41102(375)9.pdf

A Scale Model Test on Dynamic Soil-tunnel Interactions

Nina Liu1, Yang Liu2, Dongdong Han3, Yuming Men4, Jianbing Peng5

1PhD candidate, instructor Department of Geological Engineering, Changan University, 126 Yangta Road, Xian China, 710054, [email protected] 2Master degree student, Department of Geological Engineering, Changan University, 126 Yangta Road, Xian China, 710054, [email protected] 3Master degree students, Department of Geological Engineering, Changan University, 126 Yangta Road, Xian China, 710054, [email protected] 4Director, Department of Geological Engineering, Changan University, 126 Yangta Road, Xian China, 710054, [email protected] 5Director, Department of Geological Engineering, Changan University, 126 Yangta Road, Xian China, 710054, [email protected] ABSTRACT: Soil-structure interactions are important for the safety of the underground structures under earthquake loads. In this paper, model tests were conducted on 1/60 scaled model of the prototype Xian metro tunnel. The seismic response data including the acceleration, the displacement and the dynamic soil pressure between the tunnel and soil were measured by testing the scaled model on a shaking table. From the test result, it is concluded that 1) the effects of tunnel and soil interactions are important for the bulk dynamic responses; 2) the soil pressure and the acceleration changed with the seismic inputs. The soil pressure at the top of the tunnel increased while the pressure at the bottom decreased; 3) the acceleration of the tunnel is greater than the adjacent soils. INTRODUCTION

To utilize the underground space in cities, more and more research has been focused on the soil and underground structure interactions (Zhao, 1994). The field survey conducted in Xian for the past 40 years revealed 14 ground fissures crossing the current metro tunnel site (Shi et al, 2009). Xian is known for its loess deposit and has a history of strong earthquakes. Therefore, designing the infrastructure to overcome the challenging from the ground fissures, earthquake activities and loess deposit is important for the safety of the Metro system (Huang et al., 2009).

Scale model experiments were conducted to study the soil-tunnel interactions. The model was 1/60 of the prototype. Soils used in the model test were taken from the construction site. Ground motions from the El Centro earthquake, the Wenchuan earthquake, synthetic Xian earthquakes and a sinusoidal wave were individually input to the shaking table during the experiments.

85

Copyright ASCE 2010 GeoShanghai 2010 International Conference Soil Dynamics and Earthquake Engineering

Dow

nloa

ded

from

asc

elib

rary

.org

by

GA

DJA

H M

AD

A U

NIV

ER

SIT

Y o

n 03

/04/

15. C

opyr

ight

ASC

E. F

or p

erso

nal u

se o

nly;

all

righ

ts r

eser

ved.

EXPERIMENT SETUP

The size of the model box was 1.5m1.0m0.6m, which was constraint by the capacity of the shaking table. The photo of the box is shown in Fig. 1(a). In order to reduce the rigid boundary effect by the box, plastic materials, pieces of woods and nails were attached to the inside of the container. This is shown in Fig. 1(b), indicating different materials in parallel to each other. The seismic excitation by the shaking table was in the direction of walls B and D. Therefore, plastic, pieces of wood and nails were installed on Walls B and D to reduce the friction between the soil and box. There were pieces of woods and nails on Walls A and C to make sure the soil and the box are moving together. (Hashashm et al., 1998; Hou, 2005)

(a) (b) FIG. 1 (a) the model box (b) the materials inside of the box

Following the Bingham principle, the model tunnel was as close as possible to the prototype tunnel; the soil used in the model box was the same as that at the site; and most importantly, the space relationship between soil and tunnel in the model was similar to the actual conditions. The size and weight of the model were designed to the capacity of the shaking table. The size of the model tunnel was 1/60 of the actual tunnel. The material used for the model tunnel was Plexiglass, with an elastic modules , density , and poison ratio 0.35. Fig. 2 shows the model tunnel (a) and the compensator ring (d). (Li et al., 2009)

(a) (b) (c) (d)

FIG. 2 (a) the model tunnel (b) the sensor of the strain (c) the location of the strain sensors on the tunnel (d) compensators

The FIG. 3 show the dynamic sensors used in the experiments to measure the

acceleration and the soil pressure.

GEOTECHNICAL SPECIAL PUBLICATION NO. 20186


Dow

nloa

ded

from

asc

elib

rary

.org

by

GA

DJA

H M

AD

A U

NIV

ER

SIT

Y o

n 03

/04/

15. C

opyr

ight

ASC

E. F

or p

erso

nal u

se o

nly;

all

righ

ts r

eser

ved.

(a) (b) (c)

FIG. 3 (a, b) the soil pressure sensor (c) the acceleration sensor

INSTRUMENTATION

The shaking table was fabricated by the MTS Systems Corporation, Minneapolis, Minnesota USA. The size of the shaking table is 1.5m1.0m.

The dynamic measuring system for acceleration, soil pressure and displacement is IMC Mess-System GmbH. Three instruments were used to measure the strain: SDY2400, DH5923 and YE6262B Dynamic Measuring System. Figure 4 below shows the pictures of the dynamic measuring system.

