5 th INTERNATIONAL CONFERENCE ON r

EARTHQUAKE GEOTECHNICAL ENGINEERING

January 10-13, 2011
Santiago,· Chile

51CEGE
3
 GEOTECHNICAL ANO GEOPHYSICAL SITE CHARACTERIZATION
2.21 ORIENTEO TO SEISMIC ANALYSIS

An-Bin HUANG 1, Wen-Jong CHANG 2

ABSTRACT

A sound site characterization program for the purpose of geotechnical seismic analysis should provide
adequate information regarding soil stratigraphy; groundwater conditions, geostatic stresses and stress
bistory, characteristics ·of hydraulic conductivities, undrained monotonic and cyc1ic strength as well as
deformation and damping characteristics assessed in the strain range of interest. The purpose of the paper is
to describe sorne ofthe recent developments in geotechnical and geophysical site charactlrization methods
as they relate to seismic response and liquefaction potential analyses. The currently available semi-empirical
field-based as well as the critical state based methods for the evaluation ofundrained cyc1ic strength using
the site characterization test results are reviewed. Because the interpretation of in situ index tests for soil
liquefaction potential analysis remains empirical, it is imperative to calibrate these empirical methods for
local soils especially for the case of intermediate soils such as sands with fines. The critical state approach
appears promising as Ít has a sound theoretical basis and circumvents many ofthe problems associated with
empirical interpretation of in situ tests; however, it needs refinement especially in the context of inferring
soil cyc1ic strength from CPT.

Keywords: in situ test, geophysical exploration, sampling, liquefaction, soil deformability, state parameter

INTRODUCTION

Seismic analysis in Geotechnical Engmeering can cover a wide range of activities. In order to analyze the
response of a soil deposit to a seismic event in a rational manner, the required information should inc1ude
(Jamiolkowski et al., 1995):

Soil stratigraphy

Groundwater conditions

Geostatic stresses and stress history

Characteristics ofhydraulic conductivities

Undrained monotonic and cyc1ic strength

Deformation and damping characteristics assessed in the strain range of interest

Researchers have generally agreed that geotechnical and geophysical site characterization methods
are complementary to each other. The site characterization can include in situ (mechanical and
geophysical) tests and laboratory tests on natural soil samples. Near-surface geophysical tests
based on sound theoretical basis offer geotechnical engineers the opportunities to characterize the

Professor, Department of Civil Engineering, National Chiao Tung University,

e-mail: abhuang@mail.nctu.edu.tw

2
Assistant Professor, Department of Civil Engineering, National Cheng Kung University,
e-mail: wjchang@mail.ncku.edu.tw
"
63
t'

rj¡;.;:
 ";;j:;;¡¿­

sites, materials, and processes with high efficiency due to the nonintrusive nature and ability to
perform the same measurement in the laboratory and field (Stokoe and Santamarina, 2000). Soil
stratigraphy, stress states, stress history, as well as deformation and damping characteristics may be
revealed by these geophysical test methods.

Field geophysical exploration encompasses gravity, magnetic, seismic, reslstlvlty, and

electromagnetic measurements. The applicability of selective noninvasive techniques to site
characterizations are summarized in NRC (2000) and the engineering parameters inferred from
various geophysical methods are summarized in Stokoe and Santamarina (2000). In geotechnical
earthquake engineering, seismic methods are dominant because they directly measure the mechanical
properties, including material stiffness and damping in elastic or small strain level, which are required
in designs and ana1yses of geotechnical systems subjected to both monotonic and seismic loadings.
Many of the seismic-based geophysica1 exploration methods are nonintrusive cfnd can cover a large
area, which makes them cost effective in profiling the spatia1 distribution of engineering parameters
and applicable in hard-to-sample soi1s. Furthermore, seismic methods can be performed both in the
field and laboratory samp1es, which provides quantitative assessments of sample disturbance (Landon
et al., 2007), degree of saturation (Allen et al., 1980), and microstructure of soils (e.g., cementation and
anisotropy) (Belloti et al., 1996).

Taking good quality samples for cohesive soils is significantly easier than for cohesionless or granular soi1s.
Traditionally, undisturbed granular soil samples have been taken by ground freezing and dry coring. This
method can be prohibitively expensive and usually is used for critical or research projects. Successes have
been reported by Huang and Huang (2007) and Huang (2009) in the use of Laval and gel-push sampler
to take samples in granular soils with reasonable quality. By avoiding ground freezing, these methods
are significant1y more practica!. These new sampling technique developments make laboratory tests on
undisturbed soil samp1es much more feasible. With the advent ofloeal strain and bender element measurement
capabilities, valuable information on deformation and damping characteristics as well as undrained strength
of natural soils can be obtained from monotonic or cyclic laboratory shearing tests.

Because of the difficulties involved in undisturbed sampling, in situ testing plays a much more important
role in characterizing granular soil deposits especially for the case of seismic analysis. Mayne et al. (2009)
identified over thirty types of available in situ test methods. The use of in situ tests has proven significant in
the characterization of geomaterials in several aspects: (1) they can be done relatively quickly as compared
with laboratory tests, (2) results are available immediately, (3) large number of data is obtained, and (4)
vertical and lateral variability can be assessed. Seismic piezo-cone penetration (SCPTU) and seismic ftat
dilatometer tests (SDMT) (Marchetti et al., 2008) are two commonly used full-displacement penetration tests
but with added capability of shear wave velocity measurements similar to those of downhole geophysical
tests. SCPTU and SDMT can also be used to infer groundwater conditions and characteristics ofhydraulic
conductivities ofthe tested soils by monitoring the decay ofpenetration induced excess pore water pressure.
These hybrid and multi-functional test devices are especially useful in meeting the demands for seismic
analysis related site characterization. An important disadvantage for penetration tests is that t~ey usually
create complicated boundary conditions and make rigorous interpretation ofthe test results rather difficult.
For granular soils, the consideration of potentia1 seismic ftow failure or liquefaction further complicates the
content ofthe analysis. Correlations between in situ test results and soi1 cyclic strength have been proposed
for the assessment of liquefaction potential of granular soils following the simpIified procedure (Youd et
al., 2001). The approach is based on fieId observations of the performance of sand deposits that did or did
not liquefy in previous seismic events. A set of data points of earthquake induced cyclic shear stress versus
in situ test results are pIotted first. The correlation curve is established based on a borderline that separates

64
 GEOTtcHNICAL ANO GEOPHYSICAL SITE CHARACTERIZATlON ORIENTEO ro SEISMIC ANAlYSIS

the data points that correspond to liquefaction and no liquefaction. Jefferies and Been (2006) offered an
altemative way to evaluate soilliquefaction potential via state parameter. From theoretical point of view,
this is a more desirable approach as compared to the simplified procedure because of the strong correlation
between state parameter and sand dilatancy. To take full advantage of this approach however, a reliable
means to infer state parameter from in situ tests would be required.

The aim ofthis paper is to summarize sorne ofthe recent developments in geotechnical and geophysical site
characterization techniques as they relate to seismic response and liquefaction potential analyses. Subjects
to be described in the paper inelude:

Geophysical site characterization

Defonnation and damping characteristics
Undrained cyelic strength from in situ index tests
".
Undisturbed sampling in granular soils
The state parameter approach to evaluate soilliquefaction

GEOPHYSICAL SI TE CHARACTERIZATION

Field seismic testing techniques involve monitoring the partiele motions of different modes of propagation
from disturbance induced by active or passive sources. Active sources release energy from a mechanical
device or explosion to induce stress waves propagating in solid medium and a passive source used the
background noise as the excitation. The modes ofpropagation ofien used are two body waves (e.g., P-wave
and S-wave) propagating within the mass of medium and the Rayleigh wave (R-wave) existing near the
surface. Complexity of stress wave field, which depicts temporal and spatial variations of stress waves
within a medium, depends on source characteristics, soil properties, and geometry and will reflect on
recorded seismic records. Because recorded seismic data contain aH the infonnation from the complicated
wave field, all seismic testing techniques involve enhance or separation of different modes of propagation
using different testing layouts and signal processing techniques.

In situ stress-wave based techniques gain more attentions in geotechnical earthquake engineering because
they can provide infonnation oflayer thickness and dynamic soil properties (e.g., modulus and damping)
at small strain level, which are major parameters in perfonning site specific ground response evaluations
and dynamic soil structure interaction analyses. Most stress-wave based methods measure the propagating
velocities from travel time in time domain or by spectrum analysis in frequency domain. The maximum
shear modulus (GmaJ, shear wave propagatin~ veloci~, ~d strcas state of soils can be related by:
Gmax =pv, =Cs..JF(e)(;ra(;rb (1)
a a

where p is mass density, es, na, and nb are material constants, F( e) is a void ratio function, Da and q
are the principal effective stress in the direction ofwave propagation and partiele motion respectively,
and Pa is the atmosphere pressure. Incorporating G max from field testing (Gmax,field) into normalized modulus
reduction curve from laboratory tests, which is the ratio of shear strain at different shear strain level (G(Y\ab)
divided by the maximum shear modulus in the laboratory(Gmax,'ab)' the field modulus at different shear strain
level (G(Y)fiel) can be inferred by:

G( ) (G(Y)zab)G
y jield = G max,jield (2)
max,lab

65
 )',;:&1
::'l~

.ltb .::.;1' e-,n f~01 ;:h¡-.. =;~,s· (;'e·~)ít;1 ;'.11 coi Efigil ,l~)i..'?r;'-:g

A brief review of stress-wave based geophysical characterization methods is presented below. Futher details
can be found in Stokoe and Santamarina (2000), Sheriff and Geldart (1995), and Richart et al. (1970).

Intrusive seismic methods

Intrusive seismic methods require boreholes to install the source and receivers and the wave velocities are
evaluated by measuring travel time ofthe specific travel path from the source to the receiver. In geotechnical
field, cornmon types of intrusive methods are Crosshole tests, downhole and seismic cone penetration tests,
and suspension P-S logging.

Crosshole method
The crosshole method is a time-of-travel measurement where the source and receivers are placed at the
same depth in adjacent boreholes. Standard layout of crosshole methods are shown in Figure 1 and the
testing details can be found in ASTM standard D4228 (ASTM). The crosshole seismic testing requires
an in-hole source capable of generating both P-wave and S-wave propagating hof'izontally. The receivers
must be able to record particle motions in 3 orthogonal directions in arder to measure the P, SH, and SV
wave velocities. Advantages of crosshole method include high resolution for testing materials, capability of
render a tomographic image of the cross section by inclined ray paths, as well as measurements of P, SH,
and SV wave velocities at the same depth. The major disadvantage ofthe crosshole method is the time and
cost of preparing the boreholes (Stokoe and Santamarina, 2000).

fI,,<,e¡""r

~
solJrcc eorchulp, ~. ",_', ¡"
~,l'e"o es
~

~ - .. - - - - @ - . - - - . - 0
I 'C .3 ..et~cs ~
I
.. 3 oaeters I
...
(10 feet) (10 reet)

Sei~lc Reco.der

-- Sei,..,;<;
Source

Recerver:¡

...

C<loing

GrQut

Figure 1. Layout of crosshole seismic testing (ASTM).

66
 GEOTECHNICAL ANO GEOPHYSICAL SITE CHARACTERIZATlON ORIENTEO TO SEISMIC ANALYSIS

Downhole method and SCPT

In the downhole method, travel times of shortest travel path between a source on the surface
and two receivers at selected depths in a borehole are measured. Standard layout of a downhole
method is shown in Figure 2 and the testing details can be found in ASTM standard D7400
(ASTM). The average wave velocity between the two receivers is evaluated by:

v = (L 2 - L 1) / (T; - T¡) (3)

in which L and T. are distance and travel time from source to receiver i shown in Figuer 2. The
seismic cone penetration test (SCPT) (Campanella et al., 1986) is similar to the downhole seismic
test with motion sensors integrated in the cone for travel time measurement. The advantage of
downhole test is that only one borehole is required and the major disadvantage is that more source
energy is required for deep measurements.
~

AXIS OF SCPT

HEAR BEAM

SHEAR EEAU

DI

ASSUMED TRAVEL PATHS Or

SElSUIC WAVES FROM SHEAR
SCPT AT DEP1H 01 9EAM TO SEISMOMElERS IN
SCPT BOOY AT DEPTHS DI
ANO 02 EC~M:RS
01
ASlNC

SCPT Al DEPTH I:n

Rour

(a) downhole test in cased borehole (b) Seismic cone penetration test
Figure 2. Schematic of downhole seismic test and SCPT (ASTM).

