Structural Design of a
Seismic Isolated Building with a Guyed Tower
K. Saito, H. Tadokoro, M. Hayashi, T. Tsuchiya & K. Sumimura
Building Engineering Headquarters, NTT FACILITES, INC., Tokyo, Japan.
I. Shibata
Ove Arup and Partners Japan Ltd., Tokyo, Japan.
ABSTRACT: A large seismic isolated telecommunications building with a guyed mast
tower was planned to be constructed upon liquefiable subsoil. The foundation and
superstructure have to be designed taking soil condition into account. In order to
penetrate to the firm layer and to establish rigidity to the soft soil layer, a continuous wall
pile system was adopted. For seismic analysis, seven seismic waves were used to carry
out the dynamic response. The results showed a prominent response of the structure on
the higher modes for the site wave. For reducing this response, viscous wall dampers
were installed in the building. Thus the safety of the structure and the realization of the
project was achieved by the application of various structural means such as, continuous
wall piles, seismic isolators, viscous dampers and by conducting thorough seismic
analysis.
1 INTRODUCTION
This paper reports the structural design of a large
seismic isolated telecommunications building with a
unique cable-guyed telecommunication mast tower.
Because the site is reclaimed land with soft soil, the
foundation system and structural system need to be
designed considering conditions of the soil. Also,
design criteria for a building are established to be
elastic even for the biggest earthquake expected in
the construction site. To satisfy these criteria, the
building has a seismic isolation system and damping
devices are installed in its middle stories. As for the
tower, the design criteria are similarly established to
remain elastic against wind loads. The results
estimated by wind tunnel test results are extreme
because the site is near the tall buildings and the
effect of wake is large. To introduce an improved
structural performance, damping devices are also
installed in the tower. Figure 1 shows CG
perspective views of the building and the tower.
Figure 1. CG Perspective views
Paper Number 164
2 PROJECT OUTLINE :Viscous damping walls
2.1 The building
This is a 12-story building with 54.25 m in eaves
54.25m
height The building plan, with unit span of 6.4 m,
is a rectangle measuring 116.8 m in the X-
direction, by 43.2 m in the Y-direction, total floor
area is 60993.42 m2 . A seismic isolation system is
applied on this building in order to guarantee
sufficient safety and to maintain building functions
in the event of a disaster like a major earthquake.
The framework has Concrete Filled Tube columns
43.3m
and steel beams, rigid frames with bracing. Also,
viscous damping walls are arranged in the
building’s middle stories. Figure 2 shows the
typical framing elevation and plan. 116.8m
Figure 2. Framing elevation and plan
2.2 The site stratum and the foundation system
The site stratum consists of in-filled land for the first 20 m,
another 20 m of alluvium to 44 m below, and further dense
Reclaimed
gravel and clay diluvium below it. Because the reclaimed layer land
is soft and likely to liquefy during a severe earthquake, a stiff
continuous wall pile system, having a short natural period in
foundation-ground system, was applied against liquefaction
and for increasing seismic performance significantly. The
pile’s bottom level was set on a firm sand-gravel layer of Alluvium
diluvium encountered at 45 m below the surface. The stratum Sand-Clay
composition is shown in Fig ure 3; the arrangement of
continuous wall piles is shown in Figure 4 below.
Pile bottom
Wall pile Cast-in-place pile Sand-Gravel(Dg1)
Y8
Clay(Dc1)
Diluvium
Y5
Y4 Sand-Gravel(Dg2)
Y1
X1 X9X10 X18
Figure 4. Arrangement of continuous wall piles Figure 3. Stratum composition
2.3 Damping device and isolator device
Viscous damping walls are used for damping devices, of which 24 and 22 are arranged in the X- and
Y-directions, respectively, between 3 to 8 story. The damping walls are proportionally and
symmetrically arranged around the building core and periphery to expect effective performance even
when torsional vibration occurs to the superstructure and not to become obstacles in planning the
building. The device arrangement is shown in Figure 5.
The isolator system comprises 46 lead rubber bearings (LRB), 42 rubber bearings (RB), and 4 cross
linear bearings (CLB) which comes to 92 isolators in all, LRB and RB are 1000 -1500 in diameter and
280 mm in total thickness of the rubber portion. The LRBs are arranged in the building periphery,
2
Rubber bearings in this area
3F• `8 4F•`8
F F
Y8 Y8
Y5 Y5
Y4 Y4
Y1 Y1
X1 X9X10 X18 X1 X9X10 X18
LRB 1000ƒÓ 1100ƒÓ 1200ƒÓ 1300ƒÓ CLB
RB 1000ƒÓ 1100ƒÓ 1200ƒÓ 1300ƒÓ 1500ƒÓ
Figure 5. Arrangement of viscous damping walls Figure 6. Arrangement of isolators
with RBs in the central part to reduce eccentricity of the seismic
isolated story as much as possible and secure torsional stiffness
necessary for the isolated story. To prevent the building from
up-lift, CLBs are arranged right beneath the columns, where
lateral force causes high axial compression and tensile force.
