Vous êtes sur la page 1sur 10

Experimental Vibration Analysis for Civil Engineering Structures

337

Comparison of experimental and analytical data for a reinforced


concrete chimney
A.K. Antonopoulos & C.A. Syrmakezis
Institute of Structural Analysis and Aseismic Research, School of Civil Engineering, Athens, Greece

ABSTRACT: This paper presents the comparison of the results, of the experimental measurements of the constructions behavior during an actual dynamic excitation and the results of time
and frequency domain analysis of the models construction behavior by using finite element
method in a real reinforced concrete chimney 135.00 meters high.
1 INTRODUCTION
Chimneys used in the past or even those being used currently in the industry, had to be designed
elastically even for large seismic forces. The most usual materials for chimneys are concrete,
steel and brick, combined with different insulation and lining materials. These structures are
usually high and the appropriate size for each chimney is mainly selected to satisfy the draught
requirements of the boilers and the needs of the industrial processes (Pinfold, 1984).
These structures can have different forms, simple prismatic and truncated cone cantilevers, as
shown in Figure 1. Chimneys are continuous systems because mass and stiffness are distributed.
The response of a continuous system is given by the solution of the governing differential equations, which must satisfy certain prescribed boundary conditions of the system domain.

Figure 1. Main types of chimneys forms.

For these types of structures dynamic analysis is used, where superior eigenmodes are taken
into consideration to calculate the total amount of the structural response under an actual dynamic excitation. Such an analysis has to take into consideration the effects of the bending and
shear deformation, the soil-structure interaction, the effect of the rotational inertia and the effect
of insulation in the total damping coefficient of structure.
A recording process concerning the excitations as well as the response of the structure is applied first for a long period of time by using tri-axial acceleration sensors and vibration record-

338

EVACES07

ers on selected points. Both the tri-axial acceleration sensors as well as the vibration recorders
are monitored by network controllers. The processing of the results using analysis in frequency
domain is followed by the calculation of the structures dynamic characteristics, such as eigenfrequencies and eigenperiods. The damping coefficient of the structure is defined through the
displacement or acceleration measurements u1 and u1+j or 1 and 1+j, respectively, among j successive vibrations measurements from an impulsive excitation.
At the next step, the structure is discretized using a finite element mesh of three-dimensional
solid elements which reproduces satisfactorily the actual structures properties in an adequately
reliable way. The chimneys behavior is examined using time and frequency domain analysis
under an actual dynamic excitation. This process leads to the calculation of the dynamic excitations frequencies and therefore of those that cause the most unfavorable excitations.
2 DESCRIPTION OF THE STRUCTURE
A reinforced concrete chimney 135 meters high is situated in the major area of Piraeus harbor,
Greece. It was built in 1976, by a Greek industry to serve factory of fertilizers. The chimney
was in use at least for twenty years, until the closure of the factory. Nowadays the factorys area
and its surrounded are being rehabilitated. The chimney actual state is illustrated in Figure 2.

Figure 2. A reinforced concrete chimney, in Piraeus, Greece.

According to the structures drawings, found among the general files of the factory records,
the chimney considered has the form of a truncated cone shell, internal diameter of 6.40 m at the
base and of 2.90 m at the top. The thickness of shell at the base is 0.60 m and at the top is 0.20
m. The chimneys thickness and diameter are varying linearly along the structures height. As
shown in Figure 3, the interior of the shell consists of the following successive layers: 0.35 m of
glassfibre and 0.12 m of acid - resisting brick. Between these two layers of insulation and the
outer reinforced concrete shell there is a gap of 0.60 m.
The chimneys foundation is constituted by two parts of massive reinforced concrete, one of
cylindrical shape and one of truncated cone shape. The first (lower) part has a diameter of 20.00
m and is 1.60 m high. The second (higher) part has a maximum diameter of 20.00 m, a minimum of 3.00 m and is 1.40 m high.

339

Experimental Vibration Analysis for Civil Engineering Structures

Reinforced Concrete
Gap
Glassfibre
Acid-resisting Brick

0.12

0.35 0.60

0.60

Figure 3. Detailed view of this section.

