Engineering Structures 29 (2007) 15251538

www.elsevier.com/locate/engstruct

Vibration control of steel jacket offshore platform structures with damping

isolation systems
Jinping Ou a, , Xu Long a,b , Q.S. Li c , Y.Q. Xiao d
a School of Civil Engineering, Harbin Institute of Technology, Harbin, China
b School of Civil Engineering, Shenzhen University, Shenzhen, China
c Department of Building and Construction, City University of Hong Kong, 83 Tat Chee Avenue, Kowloon, Hong Kong
d Shenzhen Graduate School, Harbin Institute of Technology, Shenzhen, China

Received 31 January 2006; received in revised form 1 August 2006; accepted 29 August 2006
Available online 23 October 2006

Abstract

In this paper, a damping isolation system is developed for vibration control of steel jacket offshore platform structures. At first, a damping
isolation system composed of rubber bearings and viscous dampers is proposed for vibration mitigation of JZ20-2MUQ jacket platform located in
Bohai Sea. The influence of key parameters of the damping isolation system on the vibration suppression of the offshore structure is studied
in detail. In order to examine the vibration control effectiveness of the damping isolation system for the jacket platform, a comprehensive
experimental and numerical study is carried out. The experimental investigation is conducted on a shaking table for a 1/10-scale JZ20-2MUQ
platform model subjected to four different earthquake ground motions. The performance of the controlled structure under the actions of ice and
seismic-induced forces is then studied by numerical simulations. The experimental results are found to be in good agreement with those from
the numerical simulations. It is shown from the combined experimental and numerical study that the damping isolation system is effective in
mitigating the dynamic responses of the offshore platform structure.
c 2006 Elsevier Ltd. All rights reserved.

Keywords: Structural control; Jacket offshore platform; Damping isolation system; Shaking table test; Earthquake action; Ice-induced vibration

1. Introduction as Bohai Sea and the northern part of the Yellow Sea in
China, floating ice in sea may generate significant dynamic
The steel jacket structure which originated in the Gulf of loads on offshore platforms. The dynamic loads affect not only
Mexico and which has already spread worldwide is a typical the routine operation of an offshore platform such as drilling
type of fixed offshore platform. It is suitable to be built in water and production activities, but also the safety and serviceability
depth from a few meters to more than 300 m. Major structural of the structure. It is indispensable to reducing the overall
components of such an offshore platform are jacket, piles, and response of a jacket platform subjected to strong dynamic loads.
deck. A jacket structure which serves as bracing for the piles In general, reduction of the dynamic stress amplitude of an
against lateral loads is fixed by piles driven through the inside offshore structure by 15% can extend the service life over
of the legs of the jacket structure and into soil for many tens of two times, and can result in decreasing the expenditure on the
meters. The deck structure is fixed upon the jacket structure. maintenance and inspection of the structure.
Oceans in which offshore platforms are built present a set Vibration control technologies, which have achieved
of complicated and harsh environmental conditions. Dynamic significant success in mitigating vibrations of land-based
loads including wind, wave, current, and earthquake dominate structures under strong wind or earthquake actions over the past
the design of offshore structures. In high latitude regions, such two decades [1], have rarely been applied to offshore platforms
except in the following several cases. Vandiver and Mitome [2]
Corresponding author. Tel.: +86 451 86282094; fax: +86 451 86282094. utilized storage tanks on a fixed platform as tuned liquid
E-mail addresses: [email protected] (J.P. Ou), [email protected] dampers (TLD) to suppress wave-induced structural vibration
(Q.S. Li). response. Lee [3] inserted viscoelastic materials in jacket legs

c 2006 Elsevier Ltd. All rights reserved.

0141-0296/$ - see front matter
doi:10.1016/j.engstruct.2006.08.026
1526 J.P. Ou et al. / Engineering Structures 29 (2007) 15251538

Fig. 2. Detail of the damping isolation system.

