Shaking Table Testing of Free Field Response in ...

Shaking Table Testing of Free Field Response in Layered Granular Deposits L. Dihoru, C. A. Taylor, S. Bhattacharya & D. Muir Wood University of Bristol, United Kingdom

A. Simonelli & F. Moccia University of Sannio, Italy

G. Mylonakis University of Patras, Greece

ABSTRACT: While dynamic tests on homogeneous soil deposits have often been reported, experimental data on layered deposits is limited. This paper investigates the seismic free field response of two stratified deposits with different stiffness ratios between the top and the bottom layers. The deposits were contained in a laminar shear box and tested dynamically on a shaking table. Dynamic measurements of shearing stress and strain provided an insight into how the strain magnitude and the earthquake frequency content affected the hysteretic response of the test media. A comparison is made between the measured dynamic stiffness and the values given by previous empirical strain curves for sands. The performance of the enclosing container is also discussed.   Keywords: free field response, shear box, seismic testing

1. INTRODUCTION A synergetic relationship between earthquake engineering and soil dynamics research has developed, with soil geometric and stiffness characteristics becoming important parameters in the seismic design of structures (Gazetas 1987, Novak 1987, Badoni and Makris 1996, Tokimatsu et al 1996, Mylonakis et al 1997, Hadjian 2002, Bhattacharya 2007).The shearing behaviour of soils, in particular the shear modulus and the damping ratio were found to be the properties that affect most the dynamics of soilstructure interaction at small, medium and large strains. The measurement of these properties has been the subject of several laboratory and field techniques devised for specific problems and strain magnitudes (Richart et al 1970, Seed and Idriss 1970, Hardin and Drnevich 1972, Vucetic and Dobry 1991, Ishibashi and Zhang 1993, Darendelli 2001, Muir Wood 2007). Within the investigations up to now, the use of shaking table testing has often been exploited (Ohtsuki 1992, Meymand 1998, Prasad et al 2004, Tokimatsu and Susuki 2005, Pitilakis et al 2008). A large number of shaking table studies employed flexible container boxes designed to replicate the free field response of a soil stratum overlying a rigid base. Their role is to shear the soil in idealized manner via vertically propagating shear waves produced by the shaking table. There are certain potential advantages that a large scale shear device may have over other more conventional small, table-top devices. Firstly, a larger volume of soil can be tested, therefore the results may be more representative of the prototype field conditions. Secondly, the boundary effects are smaller and the volume of soil situated in the central part of the container is believed to reproduce reasonably well the free field dynamics. Thirdly, in principle, a shear stack can be tuned to operate over wide strain ranges with granular materials of different stiffness value (Takahashi et al 2001). This, in particular, makes the shear stack useful for studying the large-strain dynamic moduli under seismic excitation. Several laminar shear box designs have been reported for both uniaxial and biaxial loading (Dar 1993, Fishman et al 1995, Gibson 1997, Turan et al 2009). This paper investigates the dynamic response of three granular deposits in a uniaxial shear stack assembled on the shaking table at Bristol University. While there is a large body of literature on tests on homogeneous soils, experimental data on stratified deposits is scarce. This experimental programme focused on one witness homogeneous deposit and two layered deposits exhibiting a

certain stiffness ratio between the top and the bottom layers. Modal tests were carried out to explore the decoupled fundamental period for the individual layers. Real seismic records of different frequency scaling were employed in order to explore the sensitivity of dynamic response to the frequency content of the input for both the granular deposits and the enclosing container. Aspects such as container-deposit coupling and deposit mode shapes are reported.

2. EXPERIMENTAL PROGRAMME 2.1 Materials and experimental set-up for dynamic testing Three granular materials with different particle properties were employed in the study: Leighton Buzzard (LB) sand BS 881-131, Fraction B, Leighton-Buzzard sand BS 881-131, Fraction E and rubber granules type Charles Lawrence CT0515B. One monolayer configuration (E) and two layered configurations (BEE and ER) were built by pluviation. For the layered configurations, the median line of the enclosing container (Z=0.4 m) marked the interface between the two layers of contrasting stiffness. The layout and the stiffness ratio between the top and the bottom layers for the selected configurations are shown in Fig. 1. E

