Experimental Study of Seismic Soil-Structure Interaction by using

The 14th

World Conference on Earthquake Engineering October 12-17, 2008, Beijing, China

Experimental Study of Seismic Soil-Structure Interaction by using Large Geotechnical Centrifuge System

T. Kawasato1, T. Okutani1, T. Ishikawa1, T. Fujimori2, and M. Akimoto3

1 Projects Development Department, The Japan Atomic Power Company, Tokyo, Japan

2 Technical Research Institute, Obayashi Corporation, Tokyo, Japan

3 Nuclear Facilities Division, Obayashi Corporation, Tokyo, Japan

ABSTRACT :

The purpose of this study is to obtain empirical data concerning soil-structure interaction and basemat uplift phenomenon, focusing on nuclear power plants built on a hard rock site. Two kinds of vibration tests by using alarge geotechnical centrifuge system, which could realize contact pressure almost same as that of real plantswith a miniature building model, are performed. Through this experimental study, beneficial knowledgeconcerning soil-structure interaction and basemat uplift was obtained.

KEYWORDS: Soil-Structure Interaction, Basemat uplift, Embedment effect, Centrifuge test, Impedance functions, Effective input motion

1. INTRODUCTON Evaluation of dynamic responses of nuclear power plants (NPP) considering soil-structure interaction (SSI) isstrictly required for seismic design in Japan. A sway-rocking model is used often, and soil springs are estimated from the vibration admittance theory, which is based on the three dimensional wave propagation theory for theuniform half-space soil medium. Basemat uplift is also considered by using a rotational soil spring with thegeometric nonlinear characteristics. It has been requested more reasonably to evaluate basemat upliftphenomenon as the level of earthquake input motion increases accompanied with the recent development ofseismology. In the previous study, we proposed a procedure to construct a proper numerical soil model using three dimensional finite elements, in which various geometrically complex site conditions can be considered. Asfor a NPP built on a hard rock site, there is little data concerning embedment effect and uplift phenomenon,which is obtained by experimental or observational studies. The purpose of this study is to obtain empirical data, conductive to validate the proposed evaluation method mentioned above. Two kinds of vibration tests by using a large geotechnical centrifuge system, which could realize contact pressure almost same as that of realplants with a miniature building model, are performed. One is shaking test using an exciter to validate dynamicsoil springs, and the other is shaking table test to validate basemat uplift characteristics and effective input motion evaluated by the proposed method. Through this experimental study, beneficial knowledge concerningSSI phenomena for a hard rock site and effective data to validate the proposed method are obtained. 2. EXPERIMENTAL TESTS CONCERNING THE BASEMAT UPLIFT BEHAVIOR Dynamic soil spring, effective input motion, basemat-overturning-moment, basemat rotational angle and contact ratio are evaluated by the tests. Embedment effects and characteristics of basemat uplift are investigated from the test results. 2.1. Overview of Experiment Configuration of the test models are shown in Fig.2.1. Shallow embedment model (Case-1), deep embedment model (Case-2) and cavity model (Case-3) are made in a shear box for the centrifuge test. The building is modeled by steel. Considering the non-dimensional frequency (1.0), soil is made by soil cement (shear wavevelocity is 500m/s), corresponded to the actual soil of NPP.

The 14th

World Conference on Earthquake Engineering October 12-17, 2008, Beijing, China Gravitational force field of the centrifuge test is 20G. Converting the gravitational force feild to 1G based on the scaling law, the building footprint is 6*6m, weight of the building is 616tons, the embedment depth of Case-1 is 0.66m (22% of gravity center height), and that of Case-2 is 2m (66% of gravity center height). The cavity depth of Case-3 is same as the embedment depth of Case-2 (2m). In the shaking test for the building model using asmall exciter, sinusoidal vibration (2-15Hz) is applied. In the shaking table tests, excitation of sine waves(2,4,8,10Hz) and earthquake waves are applied. Acceleration of the building and the soil (13 points in Case-1,2, 16 points in Case-3), and earth pressure (15 points in Case-1,2,3) are measured. Resonance curve and dynamic soil spring are evaluated from the shaking test results using an exciter. Basemat overturning moment, basemat rotational angle, contact ratio and effective input motion of the cavity are evaluated from the shaking table testresults. The following scale is converted to the scale of 1G gravitational force field based on the scaling law.

