control loop/Analytical substructure modelling for RTHS.

3. Performance evaluation of an SDOF structure equipped with a TLD using RTHS

In this paper, to evaluate the dynamic behavior of TLDs, a comprehensive study is carried out where a wide range of
influential parameters (namely, TLD/structure mass ratio, TLD/structure frequency ratio, and structural damping ratio) are
experimentally considered using RTHS. The main focus of the study is to capture TLD/structure mass ratio effects while
other parameters are also changing. Two sets of experiments are carried out for this purpose: in the first set of experiments
the mass ratio and frequency ratio are changed to populate the test matrix while damping ratio is kept constant; setting the
frequency ratio to an optimum value while the second set of experiments investigate the effects of mass ratio together with
damping ratio on the efficiency of TLDs.

3.1. Experimental substructure

Due to test setup size and associated payload limitations, a 46 cm*31 cm*45 cm rectangular water tank was built from
plexi-glass sheets as the experimental substructure to be used during the RTHS experiments and the water height was set to
4 cm. As such, based on the linear wave theory [19], fundamental sloshing frequency of the TLD is 0.67 Hz.

3.2. Analytical substructure

A single story shear building was considered as the analytical substructure in the RTHS tests. As shown in Fig. 5, the
analytical substructure was idealized as a mass-spring-dashpot system and modelled in Simulink employing Runge-Kutta
solver provided by MathWorks.

3.3. Preliminary experiments

While running RTHS tests, several mass ratios need to be considered as listed in the test matrix. This can be achieved
using two different approaches. In approach 1,for each experiment of the test matrix, the structural properties (i.e. mass,
stiffness, damping) are chosen such that the desired mass ratio is achieved while keeping the other TLD parameters (e.g.
frequency ratio, damping ratio) as well as the uncontrolled structural response the same. This approach is very convenient
for testing as all the necessary changes are done numerically without a need to change the experimental substructure (i.e.,
the TLD dimensions and water height). However, it may cause accuracy issues in the results as a result of the scaling effects
of the test structure. Therefore the reliability of this approach needs to be verified before being used in the remainder of the
study reported here. In approach 2, the parameters of the analytical substructure are kept constant and instead the width of
the TLD is changed such that the desired mass ratio is achieved. This is not feasible using the current setup because of the
payload limit of the shake table and also difficulty of building several TLDs with different widths. However, the desired mass
ratio could still be obtained by applying an appropriate coefficient to the TLD interface force measured during test with the
assumption that identical TLDs would behave similarly if they are all placed side by side at the same location on the roof. In
other words, to achieve the desired mass ratio, a proper number of similar TLDs are considered in parallel (Fig. 6). This could
also be interpreted as changing the TLD width by the required amount. Preliminary RTHS experiments were carried out to
A. Ashasi-Sorkhabi et al. / Journal of Sound and Vibration 388 (2017) 141–153 147

Fig. 6. Multiple TLDs for larger mass ratios.

verify the results obtained from approach 1 using the results from approach 2 for different mass ratios. Table 1 summarizes
the structural properties for the analytical substructure used in these preliminary RTHS experiments. In Fig. 7, the RTHS
displacement responses obtained from both approaches are compared to the pure time history analyses of the structure
without the TLD. In these experiments, the test structure is subjected to harmonic loading tuned to the natural frequency of
the structure to excite the structure at the resonance and obtain the TLD effects accordingly. As shown in the figure there is a
good agreement between the results obtained from both approaches.

