Professional Documents
Culture Documents
Shaking Table Tests With Large Test Specimens of Seismically Isolated FBR Plants Part 3: Ultimate Behavior of Upper Structure and Rubber Bearings
Shaking Table Tests With Large Test Specimens of Seismically Isolated FBR Plants Part 3: Ultimate Behavior of Upper Structure and Rubber Bearings
Shaking Table Tests With Large Test Specimens of Seismically Isolated FBR Plants Part 3: Ultimate Behavior of Upper Structure and Rubber Bearings
PVP2009
July 26-30, 2009, Prague, Czech Republic
PVP2009-77229
ABSTRACT
This paper describes results of shaking table tests to grasp INTRODUCTION
ultimate behavior of seismic isolation system under extremely Application of seismic isolation technology to nuclear
strong earthquake motions, including failure of rubber bearings. power plants, especially FBR plants, is expected to reduce
The results of the shaking table tests are expected to be useful earthquake load to both structures and inner equipment, and to
for the design of seismically isolated nuclear facilities, facilitate the rational seismic design of the plants. On the other
especially fast breeder reactor (FBR) plants. In the test, lead hand, Japanese regulatory guide for aseismic design of nuclear
rubber bearings, of which the diameter is 505 mm and about 1/3 power reactor facilities (Nuclear Safety Commission of Japan)
scale of a prototype in planning FBR plants, are used; the test was revised in 2006. In the regulatory guide, method of
specimens are loaded by the largest three-dimensional shaking evaluation of design earthquake motions was revised and
table in E-defense of National Research Institute for Earth seismic probabilistic safety assessment (seismic PSA) of the
Science and Disaster Prevention (NIED) of Japan. Failure of plants might be introduced in the future. In the seismic PSA of
rubber bearings occurs with amplified tentative design the seismically isolated plants, it is important that evaluation of
earthquake motions. From the tests, the ultimate responses of ultimate behavior of seismic isolation system subjected to
the upper structure and rubber bearings are presented. In extremely strong earthquake motions. However, the data of the
particular, the change of floor response spectra and restoring ultimate behavior of seismic isolation system have been not
force characteristics of rubber bearings according to increase of prepared enough.
input motions is discussed. Furthermore, mechanism of the Although shaking table tests of ultimate behavior of seismic
failure of rubber bearings is investigated from the observation isolation system have been carried out [1-4], very small rubber
of failure surfaces and cut sections, static loading tests, and bearings, of which diameters were around 100mm, have been
material tests of rubber bearings. Finally, the function of seismic used in the tests. Therefore, the scale of the rubber bearings
isolation system after the failure of a part of rubber bearings is were lower than 1/15 of prototypes for seismically isolated
confirmed under the tentative design earthquake. nuclear power plants, so that the characteristics of the rubber
bearings might not be the same as those of prototypes.
500
(cm/s/s)
250
0
-250
-500
0 10 20 30 40
Time (sec)
(a) Tentative design earthquake motion (b) Moment of breaking of LRB (LRB #3& #4, Specimen-1)
Acceleration
150
0
-150 Table 3 MAXIMUM RESPONSES OF SPECIMEN-1
-300
0 10 20 30 40 Input Input Acc. at Acc. at Disp. of Shear
Time (sec) level acc. A3 A5 LRB strain of
(b) Filtered wave SH (target) [cm/s2] [cm/s2] [cm/s2] [mm] LRB [%]
1.0SH 259 208 214 60 84
Fig.3 INPUT EARTHQUAKE MOTIONS USED IN THE TEST 1.5SH 386 241 251 99 137
2.0SH 537 255 273 159 221
Failure of LRBs and Responses under Extremely Strong 2.5SH 681 312 332 221 307
