VIBRATION CONTROL OF STRUCTURE USING TUNED-MASS DAMPER(TMD)
Thesis
Submitted in partial fulfilment of the requirements
Of the degree of
Master of Technology
Submitted by
Uday Singh
(Roll No. 173040045)
Supervisor:
Prof. R.S.Jangid
Department of Civil Engineering
INTRODUCTION
A tuned-mass damper (TMD), or dynamic vibration absorber (DVA), is a most efficient passive
vibration control device consisting a mass, a spring, and viscous dampers. Since proposed in
1909, they have been widely used in power transmission, automobiles, machinery and
buildings [1]. The first theoretical investigation of TMD design was carried by Ormondroyd
and Den Hartog in 1928, and the details can be found in the text by Den Hartog [2].
The tuned mass damper is a classical engineering device consisting of a mass, a spring and a
viscous damper mounted in structure to reduce undesirable vibrations. It is also known as a
seismic damper or harmonic absorber.
Leung et al. 2009 proposed a novel evolutionary algorithm of particle swarm optimization
(PSO) to optimize the required parameters of the TMD. These optimized parameters were
obtained by minimizing some response quantities under various types of excitations. These
excitations comprised of base acceleration and external force modelled as Gaussian white-noise
random processes. It was concluded that the proposed algorithm PSO was effective to
determine optimum damping, mass ratio and tuning frequency of the TMD system and easily
programmed for practical engineering application.
Tsai et al. 1993 developed a numerical searching procedure in order to find the optimum
damping ratio and tuning frequency of TMD so as to reduce the steady-state response of
damped structure to a minimum level. The explicit formulae for these optimum parameters
were derived through a sequence of curve-fitting technique for two different harmonic
excitation sources, support motion of fixed-acceleration amplitude and support motion of
fixed-displacement amplitude. It was found that error was negligible in proposed explicit
formulae, thus provide a convenient method to computation the optimum parameters in
structural application. It was concluded that for high damped structure, the TMD was found to
be less effective for reducing the structural response. It was also found that the optimum tuning
frequency is strongly affected by the damping level of a structure whereas optimum damping
ratio of the TMD was not sensitive to the damping level of a structure.
Jangid et al. 2006 obtained optimum parameters of TMD system which was attached to a SDOF
main system for various combination of excitation and response parameter. It was applied
excitation to the main structure consists of external force and base acceleration modelled as
Gaussian white-noise random process. For minimizations of various response quantities such
as relative displacement, velocity of main mass and force transmitted to support obtained
optimum parameter such as optimum tuning frequency ratio and damping using numerical
searching technique. Response and optimum parameters of TMD for different mass ratio of
TMD and damping ratio of main structure obtained. Explicit formula for optimum tuning
frequency, damping and corresponding response were derived using curve fitting technique.
The error found in these expressions to be negligible so that these formulae convenient for
damped SDOF main structure. It was concluded that optimum tuning frequency of TMD is
affected by damping of main structure but optimum damping ratio of TMD was not affected
by damping of main structure.
[1] L.Zuo, S.A. Nayfeh, Minimax optimization of multi-degree-of -freedom tuned-mass dampers
Department of Mechanical Engineering, Massachusetts Institute of Technology, Cambridge,2003
[2] J.P.Den Hartog
Mechanical Vibration, McGraw-Hill, New York(1947)
1. Introductory remark
Rapid growth of urban areas and significant progress in the construction industry has resulted
in inclination to build tall structures. Under dynamic loads like wind or earthquakes, severe
vibrational problems are experienced by these structures. Moreover, mechanical systems
subjected to time-varying loads are also associated with undesirable vibrations. The lessening
of these structural vibrations and energy dissipation is accomplished by adopting either
strengthening or control measures. Strengthening measures, the currently standard approach,
involves designing the structural resisting elements to withstand the dynamic loads and to
achieve desired comfort levels of vibration. Strengthening of structures would however incur
excessive expenditures. Alternatively, control measures, a comparatively new approach,
involve use of supplemental control systems to damp the structural response. In this technique,
the structural motion is essentially controlled from the dynamic loadings by providing
supplemental control systems that absorb most of the input energy and only a part of this energy
is transferred to the structural system. The structural vibration control is achieved by appending
the structures with either passive or active control systems like, viscous dampers, visco-elastic
dampers, tuned mass dampers (TMD), friction dampers, shunted piezoceramic dampers and
magnetic dampers [1]. However, passive approach of energy dissipation and structural
vibration control is widely used in structural systems, over active or hybrid passive control
methods, for their simplicity in design, operation and maintenance.