3.4. Mass and frequency ratio effects on the TLD-structure interaction

In this section RTHS results of SDOF-TLD system are presented where the SDOF is subjected to harmonic force input
tuned to the natural frequency of the structure. As such the TLD-structure interaction could be captured at resonance. For
this set of experiments, the test matrix contains thirty six elements considering six mass ratios as well as six frequency
ratios. For all of the tests conducted in this section the damping ratio was maintained constant at 0.2%. In each case the
structural mass, stiffness and damping were chosen in such a way to achieve the desired mass ratio, frequency ratio and
damping ratio. To provide a fair comparison, for each frequency ratio considered, the force amplitude was adjusted to get a
constant un-controlled (i.e, without the TLD) response of the SDOF as the mass ratio varies from 0.5% to 5%. Approach 1,
explained in Section 3.3, was utilized to obtain these different mass ratio values. All the RTHS experiments were conducted
for 90 seconds and the peak values within the last 5 seconds of the tests from the uncontrolled and controlled responses
were used to calculate the reduction in the response due to inclusion of the TLD.
Figs. 8 and 9 show the reduction in the roof displacements and accelerations, respectively. As shown in these figures, by
increasing the mass ratio from 0.5% to 3%, the reduction rates increase significantly for all frequency ratios. However, for
mass ratios greater than 3%, the reduction rates appear not to have any further change. It could be observed from the results
that there is a small shift in the frequency of the coupled system (i.e. TLD-structure) with respect to the natural frequency of
the uncontrolled structure. This frequency shift could be attributed to the additional mass due to the inclusion of TLD on top
of the structure. As shown in Fig. 7, the results indicate that for mass ratios greater than 3%, although there is a global
reduction in the response, there are some local oscillations, also known as beating phenomenon. These are known to occur
when there are two closely spaced frequencies associated with the system. In addition, for each mass ratio case, maximum
reduction in response is obtained for frequency ratio values of 1 and 1.2.
In Fig. 10, the ratio of the energy dissipated by the TLD to the input energy induced by the sinusoidal force is provided for
different values of the mass and frequency ratios. The dissipated energy by the TLD is obtained by calculating the area of the
hysteresis developed by the TLD interface force of the TLD base (roof) displacement. As indicated, the maximum con-
tribution of the TLD in terms of energy dissipation occurs generally in frequency ratio values of 1.0 and 1.2. For mass ratios

Table 1
Structural properties of the analytical substructure.

Mass ratio (%) Approach 1 Approach 2

Mass (Kg) Stiffness (N/m) Damping (N s/m) Mass (Kg) Stiffness (N/m) Damping (N s/m) TLD force Coefficient

0.5 1128 19990 19.00 564 9995 9.5 0.5


1 564 9995 9.50 564 9995 9.5 1
2 282 4997.5 4.75 564 9995 9.5 2
3 188 3331.7 3.17 564 9995 9.5 3
4 141 2498.8 2.38 564 9995 9.5 4
5 112.8 1999 1.9 564 9995 9.5 5
148 A. Ashasi-Sorkhabi et al. / Journal of Sound and Vibration 388 (2017) 141–153

Fig. 7. RTHS displacement response comparisons for different mass ratios using the two approaches (frequency ratio ¼ 1.0 & damping ratio ¼ 0.2%).

Fig. 8. RTHS results for the reduction in roof displacements considering a range of mass and frequency ratios with a constant structural damping ratio of
0.2%.
A. Ashasi-Sorkhabi et al. / Journal of Sound and Vibration 388 (2017) 141–153 149

Fig. 9. RTHS results for the reduction in roof accelerations considering a range of mass and frequency ratios with a constant structural damping ratio of
0.2%.

Fig. 10. RTHS results for the ratio of the dissipated to input energy for a range of mass and frequency ratios with a constant structural damping ratio of
0.2%.

less than 3%, the peak of the graph is at the frequency ratio of 1.2; and as the mass ratio increases maximum energy
dissipation occurs at the frequency ratio of 1.0.

3.5. Mass and damping ratio effects on the TLD-structure interaction

The effects of mass ratio and the inherent structural damping on the efficiency of the TLDs are discussed in this section.
For this purpose, 35 test cases were considered in populating the test matrix including mass ratios ranging from 0.5% to 4.0%
and inherent structural damping ranging from 0.2% to 5.0%. Using approach 1, the structural properties were tuned to obtain
desired mass and inherent damping ratios while the frequency ratio was maintained constant at an optimum value of
1.2 which was determined in Section 3.4.
Figs. 11 and 12 show the displacement and acceleration reductions obtained through RTHS, respectively. In these plots
the horizontal axes show the inherent damping of the controlled structure and each line on the plots represents a particular
mass ratio. It is observed that the TLD is more effective for cases with lower structural inherent damping and as the
structural damping ratio increases the TLD ceases to become efficient in suppressing the vibrational response of the test
structure. Same trends at lower and higher damping values are valid for different mass ratios. The reduced effectiveness of
the TLD for higher inherent damping values can be explained by the fact that as the structural damping ratio increases, the
magnitude of the structural roof displacements and acceleration decrease. Since the TLD performance is highly dependent of
the structural displacements and accelerations, they become less effective for structures with higher inherent damping.
In terms of energy dissipated by the liquid damper, as shown in Fig. 13, the TLD has more contribution for lower
150 A. Ashasi-Sorkhabi et al. / Journal of Sound and Vibration 388 (2017) 141–153

Fig. 11. RTHS results for the reduction in roof displacement considering a range of mass and damping ratios with a constant frequency ratio of 1.2.