The device arrangement is shown in Figure 6.
2.4 Tower
Simply explained, the structure is a pinned column held upright Fig. 6 Typical Platform
Figure 7. Typical platform
by four pairs of stay cables, a scheme that is straightforward
and appropriate for this particular location. Weighing
approximately 1650 tonnes, the steel tower rises to a height of
145 meters above the building (see Figs. 7 and 8 for typical
TMD
platform layout and the elevation view respectively). The spinal
mast element is a 1.7 meter diameter hollow pipe with a wall
thickness that varies between 50 and 80 mm. The cables in the
tower are 200 mm parallel wire strand (PWS) cables, consisting
of 499 galvanized steel wires wrapped within a polyethylene
coating. The remaining frame elements are comprised of built-
up boxes, pipes, and H-sections. Steel cast pieces are utilized
for the cable connections. Mast cable
To stiffen the steel central mast, four sets of prestressed steel
cables and outriggers span the entire height of the tower. In
RDT
profile they form a bowstring truss, a system that works in a
similar manner as the rigging mechanism of a yacht’s mast.
During construction, the cables are slid over the saddle pieces
of each outrigger, then prestressed at the ends that are located at
the center and base of the tower. The upper pair of cables
(Cable 1) is prestressed at 5500 kN (11,000 kN per pair), and Guycable
the lower set of cables (Cable 2) has 5000 kN of prestress.
Stay cables are arranged in pairs extending in four directions
rather than the minimum of three to match the symmetry of the
building. The layout of the cables in plan stabilizes the twisting
of the tower; they also give a margin of safety in the unlikely
event that a cable should be severed. Stay cables 3 and 4 are
prestressed 6000 kN and 7500 kN respectively. With a
combination of prestress from the stays and tension truss cables,
the mast element experiences approximately 85,000 kN of axial
force.
Fig. 78.Elevation
Figure ElevationView
View
3
Damping devices were also installed. Two 7.5 tonne tuned mass dampers (TMD) were set on the top
platform of the tower(see Fig.9). Consisting of layered rubber pads and steel plates with a mass at the
top, the TMD was tuned accordingly to reduce the response. Four pairs of rotary damping tubes
(RDT) were installed in parallel with the cable truss near the center platforms to dampen the bending
movement of the tower. Figure 10 illustrates RDT. It is important to note that the effects of the
dampers were not considered in the structural design of the members, but to introduce an improved
structural performance.
Figure 9. Tuned mass damper Figure 10. Rotary damping tube
3 STRUCTURAL EXAMINATIONS OF THE BUILDING
3.1 Examination of the effect of the foundation system over the superstructure
This examination shows differences in response properties and seismic effectiveness of the foundation
system and the superstructure arising from use of continuous wall piles (hereafter, “wall pile ”) and use
of cast-in-place concrete piles (hereafter, “cast-in-place pile”). The Penzien model, continuous series
of subsoil-pile-superstructure, is used as an analytic al model. Each cast-in-place pile is arranged
directly beneath each isolator.For the superstructure, the building is modeled as an equivalent shear
model with 13 masses; the pile structure is modeled as a flexural shear model. Wall piles are
constructed on Dg1 as the pile bottom level; cast-in-place piles are constructed on Dg2. Earthquake
waves for analysis are KOKUJI 202 and Site wave. (Refer to 3.3)
As the analytical results of both models , maximum response drift in the X-direction is shown in Figure
11 and maximum response of interstory shear force is shown in Figure 12. Because of wall pile ’s high
stiffness, the drift of wall pile is one-fifth of that of cast-in-place pile. This may be constraining effects
on surrounding soil’s deformation and soil behavior is similar to the wall piles. Regarding to the shear
force of the superstructure, difference due to foundation forms is significant, but difference due to
earthquake waves is negligible and the response on continuous wall pile is small. Both models show
the similar results that the response on the site wave is largest in the middle stories, which is obviously
seen in continuous wall pile. This may depend on earthquake ground motion characteristics and
building vibration characteristics.