GL

300(6m)

300(6m)

480(9.6m)

33(0.66m) 565(11.3m)

Vs500m/s，ρ=1.6t/m3 ，Soil cement

1950(39m)

300(6m)

Earth pressure gauge

250

250(@25)

: Accelerometer

(a) Shallow embedment (Case-1)

300(6m)

300(6m) GL

300

(6m)

480(9.6m)

100(2m) 565(11.3m)


1950(39m)

300(6m)

(b) Deep embedment (Case-2)

300(6m)

300

(6m)

240(4.8m) 165(3.3m)

315(6.3m)

GL 100(2m) 565(11.3m)


1950(39m)

300(6m)

(c) Cavity (Case-3)

Figure 2.1 Configuration of test model 2.2. Shaking Tests for Building Model using Exciter Dynamic soil springs are evaluated from resonance curves obtained by the shaking test using an exciter set onthe building model. Horizontal resonance curves of the building model are shown in Fig.2.2. Natural frequencyof Case-1 is 12.7Hz, and that of Case-2 is 15.5Hz. Natural frequency and soil spring increased by the embedment effect. The peak of resonance curve in Case-2 is not so sharp as that of Case-1, and the slope of phase curve in Case-2 is not so steep as that of Case-1. The effect of radiation damping has increased by theembedment effect. Dynamic soil springs are shown in Fig.2.3. Soil spring of Case-2 is larger than that of Case-1. Soil spring has also increased by the embedment effect.

The 14th


0.0E+00

5.0E-05

1.0E-04

1.5E-04

2.0E-04

0 2 4 6 8 10 12 14 16Frequency(Hz)

Dis

plac

emen

t(m)

Horizontal displacement (RF)Horizontal displacement (1F)

0.0E+00

2.0E-05

4.0E-05

6.0E-05

8.0E-05

0 2 4 6 8 10 12 14 16Frequency(Hz)

Dis

plac

emen

t(m)


-180

-90

0

90

180

0 2 4 6 8 10 12 14 16Frequency(Hz)

Pha

se la

g(°)


-180

-90

0

90

180

0 2 4 6 8 10 12 14 16Frequency(Hz)

Pha

se la

g(°)


(a) Case-1 (b) Case-2 Figure 2.2 Resonance curves of the building model

-2.0E+07

0.0E+00

2.0E+07

4.0E+07

6.0E+07

0 2 4 6 8 10 12 14 16Frequency(Hz)

Hor

izon

tal s

prin

g(kN

/m) Horizontal(Real)

Horizontal(Imaginary)

-2.0E+07

0.0E+00

2.0E+07

4.0E+07

6.0E+07

0 2 4 6 8 10 12 14 16Frequency(Hz)

Hor

izon

tal s

prin

g(kN

/m)

Horizontal(Real)Horizontal(Imaginary)

(a) Horizontal spring (Case-1) (c) Horizontal spring (Case-2)

-1.0E+08

0.0E+00

1.0E+08

2.0E+08

3.0E+08

4.0E+08

0 2 4 6 8 10 12 14 16Frequency(Hz)

Rot

atio

nal s

prin

g(kN・

m/ra

d.) Rotational(Real)

Rotational(Imaginary)

-1.0E+08

0.0E+00

1.0E+08

2.0E+08

3.0E+08

4.0E+08

0 2 4 6 8 10 12 14 16Frequency(Hz)

Rot

atio

nal s

prin

g(kN・

m/ra

d.)

Rotational(Real)Rotational(Imaginary)

(b) Rotational spring (Case-1) (d) Rotational spring (Case-2)

Figure 2.3 Dynamic soil springs 2.3. Shaking Table Tests for Soil Model with Cavity Effective input motion in embedment foundation is derived from the spectrum ratio of response of cavity to freefield of soil model. Comparison of effective input motions for random wave (max. acc. 25gal), earthquake wave

The 14th

World Conference on Earthquake Engineering October 12-17, 2008, Beijing, China 1 (max. acc. 365gal) and earthquake wave 2 (max. acc. 533gal) are shown in Fig.2.4. As for each wave, inputmotion from 8Hz to 10Hz decrease within the range from 0.75 to 0.9. Characteristic of effective input motionhas little difference between each wave. Frequency less than 10Hz, the level of earthquake input motion has an insignificant effect on reduction of effective input motion at a hard rock site.