Fig. 12. RTHS results for the reduction in roof acceleration considering a range of mass and damping ratios with a constant frequency ratio of 1.2.

structural damping ratios. For higher structural damping ratios, since a high percentage of internal energy is dissipated by
the inherent damping, TLD efficiency is reduced. It could also be observed that TLDs with higher mass ratios are relatively
more effective.

4. Accuracy assessment of FVM/FEM method

To complement the study done by Malekghasemi et al. [33], using the experimental data obtained in Section 3.4, the
accuracy of the FVM/FEM method in modelling TLD-structure interaction was validated considering a wide range of mass
and frequency ratios. Fig. 14 displays the peak roof displacements predicted by the FVM/FEM method and the peak dis-
placement values measured during the RTHS experiments. In this figure, both experimental and analytical data sets are
plotted against the peak displacements obtained from the pure time history analysis of the uncontrolled structure. It should
be pointed out that due to the maximum stroke limit (i.e. 72.5 in. or 63 mm) of the shake table used in the study, the test
matrix was configured such that the roof displacements of the controlled structure (table displacements) during the RTHS
experiments would not exceed the aforesaid limits. In particular, for test cases with 0.5% mass ratio the contribution from
the TLD in the controlled structure was not significant enough to decrease the roof displacements considerably. Therefore,
the input force was scaled down such that the roof displacements fell within the device limits. For this reason, the roof
A. Ashasi-Sorkhabi et al. / Journal of Sound and Vibration 388 (2017) 141–153 151

Fig. 13. RTHS results for the ratio of the dissipated to input energy considering a range of mass and damping ratios with a constant frequency ratio of 1.2.

Fig. 14. FVM/FEM method verification tests.

displacements of the uncontrolled structures with 0.5% mass ratio in Fig. 14 are smaller than the rest of the experiments
associated with each frequency ratio.
As can be seen, there is a good agreement between the FVM/FEM model predictions and experimental results over a
wide range operational conditions. This confirms the FVM/FEM method as a reliable tool in capturing the TLD-structure
interaction for the cases with both weak and strong wave breaking behaviors.

5. Conclusions

This paper presents the experimental investigations to study the effectiveness of TLDs in controlling structural vibrations
using the real-time hybrid simulation (RTHS) method. As a result of the flexibility offered by the RTHS method, it was
152 A. Ashasi-Sorkhabi et al. / Journal of Sound and Vibration 388 (2017) 141–153

possible to experimentally study the TLD - structure interaction by considering a wide range for several of the influential
system parameters including TLD/structure mass ratio, TLD/structure frequency ratio, and structural damping ratio. Fol-
lowing are the summary of the key results obtained in this study:

– The experimental results showed that by increasing the mass ratio from 0.5% to %3, the TLD effectiveness in structural
response reduction is improved. However, for mass ratios beyond 3% the TLD efficiency does not improve significantly and
in some cases some un desired effects ( e.g. beating) is observed.
– It is observed that for a given mass ratio, maximum response reduction is obtained at frequency ratios 1.0 and 1.2. From
dissipated energy perspective, the highest TLD contribution is achieved when the frequency ratio is set to 1.2.
– For the mass ratio range considered, higher efficiency could be gained from the TLD if installed in lightly damped
structures as opposed to highly damped systems where the inherent damping contributes mainly in structural response
reduction.
– Using the test data collected throughout this study, the FVM/FEM method was validated as a reliable tool for modelling
the TLD-structure interaction within the frequency and mass ratio ranges considered and especially in the presence of
wave breaking.