BEE

Rubber CT0515

H/2=0.4 m

LB-Fraction E + LB-Fraction B

LB-Fraction E

H/2=0.4 m

Gbottom / Gtop= 1.75

Gbottom / Gtop= 80

LB-Fraction E H=0.8 m

ER

LB-Fraction E

Figure 1 Deposit configurations employed in testing

A mixture of LB-Fraction B and LB-Fraction E particles was employed to model the stiff bottom layer in the BEE configuration. The material properties, the pluviation parameters and the experimentally obtained bulk densities are presented in Table 1. Table 1. Material properties

1 irregular rough 1.15 0.49

Sand LB Fraction E 0.142 angular smooth 2.65 0.31

Sand Mixture LB Fraction B and Fraction E (85:15) 0.82 angular smooth 2.65 0.31

15 / 1000

40 / 1000

15 / 1000

479

1400

1942

Material Properties

Rubber Particles CT0515

Mean Diameter, D50 (mm) Particle Morphology Particle Texture Specific Gravity (g / cm3) Poisson’s Ratio Pluviation Parameters: Nozzle Diameter (mm) / Height of Fall (mm) Pluviated Density (kg / m3)

The test facility employed in this study was a large-scale laminar box (‘shear stack’) of length 1.2 m, width 0.5 m and height 0.8 m (Fig. 2, a). The box consists of eight aluminium alloy rings designed to move freely in one direction (Y), while their X and Z vibrations are restricted by a steel rigid frame and a system of bearings. The stack end walls and base have a rough surface to allow complimentary shear stresses to develop. Its side walls are lubricated to allow uniform plane strain conditions in the deposit when a structural model violating this condition is not installed. Previous investigations of the shear stack have shown that its fundamental frequency when empty is abt. 6 Hz (Dar 1993). The shear stack was installed on the platform of a six degrees of freedom shaking table. The shaking table at Bristol University has an optimal operating range of 0-50 Hz. Horizontal accelerations of up to 3.7 g (with no payload) and up to 1.6 g (with 12 tonnes payload) can be achieved. The shear stack was

installed with its long side (the shearing direction) on the Y axis of the shaking table. Four free-field accelerometers (a2-5) were embedded in sand at similar coordinates in the horizontal plane (x, y) = (0.270, 0.455) m and at vertical intervals of ∆z = 0.2 m one from another. The shear stack response was measured by four accelerometers (a6-9) attached to the outside of the box rings (Fig. 2, b). Free surface

a5 z = 0.4

z = 0.6

a6 z = 0.2 a4 a7 a3

z = 0.8

z = 0.3 z = 0.5 z = 0.7

a8

a2

a9

Interface

a1 Shaking table

(a)

z = 0.1

Y

(b)

Figure 2 The shear stack (a: general photo, b: instrumentation layout, ai = accelerometers)

2.2 Exploratory modal testing The small shear strain stiffness of the deposits was measured in pulse tests and white noise tests. Pulse testing employed 10 Hz, half-sine pulses that were generated by the shaking table in the horizontal direction. The second modal testing technique employed a random noise signal of 1-100 Hz bandwidth and RMS=50 mV generated by the shaking table in the Y direction. Frequency response functions (FRFs) were calculated by normalising the output of the receiving accelerometer by the shaking table input in the frequency domain. The peaks of the FRF functions indicate frequencies of interest for both the granular deposits and the enclosing container. The average shear wave velocity vs obtained in modal tests and the computed small strain stiffness values (G0) are shown in Table 2. The values reported for sands represent average results obtained in pulse and white noise tests in preliminary monolayer deposits. As expected, G0 increases with packing density from rubber CT0515 to the BEE sand mixture. More details on the exploratory tests can be found in Dihoru et al 2009. Table 2 Average results from modal testing M odal Te st R es ul ts

S hea r W a ve V e loci ty, v s (m / s) M ax. S he ar S ti ffnes s , G 0 (M P a) F unda m ent al P eriod , T (s ) F unda m ent al F reque ncy, H z

R ubb er P art icl es C T 0515 14 0.1

S a nd LB F ra cti on E 76 8

Sa nd M ix ture LB F ra ct ion B a nd F rac tion E (80: 20) 85 14

0.22 1 4.5 3

0.043 23.41

0.03 8 26.5 6

2.3 Seismic tests Real acceleration records of three Italian earthquakes (Friuli (1976), Irpinia (1980) and Norcia (1998)) were employed in testing. The Fourier spectra of the chosen seismic records (Tolmezzo-A270, SturnoA000, and Norcia-R090) are shown in Fig. 3. The selected inputs have different energy distribution patterns and different Fourier amplitudes. Before being applied in the experiments, the signals were amplified in the time domain in order to reach a common peak ground acceleration value of 0.3 g. The inputs were also frequency scaled at scaling factors of 2, 5 and 12 in order to shift the band of maximum seismic energy to higher frequency ranges. For example, the tests with an input scaling factor of 12 involved baseline frequencies in the 24-48 Hz range. In this manner, the influence of the frequency / energy content of the earthquake on the free field response could be investigated.