3 EXPERIMENTAL MEASUREMENTS
Tri-axial acceleration sensors and vibration recorders are installed at the mid height of the chimney and are exposed to an ambient and an impulsive excitation for a long period of time. Parameter settings trigger level to each direction, the pre-event and post-event time, and the sampling rate are applied in the vibration recorders.
Figure 4 shows a typical record of an ambient vibration of one of the two horizontal directions, which is caused by a natural wind of the area. Although the vibration level is very low
below perception threshold the signal can be measured and the eigenfrequency can be accurately determined (Syrmakezis 2005).
The analysis in frequency domain gives an additional possibility to improve the determination of the eigenfrequencies. From the amplitude spectra, as shown in Figure 5, the engenfrequencies of the structure are obtained, it provides clearly the representation of the frequency
content of the forcing function and the frequencies whose influence on the structure is the most
intense (Bracewell 2000).

Acceleration (mm/s2)

3.00
2.00
1.00
0.00
-1.00
-2.00
-3.00
0

Time (sec)
Figure 4. Ambient vibration of the structure.

10

Displacement (mm)

340

EVACES07

0.0005
0.0005
0.0004
0.0004
0.0003
0.0003
0.0002
0.0002
0.0001
0.0001
0.0000
00

20
2

40
4

606

80
8

100
10

Frequency (Hz)
Figure 5. Amplitude spectra of ambient vibration of the structure.

Impulsive excitation is a convenient way to determine the eigenfrequencies and damping of a


structure. Figure 6 shows a record of impulsive vibrations which are caused by the Schmidt
concrete test hammer during non-destructive testing.
The analysis in frequency domain gives an additional possibility to improve the determination of the eigenfrequencies. From the amplitude spectra, as shown in Figure 7, the engenfrequencies of the structure are obtained.
The damping ratio it is not possible to determine analytically for practical structures, this
elusive property should be determined experimentally. Free vibration experiments provide one
means of determining the damping. Over j cycles the motion decreases from u1 to uj+1. This ratio
is given by equation 1 (Chopra 2001).
The experiments in chimney model led to the free vibration records presented in Figure 8,
which is the same with Figure 6 but for a different scale and period of time. For damped systems the damping ratio can be determined from equation 1.

u uj
ui
u u u
= 1 2 3 ... 3
= e j
u j+1 u 2 u 3 u 4 u 4 u j+1

(1)

= (1 j) ln ( u1 u j+1 ) 2

(2)

The experiments in chimney model led to the free vibration records presented in Figure 8,
which is the same with Figure 6 but for a different scale and period of time. For damped systems the damping ratio can be determined from equation 3 (Chopra 2001).
=

u
1
ln i
2j u i+j

or

&&
u
1
ln i
2j &&
u i+j

(3)

The first of these equations is equivalent to Equation 1. The second is a similar equation in
terms of accelerations, easier to measure than displacements.
The damping coefficient of the structure is defined through the acceleration measurements 1
and 1+j among j successive vibrations and has the value of 1.5%, as shown in Figure 8.
The natural period T of the system can also be determined from the free vibration record by
measuring the time required to complete one cycle of vibration. Comparison with the natural period obtained from the calculated stiffness and mass of an idealized system shows that these
properties were calculated accurately and the idealization representing the actual structure is satisfactory.

341

Experimental Vibration Analysis for Civil Engineering Structures

Acceleration (mm/s2)

30.00
20.00
10.00
0.00
-10.00
-20.00
-30.00
0

10

80
8

100
10

Time (sec)
Figure 6. Impulse vibration of the structure.

0.0004
Displacement (mm)

0.0004
0.0003
0.0003
0.0002
0.0002
0.0001
0.0001
0.0000
0

20
2

40
60
4
6
Frequency (Hz)

Figure 7. Amplitude spectra of impulse vibration of the structure.

Acceleration (mm/s2)

30.00
20.00

=1.5%

10.00
0.00
-10.00
-20.00
-30.00
14

14.2

14.4

14.6

Time (sec)
Figure 8. Impulse vibration of the structure.