The current design practice for TLD recommends a frequency ratio of 1.0 irrespective of the mass ratio. The RTHS results
from the large experiment matrix in this study suggest that to maximize the efficiency of the TLDs in mitigating the vi-
brations (i) they should be designed with a sloshing frequency that is around 1.2 times fundamental frequency of structure,
(ii) they should have mass ratio around 3% or less with respect to the structural mass, and (iii) they should be introduced to
lightly damped structures (less than 2%).

Acknowledgements

The authors would like to thank Dr. Constantin Christopoulos for the access to the Quanser shake table and the technical
staff of the civil engineering laboratory at the University of Toronto for their assistance during experiments. The financial
support for this study from NSERC Discovery (Grant 371627-2009) is gratefully acknowledged. Any opinions, findings,
conclusions and recommendations expressed here are those of the authors and do not necessarily reflect the views of the
sponsors.

References

[1] R. Sabelli, S.A. Mahin, C. Chang, Seismic demands on steel braced frame buildings with buckling-restrained braces, Engineering Structures 25 (5) (2003)
655–666.
[2] K. Ghabraie, R. Chan, X.D. Huang, Y.M. Xie, Shape optimization of metallic yielding devices for passive mitigation of seismic energy, Engineering
Structures 32 (8) (2010) 2258–2267.
[3] Y.Q. Ni, Y. Chen, J.M. Ko, D.Q. Cao, Neuro-control of cable vibration using semi-active magneto-rheological dampers, Engineering Structures 24 (3)
(2002) 295–307.
[4] G. Yang, B.F. Spencer, J.D. Carlson, M.K. Sain, Large-scale MR fluid dampers: modeling and dynamic performance considerations, Engineering Structures
24 (3) (2002) 309–323.
[5] Y. Fujino, B. Pacheco, P. Chaiseri, L.M. Sun, Parametric studies on tuned liquid damper (TLD) using circular containers by free-oscillation experiments,
Structural Engineering/ Earthquake Engineering 5 (2) (1988) 381–391.
[6] Y. Tamura, K. Fujii, T. Sato T., Wakahara, M. Kosugi, Wind-induced vibration of tall towers and practical applications of tuned sloshing dampers,
Proceedings of Workshop on Serviceability of Buildings (Movements, Deformations, Vibrations), Ottawa, I, 1988, pp. 228-241.
[7] K.L. Shen, T.T. Soong, K.C. Chang, M.L. Lai, Seismic behavior of reinforced concrete frame with added visco-elastic dampers, Engineering Structures 17 (5)
(1995) 372–380.
[8] Y.L. Xu, B. Chen, Integrated vibration control and health monitoring of building structures using semi-active friction dampers: part i – methodology,
Engineering Structures 30 (7) (2008) 1789–1801.
[9] Y. Zhang, S. Zhu, A shape memory alloy-based reusable hysteretic damper for seismic hazard mitigation, Smart Materials and Structures 16 (2007)
1603–1613.
[10] Y.H. Chen, W.S. Hwang, L.T. Chiu, S.M. Sheu, Flexibility of TLD to high-rise building by simple experiment and comparison, Computers and Structures 57
(5) (1995) 855–861.
[11] Y.M. Kim, K.P. You, N.H. Ko, S.W. Yoon, Use of TLD and MTLD for control of wind-induced vibration of tall buildings, Journal of Mechanical Science and
Technology 20 (9) (2006) 1346–1354.
[12] Y. Tamura, K. Fujii, T. Ohtsuki, T. Wakahara, R. Kohsaka, Effectiveness of tuned liquid dampers under wind excitation, Engineering Structures 17 (1995)
609–621.
[13] Y. Fujino, P. Chaiseri, L.M. Sun, Parametric studies on Tuned Liquid Damper (TLD) using circular containers by free-oscillation experiments, Structural
Engineering 5 (1998) 381–391.
[14] D.A. Reed, J. Yu, H. Yeh, Investigation of tuned liquid dampers under large amplitude excitation, Journal of Engineering Mechanics 124 (4) (1998)
405–413.
[15] H. Gao, K.C.S. Kwok, B. Samali, Optimization of tuned liquid column dampers, Engineering Structures 19 (6) (1997) 476–486.
[16] P.A. Hitchcock, K.C.S. Kwok, R.D. Watkins, B. Samali, Characteristics of liquid column vibration absorbers (LCVA)-I, II, Engineering Structures 19 (1997)
126–144.
[17] S. Colwell, B. Basu, Tuned liquid column dampers in offshore wind turbines for structural control, Engineering Structures 31 (2) (2009) 358–368.
[18] F. Zhu, J.T. Wang, F. Jin, O. Altay, Real-time hybrid simulation of single and multiple tuned liquid column dampers for controlling seismic-induced
response, Proceeding of the sixth International Conference on Advances in Experimental Structural Engineering University of Illinois, Urbana-
Champaign, USA, 2015.
A. Ashasi-Sorkhabi et al. / Journal of Sound and Vibration 388 (2017) 141–153 153