which is provided by the damper at the isolation level,

can reduce vibration response of the platform effectively.
Unlike traditional isolation strategies, which aim at shifting
the fundamental natural frequency of a structure away from
the dominant frequencies of seismic excitations, the present
Fig. 1. Damping isolation system for jacket offshore platforms. damping isolation system aims at dissipating the vibration
energy at the isolation level.
to mitigate the vibration of an offshore structure subjected As mentioned previously, the damping isolation system
to random wave forces. Vincenzo and Roger [4] developed is composed of isolators and dampers, as shown in Fig. 2.
an adaptive control technique (Active Mass Damper) for Rubber bearing is regarded as one of the most popular isolators
suppression of vortex-induced vibrations of offshore structures. used in structural control. Various kinds of dampers have
Since 1997, aiming at mitigating the ice-induced vibration been developed and some have been applied in practice,
of offshore platforms in Bohai Sea, Ou et al. [5] numerically which include viscous, viscoelastic, magneto-rheological (MR)
investigated the ice-induced structural vibrations using in- and shape memory alloy (SMA) dampers that have good
situ measured ice force data, and also conducted numerical behavior of corrosion resistance. In addition, MR damper and
and experimental studies on ice-induced vibration control of SMA damper may be used as backup systems to restrict the
offshore structures by adding viscoelastic dampers and viscous relative displacements at the isolation level within the clearance
dampers, respectively [6,7]. Their results showed that due to provided around the structure.
the limitations on the placement positions of the dampers Stiffness and damping are important parameters for
and relatively low vibration amplitude of the platforms, the designing a damping isolation system. These parameters can
damping ratios added by the dampers were relatively small, be determined according to the vibration control objectives.
and the effectiveness of adding the dampers to the offshore
structures for the purpose of reducing their vibration responses 3. Parametric study
was not remarkable.
In order to reduce possible damages to jacket offshore The objective of the present study is to develop an efficient
platforms in harsh marine environments, the necessity of and practical damping isolation system to mitigate earthquake
carrying out further studies on developing efficient and and ice-induced vibrations of jacket offshore platforms. JZ20-
practical vibration control strategies for suppression of dynamic 2MUQ platform located in Bohai Sea is considered herein as an
responses of offshore structures should be stressed. In this example to examine the effectiveness of the proposed vibration
study, a damping isolation system is developed to mitigate control strategy for offshore structures.
the vibration of fixed jacket offshore platforms. The damping In this section, a parametric study for the jacket offshore
isolation system is composed of isolators (rubber bearings) and platform with installation of the damping isolation system is
viscous dampers. The contents of this study mainly include: presented.
(1) the investigation of the influence of the damping isolation
system parameters on vibration control of offshore platforms 3.1. Description of the jacket platform
under the actions of ice loads and earthquake excitations, (2) the
evaluation of the performance and effectiveness of the damping 3.1.1. Structure and geometric features
isolation system subjected to earthquake excitations through JZ20-2MUQ platform is located in Bohai Sea which is
shaking table tests. the southernmost sea in the Northern Hemisphere that freezes
each winter. The peak acceleration of the design earthquake
2. Damping isolation system of jacket offshore platform excitation for the offshore structure is 0.19g, and the water
depth considered in the design is 15.5 m [8].
The proposed strategy for vibration suppression of a jacket JZ20-2MUQ is a four-leg jacket platform, as illustrated
offshore platform is that a damping isolation system is placed in Fig. 3. It consists of two main components: substructure
between the bottom of the deck and above the jacket structure; and superstructure. The superstructure is supported on a deck,
such a location is called the damping isolation level in this which is fixed on the substructure. The superstructure is
paper, as shown in Fig. 1. Relatively large deformations occur composed of the living module and the utility module. The
at the isolation level, additional energy dissipation capacity living module, which is situated at the top of the platform
 J.P. Ou et al. / Engineering Structures 29 (2007) 15251538 1527

Table 1
Calculated natural frequencies of JZ20-2MUQ platform

Mode number 1 2 3 4 5 6 7 8 9 10
Frequencies (Hz) 0.633 0.738 0.788 1.834 1.857 2.558 2.609 2.614 3.024 3.320

Fig. 4. Elevation view of JZ20-2MUQ platform.

Fig. 3. Picture of JZ20-2MUQ platform.