Wall pile Cast -in-place Pile Wall pile C a s t- in-place Pile
4
Figure 11. Maximum response drift Figure 12. Maximum response shear force
3.2 Analysis model
The dynamic analytical model for the building is the equivalent shear model with 13 masses, the
isolated story’s base fixed, while restoring force characteristics are supposed to be elastic Because the
weight ratio of the tower to the building is only 2%, the tower may be considered a part of the roof.
This modeling is verified compared with response results of the dynamic analysis of tower .Damping
is the stiffness proportional type of damping ratio h=2% for the first natural mode of the building
neglecting the isolated story, damping walls are modeled as a dashpot depending on interstory velocity
. As restoring force characteristics of each isolator, LRB is Bi-Linear, RB is Linear and CLB is Bi-
Linear. The dynamic analytical model for examining the building is shown in Figure 13 and the
dynamic model for examining the tower is shown in Figure 14. Also, Table 1 shows natural period of
building model in each direction .
Including the weight of the Tower
13
K12
12
K11
11
K10
10
K9
9
K8 C1
8
K7 C1
7
K6 C1
6
K5 C1
5
K4 C1
4
K3 C1
3
K2
2
K1
1
KL KR KC
Figure 13. Dyanamic analytical model for the building Figure 14. Dynamic analytical model for the
tower (Seismic model)
Table 1. The natural period of building model
(seconds)
strain of X-direction Y-direction
Isolator 1st 2nd 3rd 1st 2nd 3rd
0% 1.489 0.537 0.320 1.441 0.528 0.320
0.5% 2.216 0.720 0.404 2.184 0.717 0.400
10% 2.808 0.769 0.414 2.783 0.764 0.408
100% 4.540 0.805 0.420 4.525 0.804 0.416
200% 4.896 0.816 0.422 4.883 0.808 0.415
3.3 Earthquake ground motions for analysis
The following waves are used as earthquake ground motion for analysis. Three earthquake waves are
conventionally observed (hereafter, “observed waves”), three simulated waves in consideration of soil
conditions at the designated site (“KOKUJI waves”), and one simulated wave in consideration of the
fault and soil conditions which is expected to affect the site the most (“site wave”).KOKUJI waves are
5
input in the base of the isolated story. KOKUJI waves are simulated earthquake ground motions based
on the spectra officially announced by the Ministry of Land(see line in Fig.15 and Fig.16),
Infrastructure and Transport and calculated by non-linear earthquake response analysis in
consideration of surface soil conditions. The maximum velocity of the observed waves on the ground
level is set to be 70 cm/s. Figure 15 and 16 show acceleration and velocity response spectra of
KOKUJI waves and site wave.
KOKUJI202 KOKUJI204 KOKUJI205 Site wave
Figure 15. Acceleration response spectra Figure 16. Velocity response spectra
3.4 Dynamic analysis results
The original project did not have the installation of viscous damping walls to this building. Figure 17
shows the maximum response interstory shear force of typical earthquake ground motion in the X-
direction on the case not considering fluctuation of isolators’ performance values (hereafter, the
“standard case”) without viscous damping walls.
According to analytical results, while all shear force is a less than the designed shear force, site wave
response differs from those of other earthquake waves. This shows the fact that the second mode
including the isolated story causes resonant vibration near the site wave’s powerful period, as shown
in Table 1 and Figure 15, 16. Even if it could satisfy design criteria without installing damping walls
in the standard case, it was difficult to satisfy criteria in considering fluctuation of isolators and other
factors, so installation of damping devices was examined and determined to use. The building is
seismic isolated structure and it’s story drift is relatively small. Therefore, Viscous damping walls
which were suitable for slight deformation were adopted.
Figure 18 show a comparison of analytical results
in the X-direction with and without installing
damping walls (maximum response shear force).
We used KOKUJI202, which was prominent
response among KOKUJI waves, and the site
wave for seismic response analysis. Installation
of damping walls shows reduced responses, by
15% in shear force and satisfies the design
criteria sufficiently even considering the
fluctuation of each device’s performance
value.
Figure 17. Maximum response shear force
6
Figure 18. Maximum response shear force
4 STRUCTURAL EXAMINATIONS OF THE TOWER
4.1 Wind analysis
Deemed to be the governing case in this tower design, in order to assess the performance of the tower,
non-linear wind time history analyses were conducted. The 500-year return period wind time history
inputs were produced by extracting data from the wind tunnel testing. Three wind types were applied
to the model for three conditions depending on the layout of the antennas and the direction of wind.