[19] L.M. Sun, Y. Fujino, M. Pacheco, P. Chaiseri, Modelling of tuned liquid damper (TLD), Journal of Wind Engineering and Industrial Aerodynamics (1992)
41–44.
[20] J.K. Yu, T. Wakahara, D.A. Reed, A non-linear numerical model of the tuned liquid damper, Earthquake Engineering and Structural Dynamics 28 (1999)
671–686.
[21] P. Banerji, M. Murudi, A.H. Shah, N. Popplewell, Tuned liquid dampers for controlling earthquake response of structures, Earthquake Engineering and
Structural Dynamics 29 (2000) 587–602.
[22] H. Kosaka, T. Noji, H. Yoshida, E. Tatsumi, H. Yamanaka, A.K. Agrawal, Damping effects of a vibration control damper using sloshing of water, Pro-
ceedings of the 10th World Conference on Earthquake Engineering. Madrid, Spain, 1992.
[23] T. Ohyama, K. Fujii, A boundary element analysis for two-dimensional nonlinear sloshing problem, Journal of Structural Engineering 35 (A) (1989)
575–584. [In Japanese].
[24] S. Kaneko, M. Ishikawa, Modeling of tuned liquid damper with submerged nets, Journal of Pressure Vessel Technology 121 (3) (1999) 334–343.
[25] B. Ramaswamy, M. Kawahara, T. Nakayama, Lagrangian finite element method for the analysis of two-dimensional sloshing problems, International
Journal of Numerical Methods Fluids 6 (1986) 659–670.
[26] J.B. Frandsen, Numerical predictions of tuned liquid tank structural systems, Journal of Fluids and Structures 20 (2005) 309–329.
[27] L.M. Sun, Y. Fujino, P. Chaiseri, M. Pacheco, The properties of tuned liquid dampers using a TMD analogy, Earthquake Engineering and Structural
Dynamics 24 (1995) 967–976.
[28] M.J. Tait, Modelling and preliminary design of a structure-TLD system, Engineering Structures 30 (2008) 2644–2655.
[29] J.K. Yu, Nonlinear characteristics of tuned liquid dampers (PhD thesis), University of Washington, 1997.
[30] A.M. Halabian, M. Torki, Numerical studies on the application of tuned liquid dampers with screens to control seismic response of structures, The
Structural Design of Tall and Special Buildings 20 (2) (2011) 121–150.
[31] J.S. Love, M.J. Tait, Non-linear multi modal model for tuned liquid dampers of arbitrary tank geometry, International Journal of Non-Linear Mechanics 46
(2011) 1065–1075.
[32] J.S. Love, M.J. Tait, Nonlinear simulation of a tuned liquid damper with damping screens using a modal expansion technique, Journal of Fluids and
Structures 26 (2010) 1058–1077.
[33] H. Malekghasemi, A. Ashasi-Sorkhabi, A. Ghaemmaghami, O. Mercan, Experimental and numerical investigations of the dynamic interaction of tuned
liquid damper-structure systems, Journal of Vibration and Control 21 (14) (2015) 2707–2720.
[34] T.J. Chung, Computational Fluid Dynamics, Cambridge University Press, Cambridge, 2002.
[35] O.C. Zienkiewicz, R.L., Taylor, The Finite Element Method, Butterworth-Heinemann, Stonham, 2000.
[36] G.R. Liu, M.B. Liu, Smoothed particle hydrodynamics: a meshfree particle method, World Scientific, Singapore, 2003.
[37] A.S. Iglesias, L.P. Rojas, R.Z. Rodriguez, Simulation of antiroll tanks and sloshing type problems with smoothed particle hydrodynamics, Ocean En-
gineering 31 (8–9) (2004) 1169–1192.
[38] S.H. Rhee, L. Engineer, Unstructured grid based Reynoldsaveraged Navier-Stokes method for liquid tank sloshing, Journal of Fluid Engineering 127
(2005) 572.