P-S logging
Welllogging or borehole logging has long history in petroleum engineering. Logging tools can be lowered
into borehole to produce the profile ofmaterial properties. The suspension P-S logging method is a relatively
new method of measuring P- and S-wave velocity profiles of soils (Nigbor and Imai, 1994) and is probab1y
the only technique that can provide high resolution wave velocity profiles for deep profiles (deeper than
200 m). The setup of a suspension logger is shown in Figure 3. A string of source and receivers is lowered
into a fluid-filled borehole. The ray path is source-fluid-surrounding material-fluid-receiver. Figure 4 shows
the time histories for S-wave profiling in southem Taiwan, where the thickness of soil deposit is greater
than 100 m. The near and far receivers are the lower (R1) and upper (R2) receivers in Figure 3. The P­
and S-wave velocities of the surrounding materials are inverted following the standard travel timeprocess
between the two receivers and the results represents the average wave velocity between the two receivers (1
m apart). The inverted wave velocity profiles are plotted in Figure 5 and the top 15 m S-wave profile agree
well with surface wave measurements in this area. The velocity profiles of soil deposits will be beneficial
to characterize seismic site response. The main limitations in borehole logging are the effect of casing and
coupling between the casing and surrounding material s on the measured response and the disturbance of
surrounding materials during boring process, which will be more significant in shallow and soft soils as
shown in Figures 4 and 5.
67
 ~
i!llút l',o~jork ~ CC¡-r'"

Arrrúred 7-CondlJotor cable

~ OYDPS-170
~ LoggeriReccfder
Cable Head // .lJ

Diskette
Head RedlJcer lilfth Data

Ji k'~'" U wer (R2)

R eceÍIIE! ~
Depth refe re¡-,;;e locati on
fur R1- R2 anatys:is: Ji 1

rrioj. point of ReceÍllers

Q~m
U;..o\ler (R 1) ----,­
R eceÍller

1D7 m 1.57 m

f:l .. r-~ J oint Deplh reference localion

10r S- R1 anatys:is : mid
1.0 m flexible point of 3.14 m S- R1 s padng
Is: ola:ion Cy'linder
1.07 m I I
Joint 3.7 m
Combined Sh and
P-lAlaVe SOlJrce (S)

Soorce Dr!ver 105m

~....
W.lght
lip
1
Overal! Lenglh ~ 5.8 m

Nct to Sc:ae

Figure 3. Setup of suspension P-S logger (from GEOVision).

68
 GEOTECHNICAl AND GEOPHYSICAl SITE CHARACTERIZATION ORIENTED TO SEISMIC ANALYSIS

near far

- 15 1 : = ;;;(;;~~
-201-1-~-"

8'
- 25 t== ==~-~~~ ~
"-'

f301 -35
3~
- 40 1 ~

-45
~~
t=b==:=:==:=:=3'~ii~{j~~.lQf2= .~-...,
- 50 1 :=~
::

-55~~
-60~
O 10 20 30 40 O 10 20 40
Time (ms) Time (ms)

Figure 4. Time histories for S-wave velocity profiling in southern Taiwan.

Velocity, m/sec
500 1000 1500 2000

00

101 ';Ii, 1,.('

r:c.
E I \: t~ I

30 1 ¡:" 1 :t

40 1 ;J' 1 1 "),

-e- S-Wave

501 '1s I ),

60' (, , ­

Figure 5. P- and S-waves profiles from suspension logger.

69

~<
':)jh .' :':'Hno!icl' ':!: C:c:~··,'

Nonintrusive seismic methods

N onintrusive seismic methods measure the wave propagation parameters on ground surface without invading
the ground by drilling. Depending on source types, nonintrusive methods can be classified as active and
passive methods. The nonintrusive, active methods include reflection survey, refraction survey, and surface
wave methods. There are many passive methods that use the background noise as the source have been
developed or under development. In reflecting the scope of the paper, microtremor measurement analyzed
with N akamura technique (1989) for identifying the site amplifications characteristics is presented.

Refleetion survey
Reflection survey is one of the most common seismic methods. The basic principal of reflection
survey and offset-time curve for a single layer is shown in Figure 6. The reflection method generally
uses the first arrival of P-wave to construct the offset-time curve to determine the thickness and
P-wave velocity of soillayers. However, interferences between reflected, refrac~d, and surface waves
increase the complexity of seismic data. Consequently, various testing layouts and signal processing
techniques have been developed to enhance the signals of specific travel paths. The complexity of wave
field has limited the applicability of the reflection survey in near-surface survey.

P,.vt:>,. VSI ~T
]Lv •
P2' Vpz. VS~

,
Figure 6. Principal of reftection survey and offset-time curve (Richart et al., 1970).

Refraetion survey
Refraction survey uses the critically refracted wave from a higher velocity layer that underlies lower
velocity sediment. Seismic refraction testing is an established geophysical method for identifying
subsurface soil stiffness and layer interface at shallow depths. The principal of refracted method and
offset-curve are shown in Figure 7. Limited by requirements of critical refraction at boundaries, the
refraction method is only applicable for stiffness increasing profiles.
s
1.. x •I

x-2H toni e

p.,. VP1
"X
Vpz>VP1 P2.vn

Figure 7. Principal of refraction survey and offset-time curve (Richart et al., 1970).

70
 GEOTECHNICAl ANO GEOPHYSICAl SITE CHARACTERIZATlON ORIENTEO ro SEISMIC ANALYSIS

Surface wave methods

Surface wave methods are the most vigorously growing seismic methods in shallow depth profiling for the
past decays. The nonintrusive, fast, and large sampling area make thern very attractive for near-surface (less
than 30 m) geotechnical site characterization as screening tools and are the only choice for hard samplel
drilling soils. The basis of surface wave methods is the dispersive characteristics of Rayleigh waves in
layered system. The phase velocity of Rayleigh wave, V R' depends primarily on the soil stiffness over a
depth approximately one wavelength. As a result, Rayleigh waves with different wavelengths will sample
different depths and the phase velocity will vary accordingly. Although several surface wave methods have
been employed in near-surface characterization, the spectral analysis of surface waves (SASW) method
(Stokoe and Nazarian, 1985) and the multi-channel analysis of surface wave (MASW) method (Park et al.,
1999) are the two most popular methods. In spite of differences in testing arrangements and data processing
procedures for different methods, surface wave methods all contain three stages, which are data acquisition,
construction of field dispersion curve, and inversion process for establishing representative shear wave
ve10city profile. r

The process ofSASW method is summarized in Figure 8 and details ofthe method can be found in Stokoe et
al. (2004). The SASW method is a simple technique required only two receivers at multiple source-receiver
configurations to construct the field dispersion curve via spectral analysis. After the construction of field
dispersion curve, iterative inversion analysis based on apparent velocities and the dynamic stiffness matrix
method (Kausel and Roesset, 1981) is performed to find the representative shear wave velocity profile that
best matches the field dispersion curve.

The MASW method uses the same data acquisition configuration as the reflection survey, which involved
multichannel recording. Then the time-space data are transformed into frequency-phase velocity dorna in
to identify the trends of dispersion from the pattern of energy accurnulation termed as phase-shift rnethod
(Park et al., 1998). This process is capable of capturing the multi-modal dispersion features. Because the
constructed field dispersion curves could contain multi-modal dispersion features, a multi-modal inversion
is adopted to infer the soillayering. The process for ID testing is shown in Figure 9. Cornparison between
MASW with multi-modal inversion process and Seismic CPT (SCPT) for a liquefaction study site at Southern
Taiwan is plotted in Figure 10 and the results show a very good agreement between the two techniques.

HVSR technique
The horizontal to vertical spectral ratio (HVSR) technique proposed by Nakarnura (1989) is one of the
effective ways of characterizing the predominant period (To) of a soil deposit or soft layer aboye bedrock.
The Nakamura technique allows one to evaluate the predominant frequency ofthe soft layer by measuring
the tremor on the surface at a single station. Because predominant period is a direct index for seismic
amplifications, spatial variations ofpredominant period will be useful in seismic microzonation. In addition,
the average shear wave velocity of the entire soil deposit can be inferred from the quarter wave law:

H
To = 4V (4)
s

where H is the soil thickness. The concept of Nakamura technique is illustrated in Figure 11. The logic
behind the HVSR technique is that the vertical motion is less sensitive to amplification than the horizontal
motion and can be seen as the bedrock motion. Therefore, the quasi transfer spectrum (QTS) can be defined
as the ratio of the horizontal over vertical motion spectra and the peak of QTS represents the predorninant
period ofthe soft layer. Figure 12 shows a HVSR rneasurement performed on a reclaimed harbor in Taiwan.
The spatial variation of predominant frequency agrees with the distribution of filled materials, in which the
gravel backfill zones shows higher predorninant frequency.

71
 ~. ~ .
.:',H~ ¡~'ltEJ!TjOtk:;r'-"'" '_-,:: ¡l,.:: '('1,'

Irrpu;t,
(a) r o
~14
D 1. ViI:m
Snpt :l,ibJdlal
lIan. or

~2 ~1
•
JJ
Randoo1~

~ I(~'a 11
t-
"";;'

-, *-
{)I'FI'
I

t.-+. ¡:~,. (c)

ítc¡f Jira 1& •

11J
Ií ..
J
~
Iltlp!I'!iM CUn<e

Ibofo-"""Hz ~.HE ~
J

f ~ .""""'*..... .,
2

el: 1
(b) ñ '0

.....
!!_.LI
6
(>

2m «ID BIXI I!OI) 1'_......,. H:.:

I'-=¡u.....,. Hz

Fip.re 7, Spectral-illllll)$ilH:H-surfac;:c-wavC$ (SASW) method: Cakutation of pb1ISC ve/.ocmc:s

..

i~
O-1m
~

Expeñmenlal [] ¡;pasil:n CuIV4;t

~
n-2m

.­ 0-..,.

-~
11~

iJ~

o • .,.

i~

i~btN\JWt D._

- . .... . .
_.
Figure 8. Process of spectral-analysis-of-surface-waves method (Stokoe et al., 2004)

72
 GEOTECHNICAL ANO GEOPHYSICAL SITE CHARACTERIZATlON ORIENTED TO SEISMIC ANAL YSIS

~­
,....
Multf:Channel Record
"../ / I "1 ...............

,,/ GI" -- o.tr1:'tt (ni)

20 «J ~I) :so 1CÚ
Q ~!.:imltrt.,1~!~~

Scimlic
~

!
S
g. fi'j~\"?~
"'7 Ii",
LI olI._

A: AIr'\\'a\"4:: D: Retkdion
B: Dl!'$tl Wave; E: R~Jh¡ctloJIIi
t:ÓO! SUt.r:1.C~ W~ \~iII M'Ildt) r: J:!.rd 8~/t¡;sf!!<Bqt'Sur(a.:ev.\1\-e
el. Sw:ru~ ~Y'e (1Iightl" Modes) G: A:.írIbiw CUlbi'II"ti, N,)jE>;o
:§ • lrlll't:v.;~"m"
/tI tl;~ fh.".~

1-D S..valoclty (Vs) PrcfU6

S-Vli1hxlty~)~r¡II~)

e:r>('n:r¡"~"")
¡ro
,=. "2:9)
! j
~(J
F
1CCjI)

D~~., lmag(t .--1-,

i1 ~ ..,~tI!1J1)

:11. 21..__
·...,t
~_----+-_.J~._
J 1. .. ­
..
1.<:;­
.8
'"
II~
\ ...
"'E
B

...
o
la
=­a
.
lO
¡f
~f' -- -j ...._+- _.
,! ,
-+.-.. .A

Fr~.hJ r..,'[t-tz)
~IL__~~~~~~

Figure 9. Process of MASW method (KGS, 2010).

73
 80 120 160 200

o -,'--1-'r--'---L--'--L-j

Silty clay (CL)

5 -
Silly sand (SM)

:§:
t'-' 10
Q
o
- Silty clay (eL)
,.
o
o
-MASW
15 -l - Silty ciay (CL)
D SCPT

o
o

,J I
80
I
I I
I
120 160 200
I
1

V, (mis)

Figure 10. Comparison ofVs profiles from SCPT and multi-modal MASW.
lI'eIIlOC
.surface waves

Fig. L Typical geological struaure of a sedimentarybasin.

'";f t··············l··············:·····I·~·:···········
~
~
i:
t
o fo 1/0 3[0
f(Hz]

Figure 11. Concept ofNakamura HVSR technique (Carniel et al., 2008).

74
 GEOTECHNICAl ANO GEOPHYSICAl SITE CHARACTERIZATION ORIENTEO ro SEISMIC ANAL YSIS

2:7

'1.5
" L~
1 .•
111'1 ~
1,1
1
n.s

Figure 12. Spatial variation of predominant frequency in a reciaimed harbor.

DEFORMATION AND DAMPING CHARACTERISTICS

For seismic analysis, the deformationand damping characteristics generally involve the relationship between
modulus/damping ratio and strain. The importance oflocal strain measurements in the determination of soil
deformation characteristics in laboratory tests has long been recognized (Burland, 1989). Severallaboratory
deformation measurement techniques with sub-micron resolutions have been developed in the past few
decades (Scholey et al., 1995). By lJlounting the measurement devices directly on the soil specimen,
deformation as well as damping characteristics under a wide strain range can be measured using a single
specimen (Toki et al., 1995). Figure 13 shows the stress, excess pore pressure and strain relationships from
a monotonic triaxial test on a clean Mai Liao Sand specimen. Local axial strain was measured using a non­
contact proximitor. Figure 13(b) depicts the part of stress-strain relationship where axial strains are less
than 10-3 %. The result is consistent with the general understanding that sand should have an elastic or linear
elastic behavior under a threshold strain of approximately 10-2% (Dobry et al., 1982 and Jardine, 1992).