Sturno A-000

0.15 0.1 0.05

Norcia R-NCB-000

0.25

Fourier Amplitude

Fourier Amplitude

Fourier Amplitude

Tolmezzo A-270 0.2

0.2 0.15 0.1 0.05

0.25 0.2 0.15 0.1 0.05

0

0 0

5

10 15 Frequency [Hz]

0 0

20

5

10 15 Frequency [Hz]

20

0

5

10 15 Frequency [Hz]

20

Figure 3 Selected seismic inputs (Fourier spectra)

3. RESULTS AND DISCUSSION 3.1 Seismic free field response The shear stack inertial response under seismic excitation is shown in Fig. 4, where the stack dynamic displacement profiles are plotted for both positive and negative shaking table displacements. In order to investigate in more detail the degree of coupling between the shearing granular deposits and the container, frequency response functions (FRFs) were computed between the accelerometer signals and the seismic input at different locations inside and outside the stack. Fig. 5 shows the FRFs inside and outside the shear stack for the ER deposit for Sturno-A000 input at a frequency scaling factor SF=2. The top half of the stack (Ring 1 and Ring 3) displays a softer response than the bottom half (Ring 5 and Ring 6) due to the very high stiffness contrast between the inside granular layers (Gbottom/ Gtop=80). The top half rubber layer becomes decoupled from the shear stack, as it can be observed in Fig. 5, b. Stack Profile, Sturno-A000, SF=2 0.8

0.8

t=9.37 s 0.7

0.7

t=9.35 s t=9.30 s

0.6

t=9.7s t=9.75s

0.6 Height (m)

t=9.25 s

0.5

t=9.20 s 0.4 t=9.15 s 0.3

Shaking Table Direction

0.2

t=9.80s

0.5 Shaking Table 0.4 Direction

t=9.85s t=9.90s t=9.95s

0.3 0.2

0.1

0.1

0 -15

-10

-5 Displacement (m)

0

0 -0.02

5 x 10

-3

-0.01

0 0.01 Displacement (m)

(a)

0.02

0.03

(b)

Figure 4 Shear stack displacement profile for Sturno-A000 input, input scale factor SF=2 (a: positive shaking table displacements, b: negative shaking table displacements)

Ring1 Ring3 Ring5 Ring7

1.12 1.1 1.08 1.06 1.04

FRF Linear Magnitude

ER Configuration, Sturno A000, SF=2 1.14 FRF Linear Magnitude

Height (m)

Stack Profile, Sturno-A000, SF=2 t=9.65s

ER Configuration, Inside Stack, Sturno A000, SF=2 1.15 Z=0 m, Free Surface Z=0.2 m Z=0.4 m Z=0.6 m 1.1

1.05

1.02 1 0

20

40 60 Frequency(Hz)

(a)

80

100

1 0

20

40 60 Frequency (Hz)

80

100

(b)

Figure 5 FRFs outside the shear stack and inside the ER deposit for Sturno-A000 input, input scale factor SF=2 (a: outside shear stack, b: inside deposit)

The rubber layer’s low stiffness is insufficient for driving the stack. In this latter case, the response is driven by the container and not by the granular deposit. It is thus confirmed that the relative stiffness between stack and the granular deposit has a paramount influence on the dynamic response and that an accurate investigation of the free field response requires a soft container and a sufficiently stiff deposit. The normalized peaks of the FRFs at various ordinates inside and outside the shear stack were employed in representing the mode shapes of the granular deposit and of the enclosing container, respectively. A comparison between the mode shapes for the monolayer E configuration and the doubly layered BEE, and between the monolayer E and the doubly layered ER is shown in Fig.6,a and Fig.6,b, respectively. The high degree of coupling between deposit and stack is evident for the E and the BEE configurations, while the ER deposit displays a very different shape from the stack. 0

Free Surface

0

Free Surface ER - Inside

0.2

BEE - Outside

E - Outside

E - Inside

0.4 BEE - Inside

0.85 0.9 0.95 Normalized Mode Shape

0.4

E - Inside

0.6

0.6 0.8 0.8

E - Outside Z (m)

Z (m)

