Vibration

theory
1
Two approaches to seismic design

Two seismic design approaches

Dynamic analysis Static analysis

2
 Dynamic analysis
In a dynamic analysis, the overall building and storey
stiffnesses and rigidities are calculated.
A specific design earthquake, including magnitude and
loading history, is selected and applied to a
mathematical model (consisting of lumped masses,
damping and spring stiffness) of the building.

The response of the system (including the

displacement and acceleration functions) is calculated
and used to determine the forces in each member as a
function of time.

3
Comments on “dynamic analysis”
The solution may rest heavily on vibrational theory,
finite element analysis and other advanced structural
techniques requiring computer analysis.
The building itself may be too simple or too
standardized to warrant the rigorous approach of the
design analysis.
Conversely, the building may be too complex and have
too many degrees of freedom to model mathematically.
In the initial design phases, the member sizes and
locations may not be known, making it difficult to
estimate stiffnesses and rigidities.
The dynamic approach is inappropriate when the
design earthquake is not known. 4
Static analysis

The equivalent lateral seismic force is calculated as

simply some fraction of the dead weight.

5
Simple harmonic motion
Ideal vibrational systems that consists of springs and
Masses and that are not acted upon by external
Disturbing forces (after an initial displacement) are
known as simple harmonic oscillators.
During steady-state motion, such oscillator move in a
Repetitive sinusoidal pattern known as simple harmonic
Motion.
Simple harmonic motion is characterized by the
Absence of a continued disturbing force and a lack of
Frictional damping.

6
Examples of simple harmonic oscillator

A slab supported
A pendulum on two massless
A mass
on a frictionless cantilever springs
hanging on
pivot an ideal spring
7
 Simple harmonic motion(contd…)
The number of variables needed to define the position
Of all parts of a system is known as the degree of
freedom.

If the oscillator is constrained to move only in one

dimension only, or alternatively, if one linear or angular
variable is sufficient to describe the position of the
oscillator, the system is known as a single-degree-of-
-freedom(SDOF) system.

The moving mass in an SDOF system is usually

concentrated at one point and is known as lumped
mass. 8
 Simple harmonic motion(contd…)
Oscillation of the SDOF system as shown in figure is
initiated by displacing and releasing the mass. The
displacement, x, is measured from the equilibrium
position. Once the system has been displaced and
released, no further external force acts on it. Because
there is no friction once it is set in motion, the mass
remains in motion indefinitely.

9
 Stiffness and flexibility

When a force, F, acts on an ideal linear spring, Hook’s

Law predicts the magnitude of the spring deflection, x.

F  kx [Hook ' s law ]

In this equation, k is the stiffness or spring constant

in N/m (lbf/ft). The stiffness is the force that must be
applied in order to deflect the spring a distance of one
unit.
10
Stiffness and flexibility(contd…)
In the spring-mass system, the spring is
undeflected until the mass is attached to it.
After the mass is attached, the spring will
deflect an amount known as the static
deflection, xstatic.
W  mg  kx static
The stiffness, k, of a beam can be calculated
as the ratio of applied force to deflection.

Flexibility is the reciprocal of stiffness. It is the

deflection obtained when a unit force is applied.
Therefore, its units are m/N (ft/lbf).
11
 Rigidity

Rigidity, R, is the reciprocal of deflection.

1
R
x

12
 Natural period and frequency
The time for a complete cycle of oscillation of an
SDOF system is known as the natural period, T, usually
Expressed in seconds.

The reciprocal of natural period is the linear natural

Frequency, f, usually called natural frequency or just
Frequency, and is expressed in Hz (I.e., cycles per
Second).
1
f
T
13
Natural period and frequency(contd…)
The natural frequency can also be expressed in
Radians per second (rad/s), in which case it is known as
the circular frequency, angular natural frequency, or
Just angular frequency, .
2
  2f 
T
For a mass on a spring, and from Hook’s law

k F
 
m x st m
14
 Damping
Damping is the dissipation of energy from an
Oscillating system, primarily through friction. The kinetic
Energy is transformed into heat. All structures have their
Own unique ways of dissipating kinetic energy, and in
Certain designs, mechanical systems known as dampers
Can be installed to increase the damping rate.