0

0.5

1

1.5

2

0 2 4 6 8 10

Random waveEarthquake wave 1Earthquake wave 2

Effe

ctiv

e in

put m

otio

n

Frequency(Hz)

Figure 2.4 Effective input motion 2.4. Shaking Table Tests for Building Model Relations between overturning moment and rotational angle (M-θ), overturning moment and contact ratio(M-η) of the basemat in the shaking table tests are investigated. M-θ and M-η relations are shown in Fig.2.5. Minimum contact ratio of Case-1 is from 40 to 60%, and that of Case-2 is from 60 to 75%. The increase of contact ratio caused by the embedment is recognized in the test results.

Figure 2.5 Relations of M-θ, M-η

3. SIMULATION ANALYSES FOR EXPERIMENTAL TESTS Simulation analyses listed below are performed based on the experimental test results mentioned in the previous section. (1) Impedance functions for the building model (2) Effective input motion for the soil model with a cavity (3) Basemat uplift characteristics for the building model

(x104)

-1.5 0.0 1.5

(x10-4)

-3.0 0.0 3.0

100

80

60

40

20

0

Overturning Moment M(kN･m)

Rotational Angle θ(rad)

Con

tact

Ratio

(%)

(x104)

1.5

0

-1.5

Over

turn

ing Mo

men

t M (kN･m)

(a) 40Hz [2Hz]

Shallow embedment (Case-1)

(x104)

-1.5 0.0 1.5

(x10-4)

-3.0 0.0 3.0

100

80

60

40

20

0



Cont

act

Ratio

(%)

(x104)

1.5

0

-1.5 Over

turn

ing Mom

ent

M (kN

･m)

Deep embedment(Case-2)

回転角(rad.)

-8.0 0.0 8.

転倒

モー

メン

ト(kN･m)

-20000.0

0.0

20000.0

転倒

モー

メン

ト（kN

・m

）

回転角(rad.)

-8.0 0.0 8.

転倒

モー

メン

ト(kN･m)

-20000.0

0.0

20000.0

転倒

モー

メン

ト（kN

・m

）接

地率

(%)

0.0

20.0

40.0

60.0

80.0

100.0

接地

率（%）

接地

率(%)

0.0

20.0

40.0

60.0

80.0

100.0

接地

率（%）

回転角(rad.)

-8.0 0.0 8.

転倒

モー

メン

ト(kN･m)

-20000.0

0.0

20000.0

転倒

モー

メン

ト（kN

・m

）

回転角(rad.)

-8.0 0.0 8.

転倒

モー

メン

ト(kN･m)

-20000.0

0.0

20000.0

転倒

モー

メン

ト（kN

・m

）接

地率

(%)

0.0

20.0

40.0

60.0

80.0

100.0

接地

率（%）

接地

率(%)

0.0

20.0

40.0

60.0

80.0

100.0

接地

率（%）

(x104)

-2.0 0.0 2.0

(x10-4)

-8.0 0.0 8.0

100

80

60

40

20

0



Con

tac

t Ra

tio

(%)

(x104)

2.0

0

-2.0

Over

tur

ning

Mom

ent M (kN･m

)

Shallow embedment (Case-1)

(x104)

-2.0 0.0 2.0

(x10-4)

-8.0 0.0 8.0

100

80

60

40

20

0



Cont

act

Rat

io (%

)

(x104)

2.0

0

-2.0 Over

turn

ing Mom

ent

M (kN･

m)

Deep embedment(Case-2)

(b) 160Hz [8Hz]

The 14th

World Conference on Earthquake Engineering October 12-17, 2008, Beijing, China 3.1. Analytical Model 3.1.1 Experimental soil medium and shear box system A three dimensional finite element (3D FE) analytical model about experimental soil medium and shear box isshown in Fig. 3.1. Solid elements are used to model the soil medium for both shallow and deep embedmentmodels (embedment depths are 3.3cm and 10.0cm for each) and the shear box (rubber layers between steelframes). Steel frames of the shear box are hollow structures, whose thickness is varying from 0.23cm to 0.6cm,but they are modeled as dense solid with equivalent density and elastic modulus.