Bender elements reported by Dyvik and Madshus (1985) provide a quick and economical determination of
maximum shear modulus (G max or G o) on soil specimens via shear wave velocity measurement. The shear
strain caused by bender element vibration is believed to be les s than 10-3% (Dyvik and Madshus, 1985). The
deformation and damping characteristics that covers the full range of strains involved in a typical seismic
analysis can be realized using cyclic triaxial tests with local strain and bender element measurements. Figure
14 shows the results from cyclic triaxial test with G o using bender element measurements on a single, silty
sand specimen from Southern Taiwan. The local strains were measured using a pair of local displacement
transducer (Goto et al., 1991). The equivalent shear modulus (Geq ) and damping ratio (h) were determined
based on a series ofunload-reload loops with different shear strain (y) amplitudes.

75
 ~

ln j ei nOi"ior.\..ll C0/11er(;;o,-" Gc ;r,r ; ."ji1-,el_':'~, ~:;

60rl----------------------~

50
o:s
~40
en
en
(1)
0.6 , r - - - - - - - - - - - - - - - - ,
..
~30
....
o
~

.~20 o:s

p.. e·· ..
.e••
0 10 ~ 0.4~ e ••

.­
en~
",.
'"
(1)
.b
O, '"O....
16
~

~
.~ 0.2 f- .,. .e • ,..

o:s
~ 12 q

~ 8
4
O.
~ "
0.1 0.2 0.3 0.4 O 0.0002 0.0004 0.0006 0.0008 0.001
Axial strain, % Axial strain, %
(a) Stress, excess pore water pressure
(b) Stress-strain relationship in srnall strain
and strain relationship
Figure 13. Monotonic triaxial compression test on a Mai Liao Sand specimen.

35 0.28
,--..
o:s C) y - Geq
~ 30 <) y-h 0.24
'--' ~
cr'
~
.Sr.
O 25 0.2 ......
G" o:s
.....
en~
;:l
o O!)
.......

.g 20 0.16.[
o
El 0<0 (l Elo:s
;:; 15
Q)
0.12 ~
~
en

<:;O
<><> 'P
1.1
.....
1:1 10
1.1

.......
° 0o 0.08 ~
ro
>- <> oo ::r;>-.
.- 5 0.04
~
6- oo~
O
0.0001 0.001 0.01 0.1
Shear strain, y (%)
Figure 14. Cyclic triaxial test with Go measurements on a single silty sand specimen.

Sorne ofthe in situ testing rnethods can also fulfill the function, or at least part of deformation characteristics
measurernents in the field. Figure 15 shows a coHection of such in situ testing methods and the range of
strains that can be used to infer deformation characteristics. The seismic geophysical tests can be used to
determined Go' The pressurerneter test (PMT) rneasures the relationship between its probe expansion pressure
and radial strain, and hence is ideal for rneasuring the shear modulus - shear strain or G-y relationship.
Using highly sensitive strain sensing arms, and unload-reload loops in PMT, the G-y relationship that refiect
76
 GEOTECHNICAl ANO GEOPHYSICAl SITE CHARACTERIZATION ORIENTE O TO SEISMIC ANALYSIS

relative1y undisturbed soil conditions can be obtained (Fahey, 1998 and Wang and O'Rourke, 2007). Figure
16 shows the results of a PMT with unload-reload loops. The test was conducted in a soft rock at the campus
ofNational Chiao Tung University in Hsin Chu, Taiwan. The PMT strain arms and expansion pressure were
monitored with fiber optic sensors. With this system, the secant shear modulus degradation that corresponds
to a minimum shear strain of 10-3% was obtained.
Reglan/or
!..-- Range far. --1_ Bearing Capacity _1
:t!GNsical l Deformation I and Stability
Testá'- Analyses Catculations
CJ '."""-,
(1)

:J
Unload-Reload PM'T.
:J '\ Flat DMT
'"O
O
, 8cr,
~
....
as ~
&"..,.o~
". l':.
1"
l1"

"~

(1)
...s:::
en
10.8 10-5
Initial Loading p;:..............

10-4 10-3 10-2

--._ Testa

10-1
..
Penetration

--- ----­
Shear Strain, 'Y8
Figure 15. Conceptual variation of shear modulus with strain level under static monotonic
and relevance to in situ tests (from Mayne et al., 2002).

5rl----------------------------~ 600,.-------------------~~-----.

4 r­
..
400 .
+ •
iJ ;t ) ro

~
6
200 ­
• +
••
••
+

~.~
.~

1 ~.- I ' /

O~ , '''ll , 'lO!: .... ;1

01
O 5 10
~
15 20 25 0.0001 1 0.01 0.1
Strain, % . 0.00 Shear strain,
Ca) PMT expanslOn (b) Modu1us degradabon
Figure 16. PMT with unload-reload loops in a soft rock.

The seismic fiat dilatometer test CSDMT) provides two data points: (1) the initial modulus Go from shear
wave velocity measurement (Equation 1), and (2) a working strain modulus corresponding to the DMT
constrained modulus MDMT' The G-y curve can be constructed by fitting "reference typical-shape" curve
through the two data points from SDMT (Marchetti et al., 2008). Similar two point approach can also be
applied using the seismic cone penetration test (SCPT). In this case, the G o from V s and a modulus va1ue
that corresponds to strain induced by cone penetration is inferred from the cone tip resistance.

77
 I j¡~on'_:!¡ I

UNDRAINED CYCLIC STRENGTH FROM IN SITU INDEX TESTS

A major concem to geotechnical engineers from a soil behavioral point of view is the potential of cyc1ic
liquefaction for a given seismic event (e.g., a design earthquake). There are general1y two avaiIabIe
approaches to assess the potential of cyc1ic liquefaction of a given soil deposit: (1) use of Iaboratory testing
on undisturbed samples, and (2) use of semi-empirical relationships that involve correlations of observed
field soil behavior with in situ index tests. Taking undisturbed samples in sand and conducting laboratory
testing can be complicated and prohibitive1y expensive. Thus, the laboratory testing approach is usually
reserved for research applications and rarely used in geotechnical engineering practice.

The semi-empirical field-based methods (Idriss and Boulanger, 2006), which evolved from the simplified
procedure by Seed and Idriss (1971) are by far the most widely used methods in assessing the cyc1ic
liquefaction potential of sand. The simplified procedure has two essential compojlents: (1) an analytical
framework to organize past case history experiences, and (2) a suitable in situ index to represent soil
liquefaction characteristics (Idriss and Boulanger, 2006). In situ penetration tests have proved to be usefuI
for representing soil liquefaction characteristics because they not only provide an indication of denseness,
but also reflect other important characteristics such as fabric, gradation, cementation, age, and stress history
(Seed, 1979).

The simplified procedure provides a boundary curve that separates cases of observed liquefaction and those
with no notable liquefaction in a two-dimensional plot of seismic loading, in terros of Cyc1ic Stress Ratio
(CSR) versus a norroalized in situ index test value. The boundary curve also serves as a correlation between
the in situ index test value and the Cyclic Resistance Ratio (CRR). The terro CRR may be considered as the
maximum CSR that a soil can resist before liquefying. Traditionally, the result ofthe liquefaction potential
analysis using the simplified procedure is presented in terms of a factor of safety (Fs) defined as the ratio
ofCRR over CSR. No soilliquefaction is predicted ifFs > 1. The assessment ofliquefaction potential in
terros of factor of safety is general1y known as the deterministic approach. In recent years, there has been an
increased effort to quantify the generally unknown degree of conservativeness that existed in the published
boundary curves and to assess the liquefaction potential in terros of probability of liquefaction (e.g., Cetin
et al., 2004 and Juang et al., 2002, 2006). Details ofthe probabilistic approach are beyond the scope ofthis
papero

Four in situ index test methods have be en identified by Youd et al. (2001) as having reached a level of
sufficient maturity for the purpose ofsoilliquefaction potential assessment under the framework ofsimplified
procedure. These tests inelude: (1) standard penetration test (SPT); (2) cone penetration test (CPT); (3)
shear wave velocity (Vs); and (4) Becker penetration test (BPT). BPT is used primarily for tests in gravely
deposits and readers interested in BPT are referred to Harder and Seed (1986).

Liquefaction Potential Assessme~t for Clean Sands

The oldest and probably the most widely used in situ index test method is the SPT. A relationship between
CRR and the SPT N value (blow count of hammer required to penetrate a split barrel sampler for 1 ft or 300
mm), corrected to a hammer energy ratio of 60% and norroalized to an effective overburden stress (Jvo') of
100 kPa (or 1 atm), [designated (N¡)6o] is used to represent the boundary curve. Figure 17 presents the CRR­
(N¡)60 correlations published in the past 3 decades for c1ean sands (fines content, FC<5%) and earthquake
events ofmagnitude M = 7.5. Fines are defined as partic1es passing #200 sieve (material < 0.075 mm). The
modifications in CRR-(N¡)60 correlations over the years general1y recognized a more significant increase of
CRR as (N ¡)60 reached values about 30.

78
 GEOTECHNICAl AND GEOPHYSICAl SITE CHARACTERIZATlON ORIENTED TO SEISMIC ANAlYSIS

The SPT is a versatile testing method applicable to soils with a wide variety of gradation and density
conditions. The equipment and skilled technicians required to perform the SPT are readily available in most
parts of the world and thus there exists a large database. A soil sample, albeit disturbed, can be retrieved
with the split barrel sampler along with the SPT. The gradation and other important basic physical soil
properties can be measured directly in the laboratory. These are important advantages in favor of the use of
SPT for soilliquefaction potential assessment. The equipment such as the rope, dimensions ofthe cathead,
type of harnmer and details in the split barrel sampler (Seed et al., 1984) can all affect the measured SPT
N values. It is imperative to follow the relevant standards when using the SPT (ASTM D1586-08a and
associated energy measurements in D4633-05).

0.6
CDSeed, 1979 ® /!
IZ>Seed and Idriss, 1982 :: ./ !
0.5 _ ~sSeedd et all., 21090841/NCEER Workshop, ~~~L. .:j
~ ee et a ., '
__ ..oJSlL........
,'.,'

11"
®Idriss and Boulanger, 2006 ': /' dJ (j) J

.-.........-....-.-.+.-....-..•....... --.+.-..........~ ... /·f/:./·:~:;,,/·~·~··

a::
a::
O
5 0.3
0.4
,:: ,

...-.-..-......,"... -~
,... ..-.-..-'-..',--'...-....-
~
Q.,
~:
: _./ ,.'
.,.,
..t6:..- .'
/ ./.,,"
~:",~:~,?""o' -.......-..-...

a::
(j)
O
0.- . _ _.
:
~4'.i' .
:." o
').:0"
o.;..¡:--- ooo
'
o:
o
0.2 --........--.•.•., .... :.-~,.e"'}~-' ¡--··-~·~·········-----:--·-·----------·Ó-----I

:;ro:...... .-" :
0.1 ~:~·:--·---o{)---·--·--:---- - üquefaction
o (bO o :
o o , • Marginal Liquefaction
o No Liq,-!efaction
o
o 10 20 30 40
(N 1)6o
Figure 17. CRR-(N 1)60 correlations published in the past 3 decades for clean sands
(after Idriss and Boulanger, 2006).

The CPT can be highly automated and yields an almost continuous stratigraphy of the soil deposit. The
quality of test data can be substantially enhanced with the addition of piezo unit(s) and seismometer(s)
(i.e., the seismic piezo-cone penetration test, SCPTU). The additional information can be helpful for
soil liquefaction potential assessment. Unlike SPT, no soil sample is retrieved in regular CPT. A soil
c1assification is required as a priori in the use of CPT for liquefaction potential analysis. Robertson and
Wride (1998) proposed a soil behavior index, le for the purpose of soil classification where

2 2]0.5
le = [(3.47 -logQt) + (logFr +1.22) (5)

Qt = (q t - avo ) / ~o (6)

Fr = [fsl(qt -Ovo)]100% (7)

in which qt = cone tip resistance corrected for pore pressure effects (Campanella and Robertson, 1988);
fs = CPT sleeve friction; crvo = in situ total vertical stress; and a'vo = in situ effective vertical stress. 1e is a soil
behavior type index that ranged from approximately 1.3 to 3.6. Lower value corresponds to clean sand and

79
':'1 ~1 in1e"'~ ·'T:OnCÚ C.C:~-,;

the soil tends to have claylike behavior when 1e > 2.6. Robertson (2009 and 2010) emphasized that 1e is a
Soil Behavior Type (SBT) c1assification indexo The le based SBT classification may not be consistent with
that from the Unified Soil Classification System (USCS).