2. Methods of control
A large number of techniques have been tried to produce better control against wind and
earthquake excitation. These can be classified into four broad categories: passive control, active
control, semi-active control and hybrid control. Each of these will be discussed in following
section.
1. Passive control
The most mechanically simple set of control schemes is enclosed in the passive control
category, which has been widely accepted for civil engineering application. According to them
a passive control system is one that does not require an external power source. All forces
imposed by passive control devices develop as direct responses to the motion of the structure.
Hence, sum of the energy of both the device and the primary system will be constant.
The main purpose of these systems is to efficiently dissipate vibrational energy, and the various
methods of achieving this can be categorized in two ways. The first method includes converting
kinetic energy directly to heat, such as through the yielding of metals, the deformation of visco-
elastic solids and fluids, or the implementation of friction sliders. The second method works
on transferring energy among two or more of the vibrational modes of the building, generally
achieved by adding a supplemental oscillator that absorbs the vibrations of the primary
structure. Tune mass damper, tune liquid damper, base isolation are example of passive system.
Passive control is the most widely-used method of controlling structural response under wind
and earthquake loading, but it has some limitations. While it is reliable and relatively straight
forward to design, passive control systems are generally only good for limited bandwidths of
dynamics input. As a result, they are susceptible to the effects of off-tuning, de-tuning, or
resonances of secondary modes.
2. Active control
Active control is a relatively upcoming subfield of structural engineering. It assures improved
response to passive systems at the cost of energy and more complex systems.
Active control system has been as any control system in which an external power source is
required to provide additional forces to the structure in a prescribed manner, by the use of
actuators. The signals are sent to control the actuators and determine the feedback from the
sensors provided on or through the structure. Due to the presence of an external power source,
the force applied may either add or dissipate energy from the structure.
In order to maximize the performance of an active system, the actuator forces must be
prescribed in real-time base on the inputs of the sensors. The direction and magnitude of these
forces can be assigned in the variety of ways, all of which have their roots in the diverse and
mathematically rich field of control engineering.
The performance of active control is quite pronounced in some cases. Due to its capability to
respond in real-time, active control eliminates most of the tuning drawbacks inherent in passive
devices. However, active control has not been exuberantly embraced by the civil engineering
community as a result of some significant limitations.
Most significant advantage of active control method is diminishing by their heavy reliance on
external power supplies. The power consumption and cost are comparatively large for output
of certain magnitude forces necessary to control large civil structures by the actuator.
Additionally, there may be situation at which the control forces are needed coincides with the
time when the power cut is the most likely, such as during an earthquake or large wind storm.
This raises question on reliability concerns.
Beyond the issue of energy supply, engineers also hesitate to embrace non-traditional
technologies for structures. The placement of sensors and the design of feedback schemes are
also beyond the scope of most practicing engineers, and poorly designed active system may
lead to deleterious energy inputs and destabilization of the primary system.
3. Semi-active control
Semi active control performs on the benefits of active control and the reliability of passive
control, which makes it a much more appealing alternative to traditional control scheme in civil
structures. Semi active control systems act on the same principle of active control system but
they differ in that their external energy requirement is smaller. These devices have an inherent
stability in terms of bounded-input and output as these do not add mechanical energy to the
primary system. Therefore, it may be viewed as controllable passive device.