Figure 3.1 Analytical model of the soil medium and the shear box

3.1.2 Building model The 3D FE building model is shown in Fig. 3.2. The FE model is composed of shell elements which are 1.6 cm thick, and it is connected with the 3D FE soil medium model by joint elements. As the initial stiffness of jointelements, 50 times as large as static stiffness of the soil medium is adopted based on the previous study (Ref 1).

Figure 3.2 Analytical model of the building model

3.2. Impedance Functions for Building Model Firstly, impedance functions are verified by comparing with the shaking table test results using an exciter.Several sine-wave force time histories with different frequencies are prepared and a series of dynamic analysesare performed by applying the sine-wave force time histories to the area, where the basemat of the buildingmodel exists, in order to obtain the responses at the soil surface. Impedance function is defined from the relationship between response displacement and applied force. The basemat area is stiffened in the analysis byusing mass-less rigid shell elements in order to provide a uniform displacement distribution under the basemat. Complex impedance functions evaluated by 3D FE model, which are compared with the experimental results,shown in Figs. 3.3 and 3.4. Fig. 3.3 shows the comparison of the shallow embedment case. Fig. 3.4 shows thecomparison of the deep embedment case. These results are shown in 1G gravitational force field convertingfrom 20G field of the centrifuge test based on the scaling law. It can be recognized that the horizontal androtational impedance functions evaluated by FE model correspond well to the experimental results, but when looking at the comparison in detail, the experimentally evaluated points of the horizontal impedance functionfor the deep embedment case seems to be a little scattering.

B300xD300xH100 or H33 [6] [6] [2] or [0.66]

Thickness: H300 [6]

1100 [22]

2250[45]

1950 [39]

800[16]

Shear box

Cavity

Unit: mm

Rubber LayerSteel Frame

Soil Medium

[ ]: Equivalent dimension under 1.0G field (m)

300[6]

300[6]

300[6]

Building model

Experimental soil model Joint elements

[ ]: Equivalent dimension under 1.0G field (m) Unit: mm

The 14th


Figure 3.3 Comparison of the experimental and analytical impedance functions (Shallow embedment)

Figure 3.4 Comparison of the experimental and analytical impedance functions (Deep embedment)

3.3. Effective Input Motion for Soil Model with Cavity Secondly, effective input motion for the soil model with a cavity is verified by comparing with the shaking tabletest results. Several sine-wave acceleration time histories with different frequencies, whose amplitude are1.0Gal, are prepared. Applying these sinusoidal acceleration time histories at the bottom of the soil mediamodeled by solid elements, maximum acceleration distribution of the surface of soil model is evaluated in eachfrequency. Fig.3.5(a) shows acceleration distribution at the surface of the soil model for sinusoidal acceleration time history of 160Hz in 20G field (8Hz in 1G field). It could be recognized that maximum response at thecavity is smaller than that of the area surrounding the cavity. Maximum response at the peripheral area of thesoil model is relatively large because the response of this area is affected by the existence of the shear box. Inthe experimental test, an accelerometer is set at the point whose distance from the center of the cavity is 48 cmalong the excitation direction (Point B in Fig.3.5(a)). This point is considered to be little affected by theexistence of both the cavity and the shear box, and it is also considered that the response of this point is treatedas that of the free field. Effective input motion at the cavity could be evaluated by dividing the maximum acceleration of Point A by that of Point B. Effective input motions for each frequency are evaluated byrepeating this procedure, using a series of acceleration time histories with different frequencies.

Figure 3.5 Comparison of experimentally and analytically evaluated effective input motion

1.46

1.43

1.40

1.37

1.34

1.31

1.28

1.25

1.22

1.19

1.16

1.13

Position A (in a cavity)

Position B (free field) ランダム波地震波1地震波2解析結果

Effe

ctiv

e In

put M

otio

n

Frequency(Hz)0 2 4 6 8 10

0.0

0.5

1.0

2.0

Experimental 1(Random wave) Experimental 2(Earthquake wave 1) Experimental 3(Earthquake wave 2) Analytical

(a) Analytical Max. Acc. Distribution (8Hz) (b) Effective input motion

回転実部回転虚部回転実部（解析)回転虚部（解析)

0 2 4 6 8 10 12 14 16振動数（Hz）

6.05.04.0

2.0

0.0-1.0

回転

ばね

(ｋN

m/r

ad)