For a given density state, qt increases with stress in a non-linear fashion. The rate of qt increase reduces as
the stress becomes larger and the dilatancy effect diminishes. A single stress normalized cone tip resistance
value is generally used to reflect the soil density state in the evaluation ofsoilliquefaction. The dimensionless
normalized cone parameter, Qtn is defined using the following format (Robertson 2010):

Qtn = [(q¡ -Ovo)jPa](Pa/dvor (8)

,
j
where Pa is the atmospheric pressure and n is the stress normalization exponent that varies with SBT. Soil
dilatancy is related to stress, grain characteristics and mineral content, as well as 'density. Recent studies
(Idriss and Boulanger, 2006; Moss et al., 2006; Cetin and Isik, 2007; and Robertson 2009) have gene rally
agreed that for c1ean sand, n = 0.5 and in clays n = 1.0. Robertson (2009) refined the correlation between n
and soil behavior type as:

'n = 0.381(IJ +0.05( dyol Pa) - 0.15 (9)

where n ~ 1.0. Figure 18 shows a selection of correlations between CRR and Qtn published within the
last decade for clean sands and earthquake events of magnitude M = 7.5. There are different levels in
conservatism among"-' these published CRR- Q tn correlations as Qtn exceeds 100.

0.9
o Non-liquefied JlJang et al., 2006 • Moss, 2003
0.8 ....• Ij Liquefied""""""''\:_'' """\""'0-"
Q.
t.. ..; • ci\./o ¡' I

cr
cr
O
0.7

0.6

0.5
., .,
.
......

.... ..• •.•...

Robertson and
Wride,.1998

.
ti:' ./
-- ..¡'-
o /./

,
.:
;

/
.• ;' --'f'
ól
/
.!í
/
I
j

Lo. .... '1ijJ¡ o .' (

o
ffi
O
0.4 ~""'';;' rot,. .. -.
. • ".~
~ .....
¡¡Q,....
"'" ¡¡ !
,
o .. 0 .. /0/2...
t::I. '/
:

.
..>/..... .
o'
,1
..

'" • 0""-'/0//
0.3 . . . . . . - -.. --ji .... ' ... --....~/...... ':6 ...........--..--..... .
.:_
¡ .. 'l ~.o ,;/7
. . . . . . . . . . .0/
....-(
/"
0.2 .,,' qP~... CJO ...... o
-<'.,,[:......ó.--~·····
o 0000 o
0.1 ?-'>.. G-.-----: 00 . o o
o
8 Idriss and Boulanger, 2006
O
O 50 100 150 200
Qtn

Figure 18. A selection of CRR-qclN correlations published recently (after Juang et al. 2006).

80
 GEOTECHNICAL AND GEOPHYSICAL SITE CHARACTERIZATlON ORIENTED TO SEISMIC ANAlYSIS

Forthe fiat dilatometer test (DMT) developed by Marchetti (1980), the horizontal stress index (Ko) has been
demonstrated to have clear correlations with Dr (Reyna & Chameau, 1991), at rest lateral earth pressure
coefficient (K) (Monaco et al., 2005) and most importantly, the age of sands (Marcehetti et al., 2008; and
Marchetti, 2010). These characteristics make DMT favourable as a viable in situ index test method for
the assessment of soil liquefaction potential. Figure 19 surnmarizes the various CRR-Ko (for earthquake
magnitude M = 7.5) for clean sands reported by Marchetti et al. (2008). The fiat dilatometer is rugged and
more capable in penetrating through dense granular materials than the cone penetrometer. The DMT control
console and its test procedure are simple and results are not operator dependent. A drawback with the use
of DMT is the lack of a large database. As in the case of CPT, no soil sample is obtained in DMT. The soil
gradation needed for liquefaction potential assessment is estimated from empirical rules.

0.5
,..
LlQUEFACTION
0.4

cr
cr 0.3

'­
o

o::

2 0.2

0.1 "Robertson and

Campanella, 1986

NO LlQUEFACTlON
o
o 2 4 6 8 10

Ko
Figure 19. Available CRR-~ correlations (after Marchetti et al., 2008).

Nonintrusive V s measurements like surface wave methods can be especially useful for sites underlain with
gravelly materials where penetration tests such as SPT or CPT are not feasible. The V s can also be easi1y
measured in a triaxial cell using bender elements. By comparing the V s from bender element test to the
CRR obtained using the same soil specimen through cyclic triaxial tests, it is possible to verify or establish
the CRR-VS1 correlations using reconstituted (Huang et al., 2005) or undisturbed samples (Baxter et al.,
2008). Figure 20 shows the CRR-VS1 (Vs normalized to a effective vertical stress of 100 kPa) correlations
published by Andrus and Stokoe (2000). An important disadvantage in the use ofVs is a lack of sensitivity
to the relative density, Dr' For a change ofD r of clean sand from 30 to 80%, the corresponding SPT N value
would increase by a factor of7.1 and qt by a factor on.3. The same D r would be expected to change the V s
by a factor of 1.4 based on available correlations (Idriss and Boulanger, 2006).

Roy (2008) compiled a databas e from 24 test sites in different parts ofthe world to analyze the correlations
among CRR, qt' and V. The data included laboratory tests on high-quality (undisturbed) soil samples, field
CPT soundings, and Vs measurements from near sampling locations. The sands had a wide range of fines
contents, age and soil grain compressibiliiy. The analysis did not indicate a coherent correlation between
Qtn or V S1 and CRR. Instead, Roy (2008) reported that the ratio of qt to maximum shear modulus (G) relates

81
reasonably with CRR using separate correlations that depend on geologic age. Two boundary curves that
separate the liquefaction region from no-liquefaction cases in the CRR-q/Go space can be identified as
shown in Figure 21; one for Holocene and one for Pleistocene soils. The maximum shear modulus (G o) is
determined from Vs measurements. The CRR-q/G t o
correlations are independent of the fines content and
soil grain compressibility. The use of CRR-q/G t o
correlations avoids the complexity of adjustment and/or
normalizations ofthe individual index parameters to account for fines contents and stress conditions. When
using the seismic cone penetration tests, V s measurements can be easily coupled with CPT directly to obtain
qt/G o values.

0.6
't~drus and Stokde, 2000)
0.5
4 ______ ~ _____ ~ __ ~ :
- '_ _ _ ~ ~ ___ "!!35:"'0 S5 _____
.'F-:_. _h._____
~ ~ ~ ~t ~% .....: _ _. __ ____
Fe ____ ~ ~ ~ ~ ~ ~ ~ ~

I j:' " !

,..
o::
0.4 1
o::
u
o 0.3
o::
Cf)
u S Olneyvillle Silt
0.2
A Fanner's Market Silt
• Niigata Sand
0.1 ~ Toyoura Sand
Q Mai Liao Sand
o
100 150 200 250

Figure 20. The CRR-V. 1 correlation proposed by Andrus and Stokoe (2000)
and other sands/silts (after Baxter et al., 2008).

--Holocene
- - - Pleistocene
Q Holocene:Liquefaction

o Holocene:No liquefaction
0.8
• Pleistocene:Liquefaction
o Pleistocene:No liquefaction
o I

•....:;. ....l·.

~ 0.6
u
.... .. "
," ,
, o
o • o ~( • • ,
o:: • • o ( -,. ,.. .~. []J ........; •...........
.. f .
.. .
(f)
• ...
o
..,.... .

U 004
\ ••", .. .? •••..
~ljt. (§ld/ o o b o
.. , . . • la.. ...,
",". ~. , .... ~~ o

.~o"~J~~~
. --~ ~ ~
0.2
o b o
o .0
o
o 0.1 0.2 0.3 004
q¡lG o

Figure 21. The CRR-q/G o correlations reported by Roy (2008).

82
 GEOTECHNICAL ANO GEOPHYSICAL SITE CHARACTERIZATlON ORIENTEO TO SEISMIC ANALYSIS

Ishihara and Harada (2008) analyzed the correlations between SPT and CPT results and their relationship
with the ratio of effective horizontal stress to vertical stress (K). The CRR values were estimated from D.r
The penetration resistance values from SPT and CPT were based on calibration chamber tests. Figure 22
shows the comparison ofthe CRR-q¡ correlations derived for three clean sands: Toyoura sand (average grain
size, Dso = 0.20mm), Da Nang sand (D so = l.13mm) and Monterey sand (D so = 0.37mm), for K=0.5. The
correlations by Robertson & Wride (1998) and AIJ (2001) are also included for reference. For these three
types of clean sands, the CRR-q¡ correlations can deviate significantIy from the published curves and among
themselves. Similarly, discrepancies can also be found among the CRR-N correlations derived for Toyoura
sand and those reported by Youd et al. (2001) and JRA (1996).

0.8
------- AIJ
_.- Robeltson & Wride, 1998
--Toyoura sand
0.6 JII'
---=UaNarigsarid -
--Mont$ry sand

riü 0.4

0.2

o
o 5 10 15 20 25
qt, MPa

Figure 22. Comparison of the CRR-qt correlations derived for three clean sands
(after Ishihara and Harada, 2008).

Liquefaction Potential Assessment for Sands with Fines

Natural sand deposits ofien contain various amounts offines (silt and clay size particles). It has been reported
that most cases of earthquake-induced liquefaction have actually occurred in silty sands (Yamamuro and
Covert, 2001). Researchers have generally agreed that as fines contents exceed 5%, relative density ceased
to be a reliable index to predict liquefaction potential (Seed et al., 1985 and Ishihara, 1993). For fine grained
soils, the cyclic resistance correlates well with the void ratio, where a lower void ratio corresponds to greater
cyclic resistance (Ishihara, 1996). There is stiU a lack of consensus as to what role the fines content plays
in relation to liquefaction.

When the CRR of a sand with fines is determined through in situ index test using simplified procedure,
the situation is more complicated as theavailable correlations are empirically derived mainly from field
observations of soil behavior following earthquakes. Although different in magnitude andlor format, most
available CRRs based on in situ index test value correlations for cohesionless silty sands suggest that a
given index test value should correspond to a higher CRR as fines content increases. Alternatively, the in
situ index test value should be increased to obtain an equivalent clean sand value. For the Vs method by
Andrus and Stokoe (2000), the adjustment in the CRR-V S1 correlations is included in Figure 20. The SPT
penetration resistance (Idriss and Boulanger, 2006) is increased to an equivalent clean sand value, (N 1)60CS
according to FC (in percent) as:

83
 1
5ih (nlmno1:CrlOi Cc:r¡i
1
l

(N) = (N) + ~ (N1 )60 J,

I 60 es I 60 (lO) ;

"'(NI)60 ~ exp(1.63 + FC + 0.1 - (FC + 0.1 )

9.7 15.7 2)
(11 )

Robertson and Wride (1998) suggested that for CPT the equivalent c1ean sand value should be deterrninedl
based on le' using the following relationship:

Qtn,cs = KcQtn (12)

For le S 1.64 Kc = 1.0 ,.. (13)

For le > 1.64

Ke = -0.403/: + 5.58/; - 21.63/; + 33.75/c -17 .88 (14)

The adjustment to account for the fines content can impact significantly the outcome ofliquefaction potential
assessment. DespÜe this significance, little explanation has been offered to justify the consideration offines
content effects (lshihara, 1993 and Ydud et al., 2001).

Figure 23 shows a series of qt profiles in dry and saturated specimens from laboratory calibration tests in
MLS reported by Huang et al. (2004). For FC of 15%, the qt profiles in dry and saturated specimens are
essentially identical upon reac,hing a stabilized value, indicating that the cone penetration is drained. When
FC exceeds 30%, the qt in saturated specimens is significantly lowered than that in dry specimens. Similar
partial drainage phenomena have also been reported by Campanella et al. (1981) for CPI in c1ayey silt.

q¡,MPa

O 5 10 15 20 O 1 2 3 4 5 O 1 2 3

~ 200
,
~ ! §
~
~ 400 __ ._..1 _____ .1. _ _ _

(D ,
c.
'­
o
.I:: ,, "
"

g. 600 ----}-cW%­
o
-sat
,
-dry
, ,
800

Figure 23. Measured qt profiles from CPT calibration tests in MLS at different fines contents

(after Huang and Hsu, 2004)

By comparing the CRR and qt from CPT calibration tests in reconstituted specimens with comparable fines
contents, density and stress states, it was possible to verify the CRR-Qtn correlation by direct comparisons

84
 GEOTECHNICAl AND GEOPHYSICAl SITE CHARACTERIZATlON ORIENTED ro SEISMIC ANAlYSIS

for MLS as shown in Figure 24 (Huang et al., 2005). The CRR values were determined based on cyclic
strength obtained from a series of cyclic triaxial tests. The results as shown in Figure 24 indicate that the
fines content adjustment becomes significant only when the fines start affecting the drainage conditions
in CPT and thus result in a group of data points with distinctly lower Qtn' A similar explanation has been
postulated by Thevanayagam and Martin (2002). The laboratory study in MLS seems to suggest that a more
effective Qtn adjustment scheme should be based on CPT drainage conditions rather than fines content.

Additional CRR- Qtn data points based on field CPTU and cyclic triaxial tests on undisturbed samples are
also included in Figure 24. The data compiled by Tokimatsu et al. (1995) were collected from 6 test sites,
where soil samples were retrieved by freeze sampling. The fines contents varied from <1 % to as much as
30%. According to Tokimatsu et al. (1995), there was a unique CRR-Qtn correlation irrespective of fines
content. For a given Qtn' the dispersiveness in CRR was attributed to changes in K and soil type, represented
as a function of minimum void ratio.