(a) (b) (c) (d)

FIG. 4 the Dynamic Measuring System(a) SDY2400 (b) DH5923 (c) YE6262B (d) IMC Mess-System GmbH TEST PROCEDURES

According to the field survey in Xian, the ground fissures run almost perpendicular to the metro tunnel. So in the model box (see Fig. 5 (a)), the tunnel was aligned perpendicular to the fissures and the direction of the shaking table. As shown in Fig. 5 (b), the ground fissure was made of silver sand. Each layer of soil to the south of the ground fissure was compacted just one time whereas each layer to the north was compacted 3 times to match the ground fissure caused by un-uniform settlement displacement in the actual site.

(a) (b) (c)

FIG. 5 (a) the location of fissure (b) the fissure filled with silver sand (c) the model box THE LOADING PROCESSES

In the test the earthquake ground input to the shaking table included the synthetic

GEOTECHNICAL SPECIAL PUBLICATION NO. 201 87


Dow

nloa

ded

from

asc

elib

rary

.org

by

GA

DJA

H M

AD

A U

NIV

ER

SIT

Y o

n 03

/04/

15. C

opyr

ight

ASC

E. F

or p

erso

nal u

se o

nly;

all

righ

ts r

eser

ved.

Xian earthquakes, El Centro earthquake and Wenchuan earthquake. The synthetic Xian earthquakes are for the geological characters of Xian. The El Centro earthquake wave was recorded from the Empire Valley, US in 1940 and was used widely in earthquake research. The record was revised to match with the geological characters of the Xian metro site. The 2008 Wenchuan earthquake was a major seismic disaster and affected most parts of China. The seismic record of Wenchuan earthquake was used to determine the model responses from the long period earthquake activity. The characteristics of 18 different seismic loading were listed in Table 1.

Table1. The loading process

No. The dynamic wave The peak acceleration of the shaking table(m/s2) 1 synthetic Xian earthquake 0.28 2 El Centro earthquake wave 0.3 3 Sine wave 1.0 4 synthetic Xian earthquake 0.5 5 El Centro earthquake wave 0.58 6 Sine wave 0.7 7 synthetic Xian earthquakes 0.9 8 El Centro earthquake wave 0.8 9 synthetic Xian earthquake 1.0 10 Sine wave 1.5 11 El Centro earthquake wave 2.0 12 synthetic Xian earthquake 2.0 13 Sine wave 2.0 14 El Centro earthquake wave 2.4 15 synthetic Xian earthquake 2.5 16 Sine Wave 2.6 17 synthetic Xian earthquake 3.0 18 Wenchuan earthquake wave 3.0

The observations during the loading period are described as follows: after 200 hours

the soil in the south part of the box was lower than the soil in the north part. As Fig. 6 (a) shows, along the ground fissure, there was a settlement caused by the movement of the south and north part of the soil.

After the 12th loading process, the settlement of the south part was measured 1.0 mm, and along the surface of the fault, some granules of soil could be observed; After the 13th loading process, the settlement of the south part became 1.5mm, and from the fault there were other fissures going out perpendicular to the south part (Fig. 6 (c)); After the 14th loading process, the settlement of the south part became 2.0mm; After the18th loading process, the soil in the south of the box cracked into several pieces (Fig. 6 (d)).



Dow

nloa

ded

from

asc

elib

rary

.org

by

GA

DJA

H M

AD

A U

NIV

ER

SIT

Y o

n 03

/04/

15. C

opyr

ight

ASC

E. F

or p

erso

nal u

se o

nly;

all

righ

ts r

eser

ved.

(a) (b) (c) (d)

FIG6 (a) the fault of the fissure (b) the damage after 12th loading process (c) the damage after 13th loading process (d) the damage after 18th loading process

There were four soil pressure sensors placed in the model box, i.e., SP001, SP003,

SP004 and SP005 attached outside of the tunnel. There were five acceleration sensors in the model box, two inside the tunnel and three in the soil as shown in FIG. 7. The locations of the soil pressures and acceleration sensors were determined from the finite element simulation of the model box to obtain the complete dynamic response at different points of the tunnel and soil (Shi et al 2009; Liu et al 2009).

(a) (b)

FIG. 7 (a) the profile of the model box (b) the location of the soil pressure and acceleration sensors

The 9th loading stage is the synthetic Xian earthquake wave and the peak acceleration of the shaking table is 1.0m/s2. The data of the 9th loading process had good quality. Therefore most analyses in this paper were based on data from this loading process.

The FIG. 8 shows the soil pressure taken during the 9th loading process. From the soil pressure time history of SP001, SP003, SP004 and SP005, the four points have the similar acceleration time history and the accelerations outside the tunnel were different from each other. The measurement indicated that the maximum soil pressure was located



Dow

nloa

ded

from

asc

elib

rary

.org

by

GA

DJA

H M

AD

A U

NIV

ER

SIT

Y o

n 03

/04/

15. C

opyr

ight

ASC

E. F

or p

erso

nal u

se o

nly;

all

righ

ts r

eser

ved.

at SP001, reaching a value of -81.8Mpa, the negative sign stands for compression. At each location the soil pressure fluctuated from time to time. The soil pressure at SP003 located at the top of tunnel became a little greater after the earthquake wave, increasing from -77Mpa to -76.5Mpa, whereas the soil pressure of SP004 decreased from -78.4Mpa to -78.6Mpa. The increase of soil pressure at SP005 was insignificant.