Table 2
Measured natural frequencies and damping ratio of JZ20-2MUQ platform
from EL. 29 m to EL. 39.758 m, is a 3-story steel space
frame structure. The utility module made of rolled shape steel Mode number 1 2 3
and round steel tube is a 2-story steel space frame from EL. Frequencies (Hz) Measured 1 [9] 0.90 1.05 1.15
16.5 m to EL. 28.5 m. The substructure is a steel tubular jacket Measured 2 [12] 0.85 1.12
structure. Under sea level, jacket has two level units at EL. Damping ratio [9] (%) 4.05.5
15.5 m and EL. 3.5 m, respectively. There are diagonal
brace members in both vertical and horizontal planes in the measured natural frequencies for the first three modes are larger
units to enhance the structural stiffness. Above sea level, the than those calculated.
jacket also has two level units at EL. 5.85 m and EL. 10 m,
respectively. There is no diagonal brace installed in the platform 3.1.3. Simplified model of the platform
from EL. 3.5 m to EL. 5.85 m for the purpose of reducing the Ou et al. [10] carried out a numerical study on the ice-
ice-induced forces acting on the platform. The elevation view induced vibrations of the JZ20-2MUQ platform with energy
of JZ20-2MUQ platform in the longitudinal (X ) direction is dissipation devices using the finite element method. According
shown in Fig. 4. to the distribution features of the mass and stiffness of the
jacket structure, the structure can be simplified as an equivalent
3.1.2. Dynamic characteristics lumped mass system with a single degree of freedom for forced
Considering the effective length, effective cross-sectional vibration analysis. The mass values of the superstructure and
area and effective moment of inertia of piles due to non- of the jacket and pile are 2576.8 t and 550 t, respectively.
elastic soilpile interaction, the dynamic characteristics of The fundamental natural frequency and damping ratio in the
JZ20-2MUQ platform were analyzed using SACS (Structure longitudinal (X ) direction are 0.87 Hz and 4%, respectively.
Analysis Computer System) software. The computational Knowing the mass, the fundamental natural frequency and
results are summarized in Table 1 [8]. The in-situ measured damping ratio, the systems stiffness k and damping coefficient
natural frequencies and damping ratios of the structure are c can be determined accordingly [11]. The parameters of the
summarized in Table 2 [9]. It is maintained that the differences structure simplified as a single degree of freedom system in X
between the calculated and measured natural frequencies are direction are obtained as follows
attributable to several reasons, including that the effective mass m = 3126.8 t, k = 9.3436 104 kN/m,
values of the structure are less than those assumed at the design
c = 1367 kN s/m.
stage and/or the effective stiffness values of the structure are
higher than those determined at the design stage due to the The jacket platform with the damping isolation system is
contribution of non-structural components. As a result, the simplified as a two degree of freedom system, as shown in
1528 J.P. Ou et al. / Engineering Structures 29 (2007) 15251538

Table 3
Parameters of ice-induced forces
Water level Failure mode of ice Ice force time history type Return period (year) Ice thickness (103 m) Maximum static ice force (kN)
EL. +4.000 m Crushing 50 400 3058
Crushing
1 176 1241
EL. +2.500 m Bending 50 400 2841
Bending
1 176 890
EL. +4.000 m Blocking Crushing 50 1000 7988

ice. On the other hand, the failure pattern is usually of crushing

when ice is in collision with the legs. It has been recognized
that when the failure pattern changes from crushing to bending,
the resultant static ice forces may be reduced significantly [5].
These two cases, i.e., ice forces acting on the jacket legs (EL.
4 m) and on the declined cones (EL. 2.5 m), were all considered
in this study. The parameters associated with the ice forces,
which are used in the numerical simulation of this study, are
given in Table 3.
The in-situ recorded time histories of the ice-induced forces
with crushing and bending failure patterns of ice were adopted
Fig. 5. Simplified model of the damping isolated structure.
in the calculation of the ice-induced responses of the structure,
Fig. 5. In the simplified model, the damping isolation system as shown in Fig. 6. The time histories of the ice forces were
is represented by an equivalent spring-damping system. The scaled up to a peak value of the maximum static ice forces listed
stiffness and damping coefficient of the system are represented in Table 3. The adjustments were made based on the return
by k2 , c2 , and the stiffness and damping of the jacket structure period and the ice failure modes.
are denoted by k1 , c1 . m 2 represents the mass of the deck
module above the damping isolation system and m 1 is the mass 3.1.4.2. Earthquake ground motions. Time history analysis of
of the jacket and piles. k2 and c2 are designed based on the the seismic responses of the platform was performed using the
vibration control objectives. El-Centro (1940, SN) and Taft (1952, N21E) earthquake ground
motion records. The peak acceleration in the accelerograms was
3.1.4. Load cases scaled to 0.2g for the seismic analysis according to the seismic
design codes of China.
3.1.4.1. Ice loads. Ice load governs the design of all main
structural members of jacket platforms in Bohai Sea [12]. On 3.2. The effects of the parameters of the damping isolation
each leg of JZ20-2MUQ platform, double-sided ice-breaking system on the vibration suppression
cones were installed near sea level for reducing ice-induced
forces on the jacket structure. When ice comes into collision 3.2.1. Relationship between the structural damping ratio and
with different parts of the cones or legs of the platform, different the parameters of the damping isolation level
kinds of failure patterns result in the ice, and thus the ice- Damping ratio of a structure is an important parameter for
induced forces would be different. For example, ice acting on the structural dynamic analysis, and it characterizes the energy
declined sides of the cones may result in bending failure of the dissipation capability of a structure.

(a) Ice-induced force (generated by bending failure pattern of ice). (b) Ice-induced force (generated by crushing failure pattern of
ice).

Fig. 6. In-situ recorded ice-induced force time histories.

 J.P. Ou et al. / Engineering Structures 29 (2007) 15251538 1529

above the damping isolation level. t is used to reflect the effect

of the stiffness of the damping isolation level on the inter-story
drift.
Figs. 8 and 9 show the effects of 1 and t on the maximum
inter-story drift of the damping isolation level under ice forces
and two earthquake ground motions, respectively. In the figures,
x1 max is the maximum inter-story drift of the damping isolation
level. The results show that x1 max decreases as 1 increases, and
x1 max increases as t increases. In addition, as t increases, the
effect of 1 on x1 max is more significant.