Similar to the seismic design criteria, all members were designed to remain elastic. Under the
assumption that the wind forces would have very little effect on the isolators of the building, the three-
dimensional model for the wind analyses did not include the building. Again the damping coefficient
for the analyses was set to 1%. The modal properties of the wind model are shown in Table 2.
Compared with the results from the seismic analyses, it can be observed that wind loads were indeed,
the governing case for this tower. The maximum tilting angle of the platforms hit a peak of
approximately 0.95 degrees. In order to further reduce the tilting angles, TMDs and RDTs were
installed in the tower. The axial forces of the stay cables are confirmed to be within allowable limits.
Table 2. Modal properties of the wind model(Tower Model)
Mode 1, X Direction Mode 2, X Direction Mode 3, X Direction
1.9434 seconds 1.0480 seconds 0.4669 seconds
7
4.2 Examination of the damping device
Case A TMD
In regards to the damping device, Time history wind force analysis MT
was carried out about the following cases of wind loads to examine
the most effective installation method (see Fig.19). Under the PL0
assumption that the wind forces would have very little effect on the
isolators of the building, the three-dimensional model for the wind PL1
analyses did not include the building.
Case A The case of installing only TMD PL2
Case BThe case of installing four RDT on TR1 line of PL6 Case B
PL3
RDT
Case C The case of installing four RDT on a cable line of PL6
PL4
Case D The case of installing 16 RDT on a cable line of PL6
Case EThe case of combining Case A and Case C PL5
Analysis results for the wind load of the above five cases are Case C, D
shown in Figure 20. RDT PL6
PL7
We found that added damping ratio is from 1 to 2% approximately by
installing TMD. Also, when RDT is installed on a cable line of the
PL8
upper PL6 where guycables show largest axial displacement when
wind load is applied, it is the most effective place as the installing PL9
location; added damping is from 1 to 2%, similarly to TMD.
Moreover, even if the number of installed RDT increases, the effect PL10
isn’t so large. Also, with an installed the damping device, results of
response analysis on effects in the seismic load are presented in Figure PL11
21. For the seismic analyses, three-dimensional model including the
building was used(see Fig.14). It can be understood that effects of the
damping device in earthquakes are la rger than in wind load. This is
because repetition components of the tower deformation in the Figure 19. The layout cases of
earthquakes are larger in comparison with those in wind load. damping device
12 12
11 11
??? without TMD and RDT
10 10
??? TMD(Case A)
9 ? ? ? RDT(Case C) 9
8 ??? TMD+RDT(Case E) 8
7 ??? damping ratio 2% 7
6 ??? damping ratio 3%
6
??? Interstory shear force for
5 design 5
4 4
3 3
2 2
1 1
0 0
0 2000 4000 6000 8000 0 2000 4000 6000 8000
(kN) (kN)
Figure 20. Shear force in wind load Figure 21. Shear force in earthquake
5 CONCLUSIONS
The results from the response analysis of the lumped mass model(Also including Penzien model)and
the three-dimensional frame model, done on the seismic isolated building with a guyed tower which
was planned to be constructed upon liquefiable subsoil, help to draw the following conclusions:
(1) As for the foundation system, a continuous wall pile foundation is chosen for the large-scale
seismic isolated building to be constructed on the reclaimed ground. Even if shear force caused by
ground drift during earthquake becomes larger compared to a cast-in-place concrete pile foundation, it
can be expected to reduce the seismic force acting on the long period structure like this seismic
isolated building.
8
(2) As for the superstructure, the response was greatly reduced by applying viscous damping walls on
the isolated structure considering the site’s soil conditions and the building’s vibration characteristics.
If earthquake ground motion characteristics coincide with the response of the building higher modes
like this building, the reduction of response by damping devices is also effective.
(3) In consideration of results of the wind tunnel test, that the wind load was predominant, and the fact
that the tower had a unique shape, various examinations were conducted for the wind in particular;
special means such as installing the damping device were taken for vibration control. As for the effects
of the damping device, response analysis was carried out to confirm its effects for wind and for
earthquake. As a result, since repetition components of tower deformation are larger in the earthquake,
damping effects were shown distinctly. Although the effects for wind were not as clear as those for the
earthquake, it was confirmed that installation of the damping device was effective for wind because it
damped approximately 1 to 2%.
This project is due for completion in June of 2004.
REFERENCES:
Recommendations of loads on building. 1993. Architectural Institute of Japan. pp.245-358.
Ohkuma et al. 1995. Simulation method of simultaneous time-series of multi-local wind forces on tall buildings
by using dynamic balance data. Journal of wind engineering and industrial aerodynamics. 54/55. 115-123.