Semi-active control relies on the reactive forces that develop due to variable stiffness or
damping devices rather than application of actuator forces. That means, by changing the
properties of these devices, using only nominal power the response of the system may be
favourably modified. As a result, semi-active control methods appear to combine the best
features of fully active and fully passive systems, leaving them as the best in term acceptance
for structural control.
The best advantage of semi-active systems is their ability to provide improved control forces
with a low demand for power. As the power can be supplied by a battery, which ensures
continued functionality even at power failure, adding reliability to any semi-active control
method. Because of these benefits that enthusiasm towards the semi-active structural control
schemes has increased in recent years, making it a viable alternative to proven passive devices.
While these advantages are in some case truly significant, semi-active control still has its
detractors. Most relevant is the need for sensors technology and computer-controlled feedback,
which is as central to semi-active controls to active control.
2. Hybrid control
Hybrid systems act on the combined use of passive and active control system. For example, a
base isolated structure which is equipped with actuator which actively controls the
enhancement of its performance.
Among these control methods, the structural passive control techniques with constant
parameters and without need of energy are the most straightforward and the focus of the present
report.
3. Tuned Mass Damper (TMD)
The tuned-mass damper (TMD) is a widely used passive vibration control device. A TMD
consists of a mass mounted on a structure via a spring system and a viscous damper, preferably
in a location where the structure’s deflections are greatest. The spring and mass are ‘tuned’ so
as to have a natural frequency close to that of the primary structure. When properly tuned, the
TMD mass oscillates in the opposite direction from the primary structure. The motion of the
mass relative to the main structure can be very large when the system is properly tuned and this
provides an opportunity to dissipate a substantial amount of energy in
the damper linking the mass to the main structure.
Figure 1.1 SDOF-TMD system
The Tuned Mass Damper (TMD) has many advantages compared with other damping devices:
compactness, reliability, low maintenance cost amongst others. Hence, in recent years it has
been widely used in civil engineering structures.
4. Need of the present study
Notwithstanding the bright prospect of TMD, there exists two vital deficiencies, including the
detuning effect and an exceedingly large stroke. The detuning effect will occur when the
frequency of TMD is not tuned to the controlled natural frequency of a structure. Due to
mistuned frequency, the vibration suppression performance of TMD will be impaired
significantly, in practical terms, meaning that TMD is not all robust. While comparing to the
detuning effect, the exceedingly large stroke of TMD will need the extraordinarily large space
and thereby its use becomes economically impractical. In practical terms, the original plan of
installing TMD for a project is often abandoned, since the stroke of TMD is too large and there
is not enough space to house it. Hence, with the growing need for TMD, it is more imperative
and of practical interest to search for TMD system with smaller stroke.
TUNED MASS DAMPERS
Introduction
Mythology
STRUCTURAL MODEL
The structural diagram of main system with TMD as shown in figure 1. The main system is
idealized as a stiffness 𝑘𝑝 , lump mass 𝑚𝑝 and damping cp. The damping ratio and natural
frequency of the main structure are 𝜉𝑝 (i.e., 𝜉𝑃 = 𝑐𝑃 /2𝜔𝑝 𝑚𝑝 ) and 𝜔𝑝 = √𝑘𝑝 ∕ 𝑚𝑝 , respectively.
Similar to the main system, the parameters of the TMD are the stiffness 𝑘𝑠 , the damping
constant 𝑐𝑠 and the mass 𝑚𝑠 . Let 𝜉𝑠 = 𝑐𝑠 /2𝜔𝑠 𝑚𝑠 and 𝜔𝑠 = √𝑘𝑠 ∕ 𝑚𝑠 are the damping ratio
and natural frequency of the TMD respectively. In present research study, two type of harmonic
excitation considered namely, acceleration applied to the main system and external force
applied to the mass, 𝑚𝑝 .