14.8

15

342

EVACES07

4 ANALYTICAL RESULTS
At a next step the structure is discretized using a finite element mesh of three-dimensional solid
elements which reproduces satisfactorily the actual structures properties in an adequately reliable way (Zienkiewicz & Taylor 1987). Taking into consideration the soil-structure interaction,
the elastic half-space is discretized using three-dimensional solid elements is shown in Figure 9.
Each node of finite elements activates three degrees of freedom, but the nodes on the circumferences of the cycle that defines the elastic half-space are constrained against horizontal translations and all the nodes at the base of the elastic half-space are considered to be constrained
against horizontal and vertical translations.
The assumed moduli of elasticity and poisson ratio of the chimney and the ground are E=32
GPa, n=0.2 and E=10 GPa, n=0.3 respectively.

2008
2007
2006

2005
2004
2003

2001

2002

Figure 9. The model of the structure.

The loads of glassfibre and acid - resisting brick have been taken into account as concentrated
masses at the nodes of the finite elements.
The eigenfrequencies were calculated through a procedure of successive parametric and probabilistic solutions. The necessity of this procedure lies on the existence of various uncertainties
that have to be taken into consideration these uncertainties are related to the geometry of the
structure, the modulus of elasticity (varies along the height), the damping coefficient, the effect
of the insulation and the poisson ratio (Antonopoulos 2005, Antonopoulos & Syrmakezis 2006).
The successive solutions reveal the extent of the influence of each parameter on the frequency of the structure. This process leads to the calculation of the dynamic excitations frequencies.

343

Experimental Vibration Analysis for Civil Engineering Structures

Table 1. Frequency of the model of the structure


Structures Modulus
Cases
of Elasticity (GPa)
32
Without insulation
34
36
32
With insulation
34
36
Table 2. Frequency of the model of the structure
Soils Characteristics
Cases
E (GPa)
1
Structures Modulus
of Elasticity 32GPa,
5
n=0.2
10

Frequency (Hz)
0.42
0.43
0.45
0.33
0.34
0.35

1.61
1.63
1.69
1.27
1.31
1.35

3.84
3.86
3.98
3.03
3.13
3.23

7.13
7.14
7.24
5.55
5,64
5.88

Frequency (Hz)
0.33
0.33
0.33

1.24
1.26
1.27

2.92
2.94
3.03

4.36
5.44
5.55

Table 3. Comparison between the measurement and analytical dynamic characteristics


Cases
Frequency (Hz)
F.E Model
0.42
1.61
3.84
7.13
Without insulation
F.E Model
0.33
1.27
3.03
5.55
With insulation
Measurement
0.41
1.55
3.71
7.08
Ambient excitation
Measurement
0.50
1.73
3.95
7.24
Impulse excitation

Table 1 shows the effects of insulation on the first four eigenfrequencies of the structure for a
variety of concretes moduli of elasticity and Table 3 shows the comparison of experimental and
analytical results for the first four eigenfrequencies of the structure.
Below, the chimneys behavior is examined using time and frequency domain analysis for the
1999 earthquake of Athens; Figure 10 depicts the ground motion recorded during this earthquake (Chopra 2001, Humar 2002).
Figure 11 is created using the Fast Fourier Transform and shows the amplitude of the acceleration for each one of the excitation frequencies. Using the amplitude spectrum are determined
those frequencies which influence most unfavorably the excitation of the structure. Figure 12
depicts the signal power of each frequency of the signal (acceleration).
Figure 13 shows the acceleration response spectrum for the 1999 earthquake of Athens
ground motion for various damping coefficients (=1%, 3%, 5%).

Acceleration (m/sec2)

3.0
2.0
1.0
0.0
-1.0

10

20

30

40

-2.0
-3.0

Time (sec)

Figure 10. Acceleration recorded during the 1999 earthquake of Athens.

344

EVACES07

Acceleration (m/s2)

0.05
0.04
0.03
0.02
0.01
0
0.0

10.0

20.0
Frequency (Hz)

30.0

40.0

Figure 11. Amplitude spectrum for the 1999 earthquake of Athens ground motion.

Acceleration (m/s2)

0.04
0.03
0.02
0.01
0
0.0

10.0

20.0
Frequency (Hz)

30.0

40.0

Figure 12. Power spectrum for the 1999 earthquake of Athens ground motion.