15
Oscillator with damping
 Damping(contd…)
There are several sources of damping.

External viscous damping is caused by the structure

Moving through surrounding air (or water, in some cases).
It is generally small in comparison to other sources of
Damping.
Internal viscous damping, commonly the only type of
Damping actually modelled,is related to the viscosity of
The structural material. It is proportional to velocity.

Body friction damping, also known as Coulomb friction,

Results from friction between members in contact. It
Includes friction at connection points. 16
 Damping(contd…)
Radiation damping occurs as a structure vibrates and
Becomes a source of energy itself. Some of the energy is
reradiated through the foundation back into the ground.

Hysteresis damping occurs when the structure yields

during reversals of the load.
For internal viscous damping, the damping force
Opposing motion is given by equation.

Fdamping  cv
The coefficient,c, is known as the damping coefficient.
17
Damping ratio
An oscillating system with a small amount of damping
Will continue to oscillate, although the amplitude of the
Oscillation will decay. Many cycles and a long time may
Elapse before the system eventually reaches the
Motionless equilibrium position. This type of system is
known as an underdamped system.

18
 Damping ratio(contd…)
A system which has a large amount of damping, when
Displaced, seems to “hang in space” taking an extremely
Long time to return to motionless equilibrium position.
This system is known as “overdamped system.

19
 Damping ratio(contd…)

In critical damping, the system is brought to equilibrium

In a minimum time without oscillation. In this case, the
damping coefficient, c, is known as the critical damping
coefficient, ccritical.

Most systems are not critically damped. The ratio of

the actual damping coefficient to the critical damping
Coefficient is known as the damping ratio, 

c

c critical
20
 Damping ratio of buildings
The exact damping ratio, , of an actual structure is
Difficult to determine. Available data on actual structures
Suggest the values given in table. There is little evidence
To support damping ratios in real structures that exceeds
15%.

Although the damping ratio is essentially constant for a

given building, the damping ratio of a particular building
Type or construction material appears to depend on the
natural period of the building. Buildings with natural
Periods of less than 1.0 s may have damping ratios two
To Three times higher than buildings with similar
Construction but natural periods greater than 1.0 s.
Building’s damping ratio, period and construction method
21
Are all related.
 Damping ratio of buildings(contd…)
Typical damping ratios
Types of construction  Types of construction 
Steel frame, welded 0.02 Steel frame, welded 0.05
connections, flexible connections, normal
walls floors, exterior cladding
Steel frame, bolted 0.10 Concrete frame, flexible 0.05
connections, normal internal walls
floors, exterior cladding
Concrete frame, 0.07 Concrete frame, concrete 0.10
flexible internal walls, or masonry shear walls
exterior cladding
concrete or masonry 0.10 Wood frame and shear 0.15
shear walls wall 22
 Damped period of vibration
The period of oscillation of a system will be slightly
Greater with damping than without it, since the damping
slows down the movement.Equations give the damped
frequency and period.

d   1   2

2
Td 
d

23
 Forced system
A forced system is an oscillatory system that is
supplied energy on a regular, irregular, or random basis.
The force that supplies the energy is known as a
forcing function. Forcing functions can be constant(I.e., a
step function), applied and quickly removed(I.e., an
impluse function), sinusoidal or random.
Examples of forced system
An example of a regularly forced system is a flexible
floor supporting an out-of-balance motor. When turning,
the motor will generate a force at a frequency
proportional to the motor’s rotational speed. An example
of a randomly forced system is a structure acted on by
wind or seismic forces. In this case, there is little or no
24
regularity to the applied forces.
 Forced system(contd…)
It is not significant whether a lateral force (e.g.,seismic
force or wind) is applied to a building directly or whether
the base moves out from under the building(e.g., as in
an earthquake). In the latter case, the equivalent lateral
force is an inertial force, but it is just as effective at
displacing the building relative to its base as any direct
force is.
The system response (I.e., the behaviour of a building)
to a force depends on the nature of the forcing function.
Unfortunately, earthquakes are never simple sinusoids
and buildings have more than a single degree of
freedom, so the determination of system response is
time-consuming and complex.Computers and numerical
techniques, however, greatly simplify the analysis. 25
 Magnification factor
When a sinusoidal forcing function with the form
F(t)=Psinft is applied to a system with stiffness k, the
steady state response will be of the form of equation
P
x ( t )    sin f t
k

