Thousand Tower & Sendai MT
Thousand Tower & Sendai MT
Technical report
Abstract
High-rise buildings with a base isolation system have been realized by investigating the aspect of practical applicability
through the identification and addressing of the difficulties involved in actual design applications. Among the
base-isolated buildings we have designed so far, the Sendai MT Building is the first base-isolated building with a height
exceeding 60 m in Japan, and the Thousand Tower was the tallest base-isolated residential Tower in Japan when com-
pleted. These examples show that the appropriate utilization of the base isolation system with high-strength materials and
a long-span structure system makes it possible to endow high-rise buildings not only with strong seismic performance but
also architectural design flexibility. This paper provides empirical evidence that the base isolation system can properly
work for high-rise buildings. According to the seismic data obtained from the seismographs in the Sendai MT Building
when the Off-Miyagi earthquake struck on May 26, 2003, the base isolation system of this high-rise building performed as
effectively as designed.
spectively, and the natural period of the base isolation Figure 3 shows that the response reduction tends to
after yielding is set to 5.0 seconds. The parameter is the decrease in proportion to the increase in the natural pe-
natural period of the superstructure, varied in the range of riod of the superstructure. However, even when the
0.2 to 5.0 seconds in 0.2 increments. Taking into account natural period is 3-4 seconds, the response displacement
the actual rate of the superstructure's weight to that of the can be reduced up to around 30-40% of the case without
base isolation story, the ratio is assumed to be Ws:Wi = the base isolation. The response displacement of the base
4:1. The dynamic analyses were carried out using the isolated superstructure remained less than the yield dis-
ground acceleration waves, El Centro (1940-NS) and the placement (δy), while the response displacement of the
artificial seismic wave of BCJ-L2 (which has a max. non-isolated superstructure often exceeded the yield
velocity of 0.574 m/s). The wave of El Centro amplified displacement. Moreover, the response displacement of
with the max. velocity of 1.00 m/s was chosen because the base isolation story is approximately 300-400 mm, so
several intensity levels (max. velocities = 0.25, 0.50, 0.75, the response of the base isolated structure is considered
1.00m/s) were studied and the intensity of 1.00 m/s suf- to be very stable.
ficiently influences all models with the varied natural Judging from these results, if the design of the yield
periods to cause large response. shear force and the stiffness after the yield of the base
Figure 3 and Fig. 4 show the response displacement of isolation story are appropriate, strong seismic perform-
the superstructure and that of the base isolation story, ance can be achieved even for high-rise buildings.
respectively, corresponding to the varied natural period
of the superstructure. In Fig. 3, the response of the ver- 3. Base isolation devices
tical axis is defined to be the ratio of the response dis-
placement to that in the case without the base isolation The improvement in dynamic performance depends on
system. the force-displacement relation in the base isolation story.
In order to reduce the dynamic response regardless of the
Shear force Shear force natural periods of superstructures, the innovative base
isolation system, called Hybrid TASS system (Hybrid
Qy
Ki TAISEI Shake Suppression system), has been developed.
Qyi This system is composed of two types of bearings, such
α yK 0 T AKED Bi-linear as rubber bearings and sliding bearings. In order to re-
Qc A duce the dynamic response due to an earthquake in a
K0
δy
high-rise building, whose natural period is basically long,
Displacemen Displacement
t the stiffness of the base isolation story needs to be so soft
Spring 1 Spring 2 that the effective natural period should be long. On the
other hand, the initial soft stiffness of the ordinal
Fig.2 Properties of springs. base-isolation system might result in poor habitability
due to swinging during strong winds (see Fig. 5).
1.0 This contradictory problem, which is denoted as
0.8 Reduction problem (2) in the introduction, can be solved with the
Response ratio
0.6
effect Hybrid TASS system. Figure 5 shows the typical
El centro '40NS 1.0 m/s
0.4
force-displacement relation of the Hybrid TASS system.
0.2
By setting the yield shear force of the base isolation story
0.0
BCJ-L2 0.574 larger than the design wind force, the initial stiffness of
0.0 1.0 2.0 3.0 4.0 5.0 the whole structure can be kept almost the same as that of
Natural period of superstructure (s)
a building without the base isolation system. At the same
Response ratio =
Response of base-isolated structure
Response of non-base-isolated structure
time, strong seismic performance in case of a huge
earthquake is obtained as the stiffness after the yield is
Fig.3 Response displacement of superstructure.
particularly soft.