[39] A. Souto-Iglesias, L. Delorme, L. Perez-Rojas, S. Abril-Perez, Liquid moment amplitude assessment in sloshing type problems with smooth particle
hydrodynamics, Ocean Engineering 33 (11–12) (2006) 1462–1484.
[40] M. Anghileri, L.M.L. Castelletti, M. Tirelli, Fluid structure interaction of water filled tanks during the impact with the ground, International Journal of
Impact Engineering 31 (3) (2005) 235–254.
[41] H.K. Versteeg, W. Malalasekera, An introduction to computational fluid dynamics: the finite volume method, Longman Scientific & Technical, 2nd ed.
Pearson Education, 1995.
[42] S.A. Mahin, P.B. Shing, Pseudodynamic method for seismic testing, Journal of Structural Engineering 111 (1985) 1482–1503.
[43] S.N. Dermitzakis, S.A. Mahin, Development of sub-structuring techniques for on-line computer controlled seismic performance testing, Earthquake En-
gineering Research Center, University of California, Berkeley, 1985 [Report UBC/EERC-85/04].
[44] M. Nakashima, H. Kato, E. Takaoka, Development of real-time pseudodynamic testing, Earthquake Engineering and Structural Dynamics 21 (1992) 79–92.
[45] T. Horiuchi, M. Inoue, T. Konno, Y. Namita, Real-time hybrid experimental system with actuator delay compensation and its application to a piping
system with energy absorber, Earthquake Engineering and Structural Dynamics 28 (1999) 1121–1141.
[46] O. Mercan, J.M. Ricles, Experimental studies on real-time pseudodynamic (PSD) and hybrid PSD testing of structures with elastomeric dampers,
Journal of Structural Engineering (ASCE) 135 (2009) 1124–1133.
[47] O. Mercan, J.M. Ricles, Stability and accuracy analysis of outer loop dynamics in real-time pseudodynamic testing of SDOF systems, Earthquake
Engineering and Structural Dynamics 36 (2007) 1523–1543.
[48] C. Chen, J.M. Ricles, Stability analysis of SDOF real-time hybrid testing systems with explicit integration algorithms and actuator delay, Earthquake
Engineering and Structural Dynamics 37 (4) (2008) 597–613.
[49] O. Mercan, J.M. Ricles, Stability analysis for real-time pseudodynamic and hybrid pseudodynamic testing with multiple sources of delay, Earthquake
Engineering and Structural Dynamics 37 (2008) 1269–1293.
[50] J. Liu, S.J. Dyke, H.-J. Liu, X.-Y. Gao, B. Phillips, A novel integrated compensation method for actuator dynamics in real-time hybrid structural testing,
Structural Control and Health Monitoring 20 (2013) 1057–1080.
[51] R. Christenson, Y.Z. Lin, A. Emmons, B. Bass, Large-scale experimental verification of semi-active control through real-time hybrid simulation, Journal of
Structural Engineering (ASCE) 134 (2008) 522–534.
[52] T. Karavasilis, J.M. Ricles, R. Sause, C. Chen, Experimental evaluation of the seismic performance of steel MRFs with compressed elastomer dampers
using large-scale real-time hybrid simulation, Engineering Structures 33 (2011) 1859–1869.
[53] B. Wu, P. Shi, J. Ou, Seismic performance of structures incorporating magnetorheological dampers with pseudo-negative stiffness, Structural Control
and Health Monitoring 20 (2013) 405–421.
[54] A. Ashasi-Sorkhabi, H. Malekghasemi, O. Mercan, Implementation and verification of real-time hybrid simulation (RTHS) using a shake table for
research and education, Journal of Vibration and Control 21 (8) (2015) 1459–1472.
[55] WinCon 3.3 ver. 1.1 User’s Manual, Quanser Consulting Inc., 2002.

Вам также может понравиться