3.2.3. The effects of the parameters of the damping isolation

system on the effectiveness of vibration control
Fig. 7. 1 versus c2 and k2 . The damping ratio of the structure and the stiffness of
the damping isolation level are two main parameters affecting
Fig. 7 presents the relationship among the first mode the effectiveness of the vibration control. The maximum
damping ratio 1 of the structure and damping coefficient c2 acceleration of the deck and the maximum displacement of the
as well as stiffness k2 of the damping isolation level based on jacket cap are of particular interest in this study. We define
the simplified model shown in Fig. 5. It can be seen that as c2
increases, 1 increases linearly. In addition, as k2 decreases, 1 xm x1cm am a2cm
Jd = , Ja = (3)
increases quickly. xm am
where xm and am are the maximum displacement and the
3.2.2. The effect of the parameters of the damping isolation maximum acceleration of the structure without the isolation
system on inter-story drift arrangements (it is called unisolated structure in this paper),
Large inter-story drift of the damping isolation level is respectively; x1cm and a2cm are the maximum displacement of
not allowed for the jacket platform to meet the drilling and the jacket cap and the maximum acceleration of the deck when
production requirements. The damping ratio and stiffness of the damping isolation system is implemented, respectively.
the damping isolation level are major parameters affecting the Figs. 10 and 11 show the effects of 1 and t on Jd and Ja under
inter-story drift of the isolation level. Based on the model shown crushing ice forces and the Taft earthquake input, respectively.
in Fig. 5, a normal period ratio t is defined as follows It can be seen from these figures that Jd and Ja increase with
the increase of t , but are less sensitive to the variation of 1 .
Td
t = (1)
Tu 4. Shaking table test for the platform model
and
,s ,s 4.1. Experimental model
k1 k2
Tu = 2 , Td = 2 (2) The shaking table tests were conducted on a shaking table
m1 + m2 m2
for a 1:10 scale model of JZ20-2MUQ jacket structure, as
where Tu is the fundamental period of the structure without the shown in Fig. 12. The model has a maximum height of 6.46 m.
isolation arrangements, and Td is the period of the structure The weights of the superstructure and the jacket structure

(a) Under ice-induced force (generated by bending failure (b) Under ice-induced force (generated by crushing failure
pattern of ice). pattern of ice).

Fig. 8. The effects of 1 and t on the maximum inter-story drift under ice forces.
1530 J.P. Ou et al. / Engineering Structures 29 (2007) 15251538

(a) El-Centro (1940, SN). (b) Taft (1952, N21E).

Fig. 9. The effects of 1 and t on the maximum inter-story drift under earthquakes.

Fig. 10. The effects of 1 and t on Jd and Ja under crushing ice forces. Fig. 11. The effects of 1 and t on Jd and Ja under Taft (1952, N21E)
earthquake.
models are 0.9 t and 0.7 t, respectively. Additional weights
with total weight of 4.2 t were uniformly distributed in the Reciprocal loading test is a standard method for determining
superstructure. mechanical properties of rubber bearings, which is usually
adopted to evaluate lateral stiffness of rubber bearings. One
4.2. Rubber bearing properties
rubber bearing was tested for its serviceability behavior under
Isolators used in the shaking table test for the platform model horizontal reciprocating distortions in four different strain
are laminated rubber bearings, which were fabricated using ranges (50, 100, 150 and 200%). A pre-load of 75 kN on
15 rubber layers and 14 mild steel layers bonded alternately. a bearing perpendicular to the displacement direction was
The bearings are 110 mm in diameter and 63.5 mm tall. The applied. The shear forcedisplacement curves with different
thickness of each rubber layer is 1.5 mm. Details of a specimen strain ranges are shown in Fig. 14. It is clear that the lateral
are shown in Fig. 13. Four identical rubber bearings were stiffness of the rubber bearing is approximately a constant
mounted between the jacket and the deck. Measurements were determined from the slope of the shear forcedisplacement
made to determine the behavior of the bearings. curves.
 J.P. Ou et al. / Engineering Structures 29 (2007) 15251538 1531

a displacement transducer. The damper force was measured

through a load cell located within the actuator.
Both the force and displacement records were gathered
through a data acquisition system. The data sampling frequency
was 100 Hz. The recorded forcedisplacement relationships
were used to investigate the mechanical properties of the
dampers.