In 1ight of the aboye findings, Huang et al. (2005) suggested that a pore pressure dissipation test during
CPTU may be used as a reference to scale the amounts of fines content adjustment. For the CPTU in
saturated specimens inc1uded in Figure 23; a dissipation test was conducted in the chamber (at respective
depths of 125,325 and 375mm for FC of 15, 30 and 50%). In a dissipation test, the cone penetration was
suspended and the dissipation of the excess pore pressure induced by cone penetration was monitored
until its full dissipation. When the FC exceeded 30%, there was a distinct increase of qt at the start of the
subsequent push. The qt setup increased furtl1er as the FC reached 50%. This phenomenon referred fo as the
qt setup was also reported by McN eilan and Bugno (1984) in their experience of CPT in offshore Galifornia
silts. The reason for setups is that partial drainage caused a lowered qt due to pore pressure accumulation.
The dissipation of pore pressure increases soil strength against cone penetration and generates the setup.
There is no obvious qt setup for the case ofFC= 15% in Figure 23, a result that is consistent with the fact that
CPT is drained as mentioned aboye.

0.6 (0.82) ... .& (0.89)

¡ F4

MLS FC,% K
• o 0.5
(Robe~on & Wride, 1998)
or (Stark & Olson, 1995) • 15
30
0.5
0.5
Á
FC=35% FC=15%
o o 1
0.4 t o
t.
15
30
a:: FC,% K
a:: YLS
u
es • 18 0.5

a:: a 43 0.5
(J)
11 89 0.5
u
Tokimatsu el al. FC, % K
¡
0.2 liI <1 0.5

• <1 1

•
Á
1-10
1-10
0.5

El >10 0.5
elean sand JI. >10 1

o +I----~----,-----~---,----_T-----
o _: 100 200 300 400
Qtn

Figure 24. Laboratory and field calibrations of CRR-Qtn correlations.

85
A series of CPTU using a standard cone (cone cross sectional area= 10 cm2) penetrating at 20mm/sec (the
standard CPTU rate), a Iarge cone (cone cross sectional area= 15cm2) penetrating at 20mm/sec (the large
CPTU), and a standard cone penetrating at 1mm/sec (the slow CPTU) were conducted at the Yuan Lin test
site. The rate of consolidation for soil surrounding a cone tip is inversely proportional to the square of
the cone diameter (Robertson et al., 1992). Therefore, changing the cone diameter can also duplicate the
effects of penetration rateo The pore pressure element was located immediately behind the cone tip, at the u2
position. Profiles ofCTPU results that include friction ratio, Rf (=~/qt xl00%) from tests at Yuan Lin site
are shown in Figure 25. The results indicated no significant differences in qt among three types of CPTU,
considering drastic differences in cone size and/or penetration rateo The slow CPTU was conducted at depth
leve1s where Laval samples were taken. The u 2 values from large CPTU were most1y identical to those
from the standard CPTU. The u2 in slow CPTU matched well with the hydrostatic pressure uo ' indicating
that 1mmlsec was slow enough to allow the penetration induced pare pressure to fully dissipate and reach
equilibrium in most parts with the surrounding hydrostatic pressure.
)J'

q,. MPa u2 • kPa R.. %

o 5 10 15 250 o 250 500 750 o 4 8
o

10

t(J)
o

15

Can e area Penetratian rate

cm 2 mm/sec
20 i--- 15 20
-10 20
-10
25

Figure 25. CPTU profiles from Yuan Lin site (after Huang, 2009).

The standard CPTU was coupled with dissipation tests at a test site in Yuan Lin, Taiwan. The results in
terms of qt profile are plotted in Figure 26 along with fines contents (from tests on SPT samples). The
comparison between Figures 23 and 26 allows the change in qt and its relationship with pore pressure
dissipation tests to be visualized. The effects of partíal drainage for CPTU in MLS were demonstrated by
the presence of significant setups following a pore pressure dissipation test as shown in Figure 23. The field
CPTU was close to drained conditions with essentially no signs of qt setup, even when the fines contents
reached as high as almost 100%. The suspension of cone penetration in field pore pressure dissipation test
caused a sharp decrease in qt and follo,Wed by a resumption of the original qt at the start of the subsequent
push. The drastic differences between CPTU in laboratory prepared, well mixed silty sand and natural silt/
sand in the fieId are likeIy due to the heterogeneity existed in natural soil. It is believed that the presence of
c1osely-spaced free draining sand Iayers made the field CPTU behave as a drained test in a siIty soil mass.
At much wider range offines contents, the lateral spread ofCRR-Qtn data points based on tests in Yuan Lin
Soil (YLS) shown in Figure 24 was less than those from tests using the reconstituted MLS specimens or
suggested by the avaiIabIe CRR-Qtn correlations. If differences in fines content are viewed as those of soil
types, it can also be concluded that the fines content effects are Iess significant than soil types for the case
ofYLS.

86
 GEOTECHNICAL ANO GEOPHYSICAL SITE CHARACTERIZATlON ORIENTED 10 SEISMIC ANALYSIS

Using bender elements, the V s can be measured on the same soil specimen of cyclic shearing test. The
CRR-V. l correlation can thus be conveniently calibrated completely based on laboratory tests (Huang et
al., 2004 and Baxter et al., 2008). Figure 20 shows CRR-VSI data points compiled by Baxter et al. (2008)
that include clean sands (Toyoura and Niigata sand), silty sand (Mai Liao sand with 0% ~ FC ~ 50%) and
non-plastic silt. The silt specimens include those from undisturbed block samples, split-barrel samples and
reconstituted samples by a modified moist tamping method. The result shows that the effects of soil type
on CRR-Vsl correlation overshadow those offines content, sample preparation methods and applications of
pre-shearing or pre-stressing.

q" MPa Fe, %

o 2 4 6 8 10 o 20 40 60 80 100

o ~::: . ¡: ¡;:: :!
--
,
~:. :.
I
.~
".
3 ....---. --:q

. "'. ~
6 ''i}­
E
t
Q)
... ¡ .,.
o
9 -.-----~---+-.--_.

1M:'" •
¡ : ; ¡
12 -.-:9;-----:----­
: :

~fi'
15 <T
-.. --­
Figure 26. Enlarged qt and fines content profiles from YLS site (after Huang, 2009).

UNDISTURBED
,
SAMPLING IN GRANULAR SOILS

Attempts oftaking high quality samples of cohesionless soi1s from be10w ground water tab1e can be traced
back by at least half a century (Singh et al., 1982). Challenges involved in taking good quality sand samples
include prevention ofthe 10ss ofsample during withdrawal and damaging soil structure during transportation.
These challenges are formidable unless the samples are taken near the ground surface or by block sampling.
Yoshimi et al. (1977) is believed to be the first among the more recent attempts in developing practical
procedures of ground freezing and dry coring for sand sampling. A column of sand is frozen in situ and then
cored out ofthe ground surface. Researchers from Japan and North America have generally considered in
situ ground freezing (Hofmann et al., 2000) and coring to be a superior method for obtaining undisturbed
samples of sand.

Driven by the demand in high-tech industry, the cost of liquid nitro gen continues to decrease. With a much
lower temperature (-196°C), the efficiency and practicality of using liquid nitrogen for ground freezing can
be much improved in contrast to the use of brine (-30°C). Provided that drainage is not impeded and that
changes in void ratio are minimized during freezing, the in situ structure can be reserved. Studies have
indicated that this structure preservation is possible if free drainage is allowed in at least one direction
during freezing (Singh et al., 1982). The reservation of soil structure is further enhanced if freezing is
conducted under a confining stress (Yoshimi et al., 1977).

87
 _[!¡-,I";;f:;':np

For silty sands, especially when fines contents are high, drainage can be significantly constrained in the
field. Ground freezing can cause void ratio changes due to frost heaving. It is possible to retrieve samples in
granular soíls by pushing a piston tube sampler under ambient temperature. However, the friction between
the sampling tube and the sUITounding soil can be excessive when fines contents are 10w. The sampling by
pushing tends to loo sen dense sand and densify loose sand (Hofmann et al., 2000). Good quality sampling
is possible iffines contents are high (Bray and Sancio, 2006) or samples are taken near the ground surface
(H0eg et al., 2000) with either a piston tube sampler or by block sampling (Baxter et al., 2008).

Huang and Huang (2007) reported the use of Laval sampler to obtain high quality silty sand samples with
fines contents ranging from 18 to 89% at Yuan Lin test site. The Laval sampler as schematically described
in Figure 27 was developed at Laval University (La Rochelle et al., 1981), originally for taking high quality
samples in sensitive c1ay. The sampler was made of two main parts; a sampling tube and an overcoring
tube. To take a sample, the drill ríg pushed the sampling tube into the bottom ofthe borehole while rotating
the overcoring tube. N o freezing was applied in the ground. The bottom of sampliIl!: tube was protruded at
20mm ahead of the steel teeth and cutters. During penetration, the head valve was kept open to allow drill
mud circulation and thus removal of soil cuttings. The Laval sample can be 450 to 550 mm long. After
a waiting period of 5 to 30 minutes, the head valve was c10sed and the bottom of the sample sheared by
rotating the inner rod. The sample was then retrieved to the ground surface.

B.Q. wire line

drilling rod
Collar
Locking pin
Internal guide
//~
Internal guide
I
----- Overcoring tube
/;.r

,g
I

~I
Steel (e,th and~ SectionA-A

Carbon steel ZW pipe

Demensions in mm

_-=----=---c~
iy
J:$--7- Steel teeth and cutters

Figure 27. Schematic view of the modified Laval sampler (modified from La Rochelle et al., 1981).

The Laval samples with FC>30% were cut into 120 to 180mm long segments and sealed in the field,
without freezing. The samples taken from soillayers with low fines contents (FC<30%) remained in the
sampling tube and kept vertical until it was completely frozen. The soil along with the sampling tube was
placed in a Styrofoam lined wooden box and gradually frozen from top of the sample by dry ice at -SOúC.
A backpressure equal to the water head within the sample was appl1ed by means of nylon tubing connected
88
 GEOTECHNICAL ANO GEOPHYSICAL SITE CHARACTERIZATlON ORIENTEO TO SEISMIC ANALYSIS

to the bottom of the sample to ensure that no water can drain under gravity. The bottom drainage and
backpressure assured that pore water drainage only due to water volume expansion during freezing (Konrad
et al., 1995). The frozen samples were stored in a freezer during shipping and laboratory storage until the
time of shearing test.

A specially designed coring device was used to cut 70rnm diameter triaxial specimens from the frozen
Laval sample, kept at -80°C by dry ice (Huang and Huang, 2007). The specimen was then placed in the
triaxia1 cell under a confining stress and thawed following the procedure suggested by Hofmann (1997). An
important advantage of Lava1 sampler is its large size. Four 70rnm diameter tri axial specimens can easily
be cored from a single Laval sample at the same depth level. The number of specimens is ideally suited
for the determination of a CRR-Ne curve through cyclic triaxial tests. As shown in Figure 28, the values of
Vs1 taken from the triaxial test specimens feH within the general range offield measurements near the Laval
sampling locations, indicating a reasonable quality ofthe soil samples.
.,.
Huang et al. (2008) reported the use of a gel-push sampler to recover high quality samples in silty sands
at a test site in Kao Hsiung of Southem Taiwan, where the fines contents varied from 5 to over 60%. The
gel-push sampler developed in Japan (Tani and Kaneko, 2006) was modified from a 75rnm Osterberg piston
sampler (also known as a Japanese sampler) as schematically shown in Figure 29. The sand sample was
obtained by pushing the gel-push sampler under ambient temperature as typically done for piston sampling
in clays. A water soluble polymeric lubricant (gel) was injected from the sampler shoe to lubricate and
alleviate friction exerted on the sampling tube as it was pushed into the sand. A shutter located at the
tip of the samp1er remained open during pushing but forced into a closed position at the end of pushing.
The closed shutter prevents the sample from falling during withdrawal. Upon withdrawal of the sampling
tube aboye ground, the ends of the tube were sealed and waxed. No freezing was applied for the sample
preservation. An accelerometer was attached to the sampling tube where the acceleration readings were
continuously recorded during shipping.

VS1 ' mIs

o 100 200 300
o

5 +----------;..........,¡."

10
E
..c
o..
a¡
o 15 ---~-------T-- J¡¡., ij t»>.
-SCPTU
----- POS Logging
20 iii lS,FC=18%
A. LS, FC=43%

25
e LS,FC:89%

Figure 28. Comparison of laboratory and field Vs1 measurements in YLS with varying fines content
(after Huang and Huang, 2007).

89
 ;, ¡ :c.:¡-¡"07jono: Con l;-,"-'::;' .' LOI [i'-!\_j'_ -~"".iÍ'.':-~ Cec-¡~~(.; ¡, ',;1........;; i..:r ,;;';i! jeE'[ ,c',l

water
injection

;JI'

Figure 29. Schematic views ofthe gel-push sampler (after Huang et al., 2008).

The soil sample extruded out ofthe gel-push sampler was trimmed to a diameter of70rnm to fit the triaxial

testing device and remove a shell of soil that was impregnated by the gel during field sampling. Figure 30

compares the V. measurements on cyclic triaxial test (CTX) specimens using bender elements and those

from the field seismic cane penetration tests (SCPTU). For the most part, the laboratory Vs falls within or

close to the range of those from SCPTU at comparable depths.

