13th World Conference on Earthquake Engineering

Vancouver, B.C., Canada

August 1-6, 2004
Paper No. 177

SIMULATION ANALYSES FOR 3-D RESPONSES OF PILE-SUPPORTED

STRUCTURE IN LIQUEFIABLE SAND SUBJECTED TO BLAST-
INDUCED GROUND MOTION

Kohji KOYAMADA 1, Katsuichiro HIJIKATA 2, Tomoaki ISHIDA 2, Hideo TANAKA 2,

Osamu KONTANI 1, Yuji MIYAMOTO 1 and Robert NIGBOR 3

SUMMARY

Simulation analyses were conducted for vibration tests on a pile-supported structure in a liquefiable sand
deposit at a large-scale mining site, in order to understand three dimensional (3-D) responses of soil-pile-
structure interaction systems under severe ground motions and to confirm the applicability of the response
analysis method to non-linear response including liquefaction. This paper describes the simulation
analysis results for the vibration tests. Firstly, site responses of the test pit were simulated using a 3-D
finite element model. Next, the soil-pile-structure responses were analyzed by a beam element model with
nonlinear springs taking into account the soil-pile interaction. Analysis results of the structure responses
and the pile stresses showed good agreement with the test results, and it was confirmed that this analysis
method was applicable for evaluating non-linear behaviors of soil-pile-structure systems during
liquefaction.

INTRODUCTION

Vibration tests of a pile-supported structure in a liquefiable sand deposit were conducted at a large-scale
mining site [1]. These tests yield the 3-D responses of the pile-supported structure under the test
conditions as well as those during an actual earthquake. It is well known that pile stresses, especially
axial forces, are greatly affected by 3-D dynamic behavior of the superstructure, and that they vary with
location in the pile arrangement [2]. It is very important to evaluate how the responses of the
superstructure affect the pile stresses and to incorporate those effects into seismic design of the pile
foundation. This paper shows simulation analyses using a beam-interaction spring model for the 3-D
dynamic behaviors of the soil-pile-structure system. The goal of this study is to achieve a better
understanding of dynamic nonlinear behaviors of soil-pile-structure interaction systems under severe

1
Kobori Research Complex Inc., Tokyo, Japan. Email: koyamada@kobori-takken.co.jp
2
Tokyo Electric Power Company, Yokohama, Japan. Email: hijikata.katsuichirou@tepco.co.jp
3
University of Southern California, Los Angels, USA. Email: nigbor@usc.edu
ground motions and to confirm the applicability of the nonlinear response analysis method to evaluate pile
stresses.

OUTLINES OF VIBRATION TESTS

Outlines of the vibration tests using ground motions induced by mining blasts are shown in Figure 1.
Vibration tests of pile-supported structures in a liquefiable sand deposit were conducted at Black Thunder
Mine of Arch Coal, Inc. Black Thunder Mine is one of the largest coalmines in North America and is
located in northeast Wyoming, USA. At the mine, there is an overburden (mudstone layers) over the coal
layers. The overburden is dislodged by large blasts called "Cast Blasts" and the rubble is removed by
huge earthmoving equipment. The ground motions caused by Cast Blasts were used for vibration tests
conducted in this research. Photo. 1 shows the situation of vibration tests.

Figure 2 shows the outline of the pile-supported structure. The test structure was supported on four piles
made of steel tube. The pile tips were closed by welding and embedded 70cm into the mudstone layer.
The top slab and the base mat were made of reinforced concrete and were connected by H-shaped steel
columns. The instrumentation is shown in Figure 3. Accelerations were measured of the structure and
one of the four piles. Accelerations in the sand deposit and in the free field were also measured in array