Earthquake Motions 3.0SH 824 480 497 275 382
Figure 4 shows the layout of failure of LRBs in the isolation 3.5SH 1054 989 1107 340 472
layer and photographs of the moment of breaking of LRBs in 4.0SH 1374 1713 1902 400 556
specimen-1 taken by east side video cameras. The failure
occurred at the corner LRBs of the isolation layer, where the Table 4 MAXIMUM RESPONSES OF SPECIMEN-2
vertical stress during earthquake motions changed larger than Input Input Acc. at Acc. at Disp. of Shear
those at the center. In specimen-1, two LRBs were level acc. A3 A5 LRB strain of
simultaneously broken in the input earthquake motion of 4.0SH; (target) [cm/s2] [cm/s2] [cm/s2] [mm] LRB [%]
1.0SH 226 207 214 51 71
LRB #3 was perfectly broken and LRB #4 was partially done at
1.5SH 343 238 241 89 124
the first rubber layer of upper side. The failure shear strain of
4.0SH 1290 1672 1828 378 525
LRBs was 556%, corresponding to horizontal displacement of
4.5SH 1235 1131 1283 402 558
400 mm. On the other hand, LRB #2 in specimen-2 was 4.8SH 1353 1661 1679 433 601
partially broken at the first rubber layer of lower side in the
motion of 4.8SH, and the failure shear strain was 601%, Figure 5 shows acceleration response spectra of 1%
corresponding to horizontal displacement of 433 mm. damping of critical at the shaking table, the lower slab (A3) and
Table 3 and Table 4 show maximum responses according to the upper slab (A5) for the input earthquake motion of 1.0 SH,
the procedures of each shaking. The specimens had effect on 3.0SH, 3.5SH and 4.0SH in specimen-1. Figure 6 shows
reduction of peak acceleration response in the earthquake acceleration response spectra at the shaking table, A3 and A5
motion of 3.0SH or less, whereas that drastically increased in the for the input earthquake motion of 1.0 SH, 4.0SH, 4.5SH and
ACC. [cm/s/s]
4.5SH
0.1 sec. These peaks of the response spectra arose as results of 4.8SH
8000
rocking vibration of the upper structure and vibration of the
wall. In particular, non-linear vibration and damage of the wall 6000
occurred in specimen-2, because the wall of specimen-1 was 4000
fixed by protection device of collapse and that of specimen-2 2000
was free. Furthermore, difference of loading history between
specimen-1 and specimen-2 hardly affected the response spectra 0
0.01 0.1 1
in comparison with those in the earthquake motion of 4.0SH. PERIOD [sec]
(a) Upper slab A5
12000 14000
1.0SH A5 12000 1.0SH A3
10000 h=0.01
3.0SH h=0.01 4.0SH
10000
ACC. [cm/s/s]
ACC. [cm/s/s]
3.5SH 4.5SH
8000 4.0SH 4.8SH
8000
6000
6000
4000 4000
2000 2000
0 0
0.01 0.1 1 0.01 0.1 1
PERIOD [sec] PERIOD [sec]
(a) Upper slab A5 (b) Lower slab A3
12000 14000
1.0SH A3 Table
10000 12000 1.0SH
3.0SH h=0.01 4.0SH h=0.01
ACC. [cm/s/s]
3.5SH 10000
ACC. [cm/s/s]
8000 4.5SH
4.0SH 4.8SH
8000
6000
6000
4000
4000
2000 2000
0 0
0.01 0.1 1 0.01 0.1 1
PERIOD [sec] PERIOD [sec]
(b) Lower slab A3 (c) Shaking table (Input)
12000
1.0SH Table Fig.6 FLOOR ACCELERATION RESPONSE SPECTRA IN
10000 SPECIMEN-2 (Damping factor h=1%)
3.0SH h=0.01
ACC. [cm/s/s]
3.5SH
8000 4.0SH
6000 Figure 7 and Figure 8 show the shear stress-shear strain
relationship of LRBs during the shaking table tests. Moreover,
4000 breaking points of LRBs are shown in the figures. In the input
2000 earthquake motion of 3.0SH, hardening of LRBs occurred a
little. According to increase of the amplitude of input
0 earthquake motion, the hardening became clearer. In specimen-
0.01 0.1 1
PERIOD [sec] 2, the hysteresis loops in the earthquake motion of 4.5SH and
(c) Shaking table (Input) 4.8SH were not symmetry with respect to the origin. From the
results, occurrence of inner damage of LRBs before the
Fig.5 FLOOR ACCELERATION RESPONSES SPECTRA
IN SPECIMEN-1 (Damping factor h=1%)
breaking is estimated.