Moreover, the Hybrid TASS system is also advanta-
600
geous in terms of costs. Figure 6 shows the comparison
between the Hybrid TASS system and another typical
of base isolation story (mm)
Earthquake Earthquake
Wind Wind
Equivalently soft
Soft stiffness Hard stiffness
Deformation Deformation
80
28
Sliding bearings
402.5
346.5
392.5
552.5
28
φ60
80
Lead damper
φ1200
φ1510
200%
φ1200 test
150%
In the design of a slender high-rise building, the uplift
due to overturning and vertical forces during a large 100% Design criteria
stand the tensile properties of large rubber bearings, Fig.9 Relation of tensile and shear strain.
tensile loading tests (Muramatsu et al. 2001) were carried
out (see Fig. 7). Figure 8 shows a test specimen with a
diameter of 1200 mm. The relation between tensile and
shear strain is shown in Fig. 9, which plots the test results 5. Design of high-rise buildings with hybrid
of these large size tests, along with the tensile and shear tass system
strain results of bearings with a smaller diameter of 500
mm from other tests (Takayama 1995). From the results 5.1 Sendai MT Building
of the bearings with a small diameter, the failure limit The Sendai MT Building is an 18-story office building
was expected to approximate the line shown in Fig. 9. with a height of 84.9 m located in Sendai City, Miyagi
However, in the tests of the large scale, the failure was Prefecture (Ogura et al. 1997a). This building is Japan's
observed at the tensile strain of 50% and the shear strain first base-isolated building with a height exceeding 60 m
of 300%, which is much less than the expected limit of (see Fig. 10 and Fig. 11).
the small scale tests. There is thus a large difference in
the failure limit due to the scale effect. The structural features of this building are as follows.
Considering this tendency, the region of tensile stain 1) Use of the high-strength materials such as concrete
less than 10% and the region of shear strain less than with compressive strength of 60 N/mm2 and SD
300% were selected as the design criteria. With these 490 re-bars for the longitudinal reinforcement of
design criteria, the response tensile strain can be kept the beams and the columns.
sufficiently lower than the failure limit.
236 T. Komuro, Y. Nishikawa, Y. Kimura and Y. Isshiki / Journal of Advanced Concrete Technology Vol. 3, No. 2, 233-239, 2005
Fc=36 N/mm2
Fc=36 N/mm2
4,080
3,950
〃
〃
〃
〃
〃
Fc=42
N/mm2
Fc=42
N/mm2
〃
84,900
74,900
〃
〃
〃
Fc=48
N/mm2
〃
〃
Fc=48 N/mm2
〃
〃
〃 Fc=60
3,950 N/mm2
7,500 1FL
GL
2,050
11,430
8 8
target seismic performance, using several earthquake 7 7
6
waves. The analysis model was a lumped-mass-model 6
5 5
4
consisting of 21 masses and the springs connecting the 4
3 (rad) 3
2 (rad)
masses are the equivalent shear type. The input levels of 2
1 10 0.002 0.004 0.006 0.008 0.01
0 0.002 0.004 0.006 0.008 0.01
the waves are categorized from level 1 to level 3, con- Max. Response Story Drift Max. Response Story Drift
sidering occurrence probability during the building usage Fig.14 Maximum response story drift.
period. The waves of levels 1, 2 and 3 are accelerations
T. Komuro, Y. Nishikawa, Y. Kimura and Y. Isshiki / Journal of Advanced Concrete Technology Vol. 3, No. 2, 233-239, 2005 237
amplified with the max. velocity of 0.25, 0.50 and 0.75 relatively slender structure. The yielding base shear co-
m/s, respectively, using the recorded waves of El Centro efficient of the base isolation and the design base shear
(1940 NS), Taft (1952 EW) Hachinohe (1968 NS) and coefficient of the superstructure are designed to be 3.5%
Sendai TH-038 (1978 EW). The artificial seismic wave and 6.5%, respectively.
of BCJ-L2 (max. velocity of 0.574 m/s) was also added The pre-cast and pre-stressed concrete beams as
to level 3. From the results of the level 1 analyses, the shown in Fig. 18 are used for the 12-m span. A
design base shear coefficient of the superstructure is pre-stressing steel tendon is arranged at the clear
defined to be 8%. Figure 14 shows the maximum re- mid-span of the beam and the longitudinal reinforcement
sponse story drift angle of levels 2 and 3. For level 2, the of the steel bars is passed through the beam and devel-
drift angles are less than 1/330 and for level 3 less than oped at the beam-column joints. Adoption of this long
1/230. These results indicate satisfaction of the seismic span system made it possible to eliminate columns in the
design targets and strong seismic performance. residential area.
High-strength concrete with strengths up to 100
5.2 Thousand tower N/mm2 is used for the columns and high-strength re-bars
Thousand Tower. a residential tower situated in Kawa- (SD 490, USD 685) are used for the longitudinal rein-
saki City, Kanagawa Prefecture, is a reinforced concrete forcement of the beams and the columns. By using these
structure with 41 stories above ground (Kawabata et al. high-strength materials, the number of structural mem-
2001) (see Fig. 15). The base isolators are placed at the bers and their cross-section can be reduced, so that the
isolation story below the first floor. Figures 16 and 17 effective architectural space can be used for the planning.
show the frame elevation and the structural plan, re- The dimensions of the typical column are 1000 x 1000
spectively. The total height is 135.0 m and the aspect mm and the dimensions of the typical beam are 650 x 850
ratio is 3.83 in the Y direction, making this building a mm.