Test Structure
Blast Area
Test Pit

Mudstone
Explosive

Coal Layer

Figure 1 Vibration Test Method at Mining Site

Photo. 1 Situation of Vibration Tests

3.0m
Top Slab(Fc=30MPa) Triaxial Accelerometer Pore Water Pressure Meter
0.5m
Uniaxial Accelerometer Strain Gage
Ｈ-shaped Column
3m
2.0m H:299mm
B:306mm
t:11mm Blast Area
Base Mat (Fc=30MPa) 3m
0.5m
GL 0.0m

Steel Tube Pile 3m

φ=318.5mm
3.0m Test Pit t=9.5mm GL -3.0m

6m GL -10.0m

0.7m GL -20.0m
2.0m
Top Slab Base Mat GL -30.0m

Figure 2 Pile-Supported Structure Figure 3 Instrumentation

configurations. Axial strains of the pile were measured to evaluate the bending moments. Excess pore
water pressures were measured at four levels in the test pit to investigate liquefaction phenomena.
Vibration tests were conducted six times. In this paper, Test-3 was chosen for detailed investigations. In
Test-3, the distance was about 140m between the blast area and the test pit and the maximum acceleration
at the ground surface was 579 Gals in the EW direction [3].

ANALYSIS METHOD

Figure 4 shows the analysis model for 3-D response of soil-pile-structure system. The soil response
analysis is conducted by a 3D-FEM effective stress analysis method [3]. The analyses were performed by
a step-by-step integration method and employed a multiple shear mechanism model for the strain
dependency of soil stiffness and Iai-Towhata model for evaluating the generation of excess pore water
pressure [4]. Table 1 shows the soil constants. The shear wave velocity was measured by PS-Logging
and the density of the saturated sand was measured by a cone penetration test. Soil nonlinearity was taken
into account for all layers and Table 2 shows the nonlinear parameter for this simulation analysis. Figures
5 and 6 show the nonlinear properties and the liquefaction curve for the reclaimed sand, respectively.
These curves are based on laboratory tests.

The super-structure is idealized by a one-stick model and the pile foundations are idealized by a four-stick
model with lumped masses and beam elements. The lumped masses of the pile foundations are connected
to the free field soil through lateral and shear interaction springs. A nonlinear vertical spring related to the
stiffness of the supported layer is also incorporated at the pile tip, as shown in Figure 7. The initial values
of the lateral and shear interaction soil springs of the pile groups are obtained using Green’s functions by
ring loads in a layered stratum [5] and they are equalized to four pile foundations. The soil springs are
modified in accordance with the relative displacements between soils and pile foundations and with the
generation of excess pore water pressures.

ANALYSIS RESULTS

3-D Responses of Liquefied Sand Deposits

Figure 8 shows the calculated time histories of the ground surface accelerations and the pore water
pressure ratios. The amplitudes of the horizontal motions became smaller due to the generation of pore
water pressure at time 2.5 seconds. However, the amplitude of the vertical motion was still large after 2.5
seconds. The analysis results are in good agreement with the test results.

Figure 9 shows the relationship between the shear stress and the shear strain at GL-1.5m in the sand. The
left figure shows the results for the EW direction and the right one shows those for the NS direction. The
blue lines show the 3-D analysis results and red ones show the 1-D analysis results. The symbols show the
occurrence time of maximum shear strain in each direction. Both occurred at almost the same time. The
maximum shear stresses of the 1-D results are higher than those of the 3-D results. This shows that the 1-
D analysis under-estimated the soil nonlinearity due to liquefaction.

Figure 10 shows the acceleration response spectrum of the ground surface in the EW direction. The blue
line and the red line show the 3-D and 1-D analysis results respectively, and green line show the test
results. All spectra have a first peak at 0.6 seconds, and the 3-D results are good agreement with the test
result. Figure 11 shows the acceleration response spectrum of the ground surface in the UD direction. All
spectra have a first peak at 0.3 seconds, and both of the 3-D and 1-D analysis results are in good
agreement with the test result.
 Lateral
Soil Spring Top Slab
P
Pmax ●

Shear Modulus Ratio G/G 0

K y Base
Analysis

Damping Factor h(%)