Input: 3.5SH Input: 4.0SH ultimate state is in good agreement with that of monotonic
4 4
loading test. Failure strain of the shaking table test is somewhat
2 2
0 0
larger than that of monotonic loading test, whereas it is almost
-2 -2
the same as that of repetition loading test.
-4 -4 8
-6 -6 4.0SH Specimen-1
6
-8 -8 Breaking Repetition loading
Monotonic loading
2 8 -2
Specimen-2 6 Specimen-2
SHEAR STRESS (MPa)
-4
Input: 1.0SH Input: 4.0SH
1 4
2 -6
0 0 -8
-2 -600 -300 0 300 600
-1 -4 SHEAR STRAIN (%)
-6 Fig.9 SHEAR STRESS - SHEAR STRAIN RELATIONSHIP
-2 -8 OF SHAKING TABLE TEST, STATIC REPETITION
-200 -100 0 100 200 -600 -300 0 300 600 LOADING TEST AND MONOTONIC LOADING TEST
SHEAR STRAIN (%) SHEAR STRAIN (%)
Figure 10 shows relationship between vertical stress and
8 8
Breaking shear strain of partially broken LRB of specimen-1 and
6 Specimen-2 6 Specimen-2
SHEAR STRESS (MPa)
Input: 4.5SH Input: 4.8SH specimen-2 in comparison with the failure condition of LRBs
4 4
obtained by static loading tests. The hysteresis curve of LRB #4
2 2
0 0
of specimen-1 in the earthquake motion of 3.5SH reaches the
-2 -2
boundary of tensile stress, whereas it is smaller than the failure
-4 -4
condition. The hysteresis curve of specimen-1 in the motion of
-6 -6
4.0SH exceeds the boundary of failure condition in both
-8 -8
compress and tensile regions; failure of LRB #4 in specimen-1
-600 -300 0 300 600 -600 -300 0 300 600 occurred in compress region. The hysteresis curve of LRB #2 of
SHEAR STRAIN (%) SHEAR STRAIN (%) specimen-2 in the earthquake motion of 4.0SH is quite similar to
that of specimen-1; they were symmetrical with respect to the
Fig.8 RELATIONSHIP BETWEEN SHEAR STRESS AND
SHEAR STRAIN OF LRBs IN SPECIMEN-2 vertical axis. Nevertheless, failure of specimen-2 in the motion
of 4.0SH did not occur, because the hysteresis curve may be a
Comparison between Results of Static Failure Tests and little smaller than that of specimen-1 in compress region. The
Those of Shaking Table Tests hysteresis curve of specimen-2 in the motion of 4.8SH exceeds
Static loading tests of LRBs were performed before the the boundary of failure condition; failure of LRB occurred in
shaking table tests. The static loading tests were both compress region. However, input earthquake motion increased,
examinations of the design performance and static failure tests. the increase of vertical stress in tensile side was relatively small.
5 Input: 3.5SH Failure surfaces of broken LRBs in the shaking table tests
were observed and classified according to failure modes of
0
rubber and bond in them. Figure 11 shows failure surfaces and
-5 the results of classified failure for LRB #3 and #4 of specimen-
1. The states of the failure surfaces at the starting point of the
-10 breaking were failure of rubber material; the state of broken
LRB of specimen-2 was the same. Therefore, failure of LRBs
-15
occurred by failure of rubber material, because of large shear
-20 deformation. The integrity of the bond between rubber sheets
-600 -400 -200 0 200 400 600 and inner steel plates of LRBs was confirmed.