4.3.2. Experimental results

Two dampers were tested under a temperature of about
20 C (normal room temperature). The harmonic motion of
the piston rod varied in the range of 0.12 Hz with amplitude
variation from 6 to 12 mm. Fig. 16 shows six recorded
forcedisplacement loops for one damper. Several features are
observed from these loops:
(a) The devices exhibit, in addition to viscous behavior,
Fig. 12. The platform model on shaking table.
frictional behavior that originates in the seals. The friction
A compression test was also carried out to assess the force was found to be about 1 kN on average through the
vertical stiffness of the rubber bearings. The test results show static tests.
that the vertical stiffness of the rubber bearing is more than (b) The loop curves resemble elliptical shapes, implying the
1000 times greater than its lateral stiffness. The large vertical linear feature of the damper force.
stiffness enables the rubber bearing to withstand heavy vertical
loading. 4.3.3. Mathematical modeling of the dampers
An acceptable model for describing the behavior of the
4.3. Viscous damper properties tested dampers is the viscous model (Eq. (4)). The model is
valid for dampers with viscous behavior, for example fluid
4.3.1. Experimental setup
dampers. The viscous models were calibrated first for the tested
Two fluid viscous dampers operating on the principle of flow
dampers and then used in the prediction of the responses of the
of viscous fluid through orifices were utilized in the damping
platform structure with the damping isolation system.
isolation level as vibration energy absorbing devices. Fig. 15
shows the fluid viscous dampers used in the shaking table test. )sgn(x)
f v = ( f 0 + c0 |x| (4)
Dimensions of the dampers are: diameter of piston rod, cylinder
and orifice are 45 mm, 100 mm and 2.9 mm, respectively; stroke where f v is the force of the viscous dampers, f 0 is the friction
is 20 mm. Steel cylinders were filled with silicone oil. force that originates in the seals, c0 is the damping constant of
It is required to conduct tests for investigating the the dampers, x is the piston moving velocity against cylinder,
performance of the dampers. The tested dampers were and is an exponential coefficient. c0 and are determined by
connected to a hydraulic actuator that applied a dynamic force fitting the experimental data. Fig. 17 presents the relationship
along the dampers major axis. The force was applied in such a between the maximum force and the peak velocity for the tested
way that the piston rod moved in harmonic motion with specific dampers. Evidently the behavior of the dampers can be modeled
amplitude and frequency. The displacement was measured by by Eq. (4) with = 1 and c0 = 153 kN s/m.

(a) Laminated rubber bearing. (b) Detail of the rubber bearing.

Fig. 13. Laminated rubber bearing.

1532 J.P. Ou et al. / Engineering Structures 29 (2007) 15251538

(a) = 50%. (b) = 100%.

(c) = 150%. (d) = 200%.

Fig. 14. Shear forcedisplacement loops of the laminated rubber bearing.

(a) Viscous dampers. (b) Detail of a damper.

Fig. 15. Viscous dampers.

4.4. Shaking table test displacements. An extra accelerometer was placed at the center
of the shaking table to monitor the acceleration response of the
4.4.1. Experimental procedures shaking table.
Shaking table tests were carried out in the Structural A total of 17 channels of data were recorded in the shaking
Engineering Laboratory of Harbin Institute of Technology. The table test, including 1 channel for monitoring the shaking
shaking table has a dimension of 3 m 4 m. The maximum table response, 12 channels for measuring the model responses
acceleration of 1g and maximum stroke of 250 mm can be and 4 channels for recording the dampers responses. Fig. 18
achieved. The frequency of the input waves can vary from 0.1 illustrates the arrangements of the experimental setup.
to 25 Hz. The shaking table was driven along the longitudinal
(X ) direction of the platform model.
Displacement transducers were installed at the center of 4.4.2. Input earthquake excitations
four elevations of the model to measure the displacements Four earthquake records: El-Centro (1940, SN), Taft (1952,
relative to the shaking table, and accelerometers were mounted N21E), Tianjin (1976, EW) and Kobe (1995, NS) were used as
symmetrically at the two sides of four elevations of the model the input earthquake excitations for the shaking table test. For
to capture acceleration response data. Two load cells and two each record the time scale was compressed by a factor of 4 to
displacement transducers were placed along the major axis satisfy the frequency similarity requirement for the model test.
of the two dampers to measure the damper forces and axial The measured maximum acceleration outputs on the shaking
 J.P. Ou et al. / Engineering Structures 29 (2007) 15251538 1533

Fig. 16. Recorded forcedisplacement loops of VD1.

Fig. 17. Forcevelocity relationship of the viscous dampers.

table were 0.35g for El-Centro, 0.31g for Taft, 0.32g for Kobe
and 0.25g for Tianjin seismic excitations. Fig. 18. Instrumentation diagram for the shaking table test.
1534 J.P. Ou et al. / Engineering Structures 29 (2007) 15251538

(a) El-Centro (1940, SN). (b) Kobe (1995, NS).

(c) Taft (1952, N21E). (d) Tianjin (1976, EW).