H-D Model
● Model
Pmax= Pmax 0( σ '/ σ0')
1/2
K = K 0 (σ '/ σ 0')
Laboratory
σ' = σ 0'-u
Test
σ' : Effective Stress
σ 0 ' : Initial Effective Stress
u : Pore Water Pressure

Vertical
Soil Spring 杭
PV Shear Strain (%)
P V max Soil Response
Fig. 5. Nonlinear properties
KV of reclaimed sand
v Soil Response
H-D Model 3-Dir.
Input Motion

Fig. 4. Three-dimensional analysis model

for soil-pile-structure system
Cyclic Shear Stress Ratio
Analysis
Model

Table 1: Soil Properties for simulation analysis

Unit Laboratory
Thickness Vs Vp
Soil type Weight Test
(m) (m/s) (m/s)
(kN/m3)
Sand
3.0 18.9 80 1530
(Test Bed)
Cyclic Number
Clay 2.0 16.7 200 400
Fig. 6. Liquefaction curve
Mudstone 5.0 18.6 320 1240

200
Table 2: Nonlinear Parameters
Maximum Axial Force 137kN
Axial Force (kN)

150
Reclaimed Sand
Reference Strain : 0.034%
100 Test-1
Maximum Damping Factor : 28% Test-2
Test-3
Liquefaction Parameter Test-4
50 Test-5
W1=1.15, S1=0.005, P1=0.5, Test-6
Analysis Model
P2=1.12, C1=1.6, Initial Vertical Spring
Phase Transformation Angle=28deg. 0
0.0 0.5 1.0 1.5
Vertical Disp.(cm)
Clay and Mud Stone
Reference Strain : 0.17%
Fig. 7. Relationship between vertical
Maximum Damping Factor : 25% displacements and axial forces at pile head
 150. (cm/s2) Test Analysis

500

-150. Ground Surface Acc. (EW) Analysis Results

h=0.05
(cm/s2) 400 3-D
150.

Acc.Response(Gal)
1-D (EW)
300 Test Results

-150. Ground Surface Acc. (NS)

(cm/s2) 200
500.

100

-500. Ground Surface Acc. (UD)

2. 0
0.02 0.05 0.10 0.20 0.50 1.00 2.00

Period(sec)
P.W.P.R. (GL-0.6m)
-2. (sec.)
8.0 Fig. 10. Comparisons of acceleration
response spectrum of ground surface in EW
direction
Fig. 8. Comparisons of time-histories of ground
surface accelerations and excess pore water
pressure ratios at GL-0.6m for Test-3
2500
Analysis Results
h=0.05
2000 3-D
Acc.Response(Gal)

Maximum Shear Strain 1-D (EW)

3-D 1-D(EW)
1500 Test Results
3 3
Shear Stress(kN/m2)

2 2 1000
1 1

0 0 500

-1 -1
0
-2 -2 2.1 sec.
2.2 sec. (EW) (NS) 0.02 0.05 0.10 0.20 0.50 1.00 2.00
-3 -3
-1.0 -0.5 0.0 0.5 1.0 -1.0 -0.5 0.0 0.5 1.0 Period(sec)
Shear Strain(%) Shear Strain(%)
Fig. 11. Comparisons of acceleration response
Fig. 9. Relationship between shear stress and spectrum of ground surface in UD direction
shear strain at GL-1.5m in sand bed

Dynamic Responses for Test Structure

Figure 12 compares the calculated time-histories of acceleration for the test structure with the test results.
The horizontal motions for the top slab of the test structure have almost the same amplitudes in the EW
and NS directions, and are different from the records for ground surface shown in Figure 8. However, the
vertical motion for the base mat of the test structure is almost the same as that for the ground surface
shown in Figure 8. The analysis results are in good agreement with the test results not only in the
horizontal directions but also in the vertical direction.