0.2

1

0.8 0.85

ER - Outside 0.9 0.95 Normalized Mode Shape

(a)

1

(b)

Figure 6 Normalized mode shapes for E and BEE deposits (a) and E and ER deposits (b), respectively, inside and outside the shear stack (Sturno-A000 input, input scale factor

SF=2). 3.2 Dynamic strain and hysteretic response for layered deposits In general, the shear loading of granular deposits during an earthquake involves stress reversal, with varying amplitude and frequency. A classic investigation of shear modulus and damping in soils (Hardin and Drnevich 1972) presented the complex set of factors influencing these two dynamic properties. For non-cohesive soils (i.e. ’clean sands’), the relative influence of strain amplitude, effective mean principal stress and void ratio on dynamic stiffness is very high. One of the objectives of the present study was to determine the influences of dynamic strain magnitude, initial stiffness (void ratio) and input frequency content on dynamic stiffness. In this study, the displacement, the shear strain and the shear stress time histories were directly evaluated by integrating the measured acceleration time histories. Cumulative trapezoidal integration was combined with filtering for elimination of baseline drifts. A high-pass Butterworth filter of 5th order and a cut-off frequency of 0.5 Hz was employed to eliminate the low-frequency signal components. The evaluation of strain at the interface between the sandwich layers (Z=0.4 m) for the BEE and the ER deposits, calculated for two different reference ordinates (Z=0 m and Z=0.2 m, respectively) is shown in Fig. 7. It was interesting to see how the width of the shearing layer influences the value of the strain magnitude at the interface. It was found that the interface strains calculated against the two selected reference ordinates were similar for the BEE configuration, but very different for the ER deposit. The ER interface strains calculated against the free surface reference level were unrealistically large, due to the big displacements of the free rubber surface. It was concluded that in the ER case, the free surface could not be taken as a reliable reference for computing strain. The evaluation of dynamic stiffness for the three deposit configurations was carried out by representing the stress-strain loops for certain cycles extracted from the time histories of the seismic inputs. The dynamic strain (γ) measurements in the seismic tests ranged up to a maximum value of 6x10-3. Fig. 8 shows the hysteretic loop for two different input cycles of the time history. The observed hysteretic response for the E monolayer configuration shows a small difference in the shear moduli calculated below and above the interface level (Z=0.4 m). The link between the frequency content of the input and the measured dynamic stiffness moduli was found to be significant. Higher input scaling factors (e.g. SF=12) shifted the earthquakes energy to the higher frequency range (24-48 Hz). Higher

frequency cycling led to higher dynamic stiffness values. An average stiffness modulus measured for Sturno-A000, SF=2 was abt. 2.6 MPa, while the modulus measured at SF=12 was in the 4-6 MPa range. Both tests employed the same seismic input (the same number of loading cycles N), but different frequency scaling factors (SF). 1

x 10

BEE, Sturno-A000, SF=2

-3

x 10

layer: 0.2m-0.4m layer: 0m-0.4m

0.5

1 Strain

Strain

BEE ,Tolmezzo-A270, SF= 2

-3

2

layer: 0.2m - 0.4m layer: 0m -0.4m

0

-0.5

0

-1

-1

8

10

12 Time (s)

14

-2 8

16

9

10

11 Time (s)

(a)

12

13

14

(b)

ER, Sturno-A000, SF=2

ER, Tolmezzo-A270, SF=2

0.1

0.03

layer:0.2m-0.4m layer:0m-0.4m

0.02

0.05 Strain

Strain

0.01

0

0 -0.01

-0.05 -0.02

layer: 0.2m-0.4m layer: 0m -0.4 m -0.1 8

10

12 14 Time (s)

16

18

-0.03 8

10

12

(c)

14 Time (s)

16

18

20

(d)

Figure 7 Strain in soil at interface level (Z=0.4 m) for two deposit configurations (BEE (a,b), ER (c,d) and two seismic inputs (Sturno-A000 (a,c), Tolmezzo-A270 (b,d), SF=2) Monolayer E Configuration, Tolmezzo-A270, SF=2

1200

1000

Gsec-bottom=1.3 MPa 1000 t1=11.00 s t2=11.32 s 800 Shear Stress (Pa)

Shear Stress (Pa)

Gsec-top=0.89 MPa t1=10.11 s 500 t2=10.5 s 0

-500 -1000 -1500 -3

600 400 200 0 -200

-2

below median line (Z=0.4-0.6 m) above median line (Z= 0.2-0.4 m) -1 0 1 -3 Shear Strain x 10