Fig. 19. The 5% damped response spectra under the four earthquake excitations and the corresponding excitations generated on the shaking table.

Fig. 19 shows the 5% damped response spectra of a single

degree of freedom system subjected to the four earthquake
excitations and those under the corresponding excitations
generated on the shaking table. It is evident that the two sets
of data are in good agreement, thus confirming that the seismic
motions produced by the shaking table are consistent with the
desired ground excitations.

4.4.3. Experimental results

The structural models without the isolation arrangements
(called the unisolated structure), with the rubber bearings
(called the isolated structure), and with the rubber bearings
and the dampers (called the damping isolated structure) were
Fig. 20. Picture of the platform model with the rubber bearings and dampers.
tested. Installation of the rubber bearings and the dampers on
the platform model is shown in Fig. 20. the unisolated, isolated and damping isolated structures under
The measured maximum displacement of the jacket cap the Taft earthquake input are shown in Fig. 21. The results
(EL. +4.100 m), the maximum acceleration of the deck (EL. presented in Table 4 and Fig. 21 demonstrate significant
+4.210 m) and the maximum inter-story drift of the damping response reduction when the rubber bearings and the dampers
isolate level under the four ground motions are presented in were added to the jacket structure. For the isolated structure
Table 4. Considering the effect of torsional movement of the case, reductions in the deck acceleration and the jacket cap
deck, the maximum acceleration of the deck was determined relative displacement are significant, but larger response of the
by adopting the mean value of the maximum accelerations inter-story drift than the damping isolated structure case was
measured at the two sides of the deck. observed. For the isolated structure under the four earthquake
Comparison of the acceleration time history of the deck inputs, both the maximum deck acceleration and the maximum
and the displacement time history of the jacket cap between jacket cap relative displacement were reduced by 71% on
 J.P. Ou et al. / Engineering Structures 29 (2007) 15251538 1535

Table 4
Summary of the experimental results for the jacket model under the four earthquake excitations

Earthquake Peak acceleration Maximum displacement of jacket Maximum inter-story drift of Maximum acceleration of
(gal) cap (103 m) isolate level (103 m) deck (103 m/s2 )
Unisolated 9.0 5465.8
2.6 1945.1
El-Centro 352 Isolated 9.6
(0.71)a (0.64)a

Damping 2.9 1870.7

2.2
isolated (0.68)a (0.66)a

Unisolated 11.9 7286.1

3.9 2685.6
Taft 314 Isolated 14.0
(0.67)a (0.63)a

Damping 6.2 3558.7

4.9
isolated (0.48)a (0.51)a

Unisolated 12.9 7721.4

3.1 2091.3
Kobe 322 Isolated 10.0
(0.76)a (0.73)a

Damping 5.8 3376.3

3.7
isolated (0.55)a (0.56)a

Unisolated 12.9 7857.9

3.9 1443.2
Tianjin 249 Isolated 15.1
(0.70)a (0.82)a

Damping 6.7 4066.7

4.9
isolated (0.48)a (0.48)a
a The measure of the vibration control effectiveness = (Unisolated Isolated)/Unisolated.

Table 5
Dynamic characteristics of the structure model

Unisolated structure Isolated structure Damping isolated structure

The fundamental frequencies (Hz) 3.61 1.95 2.93
Damping ratios (%) 1.29 3.39 5.58

average, and the mean value of the maximum inter-story natural frequencies of the unisolated, isolated and damping
drift of the isolation level was found to be 12.2 103 m. isolated structures were also measured on the shaking table
For the damping isolated structure under the four earthquake excited by a sweep excitation, and the results were found to
excitations, the maximum deck acceleration and the maximum be quite close to those obtained in the shaking table tests
jacket cap relative displacement were reduced by 55% on under the earthquake excitations. The first mode damping ratios
average, and the mean value of the maximum inter-story drift of the unisolated, isolated and damping isolated structures
was 3.9 103 m. were obtained using the random decrement method [1316].
The damper force versus the inter-story drift of the isolation The fundamental natural frequencies and damping ratios are
level for the case of the Taft earthquake action is shown in summarized in Table 5. It is noted that the presence of the
Fig. 22. The loop curves exhibit storage stiffness, which are rubber bearings and the dampers has an effect on the structural
somewhat different from the test results of a separated viscous natural frequency and also increases the damping ratio as
damper. This may be attributed to the existence of position expected. The dampers added both stiffness and damping to the
errors in the damper installation. structure. That is, the dampers behave effectively as viscoelastic
energy dissipation devices.
4.4.4. Dynamic characteristics
4.4.5. Numerical analysis
The dynamic characteristics of the platform model were
determined from the shaking table test data. The spectrum 4.4.5.1. Computational model. The platform model (with all
analysis was performed for the measured response data of the additional weights) weighs 5.35 t. Taking advantage of the fact
unisolated, isolated and damping isolated structures under the that the majority of the masses were concentrated at the deck
four earthquake motion inputs. The results demonstrate that module, the unisolated structure can be modeled as a lumped
the structure primarily responds in its fundamental mode. The mass system with a single degree of freedom to simplify the
1536 J.P. Ou et al. / Engineering Structures 29 (2007) 15251538