Figure 13 shows the displacement orbit in the EW and NS directions for the top slab and the ground
surface. The horizontal motions of the ground surface had an almost circular orbit. On the other hand, the
top slab had an elliptical orbit and amplitudes for the EW direction became larger than those for the NS
direction due to the different vibration property of the test structure. The analysis results are in good
agreement with the test results, and it is confirmed that this analysis method is applicable to evaluate the
3-D responses of pile-supported structures in liquefied sand deposits.

Test Analysis Maximum Displacement

300. (cm/s2)
Test Analysis

EW-Dir. 3 3
-300.
2
NS 2
300. (cm/s2)
1 1

NS (cm)
NS (cm)
0 0

-300. NS-Dir. -1 -1
EW
-2 -2
500. (cm/s2) EW
NS
-3 -3
-3 -2 -1 0 1 2 3 -3 -2 -1 0 1 2 3
EW (cm) EW (cm)
UD-Dir. (Ground Surface)
-500. (Top Slab)
(sec.)
8.0

Fig. 13. Relationship between

Fig. 12. Comparisons of calculated time- displacements in EW and NS directions of
histories of accelerations for top slab top slab and ground surface

Bending Moments and Axial Forces for Pile Foundation

The distributions of maximum pile stresses, bending moments and axial forces, are shown in Figure 14.
Bending moments became larger at the pile head as well as at the interface between the reclaimed sand
and the supporting layer. The calculated maximum bending moments at pile heads are almost the same in

Pile-4 Pile-1
N 2.9sec Analysis Test
0.0 0.0 W E 30. (kN.m) Occurrence time
S of maximum value
-0.5 -0.5 Pile-3 Pile-2
3-D(EW)
-1.0 -1.0 -30. Bending Moment (EW)

-1.5
3-D(NS) -1.5 (kN.m)
30.
GL(m)

3-D
-2.0 Test(EW) -2.0 (Pile-1,2,3,4)
-2.5 -2.5 -30. Bending Moment (NS)
1-D(EW) 2.8sec
(kN)
2.0sec
-3.0 -3.0 100.

-3.5 -3.5

-4.0 -4.0 -10 0. Axial Force

0 10 20 30 40 0 100 200 (sec.)
(kN*m) (kN) 8.0
Bending Moment Axial Force

Fig. 14. Comparisons of the maximum bending Fig. 15. Comparisons of calculated time-
moments and the maximum axial forces histories of bending moments and axial force
by the location in pile arrangement at pile head of Pile-1
(Axial force : compression(+), tensile(-))
the EW and NS directions, since the maximum acceleration of the superstructure were almost the same in
both directions, as shown in Figure 12. The calculated maximum axial forces in the four piles are almost
the same and about 90kN. The 1-D analysis result became smaller than the 3-D analysis results.

The time histories of the pile stresses at pile heads are shown in Figure 15. The analysis results are in
good agreement with the test results, which indicates that this analysis method is applicable to evaluate
pile stresses during liquefaction. The maximum bending moments occurred at 2.9 seconds in the EW
direction and at 2.0 seconds in the NS direction. These times correspond closely with the superstructure
responses, as shown in Figure 12. The time history of axial force at the pile head is similar with that of
the bending moment in the EW direction, and it is different with that of the superstructure response in the
UD direction shown in Figure 12.

Figure 16 shows the relationship between the bending moments and axial forces at the pile head for the
Pile-1 and Pile-4. At the occurrence time of bending moments in EW direction, the time of 2.9 seconds,
compression force (74kN) arose in Pile-1, and tensile force (37kN) was caused in Pile-4. The tilts of the
loop became inversely in Pile-1 and Pile-4, it means that axial forces mainly caused by rotational
displacement of super-structure and the effects by the vertical motion of the super-structure was small. An
area of the loop by 1-D analyses became smaller than that of the loop by 3-D analyses. It suggests that a
great care is needed in applying the 1-D analyses to evaluate pile stresses on seismic design.

Maximum Bending Moment in EW-Dir.