-400 -600 -1

-0.5

0

0.5 Shear Strain

1

Figure 8 Observed hysteretic response of soil in the shear stack for monolayer E configuration, input: Tolmezzo-A270, input scaling factor SF=2

1.5

2 x 10

-3

The dynamic hysteretic response retains some memory of the initial stiffness of the deposit as it becomes evident for the two layer BEE configuration. Fig. 9 shows lower dynamic stiffness for the lower density layer above the median line (Z=0.4 m). The measured dynamic stiffness values from the seismic tests were compared to a standard degradation curve for sands (Seed and Idriss 1970). Fig. 10 shows that the RELUIS experimental values are in reasonable agreement with the classic degradation curve for the monolayer E configuration. The dynamic moduli measured for the bottom layer in the BEE double layer configuration lie slightly lower under the Seed & Idriss 1970 curve.

1500

Bilayer BE-E Configuration, Tolmezzo-A270,SF=2 BE- below median line (Z=0.4-0.6 m) E - above the median line (Z=0.2-0.4 m)

Bilayer BE-E Configuration, Tolmezzo-A270, SF=2 1500

1000

Shear Stress (Pa)

Shear Stress (Pa)

1000 500

Gsec 1

0 -500 Gsec-top= 0. 57 MPa Gsec-bottom=2.02 MPa t1=9.75 s t2=10.25 s

-1000 -1500 -1.5

-1

-0.5

0 0.5 Shear Strain

1

1.5

2 x 10

Gsec-top=0.89 MPa Gsec-botom=2.29 MPa t1=10.83 s t2=11.05 s

500 0 -500 -1000 -1

-0.5

0 Shear Strain

-3

0.5

1 x 10

-3

Figure 9 Observed hysteretic response of soil in the shear stack for BEE configuration, input: Sturno-A000, input scaling factor SF= 2. Bilayer: E - top, BE - bottom BE Shear Stiffness Go-bottom=14 MPa

Monolayer E Go=9 MPa

v

v

S12

N2

N2 S12

N2

T12 T2

S,T,N – RELUIS(2009) Seed & Idriss (1970)

(a)

S12

T2 T2

S5 S2 S,T,N – RELUIS(2009) Seed & Idriss (1970)

T2

T2

(b)

Figure 10 Comparison between test results and a standard stiffness degradation curve for = monolayer E (a) and bilayer BEE (b) deposits. Notation: S (Sturno), T (Tolmezzo), N (Norcia). Ex: S12 = Test Sturno-A000, input scaling factor SF=12.

4. CONCLUDING REMARKS Stiffness measurements at small strains via modal testing and dynamic stiffness measurements at large dynamic strains (γ max=6x10-3) were reported for one monolayer and two doubly layered granular deposits. The initial stiffness of the test deposits was mainly influenced by the packing density for sands (E and BEE configurations) and by the packing density and particle intrinsic stiffness for the rubber CT0515 layer. The low stiffness rubber CT0515 layer became decoupled from the input motion and displayed large lateral displacements and strain magnitudes in all seismic tests. For this particular layer, the free surface measurements were unreliable in strain calculations. The sand deposits (monolayer E and the two layered BEE) drove the shearing response of the shear stack and exhibited motion coupling inside and outside the shear stack across the vertical direction. The shear stress-shear strain results for the sand deposits compare well with previous measurements made for Leighton Buzzard sands (Seed and Idriss 1970). The dynamic hysteretic loops obtained in the experiments show some memory of the initial stiffness of the deposits, as it became evident for the two layered BEE configuration (higher secant stiffness for the dense/stiff bottom layer).The link between the hysteretic response of sands and the frequency content of the seismic input was investigated. Higher frequency (24-48 Hz) loading cycles resulted in higher values of measured dynamic stiffness in both the monolayer and the stratified deposits. ACKNOWLEDGEMENT The financial support from RELUIS (Rete di Laboratori Universitari Ingegneria Seismica) for conducting this research is gratefully acknowledged.