Here P/k is the static deflection, xstatic, that is

experienced if a constant force P is applied to the
system.  is dynamic magnification factor that depends
on all other characteristics of the system,e.g. natural and
forcing frequencies, the mass in motion, the amount of
damping.
26
 Resonance
For a given system,the dynamic magnification factor,
, can be less than or greater than unity,depending on
the ratio of the natural and forcing frequencies. Fig
illustrates how the magnification
factor varies for different
Frequency ratios. At one point,
corresponding to where
the forcing function frequency
equals the natural frequency
of the system, the magnification
factor is very large(theoretically
infinite for undamped systems). Such a condition is
known as resonance. The ratio f/ must be greater
than 2 for  to drop below 1.0 27
 Examples of resonance

The 1985 Mexico City magnitude 8.1 earthquake

occurred on September 19, with a 7.5 aftershock
occurring the next day. Approximately 400 buildings
were destroyed and 700 were damaged. The death toll
was over 5000. The earthquake consisted of
(approximately) twenty 0.18g pulses coming every 2s
(the natural period of the ground). This coincided with
the period for buildings in the 7-20 storey range. The
resulting resonance related yielding was the primary
cause of structural failure. Quality of construction was
not a major factor in the widespread destruction.

28
 Examples of resonance(contd…)

Resonance is considered a prime factor in the

collapse of the Oakland Interstate 880 Cypress structure
during the October 17, 1989 Loma Prieta earthquake.
The structure had a natural frequency of 2-4 Hz, which
coincided with the 3-5 Hz natural period of the deep
mud that underlaid piles that supported portions of the
freeway that collapsed. The depth of the mud and the
length of the piles varied between 20-80 ft. The natural
period also varied. Portions of the freeway built on
harder alluvial sediments remained standing.

29
 Impulse response-undamped system
Seismic energy is applied to a structure in a
nonregular manner. A Fourier analysis can be used to
analyse the structure response, it is also possible to
break the irregular seismic loading into a series of short
duration rectangular impulses. An impulse is a force,F,
that is applied over a duration, dt, that is much less than
the natural period,T, of the structure. The product Fdt is
1.0 for a unit impulse.
From equation, the same response will be achieved
from all short duration impulses (sine, rectangular,
square, triangular, random, etc.) that has the same
value of  Fdt . Notice that the response is sinusoidal even
though the loading is not
x(t) 
1
 Fdt sin t (undamped )
m
30
 Multi degree of freedom systems
A system with several lumped masses, such as a
building with multiple concrete floors supported by steel
columns is multi degree of freedom(MDOF) system.
A MDOF system has as many ways of oscillating as
there are lumped masses. These “ways” are known as
modes.
Each mode has its own characteristics mode shape
and natural frequency of vibration, each being some
multiple of the previous mode’s frequency.
The mode with the longest period is known as the first
or fundamental period. Higher modes have higher
frequencies (smaller periods), and the period decreases
rapidly from the fundamental mode. 31
Multi degree of freedom systems(contd…)
For example, for a typical high rise building with a
uniform plan view and a moment resisting frame, the
decrease is in the order of 1, 1/3, 1/5, 1/7, 1/9 and so on.