Acceleration (m/sec2)

10.00
8.00
6.00
4.00
2.00
0.00
0.0

0.5

1.0
Time (sec)

1.5

2.0

Figure 13. Acceleration response spectrum for the 1999 earthquake of Athens ground motion for various
damping coefficients (=1%, 3%, 5%).

345

Experimental Vibration Analysis for Civil Engineering Structures

Displacement (m)

0.2
0.1
0
-0.1
-0.2
0

4
6
Time (sec)

10

Figure 14. Displacement Response at the top of the structure for =1%, 3%, 5% damping coefficients.

Figure 14 shows the displacement response at the top of the structure for various damping
coefficients (=1%, 3%, 5%). The changes of the response prove the major importance of the
damping coefficient. It is quite clear that when it comes to existing structures, it is vital to confirm the analytical data concerning the damping coefficient with experimental measurements in
order to minimize the uncertainties included and the possibility of fatal errors as well.
5 CONCLUSIONS
In this paper the experimental measurements of the chimney and the dynamic analysis in the
time and frequency domain have been presented and compared.
The processing of the results using analysis in frequency domain is followed by the calculation of the structures dynamic characteristics, such as eigenfrequencies and eigenperiods. Comparative calculation of the frequencies of the structure for the two different dynamic excitations
took place. Ambient excitation is caused by wind forces and impulse excitations are caused by
Schmidt hammer impact at the bottom of the structure.
The damping coefficient of the structure is defined through the displacement or acceleration
measurements u1 and u1+j or 1 and 1+j, respectively, among j successive vibrations measurements from an impulsive excitation. The impulse excitations produce a diminishing response,
because their influence is not continuous like the winds, and therefore allow the calculation of
the damping coefficient of the structure.
Using a three dimensional mesh of finite elements, the chimney was discretized, and the eigenfrequencies were calculated through a procedure of successive parametric and probabilistic
solutions. The necessity of this procedure lies on the existence of various uncertainties that need
to be taken into consideration. These uncertainties are related to the geometry of the structure,
the modulus of elasticity (varies along the height), the damping coefficient, the effect of the insulation and the poisson ratio. The successive solutions reveal the extent of the influence of each
parameter on the frequency of the structure. This process leads to the calculation of the dynamic
excitations frequencies.
The aim of the frequency domain analysis is to determine the frequencies of the signal which
cause the most unfavorable excitations to the structure. In this paper the accelerogram which
provided the signal is the one of the 1999 earthquake of Athens. The frequency and power spectra are determined with the help of the Fast Fourier Transform.
The acceleration response spectrum derives as well from the accelerogram of the 1999 earthquake of Athens and is presented for a variety of damping coefficients.

346

EVACES07

The need for a realistic identification of the structure and error limitation is very important.
Therefore results of both experimental and analytical methods as well as their comparison
should be performed before any intervention measures are taken.
REFERENCES
Antonopoulos, A.K. 2005. Comparative Explorations of Analysis Methods for the Dynamic Behavior of
Structure with Continuously Contributed Mass and Stiffness. Master Thesis.
Antonopoulos, A.K. & Syrmakezis, C.A. 2006. Dynamic Behaviour of Castellated Reinforced Concrete
Structures with Continuously Distributed Mass and Stiffness: Comparative Exploration Design Methods. Sixteenth Greek Conference on Reinforced Concrete Structure. Volume 4. Pages: 3-14.
Bracewell, N.R. 2000. Fourier Transform and its Applications. New York: McGraw-Hill Companies.
Chopra, A.K. 2001. Dynamics of Structures. New Jersey: Prentice Hall.
Humar, J.L. 2002. Dynamics of Structures. Tokyo: Balkema Publisher Lisse.
Syrmakezis C.A. 2005. Seismic Protection of Historical Structures and Monuments. Structural Control
and Health Monitoring. Volume 13. Issue 6. Special Issue: Third European Conference on Structural
Control. Pages: 958-979.
Pinfold, G.M. 1984. Reinforced Concrete Chimneys and Towers. London: Palladian Publications LTD.
Zienkiewicz, O.C. & Taylor, R.L. 1987. The finite Element Method, Basic Formulation and Linear Problems. London: McGraw-Hill Book Company.

Vous aimerez peut-être aussi