Vs,mls
100 150 200 250 300

10

--SCPTU

13 +-----m-~b-- .... ---h CTX

E 16

~ID
o 19

22

25

Figure 30. Comparison between the Vs measurements from bender elements and thos~ from SCPTU
(after Huang et al., 2008).

90
 GEOTECHNICAl AND GEOPHYSICAl SIlE CHARACTERIZATlON ORIENTED TO SEISMIC ANAL YSIS

THE CRITICAL STATE APPROACH TO EVALUATE SOIL LIQUEFACTION

In contrast to the semi-empirical, field-based methods in simplified procedures, the critical state approach
to cyclic strength of granular materials has a sound theoretical framework. The critical state approach is
anchored to the state parameter (\ji). The state parameter is the void ratio difference between the current
state ofthe soil and the critical state at the same effective mean normal stress (P'). The more negative state
parameter corresponds to higher soil dilatancy in shearing. Figure 31 shows results from cyclic triaxial tests
on 13 sands compiled by Jefferies and Been (2006). Consistent with the correlation between state parameter
and soil dilatancy, the data show that the cyc1ic strength (CRR) increases as \ji becomes more negative. The
state parameter is soil fabric independent. The cyc1ic strength however, is soil fabric dependent. The effects
of soil fabric show up when specimens prepared by different methods that result in different cyclic strength
(Huang et al., 2004). The scatter of test data in Figure 31 is believed to have been caused by variations in
soil fabrics originated in different specimen preparation methods involved in the data ofF~ure 31 (Jefferies
and Been, 2006).

0.6

0.5 -1- ... ········t··············~ ...•............. L . ..

DA -+ .............. ~············f············· .,.. .

~0.3
Ü
···t··············:'··········-.,;·y··········+··········.....
~.~
0.2
,
+·············:·~·~·'V
,.1'-. ....
"
0.1 ,.~ ~
........ ....... .... : . . . . . . . . . . . . . . ..

0.1 o -0.1 -0.2 -0.3 -OA

State parameter, 'V

Figure 31. Cyclic strength as a function of state parameters for 13 sands

(after Jefferies and Beeo, 2006).

Traditionally, studies on the effects of fines on cyclic strength of granular soils have been based on density
state (Polito, 1999; Thevanayagam et al., 2002; and Rahman et al., 2008). It can be advantageous to extend
the idea of correlating cyclic strength wi,th state parameter into sands with fines. The strength, inc1uding
cyclic strength, is a function of density and stress states which are both covered in state parameter. Plots
between CRR and 'l' from tests from cyclic triaxial tests on reconstituted Mai Liao Sands(MLS) are
shown in Figure 32. The specimens were prepared by moist tamping (MT), dry deposition (DD) and water
sedimentation (WS) methods. The test data generally have coefficients ofcorrelations in excess ofO.75 with
their corresponding exponential curve fits. The range of state parameters shifts significantly towards the
positive side that reflects the compressive nature of MLS grains. For data of the same specimen
preparation group, those with high fines contents are clustered towards the positive side of the
state parameter axis. This is again a reflection of the higher compressibility associated with sand/
silt mixtures. For specimens with the same state parameter, CRR of MT specimens are the highest

91
and those ofDD specimens are the lowest. This is consistent with the earlier findings reported by
Huang et al. (2004) which indicated that for specimens with the same void ratio, CRR from MT
specimens are the highest and those from DD specimens are the lowest. Therefore, in addition to
the consideration of stress state, the use of state parameter has the additional advantage of reflecting
soil grain characteristics. These _are desirable features that the concept of density state lacks.

0.6
SPM fe' %
O MT O CRR=O.307*exp(-3.408'P
EB MT 15
Rl = 0.758
0_5 f- +
MT 30
O DD O
EE DD 15 E!l

0'[ ~ ;" ,.
DD 30
WS O ,
2u \l WS 15 e
o ,,/Itll/
'
1> WS 30 e
0.3 11 6/

+~
+
,';o-B
,,")Y'~ 'V
I

0.2 f- +'83, L!.l~ ¡jJ

4 ¡¡,.:t(~ o CRR=0.224*exp(-2.653'i')
-t' o R2 = 0.752
[>

O.ll~~--~~--~--~~--~~---L--~~~
03 0.2 0.1 o -0.1 -0.2 -03
State parameter, 'P

Figure 32. Correlation between cyclic strength and state parameter for MLS.

Attempts have been made to infer state parameter from in situ tests. By analyzing a series of calibration
chamber CPT data, Been et al. (1986; 1987) and Jefferies and Been (2006) demonstrated that the normalized
cone tip resistance Qp (= (qt-p)/p') in log scale has a linear relationship with 'V as shown in Figure 33. The
trend line has a simple exponential form:

Qp = kexp( -m1jJ) (15)

This empirically derived equation based on dimensional grounds was consistent with the cavity expansion
solution proposed by Carter et al. (1986). The values of k and m in Equation (15) correspond, respecti vely
to the intercept and slope of the trend lines in Figure 33. These values are sand-specific and as such are
functions of intrinsic properties of the sands. In addition, k and m should relate to slope of the critical the
state Ene (A).

The validity ofEquation (15) was challenged by Sladen (1989) who indicated that k and possibly m were
functions ofp'. Numerical framework for evaluating 'V from CPT has been developed (Shuttle and Jefferies,
1998; Ghafghazi and Shuttle, 2008). This framework was based on cavity expansion analysis and NorSand
numerical model. Based on their analysis and those of Carter et al. (1986), the stress level bias pointed out
by Sladen (1989) can be properIy addressed by treating k and m as functions ofrigidity index (Ir) defined
as Ir = G/p' o' Robertson (2010) proposed an empirical chart to estímate 'V based on Qtn and Fr as shown in
Figure 34. According to Robertson, the contours of\jl plotted in the Qtn -Fr space are very similar to those
ofQ tn,es .

92
 GEOTECHNICAl ANO GEOPHYSICAl SITE CHARACTERIZATION ORIENTEO TO SEISMIC ANALYSIS

1000
+Montery
·Tícino
....... Hokksund

*Otlawa
*Reid Bedford
+Hilton Mines
-i-Erksak 355/3
*Syncrude tailings
-Yatesville Silty Sand
-o. 100
+-Chek Lap Kok
~ *West Kowloon
ó=
11
el.
a
~;

10 I

-0.3 -0.2 -0.1

""

o 0.1
""
State para meter, '1'

Figure 33. The Qp-'1' trends for different sands (after Jefferies and Been, 2006).

State Pararneter, \V

r:J
I:f
~.
~.
UJ

~
j
~

1 10

NORMAUZED menON RAllO, Fr

Figure 34. Contours of", in Qto -Fr space (Robertson, 2010)

Taking advantage of Qp -\ji correlation, Jefferies and Been (2006) demonstrated the potential of expressing
the liquefaction boundary curve in terms of \ji for clean sands. Assuming a set of critical state parameters
expected for typical clean sands, a'vo = 100kPa and Ko =0.7, a series ofQp and \ji were computed using the
critical state based numerical framework and Equation (15). The \jI-based boundary curve that separates
liquefaction from no liquefaction is a simple exponential function as shown in Figure 35.

93
5th
~
¡ní~~!nQjjonu¡ C':";~-': "";;,''''::l-:nicoi i:ng:nt.':'O!iI'IQ

0.5

0.4 .............. ···f······~:-\i······i······l .......... , ................. .

, ,
",:
-+i
0.3
: ~ ~."
....•........... ..i
A
: 0" .:
o::
o:: -\. ~¡ ~\2:
ü .'" : ~ ~~, : O A
50.2 ···············'\:~·~~····;·¿.r··=···=··=··==
o:: ~ ~ . --.a .. O: O Liquefaction
(f)
Ü 1.l
... L.,.......~, t ° :0
¡fJ .t> : A
• Stark & Olson, 1995
. : : ASuzukietal., 1995
0.1 ........... ~ ... : ... ;.0 ..
No liquefaction :;
° Stark & Olson, 1995
o I O.03exp(-11'1') A Suzukietal.,1995

O'. -0.1 -0.2 -0.3 -0.4

State parameter, 'l'

Figure 35. Liquefaction boundary curve expressed in terms of", (after Jefferies & Been, 2006).

Figure 36 shows qt versus p' from a set of calibration chamber tests using Da Nang sand. For a given state
parameter, the dilatancy remains constant, qt should increase linearly with p'. The rate of qt increase with p'
is k exp( -mljJ) according to Equation (15). This justifies the use of a linear stress normalization of Qp' Da
Nang Sand is a uniformly graded clean quartz sand with distinct dilatant behavior. As shown in Figure 36,
there is a consistent correlation. between Q p and 'Psimilar to those ofFigure 33. Mai Liao Sand with 15%
offines, on the other hand is relatively compressible.
.
The q t is much more sensitive to stress than 'P. The
data points of qt versus 'P tend to cluster together despite of the variation in 'P as shown in Figure 37. This
phenomenon is also reflected in a poor correlation between Qp and'P as demonstrated in Figure 38.

qt,MPa
o 10 20 30 40 50
O.
State Parameter
('P)
40 • -0.040

80
~\\\~ <>
•
O
-0.097
-0.141
-0.184
~

~. (.; -0.228
-p..

120

160

200 !L ----------------------------~

Figure 36. Increase of qt with p' for Da Nang Sand.

94
 GEOTECHNICAL AND GEOPHYSICAl SITE CHARACTERIZATlON ORIENTED TO SEISMIC ANAlYSIS

<>

200

~
~

p.. 100
cf
'--'

<>
50
".

20~' __- L__- L__- L__- L__J -_ _J -_ _~_ _~_ _~~

-0.25 -0.2 -0.15 -0.1 -0.05 o
'P
Figure 37. Correlation between Qp and '1' for Da Nang Sand.

The critical state approach can have many advantages over the semi-empirical field observation based
·simplified procedures
.
in liquefaction potential analysis. The effects of confining stress on CRR (i.e., the K cr
effect) are inc1uded in the CRR-\ji correlation. The cone tip resistance, Qp is linear1y normalized with p' , and
thus avoids the potential error associated with the exponential stress normalization typically used to obtain
Qtn. As long as CPT remains drained, the fines content is part ofthe soil intrinsic properties (i.e., grain size,
gradation, mineralogy, interpartic1e frietion etc.). Their effeets are reflected in the critical state parameters
and eoefficients ofk and m. Potential eonfusion in the fines content adjustment, which is an integral part of
the simplified procedure, can thus be minimized if not avoided.

qt>MPa

O 5 10 15 20 25 30

O,
. .•
-

O' e
State Parameter
( 'fI )

~
,JL 0+ +e

~
•
o
ti)

00 ca
•
O

•<>
e
0.091
0.042
-0.001
-0.061
-0.127
-o."

2401­

3201­
• ••
• • <>. <>

4001L-------------------------------~

Figure 38. Increase of qt with p' for Mai Liao Sand.

95
[)"lh
I
:r,jefnct:On(l~ C":C;-,I :

500"--------------~------------~

~ I •
2001­
.',
,
,
~ lOO· l',

'el
50

,.
20,~ __ J __ __ L_ _- J____~__~__~__~~~

-0,2 -O.l O 0.1 0,2

'P
Figure 39. Correlation between Qp and '1' for
.
Mai Liao Sand.

Yu (2004) reported the use ofKD from DMT to infer state parameter as shown in Figure 40, By coupling
Figures 35 and 40, Marchetti (2010) experimented the possibility of relating CRR with KD via state
parameter. The result shown in Figure 41 indicates that the CRR inferred from K D via state parameter tends
to be umealistically low. If used for liquefaction potential analysis, the procedure would yield a rather
conservative result. Similarly, by coupling Figures 35 and 37, it is possible to establish the relationship
between CRR and Qp for Da Nang Sand as shown in Figure 42. A comparison with the correlation by
Robertson and Wride (1998) (assuming cr'h o = 0.50" vo ) also shows that the CRR-Qp correlation via state
parameter tends to be conservative.

16.1-------------------------------------------------.

•
....
Hokksund sand
Kogyuksand
Reid Bredford sand
12 *x Ticino sand
... ... ... ~
...
~
~ 8
:::.:::
~~ x I
x* •
4
*x

O~I ____L-__ ~ ____ ~ ____L __ __ L_ _ _ _ ~ _ _~L__ __ L_ _ _ _~_ _~_ _ _ __ L_ _~

-0,2 -0.15 -0.1 -0,05 O 0.05 0.1

State Pararneter, ':l'
Figure 40. Infer state parameter from KD of DMT (Yu, 2004).

96
 GEOTECHNICAl AND GEOPHYSICAl SITE CHARACTERIZATlON ORIENTED TO SEISMIC ANALYSIS

0.5 - - - - - - - " T " " T - - - - - - - - - ,

'-0

0.4

0.3
~ M~chetti, 1 ) 1 ~I
I ~evna
L l
Chameau. 1 91

~
U
0.2

0.1
,.
2 4 6 8 10
Horizontal Stress Index, K D
Figure 41. Correlations between CRR and KD ofDMT (after Marchetti, 2010).