3-D 1-D(EW)
150 150
100 74kN 100
Axial Force(kN)

50 50
0 0
-50 -50
37kN
-100 -100
(Pile-1) (Pile-4)
-150 -150
-50 0 50 -50 0 50
Bending Moment(kN*m) Bending Moment(kN*m)
Fig. 16. Relationship between bending moments and
axial forces at pile head of Pile-1 and Pile-4
(Axial force : compression(+), tensile(-))

CONCLUSIONS

(1) 3-D responses of a pile-supported structure in liquefied sand deposits have been obtained from
vibration tests using ground motions induced by large-scale blasts. Simulation analysis results were in
good agreement with the test results for the responses of the superstructure and pile stresses due to
liquefaction.
(2) The maximum 1-D shear stresses and acceleration response spectrum in the horizontal direction were
higher than the 3-D results. This shows that the 1-D analysis under-estimated the soil nonlinearity due
to liquefaction.
(3) The horizontal motions for the top slab of the test structure were very different from those at the
ground surface due to liquefaction. However, the vertical motion for the base of the test structure was
almost the same as that for ground surface regardless of the liquefaction
(4) The pile bending moments became larger at the pile head and at the interface between the reclaimed
sand and the supported layer. The bending moments at the pile head were greatly affected by the 3-D
responses of the superstructure and the maximum bending moments occurred at almost the same time
as the maximum acceleration of the superstructure.
(5) The pile axial forces varied greatly with the location in the pile arrangement, because they were caused
by the rotational displacement of the superstructure, and because the effects of the vertical motion of
the structure were small. For the bending moments and axial forces at the pile head, the area of the
loop by 1-D analyses became smaller than that by 3-D analyses. This suggests that a great care needs
to be taken in applying the 1-D analysis to evaluate pile stresses for seismic design.

ACKNOWLEDGMENTS

We would like to express our deep appreciation to Dr. Aoyama, Professor Emeritus of the University of
Tokyo, for his guidance throughout this experimental research. We would also like to thank the
management and staff of Arch Coal's Black Thunder Mine for their assistance throughout the project.
Special thanks go to Paul Lang, Steve Beil and Gordon Shinkle. They were always very generous and
provided us with support in many ways. Without their help, this work would not have been possible.

REFERENCES

1. Hijikata, K., Ishida, T., Tanaka, H., Koyamada, K., Miyamoto, Y., Kontani, O., and Nigbor, R., "
Experimental Study on Soil-Pile-Structure Interaction in Liquefiable Sand Subjected to Blast-
Induced Ground Motion", Proceeding of the 13th World Conference on Earthquake Engineering
(13WCEE), Vancouver, B.C., Canada, August 1-6, 2004, Paper No. 190
2. Saito, H., Tanaka, H., Ishida, T., Koyamada, K., Kontani, O. and Miyamoto, Y., “Vibration Test of
Pile-Supported Structure in Liquefiable Sand Using Large-Scale Blast, -Simulation analyses for
vibration tests of soil-pile-structure-”, Journal of Structural and Construction. Engineering,
Architectural Institute of Japan, No. 557, 2002, pp.85-92 (in Japanese)
3. Kontani, O., Tanaka, H., Ishida, T., Miyamoto, Y. and Koyamada, K., "Simulation Analyses of
Seismic Tests of a Pile-Supported Structure in Liquefiable Sand Using Large-Scale Blast
Excitation," Proceeding of the 16th International Conference on Structural Mechanics in Reactor
Technology (SMiRT-16), 2001, Washington, Paper No.1492
4. Iai, S., Matsunaga, Y. and Kameoka, T., “Strain Space Plasticity Model for Cyclic Mobility”, Soils
and Foundation, Vol.32, No.2, pp.1-15, 1992.6
5. Kausel E. and Peek R., “Dynamic Loads in the Interior of a Layered Stratum- An Explicit Solution”,
Bulletin of the Seismological Society of America, Vol. 72, 1982, pp. 1459-1481