SHEAR STRAIN (%)
10
Specimen-1 LRB #4 Failure in static test
VERTICAL STRESS (MPa)
5 Input: 4.0SH
-5
-10
Breaking
-15
-20
-600 -400 -200 0 200 400 600 Starting point of Breaking
SHEAR STRAIN (%) (a) LRB #3 of specimen-1
10
Specimen-2 LRB #2 Failure in static test
VERTICAL STRESS (MPa)
5 Input: 4.0SH
No breaking area
0
-5
-10
-15
-20
-600 -400 -200 0 200 400 600 Starting point of Breaking
SHEAR STRAIN (%) (b) LRB #4 of specimen-1
10 Fig.11 FAILURE SURFACE AND FAILURE MODE (black:
Specimen-2 LRB #2 Failure in static test
failure of rubber including no breaking, white: failure of bond)
VERTICAL STRESS (MPa)
5 Input: 4.8SH
ACC. [cm/s/s]
out to investigate the response, especially the function of Specimen-1
1500
response reduction. Before the shaking, it was confirmed that A3
LRBs kept the function of support of the upper structure by the h=0.01
1000
inspection after their failure, because of no gap in their breaking
points. 500
Table 7 and Table 8 show maximum responses of
specimens before and after the failure of LRBs. The difference 0
of the input accelerations was very small in each specimen. 0.01 0.1 1
Response acceleration at lower slab A3 after the failure was a PERIOD [sec]
little smaller than that before the failure. On the other hand, (a) Specimen-1
shear strain of LRBs after the failure was larger than that before 2500
the failure.
Figure 14 shows acceleration response spectra at A3 of Before Failure
2000 After Failure
each specimen. The peak period of the spectra after the failure
ACC. [cm/s/s]
became longer than that before the failure. Since LRBs were 1500 Specimen-2
damaged by extremely strong earthquake motions, their A3
stiffness after the failure slightly decreased. 1000 h=0.01
Nevertheless, the difference of the responses of upper
structure before and after the failure was small. From the 500
results, it was confirmed that the function of response reduction
in the specimens after the failure remained. 0
0.01 0.1 1
Table 7 MAXIMUM RESPONSES OF SPECIMEN-1 BEFORE
PERIOD [sec]
AND AFTER THE FAILURE OF LRBs (b) Specimen-2
Before failure After failure Fig.14 ACCELERATION RESPONSES SPECTRA AT THE
Input acceleration [cm/s2] 533 520 LOWER SLAB A3 BEFORE AND AFTER THE FAILURE
Acceleration at A3 [cm/s2] 203 191 OF LRBs UNDER THE TENTATIVE DESIGN
Shear strain of LRBs [%] 69 106 EARTHQUAKE MOTION (Damping factor: h=1%)
ACKNOWLEDGMENTS
The authors would like to express their applications to
Prof. T. Fujita, University of Tokyo, for his helpful suggestions.
The shaking table test was carried out with the cooperation of
National Research Institute of Earth Science and Disaster
Prevention, and the static loading and material tests of lead
rubber bearings were performed with the cooperation of
Bridgestone Corporation.
REFERENCES
[1] Moteki, M. et al., ”Shaking table test on ultimate behavior
of seismic isolation system Part 1: Outline of the test and
response of superstructure, 10th World Conference on
Earthquake Engineering (10th WCEE), 1992.
[2] Ishida, K. et al., ”Shaking table test on ultimate behavior of
seismic isolation system Part 2: Response behavior of rubber
bearings, 10th World Conference on Earthquake Engineering
(10th WCEE), 1992.
[3] Ishida, K. et al., ”Recent Results of Seismic Isolation Study
in CRIEPI –Tests on Seismic Isolation Elements, Vibration
tests and Observations-, IAEA Specialist Meeting (San Jose,
California, U.S.A.), 1992.
[4] Watanabe, Y. et al., “Shaking Table Test on Failure
Characteristics of Base Isolation System”, Trans. of SMiRT-
15, 1999, K16/4.
[5] Kitamura, S., et al., “Shaking Table Tests with Large Test
Specimens of Seismically Isolated FBR Plants, Part 1:
Response Behavior of Test Specimen under Design Ground
Motions“, ASME PVP09, 2009.
[6] Inaba, S., et al., “Shaking Table Tests with Large Test
Specimens of Seismically Isolated FBR Plants, Part 2:
Damage Test of Reinforced Concrete Wall Structure“,
ASME PVP09, 2009.
[7] Ohtani, K. et al., “World’s Largest Shaking Table Takes
Shapes in Japan (The 3rd Report)”, SMiRT 17 2003, K12-1.
[8] Kajiwara, K. et al., "Shaking table and activities at E-
Defense", First European Conference on Earthquake
Engineering and Seismology, Geneva, Switzerland, 2006,