(a) (b)

(c) (d)

FIG8 (a) the soil pressure of SP001 (b) the soil pressure of SP003 (c) the soil pressure of SP004 (d) the soil pressure of SP005

FIG. 9 below shows the acceleration data taken from the 9th loading process.

(a) (b)

(c) (d)

FIG. 9 (a) acceleration of A001 (b) acceleration of A002 (c) acceleration of A006 (d) acceleration of A007

The acceleration sensors A001 and A006 were on the same side of the box, where A001 being inside the tunnel and A006 in the soil. The acceleration time history of A001 was similar to that of A006, but the magnitude of peak accelerations of A006 was smaller than those of A001. It showed that the accelerations of the tunnel were bigger than the accelerations of soil around the tunnel. The acceleration sensor A002 was on the shaking table, the time history of A002 showed the motion of the shaking table. Comparing A002 with the time history of synthetic Xian earthquake wave indicated that the ground motion followed the synthetic Xian earthquake wave. This validated that the shaking table inputted the earthquake excitation correctly. A006 and A007 were on the opposite sides of the tunnel with equal distance to the tunnel, from the data of FIG.9 (c) and (d), the accelerations of A006 and A007 are similar to each other.



Dow

nloa

ded

from

asc

elib

rary

.org

by

GA

DJA

H M

AD

A U

NIV

ER

SIT

Y o

n 03

/04/

15. C

opyr

ight

ASC

E. F

or p

erso

nal u

se o

nly;

all

righ

ts r

eser

ved.

CONCLUSIONS

The results of the experiments indicated: 1. The model in the test matches with the reality of the site. The materials and the

conditions in the model closely represent the prototype metro tunnel. The test would indicate that the interaction between the soil and underground structure was as expected.

2. The soil pressure on the tunnel and in the soil changed with time. The time intervals of pressure change were similar to that of the earthquake loading. The soil pressure at the top of the tunnel increased while the pressure at the bottom decreased.

3. The acceleration of the tunnel is greater than the adjacent soils.

ACKNOWLEDGMENTS

Thanks to the National Natural Science Foundation (40772183, 40534021) and the Shaanxi Natural Science Foundation (2005D04), China Geology Survey (1212010641403) and Youth Foundation of Changan University (0305-1001). REFERENCES Zhao, Y. (1994). Random vibration for seismic analysis of multiply supported nuclear

piping. Case Western Reserve University. Shi Yuling, Men Yuming, Pen Jianbing, Liu Yang et al. (2009). Analysis on Xi an

ground-fissure destruction to Chang an Road overpass. The Chinese Journal of Geological Hazard and Control. Vol. 20(2), pp. 65-69.

Huang Qiangbing, Peng Jianbing, Fan Hongwei, Yang Peimin, Men Yuming (2009). Metro tunnel hazards induced by active ground fissures in Xian and relevant control measures. Chinese Journal of Geotechnical Engineering. Vol. 31(10), pp. 1525-1532.

Huo, H. (2005). Seismic design and analysis of rectangular underground structures, Purdue University.

Hashashm, Y.M.A., Tseng, W.S. and Krimotat, A. (1998). Seismic Soil-structure Interaction Analysis for Immersed Tube tunnel Retrofit. Geotechnical Earthquake Engineering and Soil Mechanics, Vol. III 2, pp. 1380-1391.ASCE Geotechnical Special Publication No.75.

Li Jie; Yue Qingxia; Chen Jun (2009) Research on shaking-table test and finite element numerical simulation of utility tunnel. Journal of Earthquake Engineering and Engineering Vibration. Vol 29 (4), pp. 41~45.

Shi Yuling; Men Yuming; Peng Jianbing; Huang Qiangbing; Liu Hongjia; Chen Liwei (2008). Numerical simulation of ground-fissure cracking and extending in Xi an. Hydrogeology & Engineering Geolog. Vol. 2008(6), pp. 56-60.

Liu Nina; Men Yuming; Liu Yang (2009). Model Test of Soil and Metro-tunnels Interaction in Earthquake Activities. Journal of Earth Sciences and Environment Vol 31(3), pp. 295-298.



Dow

nloa

ded

from

asc

elib

rary

.org

by

GA

DJA

H M

AD

A U

NIV

ER

SIT

Y o

n 03

/04/

15. C

opyr

ight

ASC

E. F

or p

erso

nal u

se o

nly;

all

righ

ts r

eser

ved.

41102(375)9.pdf

Documents

soil interactions

xian china

soilstructure interactions

dynamic soil pressure

earthquake engineering

changan university

prototype xian metro

model box