32
 Response of MDOF

Each modal frequency results in a specific mode

shape, However, an earthquake contains waveforms
with varied frequency content; therefore, all of the
modes may be present simultaneously in an
earthquake. This makes it difficult to determine the
building’s response.

Since MDOF response can be determined as the

superposition of many SDOF responses, matrix analysis
(on a computer) can be used to evaluate MDOF
systems based on the equivalent SDOF performances.
33
 Response of MDOF(contd…)
As with SDOF systems, considerable simplifications
can be achieved by limiting the analysis to the maximum
deflections. However, even this simplification requires a
probabilistic analysis because the modal maxima do not
occur at the same time, nor do they necessarily have
the same sign.
Various approximation formulas are used to combine
the modal maxima, and the sum of the squares
approximation is commonly quoted. If the maximum
displacements, i, are known for the first n modes for
some particular point(e.g. the top storey), equation
usually gives a conservative estimate of the total
displacement.
t   2
i 34
 Response of MDOF(contd…)
The method of combining modal responses by taking
the square root of the sum of the squares is referred to
in the UBC as the SRSS method.because the modal
maxima do not occur at the same time, nor do they
necessarily have the same sign.
Theoretically, all mode shapes must be included in the
summation, but, in practice, most of the vibration energy
goes into the first three to six modes, and higher modes
can be disregarded. (with the use of a computer,
however, there is no need to stop with such a small
number of modes).

35
 Response of MDOF(contd…)
Since the lower modes dominate, the response
spectra for MDOF systems are similar to those of SDOF
systems. For short periods(e.g., less than 1s), the
MDOF response is usually slightly less than the first
mode SDOF systems. For periods exceeding 1s, the
response usually slightly exceeds SDOF response.

The UBC requires that all significant modes be

included. This can be accomplished by making sure that
for all modes considered, at least 90% of the mass of
the structure is included in the calculation of response
for the horizontal direction being investigated.
36
 Mode shape factor
The modes shape factor, , are relative numbers that
represents the ratios of the storey deflections (from the
equilibrium position) to some common basis,usually the
deflection of the first or last storey.Since mode shape
factors are relative, they can usually be determined by
initially assuming a value of one of the deflections.
xi
i 
x1
In some cases, the mode shape
factor are normalized by dividing by
 i i . Then the following
m  2

equation will be valid

 i i 1
m  2
37
 Rayleigh method
A significant computational burden is required for full
dynamic analysis for even a simple MDOF system.
Sometimes it is unnecessary or inappropriate to carry
out such sort of analysis. For such situations, it is
possible to use iterative procedures like Rayleigh
method.
In this method, first a mode shape is assumed. Even
poor initial assumptions converge rapidly to the correct
answer.
Then the maximum kinetic energy is set equal to the
maximum potential energy.

Eventually, the mode shape is calculated and used as

the starting point for the subsequent iteration. 38
 Stodola method
The stodola method consists of the following steps.
Step-1: Assume a mode shape. That is, assume a
deflection, x, for each mass.
Step-2: Compute the inertial forces for each mass
from equation.
Finertial  m2x
Step-3: Compute the spring forces on each mass as
the sum of the inertial forces acting on the springs.

Step-4: Compute the spring deflections.

Step-5: Calculate the mode deflections from the spring

deflections. Repeat from step as required. 39
 Participation factor

The participation factor, j, is the fraction of the total

building mass that acts in any particular mode, j. It can
be used to calculate the storey drift, x.

The denominator of this equation is the same as the

previous equation and will be equal to 1.0 if normalized
mode shape factors are used

j 
 m i ij

m  i
2
ij

x  S d 
40
 Participation factor(contd…)

The participation factor can also be used to calculate

the floor force, Fx, that acts at storey x (i.e., the force
that acts at that level.) and the cumulative storey shear,
Vx, that acts at that level and above. This can be done in
two ways, one method derived from Hook’s law and
using the spring constant, and the other method derived
from Newton’s law and using the mass.

Fx  mS a   kSd 

41