3.0

1.0

×108

0 2 4 6 8 10 12 14 16振動数（Hz）

0 2 4 6 8 10 12 14 160 2 4 6 8 10 12 14 16振動数（Hz）

6.05.04.0

2.0

0.0-1.0

回転

ばね

(ｋN

m/r

ad)

3.0

1.0

×108

6.05.04.0

2.0

0.0-1.0

回転

ばね

(ｋN

m/r

ad)

3.0

1.0

×108 Experimental (Real) Experimental (Imag) Analytical (Real) Analytical (Imag)

Rot

atio

nal K

R (k

Nm

/rad)

Frequency(Hz)

水平　実部水平　虚部水平実部（解析)水平虚部（解析)

0 2 4 6 8 10 12 14 16振動数（Hz）

×107

8.0

6.0

4.0

2.0

0.0

-2.0

水平

ばね

(ｋN

/m)

0 2 4 6 8 10 12 14 16振動数（Hz）

0 2 4 6 8 10 12 14 160 2 4 6 8 10 12 14 16振動数（Hz）

×107

8.0

6.0

4.0

2.0

0.0

-2.0

水平

ばね

(ｋN

/m)

×107

8.0

6.0

4.0

2.0

0.0

-2.0

水平

ばね

(ｋN

/m)

8.0

6.0

4.0

2.0

0.0

-2.0

8.0

6.0

4.0

2.0

0.0

-2.0

水平

ばね

(ｋN

/m)

Experimental (Real) Experimental (Imag) Analytical (Real) Analytical (Imag)

Frequency(Hz)

Hor

izon

tal K

H (k

N/m

)

回転実部

回転虚部

回転実部（解析)

回転虚部（解析)

6.05.04.0

2.0

0.0-1.0

回転

ばね

(ｋN

m/r

ad)

3.0

1.0

×108

0 2 4 6 8 10 12 14 16振動数（Hz）

6.05.04.0

2.0

0.0-1.0

回転

ばね

(ｋN

m/r

ad)

3.0

1.0

×108

6.05.04.0

2.0

0.0-1.0

回転

ばね

(ｋN

m/r

ad)

3.0

1.0

×108

0 2 4 6 8 10 12 14 16振動数（Hz）

0 2 4 6 8 10 12 14 160 2 4 6 8 10 12 14 16振動数（Hz）


Rot

atio

nal K

R (k

Nm

/rad)

Frequency(Hz)

水平　実部

水平　虚部

水平実部（解析)

水平虚部（解析)

0 2 4 6 8 10 12 14 16振動数（Hz）

×107

8.0

6.0

4.0

2.0

0.0

-2.0

水平

ばね

(ｋN

/m)

0 2 4 6 8 10 12 14 16振動数（Hz）

0 2 4 6 8 10 12 14 160 2 4 6 8 10 12 14 16振動数（Hz）

×107

8.0

6.0

4.0

2.0

0.0

-2.0

水平

ばね

(ｋN

/m)

×107

8.0

6.0

4.0

2.0

0.0

-2.0

水平

ばね

(ｋN

/m)

8.0

6.0

4.0

2.0

0.0

-2.0

8.0

6.0

4.0

2.0

0.0

-2.0

水平

ばね

(ｋN

/m)

Hor

izon

tal K

H (k

N/m

)


Frequency(Hz)