Fig. 21. Measured time histories of the displacement of the jacket cap and the Fig. 23. Computational time histories of the jacket cap displacement and the
acceleration of the deck under Taft earthquake input. deck acceleration under Taft earthquake excitation.

be written as
M X + C X + K X + Bs f v = M I u g (5)
h i h i
where M = 0.45 0
0
4.9 t, K =
3764.5 1012
1012 1012 kN/m, and C
represent respectively the mass, stiffness and damping matrices,
and X , X and X denote the vectors of relative displacement,
velocity and acceleration, respectively. Rayleigh proportional
damping was adopted herein, and then C was determined
based on the experimentally determined values of the damping
ratios of the isolated structure. f v represents the damper force
determined by the viscous model described by Eq. (4) with
T
= 1.0, cv = 306 kN s/m and f 0 = 2 kN. Bs = 1 1 is


the vector associated with the damper position.

Fig. 22. Damper force-drift loop under Taft earthquake excitation.

4.4.5.2. Results and comparison. The numerical simulations
were carried out for the unisolated, isolated and damping
analysis. The natural frequency and critical damping ratio of the isolated structures under the four earthquake motion inputs. The
unisolated model in the longitudinal (X ) direction were found computational results are presented in Table 6. Fig. 23 shows
to be 3.61 Hz and 1.29%, respectively. The structural stiffness k the comparison of the time history of the deck acceleration and
and damping coefficient c were determined to be 2752.5 kN/m the time history of the jacket cap relative displacement among
and 4.83 kN s/m, respectively. the results for the cases of the unisolated, isolated and damping
The damping isolated structure was simplified as a two isolated models under the Taft earthquake excitation.
degree of freedom system. The parameters of the damping The computational response results were compared with
isolated structure model are: m 1 = 0.45 t, m 2 = 4.9 t; k1 = the experimentally obtained maximum jacket cap relative
2752.5 kN/m and k2 = 1012 kN/m that was obtained based displacements, maximum deck accelerations and maximum
on the experimentally determined value of the fundamental inter-story drifts. The differences among these were found
frequency of the isolated structure. to be 9.3%, 9.1% and 23%, respectively, indicating that the
In order to numerically simulate the seismic responses of agreement between the numerical results and experimental data
the damping isolated structure model, its motion equation can are acceptable for engineering applications. The reason for the
 J.P. Ou et al. / Engineering Structures 29 (2007) 15251538 1537

Table 6
Summary of the computational results under the four earthquake excitations

Earthquake Peak acceleration Maximum displacement of jacket Maximum inter-story drift of Maximum acceleration of
(gal) cap (103 m) isolate level (103 m) deck (103 m/s2 )
Unisolated 9.5 4882.2
3.1 2008.3
El-Centro 352 Isolated 9.7
(0.67)a (0.59)a

Damping 3.2 1649.0

1.5
isolated (0.66)a (0.66)a

Unisolated 13.9 7144.8

4.8 2591.1
Taft 314 Isolated 14.8
(0.65)a (0.64)a

Damping 5.8 3069.9

2.8
isolated (0.58)a (0.57)a

Unisolated 13.2 6803.4

3.6 2081.9
Kobe 322 Isolated 9.9
(0.73)a (0.69)a

Damping 6.2 3247.9

2.2
isolated (0.53)a (0.52)a

Unisolated 14.7 7579.1

3.7 1892.6
Tianjin 249 Isolated 9.9
(0.75)a (0.75)a

Damping 6.4 3306.3

3.0
isolated (0.56)a (0.56)a
a The measure of the vibration control effectiveness = (Unisolated Isolated)/Unisolated.