REFERENCES Badoni D., Makris N. (1996). Nonlinear response of single piles under lateral inertial and seismic loads. Soil Dyn Earthquake Eng. 15, 29-43. Bhattacharya S.A (2007). Review of methods for pile design in seismically liquefiable soils. In: Bhattacharya S., editor. Design of foundations in seismic areas: principles and applications. National Information Centre of Earthquake Engineering. Indian Institute of Technology, Kanpur. Dar A.R. (1993). Development of a flexible shear stack for shaking table testing of geotechnical problems. PhD Thesis, University of Bristol. Darendelli, M.B. (2001). Development of a new family of normalized modulus reduction and material damping curves. PhD Thesis. Univ. of Texas at Austin, Austin, TX. Dihoru, L., Bhattacharya, S., Taylor, C.A., Muir Wood, D., Moccia, F., Simonelli, A.L. & Mylonakis, G. (2009) Experimental modelling of kinematic bending moments of piles in layered soils. In (Eds) Kokusho, Tsukamoto & Yoshimine, Proc. Performance-based design in earthquake geotechnical engineering, Tokio, Japan. Fishman K.L., Mander J.B. & Richards R. (1995). Laboratory study of seismic free-field response of sand. Soil Dyn Earthquake Eng. 14, 33-43. Gazetas G. (1987). Soil dynamics: an overview. In: Banerjee P.K. & Butterfield R., editors. Dynamic Behaviour of Foundations and Buried Structures. Elsevier Science Publishing Co. Gibson A.D. Physical scale modelling of geotechnical structures at one-g. (1997). Ph.D. Thesis, California Institute of Technology, Pasadena, CA Hadjian, A.H. (2002).Fundamental period and mode shape of layered soil profiles. Soil Dyn Earthquake Eng. 22, 885-891. Hardin, B.O., Drnevich V.P. (1972). Shear modulus and damping in soils. J. Soils Mech. Found. Div.a. 98(7). 667-692. Ishibashi, I., Zhang, X.J. (1993) Unified dynamic shear moduli and damping ratios of sand and clay. Soils Found. 33(1), 182-91. Meymand P.J. (1998). Shaking table scale model tests of nonlinear soil-pile-superstructure interaction in soft clay. Ph.D.Thesis. Berkeley, CA: The University of California, Berkeley. Mylonakis G., Nikolaou A., Gazetas G. (1997). Soil-pile-bridge seismic interaction: kinematic and inertial effects. Earthquake Eng Struct Dyn. 26, 337-59. Muir Wood, D. (2007). Modelling of dynamic soil problems. Keynote lecture: 4th International Conference on Earthquake geotechnical engineering, Thessaloniki. Novak M. (1987). Experimental studies of the dynamic behaviour of piles and pile groups. In: Banerjee P.K. & Butterfield R., editors. Dynamic Behaviour of Foundations and Buried Structures. Elsevier Science Publishing Co. Ohtsuki A., Hirota M., Ishimura K., Yokoyama K., Fukutake K. (1992). Verification of two-dimensional nonlinear analysis of sand-structure system by examining the results of the shaking-table test. Earthquake Eng & Structural Dynamics. 21(7), 591-607. Pitilakis D., Dietz M, Muir Wood D., Clouteau D., Modaressi A. (2008). Numerical simulation of dynamic soilstructure interaction in shaking table testing. Soil Dyn Earthquake Eng 28, 453-67. Prasad S.K., Towhata I., Chandradhara G.P., Nanjunaswamy P. (2004). Shaking table tests in earthquake geotechnical engineering. Curr Sci.87(10), 1398-404. Richart, F.E, Jr., Hall, J.R., Jr., Woods, R.D. (1970). Vibrations of soils and foundations. Prentice-Hall, Englewood Cliffs, N.J. Seed, H.B., Idriss, I.M. (1970). Soil moduli and damping factors for dynamic response analysis. Rep. No. EERC 70-10. Earthquake Engineering Research Centre, Berkeley, CA. Takahashi A., Takemura J., Susuki A., Kusakabe O. (2001). Development and performance of an active type shear box in a centrifuge. Int. J. Phys. Model Geotec. 1(2), 1-18. Tokimatsu K., Mizuno H., Kakurai M. (1996). Building damage associated with geotechnical problems. Special issue Soils and Foundations, Japanese Geotechnical Society, 219-34. Tokimatsu K., Susuki, H. (2005) Effect of inertial and kinematic interactions on seismic behaviour of pile foundations based on large shaking table tests. Proc. 2nd CUEE Conf Urban Earthquake Eng, Tokio Institute of Technology. Turan A., Hinchberger S., Naggar H.E. (2009). Design and commissioning of a laminar soil container for use on small shaking tables. Soil Dyn Earthquake Eng. 29, 404-414. Vucetic, M., Dobry, R. (1991). Effect of soil plasticity on cyclic response. J. Geotech. Eng. 117(1), 89-107.