The 14th

World Conference on Earthquake Engineering October 12-17, 2008, Beijing, China Effective input motion for the cavity evaluated by 3D FE model, which are being compared with the shakingtable test results, are shown in Fig. 3.5(b). This result is also shown in 1G gravitational force field convertingfrom 20G field of the centrifuge test based on the scaling law. It can be recognized that the effective inputmotion at the cavity evaluated by 3D FE model corresponds to the experimental result very well. As the excitation frequency becomes higher, effective input motion at the cavity decrease. It is admitted that the inputmotion to the building model decreases to 0.8 or 0.9 times of the original input motion near the excitation frequency of 10 Hz. 3.4. Basemat Uplift Characteristics for Building Model Finally, basemat uplift characteristics for the building model are verified by comparing with the shaking tabletest results. Dynamic analyses for the building-soil model system are performed by applying sinusoidal acceleration time histories at the bottom of the soil medium modeled by solid elements. Basemat overturningmoment, basemat rotational angle, and contact ratio are evaluated as the indices of basemat upliftcharacteristics. Basemat overturning moment M is obtained from the horizontal acceleration distribution of thebuilding model, same as the M obtained from the test where acceleration distribution is measured byaccelerometers attached to the building model in the experimental tests. Basemat rotational angle θ is obtained from vertical response displacement distribution of the basemat. Contact ratio η is obtained from the response of joint elements. Figs.3.6 through 3.7 show the comparison between experimental and analytical results concerning M-θ and M-η relations of the basemat. The maximum acceleration of input sinusoidal motion for each excitationfrequency is decided that the minimum contact ratio becomes about 60% in the shallow embedment case.Fig.3.6 shows the results for the shallow embedment case and Fig.3.7 shows the results for the deep embedmentcase. These results are also shown in 1G gravitational force field converting from 20G field of the centrifugetest by the scaling law. From these figures, it can be recognized that the analytical results by FE model correspond to the experimentalresults qualitatively well. For the quantitative comparison, the maximum amplitude of over turning moments ofboth experimental and analytical results are almost same, but the gradient of experimentally evaluated M-θcurve is lower than that of analytically evaluated M-θ curve. The soil cement used as experimental soil mediumis a strain-dependent material unlike the bedrock of the actual nuclear power plants, and has a nature that the stiffness becomes smaller as the response strain becomes larger. This may be one of the reasons of thedifference between experimental and analytical gradients for M-θ curve. Moreover, regarding the difference of the embedment depths of the building model, it is confirmed that the response of the building model apparentlytends to decrease in quantity by considering embedment effects, and this tendency becomes more remarkable asthe excitation frequency is increasing. It is also confirmed that this result is harmonized with the discussion of effective input motion for the cavity soil model discussed in the previous section.

Figure 3.6 M-θ, M-η by experiment and analysis (Shallow embedment)

3.03.0 0.00.0 3.03.0回転角（回転角（radrad））

3.03.0 0.00.0 3.03.0回転角（回転角（radrad））

(x104)

-1.5 0.0 1.5

(x104)

-1.5 0.0 1.5

(x10-4)

-3.0 0.0 3.0

(x10-4)

-3.0 0.0 3.0

100

80

60

40

20

0

100

80

60

40

20

0

(x104)

1.5

0

-1.5

Overturning Moment M(kN･m) Overturning Moment M(kN･m)

Rotational Angle θ(rad) Rotational Angle θ(rad)

Contact

Ratio (%)

Contact

Ratio (%)

Overturning Moment M (kN･m) (x104)

1.5

0

-1.5

Overturning Moment M (kN･m)

(a) 40Hz [2Hz]

.0.0 0.00.0 8.08.0回転角（回転角（radrad））

××1010--44

××101044

回転角(rad.)

-8.0 0.0 8.

転倒

モー

メン

ト(kN･m)

-20000.0

0.0

20000.0

転倒

モー

メン

ト（kN

・m

）

回転角(rad.)

-8.0 0.0 8.

転倒

モー

メン

ト(kN･m)

-20000.0

0.0

20000.0

転倒

モー

メン

ト（kN

・m

）接

地率

(%)

0.0

20.0

40.0

60.0

80.0

100.0

接地

率（%）

接地

率(%)

0.0

20.0

40.0

60.0

80.0

100.0

接地

率（%）

.0.0 0.00.0 8.08.0回転角（回転角（radrad））

××1010--44

××101044

回転角(rad.)

-8.0 0.0 8.

転倒

モー

メン

ト(kN･m)

-20000.0

0.0

20000.0

転倒

モー

メン

ト（kN

・m

）

回転角(rad.)

-8.0 0.0 8.