Response to base acceleration
Consider to a main structure with TMD subjected to base acceleration. Let displacement of the
main structure and TMD relative to ground are 𝑢𝑝 and 𝑢𝑠 respectively. The governing equation
of motion subjected to ground acceleration are expressed as
𝑚𝑝 0 𝑢̈ 𝑐 +𝑐 −𝑐𝑠 𝑢̇ 𝑃 𝑘𝑃 + 𝑘𝑠 −𝑘𝑠 𝑢 𝑚
[ ] {𝑢̈ 𝑠 𝑝 } + [ 𝑃−𝑐 𝑠 ] {𝑢𝑠 𝑝 } = -{𝑚𝑠 𝑝 }𝑢̈ 𝑔
𝑜 𝑚𝑠 𝑠 𝑐𝑠 ] {𝑢̇ 𝑠 } + [ −𝑘𝑠 𝑘𝑠
where 𝑢̈ 𝑔 is the base acceleration applied to base of combined system. The applied excitation
is harmonic, i.e. 𝑢𝑔 = 𝐻ⅇ ⅈ𝜔𝑡 .Let 𝑢𝑝 = 𝐴ⅇ ⅈ𝜔𝑡 and 𝑢𝑠 = 𝐵ⅇ ⅈ𝜔𝑡 , the steady state response of the
system can be solved from
Where 𝛾 = 𝑚𝑆 ∕ 𝑚𝑝 ,is the ratio of damper mass to main mass. The ratio of displacement
amplitude of the main structure to amplitude of applied can be solved as
|𝑢𝑝 | 2 [𝑓2 (1+𝛾)−𝑔2 ]2 +4𝑔2 𝑓2 𝜉𝑠2 (1+𝛾)2
= 𝑔√ 2 2
𝐻 [𝛾𝑓2 𝑔2 −(𝑔2 −𝑓 𝑧 )(𝑔2 −1)+4𝜉𝑝 𝜉𝑠 𝑓𝑔2 ] +4𝑔2 [𝑓𝜉𝑠 (𝑔2 +𝛾𝑔2 −1)+𝜉𝑝 (𝑔2 −𝑓2 )]
Where 𝑔 = 𝜔/𝜔𝑝 and 𝑓 = 𝜔𝑠 ⁄𝜔𝑝 represent the non-dimensional input frequency and non-
dimensional tuning frequency respectively. The equation is a function of 𝑓, 𝜉𝑠 , 𝜉𝑝 , 𝑔 and 𝛾.
These parameters can affect the magnitude of vibration response.
Optimum parameters for undamped systems
Den Hartog’s procedure [2] used to find the optimum parameter of tuned mass damper for
undamped structure subjected to base acceleration at the base. For undamped main
structure (i.e. 𝜉𝑝 = 0), vibration response of main structure becomes
𝑔2√𝑓 2 (1 + ϒ) − g2]2 + 4𝑔2𝑓2
𝑥=
2𝑎
120
100
80
up/H
0
60
0.1
0.3
40
20
0
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6
Input frequency
7
4
up/H
0
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8
Input frequency
Optimum tunig vs Mass ratio
1.01
1
0.99
Optimum tuning
0.98
0.02
0.97
0.01
0.96
0.05
0.95
0.94 0.1
0.93 0
0.92
0 0.02 0.04 0.06 0.08 0.1 0.12
Mass ratio
Optimum damping vs mass ratio
0.2
0.18
0.16
Optimum damping
0.14
0.12 0.01
0.1 0.02
0.08 0.05
0.06
0.1
0.04
0.02 0
0
0 0.02 0.04 0.06 0.08 0.1 0.12
Mass ratio
mass ratio vs minimax amplitude
25
20
Minima Amplitude
15 0.01
0.02
10 0.05
0.1
5
0
0
0 0.02 0.04 0.06 0.08 0.1 0.12
Mass ratio
Input frequency ratio vs Mass ratio
1.2
0.01
1 0.01
Input frequency ratio
0.8 0.02
0.02
0.6
0.05
0.4 0.05
0.2 0.1
0.1
0
0 0.02 0.04 0.06 0.08 0.1 0.12 0
Mass ratio 0