Co n c r e t e
Co l u mn Gi r d e r
t ▽R F L
▽4 1 F L
Fc30 Fc30
Fc36
127750
135000
Fc36
Fc48
Fc48
Fc60
Fc80
▽1 F L
Fc100
Fc30
▽B 1 F L Fc60 Fc60
Fc30
FL
PC cable
PCケーブル
FL
In-situ concrete
Half precast
ハーフPCa床板 slab
後打部
Spiral stirrup
スパイラルスタ-ラップ
(ウルボン)
The dynamic analyses were conducted using several non-isolated model, the responses are effectively reduced
artificial waves (see Table 1). Artificial wave H is based by the base isolation system.
on the seismic activity and the ground conditions at the Using the shear-flexural model under the simultaneous
building location. The analysis model is the equivalent application of horizontal and vertical excitations, the
shear system, composed of concentrated masses and influences of the rocking effect and the response tensile
story shear springs. The relation between the lateral stress of the rubber bearings were studied additionally.
resistance and the story drift are derived from a Under the Level 3 seismic motion, tensile stress occurs in
non-linear static incremental loading analysis, and ap- the 45 degree direction case, although the corresponding
proximated to a tri-linear relation. The degrading stiff- tensile strain is approximately 1.7%, which satisfies the
ness model proposed by Takeda is used for the hysteresis design criteria.
model. The damping of the superstructure is assumed to
be proportional to the instantaneous stiffness, and to be 6. Seismic record of Off-Miyagi quake, May
3% with respect to the superstructure's fundamental 26, 2003
natural period.
The level 2 maximum response of the isolation story in The seismic data were obtained from the seismographs
the Y direction is shown in Fig. 19 and the level 2 set at the isolation story and the 1st, 10th and 18th stories
maximum response story drift angle in the Y direction is of the Sendai MT Building, when the Off-Miyagi
shown in Fig. 20. Compared with the response of the earthquake occurred on May 26, 2003 The maximum
accelerations at these floors are listed in Table 2 and the
acceleration waves in the E-W direction are shown in Fig.
Table 1 Seismic waves.
21. The accelerations of the isolation story are thought to
Level 1 Level 2 Level 3 be very close to that of the ground, because the stiffness
(m/s2) (m/s2) (m/s2) of the basement is very large. Compared with the accel-
BCJ-L2 - 0.574 - erations of the isolation story and the 1st story, reduction
Artificial wave H 0.237 0.474 0.595 in the horizontal accelerations (NS and EW directions)
was found.
Shear force [kN]
18
Record
18
Record
1. Application of base isolation to high-rise buildings is
17 17
16 16 innovative, calling for a design approach that differs
15 15 from the procedures commonly applied to low-rise
14 14
13 13 types of construction.
12
11
12
11
2. Several actual design problems have been resolved for
10 10 the applicability of the base isolation system to
9 9
8 8 high-rise buildings.
7 Analysis 7 Analysis 3. Using high-strength materials and a long-span struc-
6 6
5 5 ture system, the Sendai MT Building and Thousand
4 4 Tower have not only high seismic performance but
3 3
2 2 also design planning flexibility.
1 (mm/s 2) 1 (mm/s 2) 4. The recorded data of the seismographs and the analy-
0 0
0 500 1000 1500 0 500 1000 1500 ses show that the base isolation system acted effec-
E-W Direction N-S Direction tively in the Sendai MT Building when the
Fig.23 Comparison of analyses and records. Off-Miyagi earthquake occurred on May 26, 2003.
S References
Kawabata, I., Takayama, M., Nishikawa, Y., Kimura,
Y.,Yamazaki, E. and Isshiki, Y. (2001). “Structural
Max.= 18.8(mm) Design of High-rise Building with Base Isolation
(Time:27.92sec)
System Using Elastic Sliding Bearings and Rubber
E W
Bearings.” AIJ Journal of Technology and Design,
10 ㎜ (12), 99-104.
20 ㎜ Muramatsu, Y., Nishikawa, I., Kawabata, I., Takayama,
M. and Kimura, Y. (2001). “Tensile Property of
Large-sized Natural Rubber Bearing.” AIJ Journal of
Fig.24 Analyzed displacement track. Technology and Design, (12), 53-56.
Ogura, K., Kawabata, I., Hara, T., Komuro, T., Hara, K.
Figure 22 shows the observed track of the relative and Terashima, T. (1997a). “Sendai MT Building.”
displacement between the 1st floor and the floor below, Building Letter, 1-8.
which means the displacement of the isolation story. Ogura, K. Kawabata, I., Komuro, T., Soya, K. and
From this result, the maximum displacement of the iso- Terashima, T. (1997b). “Seismic Response Character-
lation story is found to be about 20 mm and there is no istics of High-rise Buildings with Base Isolation
final displacement. Considering with the relation be- System.” AIJ Journal of Technology and Design, (5),
tween the force and the displacement in the design, slid- 47-51.
ing occurred and the isolation system is considered to Takayama, M. (1995). “Ultimate Capacity of Natural
have functioned well. Rubber Bearings Used in Seismic Isolation System.”
Dynamic analyses using the design model were carried AIJ Journal of Technology and Design, (1), 160-165.
≒20mm