転倒

モー

メン

ト(kN･m)

-20000.0

0.0

20000.0

転倒

モー

メン

ト（kN

・m

）接

地率

(%)

0.0

20.0

40.0

60.0

80.0

100.0

接地

率（%）

接地

率(%)

0.0

20.0

40.0

60.0

80.0

100.0

接地

率（%）

(x104)

-2.0 0.0 2.0 (x104)

-2.0 0.0 2.0

(x10-4)

-8.0 0.0 8.0

(x10-4)

-8.0 0.0 8.0

100

80

60

40

20

0

100

80

60

40

20

0



Contact Ratio (%)

Contact Ratio (%)

(x104)

2.0

0

-2.0

(x104)

2.0

0

-2.0

Overturning Mom

ent M (kN･m)

Overturning Mom

ent M (kN･m)

Analysis Experiment

(b) 160Hz [8Hz]

Analysis Experiment

The 14th


Figure 3.7 M-θ, M-η by experiment and analysis (Deep embedment)

4. CONCLUSIONS In this paper, the experimental studies concerning soil-structure interaction at a hard rock site are performed focusing on basemat embedment effects and basemat uplift behavior, in order to verify the knowledgeconcerning basemat embedment effects obtained in the previous analytical study, and to evaluate basemat upliftbehavior quantitatively. Three kinds of vibration tests, which are concerning basemat embedment effects anduplift phenomena, are performed. Those are (1) Shaking tests for building model using an exciter settled on thebuilding model, by which impedance functions of the experimental soil model are evaluated, (2) Shaking table tests for soil with cavity, by which the reduction effect of the input motion for the cavity of soil are confirmed, and (3) Shaking table tests for building model, by which base uplift characteristics and embedment effects are confirmed. Experimental soil model whose shear wave velocity is 500m/s, which is appropriate to capture dynamicbehavior of actual NPP around the non-dimensional frequency 1.0, is used and the experimental tests in 20Ggravitational force field are performed by the large geotechnical centrifuge system. Basemat embedment effects and geometrical nonlinearity of basemat uplift could be confirmed from impedancefunctions and basemat uplift characteristics of the various experimental studies performed by varying theembedment depth of the building model. These correspond with the knowledge obtained from the previousanalytical studies, and effectiveness of calculation model for basemat uplift behavior concerning NPP built on ahard rock site, which is proposed in the previous study (Ref 1), is verified. ACKNOWLEDGEMENT The authors wish to express cordial gratitude to Dr. Takao Nishikawa, Professor emeritus at TokyoMetropolitan University, for many useful advices and suggestions given to this study. REFERENCES 1) T.Kawasato, T.Okutani, O.Kurimoto, M.Akimoto (2007). A study on Evaluation of Seismic Response considering Basemat Uplift for Soil-building System using 3D FEM. Trans. of SMiRT19, Vol.K, K06/4. 2) Tanaka,H., Maeda,I., Moriyama,K, Watanabe,S. (1995). Study on horizontal-vertical interactive SR model for basemat uplift (Part 1 & Part 2). Trans. of SMiRT13, Vol.K, 43-54. 3) Japan Electric Association, Technical Guidelines for Aseismic Design of Nuclear Power Plants, JEAG 4601-1991 Supplement (in Japanese) 4) Tajimi,H. (1968), Interaction of Building and Ground, Earthquake Engineering, Shokokusha Publishing Co.

00 0.00.0回転角（回転角（radrad））

3.3.

00 0.00.0回転角（回転角（radrad））

3.3.

(x104)

-1.5 0.0 1.5

(x104)

-1.5 0.0 1.5

(x10-4)

-3.0 0.0 3.0 (x10-4)

-3.0 0.0 3.0

100

80

60

40

20

0

100

80

60

40

20

0



Con

tact

Ra

tio (%

)

Cont

act

Ratio

(%)

(x104)

1.5

0

-1.5

(x104)

1.5

0

-1.5 Ove

rturn

ing

Moment M (kN･

m)

Over

turn

ing Mom

ent

M (kN

･m)

Analysis Experiment

(a) 40Hz [2Hz]

8.0 0.0 8.回転角（rad）


Conta

ct Ra

tio

(%)

(x104)

-2.0 0.0 2.0

(x104)

-2.0 0.0 2.0

(x10-4)

-8.0 0.0 8.0

(x10-4)

-8.0 0.0 8.0

100

80

60

40

20

0

100

80

60

40

20

0



Cont

act

Rat

io (%)

(x104)

2.0

0

-2.0

(x104)

2.0

0

-2.0 Overt

urnin

g Mom

ent M (kN

･m)

Overt

urnin

g Mom

ent M (kN

･m)

Analysis Experiment

(b) 160Hz [8Hz]

Documents

Experimental Study of Seismic Soil-Structure Interaction by using