0.6 ' - 0- - - - - - - - - - - - - - - - - ,

Robertson & Wride I

(1998) I

004 ~,'
I

~
U I
I

0.2

O 100 200
'"
OLI__~~___ L_ _~~_ __L__~~___L~
CRR = 0.03exp(-11 'P)

300 400 500

Qp
Figure 42. Correlations between CRR and Qp .

CONCLUDING REMARKS

Significant deveIopments have be en made in fieId testing, soil sampling and Iaboratory testing techniques
in the past few decades. Geotechnical and geophysical site characterization methods are compleinentary
to each other. Geophysical methods can provide soil stratigraphy, elastic/small strain soil stiffness, and
damping/attenuation characteristics. Nonintrusive geophysical methods are especially attractive for hard-to­
sample soils. Hybrid or multi-function in situ testing methods such as SCPTU or SDMT enable penetration
tests that provide information related to soil stiffness under different strain levels, hydraulic conductivity as
well as soil stratigraphy. Soil samples with reasonable quality can be retrieved under ambient temperature
using Laval or gel-push sampler. Using local strain measurements, deformation and damping characteristics

97
 G

that cover a wide range of strains can be measured in laboratory tests on a single specimen. With these
new developments, efficient and cost effective testing/analysis frameworks are available to perforrn
geotechnical and geophysical characterization for seismic analysis. Essentially all information as outlined
by Jamiolkowski and Lo Presti (1995) can be covered under these frameworks. The interpretation of in
situ index tests for soilliquefaction potential analysis remains empirical. It is imperative to calibrate these
empirical methods for local soils. This is especially true for the case of intermediate soils such as sands with
fines. The critical state approach appears promising as it has a sound theoretical basisand it is possible to
circumvent many of the problems associated with empírical interpretation of in situ tests, such as the stress
normalization of cone tip resistance. The method however, needs refinement especially in the context of
inferring soil cyclic strength from CPT.

AKNOWLEDGEMENTS

Parts ofthe research presented in the paper were funded by the National Sciertce Council ofTaiwan,
ROC under contracts 97-2221-E-009-128, 96-2221-E-009-005, 95-2221-E-009-202 and 94-2211­
E-009-043. Field sampling at Kao Hsiung test site was funded by Taiwan Construction Research
Institute, Taipei, Taiwan. The authors gratefully acknowledge their support.

REFERENCES

AlIen, N. F., Richart, F. E. Jr. an,d Woods, R. D. (1980). "Fluid wave propagation in saturated and nearly
saturated sands". Joumal ofGeotechnical Engineering, Vol. 106, No. GT3, pp. 235 - 254.
Amerícan Society for Testing and Materials (ASTM), ASTM Intemational, West Conshohocken, PA,
2003.
Andrus, R. D. and Stokoe, K.H., Il. (2000). "Liquefaction resistance of soils from shear-wave velocity".
Joumal of Geotechnical and Geoenvironmental Engineering, Vol. 126, No. 11, pp.1015 - 1025.
Architectural Institute of Japan (AIJ). 2001. Recornmendations for design ofbuilding foundations.
Baxter, C. D. P., Bradshaw, A. S., Green, R. A., and Wang, J. H. (2008). "Correlation between cyclic
resistance and shear-wave velocity for Providence silts". Joumal ofGeotechnical and Geoenvironmental
Engineering, Vol. 134, No. 1, pp. 37 - 46.
Been, K., Crooks, J. F. A., and Jefferies, M. G. (1986). "The cone penetration test in sands: part 1, state
parameter interpretation". Geotechnique, Vol. 36, No. 2, pp. 239 - 249.
Been, K., Jefferies, M. G., Crooks, J. F. A., and Rothenberg. L. (1987). "The cone penetration test in sands:
part Il, general inferenceofstate". Geotechnique, Vol. 37, No. 3,pp. 285 - 299.
BelIoti, R., Jamiolkowski, M., Lo Presti, D. C .P., and O'Neill, D. A. (1996). "Anisotropy of small
strain stiffness of ticino sand". Geotechnique, Vol. 46, No. 1, pp. 115 - 131.
Bray, 1. D. and Sancio, R. B. (2006). "Assessment ofthe liquefaction susceptibility offine-grainedsoils".
Joumal of Geotechnical and Geoenvironmental Engineering, Vol. 132, No. 9, pp. 1165 - 1177.
Burland, J. B. (1989). "Small is beautiful-the stiffne.ss of soils at small strains". Canadian Geotechnical
Joumal, Vol. 26, No. 4, pp. 499 - 516. .
Campanella, R. G., Robertson, P. K. and Gillespie, D. (1981). "In-situ testing in saturated silt (drained or
undrained?)", 34th Canadian Geotechnical Conference, Fredericton, New Brunswick.
Campanella, R. G., Robertson, P.K., and Gillespie, D. (1986). "Sesimic cone penetration test". Use of
in Situ Tests in Geotechnical Engineering, Proceedings, In Situ '86, ASCE Geotechnical Specialty
Publication No. 6, Samuel P. Clemence Ced.), Balcksburg, VA, pp. 116-130.
Campanella, R. G. and Robertson, P. K. (1988). "Current status of piezocone test". Proc. Intemational
Symposium on Penetration Testing, ISOPT-1, Orlando, Vol. 1, pp. 93-116, Balkema Pub., Rotterdam.

98
 GEOTECHNICAL AND GEOPHYSICAL SlrE CHARACTERIZArlON ORIENTED ro SEISMIC ANAlYSIS

Carniel, R., Malisan, P., Barazza, F., and Grimaz, S. (2008). "Improvement ofHVSR technique by wavelet
analysis". Soil Dynamics and Earthquake Engineering, Vol. 28, pp. 321 - 327.
Carter, 1. P., Booker, 1. R., and Yeung, S. K. (1986). "Cavity expansion in cohesive frictional soils".
Geotechnique, Vol. 36, No. 3, pp. 349 - 358.
Cavallaro, A, Grasson, S., and Maugeri, M. (2006). "Clay soil characteristics by the new seismic dilatometer
Marchetti test (SDMT)". Proc. 2nd International Conference on the Flat Dilatometer (DMT2006),
pp.261-268.
Cetin, K. O., and Isik, N. S. (2007). Probabilistic assessment of stress normalization for CPT data. Journal
ofGeotechnical and Geoenvironmental Engineering, Vo1.133, No.7, pp.887-897.
Cetin, K. O., Seed, R. B., Kjureghian, A D., Tokimatsu, K., Harder, L. F., Kayen, R. E., and Moss, R. E.
S. (2004). "Standard penetration test-based probabilistic and deterministic assessment of seismic soil
liquefaction potential". Journal of Geotechnical and Geoenvironmental Engineering, Vol. 130, No. 12,
pp. 1314-1340.
Dobry, R., Ladd, R. S., Chung, R. M. and Powell, D. (1982). Prediction ofpore water prersure buildup and
liquefaction of sands during earthquakes by the cyclic strain method. N ational Bureau of Standards
Building Science Series 138.
Dyvik, R. and Madshus, C. (1985). "Laboratory measurement of G max using bender element". Proc. AS CE
Annual Convention: Advance in the art oftesting spils under cyc1ic conditions, Detroit, pp. 186 - 196.
Fahey, M. (1998). "Deformation and in situ stress measurement". Proc. 1st International Conference on
Geotechnical Site Characterization (ISC-1), Atlanta, USA, pp.49-68.
Ghafghazi, M., and Shuttle, D. (2008). "Interpretation of sand state from cone penetration resistance".
Geotechnique, Vol. 58, No. 8, pp. 623 - 634 .
. Goto, S., Tatsuoka, F., Shibuya, S., Kim, Y~S., and Sat(), T., (1991). "A Simple gauge for local small strain
measurements in the laboratory". Soils and Foundations, Vol. 31, No. 1, pp. 169 - 180.
Harder, L.F., Jr., and Seed, H.B. (1986). Determination of penetration resistance for coarse grained soils
using the Becker harnmer drill. Rep. USB/EERC-86/06, Earthquake Engineering Research Center,
University of California at Berkeley.
H0eg, K., Dyvik, R. and Sandb:ekken, G. (2000). "Strength of undisturbed versus reconstituted silt and
silty sand specimens". Journal of Geotechnical and Geoenvironmental Engineering, Vol. 126, No. 7,
pp. 606 - 617.
Hofmann, B.A (1997). In situ ground freezing to obtain undisturbed samples ofloose sand for liquefaction
assessment. Ph.D. thesis, Department of Civil and Environmental Engineering, University of Alberta,
Edmonton, Alberta, Canada.
Hofmann, B. A., Sego, D. C., and Robertson, P. K. (2000). "In situ ground freezing to obtain undisturbed
samples ofloose sand". Canadian Geotechnical Journal, Vol. 126, No. 11, pp. 979 - 989.
Huang, Y. T, Huang, A. B., Kuo, Y C., and Tsai, M. D. (2004). "A laboratory study on the undrained strength
ofa silty sand from Central Western Taiwan". Soil Dynamics and Earthquake Engineering, Vol. 24, No.
9-10, pp. 733 - 743.
Huang, A B., and Hsu, H. H. (2004). "Advanced calibration chambers for cone penetration testing in
cohesionless soils". Proc. ISC-2 on Geotechnical and Geophysical Site Characterization, Porto, Portugal,
VoLl, pp.147-167.
Huang, A B., Huang, Y. T. and Ho, F. J. (2005). "Assessment of liquefaction potential for a silty sand in
Central Western Taiwan". Proc. XVI ICSMGE Osaka, pp. 2653 - 2657.
Huang, A B., and Huang, Y T. (2007). "Undisturbed sampling and laboratory shearing tests on a sand with
various fines contents". Soils and Foundations, Vol.47, No. 4, pp. 771 - 781.
Huang, A B., Tai, Y. Y, Lee, W. F., and Ishihara, K. (2008). "Sampling and field characterization of the
silty sand in Central and Southern Taiwan". Proc. 3rd International Conference on Site Characterization
(ISC-3), Taipei, Taylor & Francis, pp. 1457 - 1463.

99
Huang, A. B. (2009). "Lessons leamed from sampling and CPT in silt/sand soils". Intemational Conference
on Performance-Based Design in Earthquake Geotechnical Engineering - from case history to practice,
Tokyo.
Huang, A. B., Tai, Y Y, Lee, W. F., and Huang, Y T (2009). "Fie1d evaluation ofthe cyclic strength versus
cone tip resistance corre1ation in silty sands". Soi1s and Foundations, Vol. 49, No. 4, pp. 557-568.
Idriss, 1. M. and Boulanger, R. W. (2006). "Semi-empirica1 procedures for evaluating liquefaction potential
during earthquakes". Soi1 Dynamics and Earthquake Engineering, Vol. 26, pp. 115 - 130.
Ishihara, K. (1993). "Liquefaction and flow failure during earthquakes". Geotechnique, Vol. 43, No. 3, pp.
351 - 415.
Ishihara, K. (1996). "Soil behaviour in Earthquake Geotechnics". Oxford University Press Inc., New York.
Ishihara, K. and Harada, K. (2008). "Effects of lateral stress on relations between penetration resistances
and cyclic strength to liquefaction". Proc. 3rd Intemational Conference on Site Characterization (ISC­
3), Taipei, pp. 1043 - 1050.
Iwasaki, T. and Tatsuoka, F. (1977). "Effects of grain size and grading on dynamic sfiear modulus ofsands".
Soils and Foundations, Vol. 17, No. 3, pp. 19 - 35.
Jamiolkowski, M., Lo Presti, D. C. F., and Pallara, O. (1995). "Role of in situ testing in geotechnica1
earthquake engineering". Proc. Third Intemationa1 Conference on Recent Advances in Geotechnica1
Earthquake Engineering and SoU Dynamics, State of the Art 7, St. Louis, Missouri, April 2-7, 1995,
vol. II, pp. 1523 - 1546.
Japan Road Association (JRA) (1996). Specifications for Highway Bridges, Part III, Seismic designo
Jardine, R. J. (1992). "Sorne observations on the kinematic nature ofsoi1 stiffness". Soi1s and Foundations,
Vol. 32, No. 2, pp. 111 - 124.
J efferies, M. G., and Been, K. (2006). "Soilliquefaction - A critical state approach". Taylor·and Francis.
Juang, C. H., Jiang, T., and Andrus, R. D. (2002). "Assessing probability-based methods for liquefaction
potential evaluation". Jouma1 of Geotechnica1 and Geoenvironmenta1 Engineering, Vol. 128, No. 7, pp.
580 - 589.
Juang, C. H., Fang, S. Y and Khor, E. H. (2006). "First-order reliability method for probabilistic 1iquefaction
triggering analysis using CPT". Jouma1 ofGeotechnica1 and Geoenvironmental Engineering, Vol. 132,
No. 3, pp. 337 - 350.
Kause1, E. and Roesset J. M. (1981). "Stiffness matrices for 1ayered soils". Bulletin of the Seism010gical
Society of America, Vol. 71, NO. 6, pp. 1743 - 176l.
Kokusho, T (1980). "Cyclic triaxia1 test of dynamic soi1 properties for wide strain range". Soils and
Foundations, Vol. 20, No. 2, pp. 45 - 60.
Konrad, J. M., St-Laurent, S. Gi1bert, F., and Lerouei1, S. (1995). "Sand sampling below the water table.
using the 200 mm diameter Laval sampler". Canadian Geotechnica1 Jouma1, Vol. 32, pp. 1079 - 1086.
Kotani, N., Ishihara, K., Imamura, S., Hagiwara, T, and Tsukamoto, Y (2002). "Centrifuge experiments on
group pi le effects under lateral !ll0vement ofliquefied ground". ProC. JSCE Annua1 Conference, 57, pp.
1243 - 1244 (in Japanese).
La Rochelle, P., Sarrai1h, 1., Tavenas, F., Roy. D, and Larouei1, S. (1981). "Causes ofsamp1ing disturbance
and design of a new samp1er for sensitive soi1s". Canadian Geotechnical Joumal, Vol. 18, No. 1, pp.
52 - 66.
Landon, M. M., DeGroot, D. 1., and Sheahan, T. C. (2007). "Nondestructive sample quality assessment
of a soft c1ay using shear wave ve1ocity". Joumal of Geotechnical and Geoenvironmenta1 Engineering,
Vol. 133, No. 4, pp. 424 - 432.
Marchetti, S. (1980). "In situ tests by F1at Di1atometer". J ouma1 of Geotechnica1 Engineering, Vol. 106, No.
GT3, pp. 299 - 32l.
Marchetti, S. (1982). "Detection of 1iquefiab1e sand 1ayers by means of quasi-static penetration probes".
ESOPT II Amsterdam, Vol. 2, pp. 689 - 695.