differences may be attributed to the fact that the computational maximum inter-story drift increases, and the effect of
results were obtained based on the simplified model and in the damping ratio on the reduction of the drift is more
addition, the effects of earthquake-induced torsional moments significant.
were not considered in the simulation. 3. As the normal period ratio increases, the effects of the
vibration reduction in the maximum relative displacement
5. Conclusion of the jacket cap and the maximum acceleration of the deck
increase, but the reductions are less sensitive to the change
In the present study, a damping isolation system was of the first mode damping ratio.
developed to mitigate earthquake and ice-induced vibrations 4. The experimental and computational results for JZ20-
of jacket offshore platforms. Taking JZ20-2MUQ platform in 2MUQ platform demonstrate that the damping isolation
the Bohai Sea as an example to examine the effectiveness of strategy is a very effective way to reduce the seismic
the proposed vibration control strategy for offshore structures, responses of the jacket offshore platform. For the El-Centro,
a combined experimental and numerical study was conducted Taft, Kobe and Tianjin earthquake excitations, both the deck
with consideration of ice and seismic actions on the platform. maximum acceleration and the jacket cap maximum relative
The major conclusions of this study are summarized below: displacement were reduced by 55% on average, and the
mean value of the maximum inter-story drift of the damping
1. The developed damping isolation system placed between the
isolation lever was 3.9 103 m.
deck and the jacket structure is composed of rubber bearings
5. The results from the shaking table test were in good
and viscous dampers. Since the isolation system provides
agreement with those from the numerical analysis, thus
low lateral stiffness relative to that of the superstructure,
indicating that the simplified computational model is
causing the isolated structure to primarily deform at the
applicable to engineering practices.
isolation level, viscous dampers placed at the isolation level
were used to dissipate the vibration energy to suppress
Acknowledgments
the dynamic responses of the structure effectively. Unlike
traditional isolation techniques, the damping isolation This study was sponsored by the Hi-Tech Research and
system aims at dissipating the vibration energy. Development Program of China (863 Program) (Grant No.
2. As the damping ratio of the first mode increases, the 2001AA616120). The work described in this paper was also
maximum inter-story drift of the damping isolation level partially supported by the National Natural Science Foundation
decreases. As the normal period ratio increases, the of Chinas Outstanding Overseas Scientist Award (Grant No.
1538 J.P. Ou et al. / Engineering Structures 29 (2007) 15251538

550856). The authors would like to thank Prof. Q.J. Yue of [8] JZ20-2MUQ platform structure design report. China Offshore Oil
Dalian University of Technology for kindly providing several Engineering Co., Ltd. 1990.
[9] Wang LY. The principle of tuned liquid dampers and study on application
records of ice force.
upon JZ20-2MUQ jacket platform for mitigating vibration. Research
report. Dept. of Ocean Eng. And Naval Arch, Tianjin University; 1994
References [in Chinese].
[10] Ou JP, Wei JS, Yuan CF, Wu B. Numerical simulation of ice-induced
[1] Housner GW, Bergman LA, Caughey TK, Chassiakos AG, Claus RO, vibration control for JZ20-2MUQ jacket platform structure. Research
Masri SF. Structural control: Past, present, and future. Journal of report. Harbin University of Civil Engineering & Architecture; 1997 [in
Engineering Mechanics ASCE 1997;123:897908. Chinese].
[2] Vandiver JK, Mitome S. Effect of liquid storage tanks on the dynamics [11] Meirovitch L. Elements of vibration analysis. McGraw-Hill, Inc.; 1975.
response of offshore platform. Applied Ocean Research 1979;1:6774. [12] 19941995 Annual report of ice forces measurement on JZ20-2MUQ
[3] Lee HH. Stochastic analysis for offshore structures with added mechanical jacket platform. Bohai Inspection Company for Offshore Structure.
dampers. Ocean Engineering 1997;24(9):81734. Institute of Engineering Mechanics, China Seismological Bureau. 1995
[4] Vincenzo G, Roger G. Adaptive control of flow-induced oscillation [in Chinese].
including vortex effects. International Journal of Non-Linear Mechanics [13] Li QS, Xiao YQ, Wong CK, Jeary AP. Field measurements of wind effects
1999;34:85368. on the tallest building in Hong Kong. The Structural Design of Tall and
[5] Ou JP, Duan ZD, Xiao YQ. Ice-induced vibration analysis of JZ20-2MUQ Special Buildings 2003;12(1):6782.
offshore platform using in-situ ice force histories. The Ocean Engineering [14] Li QS, Xiao YQ, Wong CK, Jeary AP. Field measurements of typhoon
1999;17(2):708. effects on a super tall building. Engineering Structures 2004;26:23344.
[6] Ou JP, Xiao YQ, Duan ZD, Zou XY, Wu B, Wei JS. Ice-induced [15] Li QS, Wu JR, Liang SG, Xiao YQ, Wong CK. Full scale measurements
vibration control of JZ20-2MUQ platform structure with viscoelastic and numerical evaluation of wind-induced vibration of a 63-storey
energy dissipators. The Ocean Engineering 2000;18(3):914. reinforced concrete tall building. Engineering Structures 2004;26:
[7] Ding JH. Theoretical and experimental study on structural vibration- 177994.
repressed system using viscous fluid dampers. Ph.D. dissertation. Harbin [16] Li QS, Xiao YQ, Wong CK. Full-scale monitoring of typhoon effects on
Institute of Technology; 2001 [in Chinese]. super tall buildings. Journal of Fluids and Structures 2005;20:697717.