100
 GEOTECHNICAl AND GEOPHYSICAl SITE CHARACTERIZATION ORIENTED TO SEISMIC ANAlYSIS

Marchetti, D., Marchetti, S., Monaco, P., and Totani, G. (2008). "Experience with Seismic Dilatometer
(SDMT) in various soil types". Proc. 3rd International Conference on Geotechnical and Geophysical
Site Characterization (ISC-3), Taipei, Taiwan.
Marchetti, S. (2010). "Sensitivity of CPI and DMI to stress history and aging in sands for liquefaction
assessment". Proc. CPT 10 (www.cptlO.com). Huntington Beach, California.
Mayne, P. W., Christopher, B. R., and DeJong, J. (2002). Subsurface Investigations - Geotechnical
Site Characterization Reference Manual. U.S. Department of Transportation, Federal Highway
Administration.
Mayne, P.W., Coop, M.R., Springman S.M., Huang, A.B., Zornberg, J.G. (2009). "Geomaterial behavior
and testing". Proc. XVII ICSMGE, Alexandria, Egypt, Vol. 4, pp. 2777 - 2872.
McNeilan, T. W., and Bugno, W. T. (1984). "Cone penetration test results in offshore California silts".
Symposium on Strength Testing of Marine Sediments. Laboratory and In-Situ Measurements, San
Diego, ASTM STP 883, pp.55-7l.
Monaco, P., Marchetti, S., Totani, G., and Calabrese, M. (2005). "Sand liquefiability a~essment by Flat
Dilatometer Test (DMT)". Proc. XVI ICSMGE, Osaka, Vol. 4, pp. 2693 - 2697.
Moss, R. E. S., Seed, R. B. and Olen, R. S. (2006). "Normalizing the CPT for overburden stress". Journal of
Geotechnical and Geoenvironmental Engineering, Vol. 132, No. 3, pp. 378 - 387.
Nakamura, Y. (1989). "A method for dynamic characteristic estimation of subsurface using microtremor
on the ground surface". Q Rep RaiIway Tech Res Inst Vol. 30, No. 1, pp. 25 - 33.
Nigbor, R. L. and Imai T. (1994). "The suspension P-S velocity logging method". Geophysical
Characteristics of Sites, ISSMFE, Technical Committee 10 for XIII ICSMFE, International
Science Publishers, New York, pp. 57-63.
NRC (2000). "Seeing into the Earth". Committee for Noninvasive Characterization of the Shallow
Subsurface for Environmental and Engineering Applications, P.R. Romig, Chair.
Park, C. B., Miller, R. D., and Xia, J. (1999). "Multichannel analysis of surface waves". Geophysics, Vo.l.
64, No. 3, pp. 800 - 808.
Po lito, C. P. (1999). The effects of nonplastic and plastic fines on the liquefaction of sandy soils. Ph.D.
Thesis, Virginia Polytechnic Institute and State University, Blacksburg, Virginia.
Rahman, M. M., Lo, S. R., and Gnanendran, C. T. (2008). "On equivalent granular void ratio and steady
state behavior of loose sand with fines". Canadian Geotechnical J ournal, Vol. 45, pp.1439 - 1456.
Reyna, F. and Chameau, J. L. (1991). "Dilatometer based liquefaction potential of sites in the Imperial
Valley". Proc. 2nd Int. Conf on Recent Advances in Geot. Earthquake Engrg. and Soil Dyn. St. Louis,
pp. 385 - 392.
Richart, F. E., Hall, J. R., and Woods, R. D. (1970). "Vibrations of SoiIs and Foundations", Prentice
Hall, Eglewood Cliffs.
Robertson, P. K. and Campanella, R. G. (1986). "Estimating liquefaction potential of sands using the flat
plate dilatometer". Geotechnical Testing Journal, Vol. 9, No. 1, pp. 38 - 40.
Robertson, P. K., Sully, J. O., Woeller, D. J., Lunne, T., Powell, J. J. M., and Gillespie (1992). "Estimating
coefficient of consolidation from piezocone tests". Canadian Geotechnical Journal , Vol. 29, pp. 539 ­
550.
Robertson, P. K. and Wride, C. E. (1998). "Evaluating cyc1ic liquefaction potential using the cone penetration
test". Canadian Geotechnical Journal, Vol. 35, pp. 442 - 459.
Robertson, P. K. (2009). I"nterpretation of cone penetration tests - a unified approach". Canadian
Geotechnical Journal, Vol. 46, pp. 1337 - 1355.
Robertson, P. K. (2010). "Evaluation of flow liquefaction and liquefied strength using the c~ne penetration
test". Journal of Geotechnical and Geoenvironmental Engineering, Vol. 136, No. 6, pp. 842 - 853.
Robertson, P. K. (2010). "Estimating state parameter and friction angle in sandy soils from CPT". Proc.
CPTlO (www.cptlO.com). Huntington Beach, California.

101
 I
:r-~Iemcj io'-, - ,::'¡-:!<::-;'Ü¡ ',(

Rollins, K., Evans, M., Diehl, N. and Daily, W. (1998). "Shear modu1us and damping relationships for
gravels". Journal of Geotechnical and Geoenvironmental Engineering, Vol. 124, No. 5, pp. 396 - 405.
Roy, D. (2008). "Coupled use of cone tip resistance and small strain shear modulus to assess liquefaction
potential". Journal ofGeotechnical and Geoenvironmenta1 Engineering, Vo1.134, No. 4, pp. 519 - 530.
Seed, H. B. and Idriss, I. M. (1971). "Simplified procedure for evaluating soi11iquefaction potential". Journal
ofSoil Mechanics Division, Vol. 97, No. SM9, pp. 1249 - 1273.
Seed, H. B. (1979). "Soil1iquefaction and cyclic mobility evaluation for 1evel ground during earthquakes".
Journa1 ofthe Geotechnical Engineering Division, Vol. 105, No. GT2, pp. 201 - 255.
Seed, H. B., Tokimatsu, K., Harder, Jr. L. F. and Chung, R. (1984). The influence ofSPT procedures on soil
liquefaction resistance eva1uations, Report No. UCB/EERC-84115, University of California, Berke1ey.
Seed, H. B., Tokimatsu, K., Harder, L.F., and Chung, R.M. (1985). "Influence of SPT procedures in soil
liquefaction resistance evaluation". Journal of Geotechnical Engineering, Vol. 111, No. 2, pp. 1425
-1445.
Scho1ey, G. K., Frost, 1. D., Lo Presti, D. C. F., and Jamio1kowski, M. (1995). "Instrum¿ntation for measuring
local strains during triaxial testing of small specimens". ASTM Geotechnical Testing Journal, Vol. 18,
No. 2, pp. 137 - 156.
Sheriff, R. E. and Geldart, L. P. (1995). "Exploration Seismology", 2 nd edition. Cambridge University
Press lnc.
Shuttle, D. A. and Jefferies, M. G. (1998). "Dimensionless and unbiased CPT interpretation in sand".
International Journal ofNumerical and Ana1ytica1 Methods in Geomechanics, Vol. 22, pp. 351 - 391.
Singh, S., Seed, H. B., and Chan, C. K. (1982). "Undisturbed sampling of saturated sands by freezing".
Journal ofthe Geotechnical Engineering Division, Vol. 108, No. 2, pp. 247 - 264.
Sladen, 1. A. (1989). "Prob1ems with interpretation of sand state from cone penetration test". Geotechnique,
Vol. 39, No. 2, pp. 323 - 332.
Stark, T. D., and Olson, S. M. (1995). "Liquefaction resistance using CPT and field case histories". Journal
ofGeotechnical Engineering Division, Vol. 121, No. 12, p. 856 - 869.
Stokoe, K. H. II, and Nazarian, S. (1985). "Use ofRayleigh waves in liquefaction studies". Proceedings,
Measurement and Use of Shear Wave Velocity for Evaluating Dynamic Soil Properties. Geotechnical
Engineering Division, ASCE, pp. 1 - 17.
Stokoe, K H., JI, and Santamarina, J. C. (2000). "Seismic-wave based testing in geotechnical
engineering". Plenary Paper, Intemational Conference on Geotechnical and Geological . ir
Engineering, GeoEng 2000, Melboume, Australia, pp.1490 - 1536.
Stokoe, K H., JI, Joh, S. H., and Woods, R. D. (2004). "Sorne contributions ofin situ geophysical
measurements to solving geotechnical engineering prob1ems". Proceedings, ISC-2 on Geºt~chnical andf
Geophysical Site Characterization;Millpress, Rotterdam. - . -
Susuki, Y., Tokimatsu, K., Koyamada, K., Taya, Y., and Kubota, Y. (1995). "Field correlation of soil
liquefaction based on CPT data". Proc. International Symposium on Cone Penetration Testing, CPT'95,
Vol. 2, pp. 583 - 588.
Tani, K, and Kaneko, S. (2006). "Method of recovering undisturbed samples using water-soluble thick
polymer". Tsuchi-to-Kiso, Journal of JapaneseGeotechnical Society, Vol. 54, No. 4, pp. 145 - 148. (in
Japanese)
Thevanayagam, S., Shenthan, T., Mohan, S., and Liang, J. (2002). "Undrained fragility of clean sands, silty
sands, and sandy silts". Journal of Geotechnical and Geoenvironmental Engineering, Vol. 128, No. 10,
pp. 849 - 859. -
Thevanayagam, S. and Martin, G. R. (2002). "Liquefaction in silty soils - screening and remediation issues".
Soil Dynamics and Earthquake Engineering, Vol. 22, pp. 1035 - 1042.
Toki, S., Shibuya, S. and Yamashita, S. (1995). "Standardization of laboratory test methods to determine
the cyclic deformation properties of geomaterials in Japan". Proc., 1st Pre-Failure Deformation of
Geomaterials, Sapporo, Japan, Vol. 2, pp. 741 - 784.

102
 GEOTECHNICAL ANO GEOPHYSICAL SITE CHARACTERIZATlON ORIENTED ro SEISMIC ANAL YSIS

Tokimatsu, K, Suzuki, Y, Taya, Y, and Kubota, Y. (1995). "Correlation between líquefaction resistance of
in-situ frozen samples and CPT resistance". Proc. 10th Asían Regional Conference on Soil Mechanics
& Foundation Engineering, Beijing, pp. 493 - 496.
Wang, Y, and O'Rourke, T. D. (2007). "Interpretation of secant shear modu1us degradation characteristics
from pressuremeter tests". Joumal of Geotechnical and Geoenvironmental Engineering, Vol. 133, No.
12, pp. 1556 - 1566.
Yamamuro, J. A. and Covert, K. M. (2001). "Monotonic and cyc1ic liquefaction of very lose sands with
high silt content". Joumal of Geotechnical and Geoenvironmenta1 Engineering, Vol. 127, No. 4, pp.
314 - 324.
Yoshimi, Y., Hatanaka, M., and Oh-Oka, H. (1977). "A simple method for undisturbed sand sampling by
freezing". Proc. 9th Intemationa1 Conference on Soi1 Mechanics and Foundation Engineering, Specia1
Session 2, Tokyo, Japan, pp. 23 - 28.
Youd, T. L. et al. (2001). "Liquefaction resistance ofsoi1s: surnmary report from the 1996 NCEER and 1998
NCEER/NSF workshops on eva1uation ofliquefaction resistance of soils". Joumal of:6eotechnical and
Geoenvironmental Engineering, Vol. 127, No. 10, pp. 817 - 833.
Yu, H. S. (2004). "In situ soil testing: from mechanics to interpretation". 1st J. K. Mitchell Lecture, Proc.
2nd Int. Conf. on Site Characterization ISC-2, Porto. Vol. 1, pp. 3 - 38.

-'.",.

103