Coastal Education & Research Foundation, Inc.

Evaluation of Liquefaction Potential of Port Structures with Earthquake Magnitude Adjustment

Author(s): Duhee Park, Dong-Yeop Kwak, Chang-Koo Cho and Byung-Sik Chun
Source: Journal of Coastal Research, Special Issue No. 56. Proceedings of the 10th International
Coastal Symposium ICS 2009, Vol. II (2009), pp. 1035-1039
Published by: Coastal Education & Research Foundation, Inc.
Stable URL: http://www.jstor.org/stable/25737944
Accessed: 31-10-2015 13:52 UTC

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at http://www.jstor.org/page/
info/about/policies/terms.jsp

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content
in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship.
For more information about JSTOR, please contact [email protected].

Coastal Education & Research Foundation, Inc. is collaborating with JSTOR to digitize, preserve and extend access to Journal
of Coastal Research.

http://www.jstor.org

This content downloaded from 128.111.121.42 on Sat, 31 Oct 2015 13:52:55 UTC
All use subject to JSTOR Terms and Conditions
 ofCoastalResearch I
[Journal SI 56 | 1035-1039 | ICS2009(Proceedings) | Portugal ] ISSN0749-0258 |

Evaluation ofLiquefaction Potential of Port Structures with Earthquake

Magnitude Adjustment
Duhee Parkf, Dong-Yeop Kwak j, Chang-Koo Chooo and Byung-Sik Chun?
tDept. ofCivil & Env. Eng. %Dept. ofCivil & Env. Eng. ooDept. of Civil & Env. Eng. ? Dept. ofCivil & Env. Eng.
Hanyang University, Seoul Hanyang University, Seoul Hanyang University, Seoul Hanyang University, Seoul

133-791, Korea 133-791, Korea 133-791, Korea 133-791, Korea

dpark@hany ang. ac .kr [email protected] petercck@poscoenc. com [email protected]

ABSTRACT
^^^^^^^^^^^^^^
,!!!!!!!!:.
f Park, D., Kwak, D.-Y., Cho, C.-K., and Chun, B.-S., 2009. Evaluation of liquefaction potential of port

W^g^SgJh^ structureswith earthquake magnitude adjustment (Proceedings of the 10th International Coastal
-
Symposium),1035 1039.Lisbon, Portugal, ISSN 0749-0258.
^S^B^S^
The port structures and breakwaters are frequently exposed to failure under sever seismic loading. One of the
main causes of failure is loss of bearing capacitydue to liquefaction,
which is caused by build-up of excess pore
water pressure under repeated dynamic loading. It is thus important to accurately estimate liquefaction
susceptibilityof port structures.Liquefaction susceptibility is commonly determinedby simplifiedmethod
relating the equivalent shear stress and resistance determined by the standard penetration test. The equivalent
shear stress is highly dependent on the duration of the groundmotion, which in turn is determinedby the
magnitude and distance of the earthquake. This study performs deaggregation analysis to determine the
representative earthquake scenarios and ground motion time histories for selected probabilities of occurrence and
the results are applied to simplified and advanced liquefaction analyses at ports located at southern parts of
Korea. Analyses indicate the importance of adjusting the effective stress with the magnitude of the earthquake.

ADITIONAL INDEX WORDS: Liquefaction,Deaggregation, Magnitude adjustment,Effective-stresssite

response analysis, Excess pore water pressure

INTRODUCTION LIQUEFACTION POTENTIAL EVALUATION

Port facilitiesare oftenbuilt over very loose sandy soil,making PROCEDURE
themvulnerable tomajor damage under liquefactionduring an Various methods have been to evaluate the
developed
earthquake. Main causes of failure are loss of bearing capacity, liquefaction potential. The most widely used method in
lateral spreading, or large settlement. It is important to accurately
engineeringpractice is the simplifiedprocedureproposed by Seed
estimate liquefaction susceptibility of port structuresand to and Idriss (1971). The method, often termedas the "cyclic stress
provide adequate remediation plans if necesaary. method", compares the earthquake induced loading, represented in
It is a common practice to evaluate the liquefaction potential terms of cyclic stress ratio (CSR), to the liquefaction resistance,
with the simplifiedmethod proposed by Seed and Idriss (1971).
representedin termsof thecyclic resistingratio (CRR).
The procedure relates the equivalent cyclic shear stress caused by CSR is calculated by thefollowingequation:
the earthquake loading to the resistance determined by the
standard test (SPT) in evaluation of the liquefaction
CSR = = 0.65
penetration
I /<x (i)
susceptibility. The equivalent shear stress is highly dependent on
(rja) (adepih
g) (<r )
the frequency content and duration of the ground motion, which in
turn is determined by the magnitude and distance of the where rd
=
shear stress, adepth
=
maximum acceleration at the depth
earthquake. However, the magnitude and distance of the =
at which liquefaction susceptibility is evaluated, crv total vertical
representative earthquake is not defined inKorea, and a magnitude =
stress, <j\ effective vertical stress. adepth in the equation is either
of 6.5 is used in the seismic design guideline (Moct, 2005) determined from a total-stress site response or by
analysis,
without a proper evaluation.This studyperformeddeaggregation
calculatingpeak acceleration at the surfaceand thenmultiplying it
to determine the representative The determined
magnitude. newly to a reduction factor (e.g. depth dependent reduction factor
magnitudeswere applied in the simplifiedapproach and theeffect developed by Seed and Idriss, 1971).
of the magnitude adjustment based on deaggregation was CRR represents the minimum CSR required to cause
evaluated and used it to estimate the liquefactionsusceptibilityof
liquefactionof a given SPT resistance,or (Ni)60.A modified plot
ports at Pohang and Kwangyang, located in the southern coast of
of CRR is shown in Figure 1,which is proposed by Seed et ah
Korea. The accuracy of the magnitude is evaluated
adjustment
(1985). If the CSR is above the CRR line, then the soil is
with effective-stress nonlinear site response
through comparison
susceptible to liquefaction. If it is below theCRR line, then the
analysis.
liquefaction is not likely to occur. CRR is also a functionof the
fine content, since the fine content controls the permeability of
soils and hence, the amount of excess pore pressure build-up. Note

JournalofCoastal Research, Special Issue 56, 2009

1035

This content downloaded from 128.111.121.42 on Sat, 31 Oct 2015 13:52:55 UTC
All use subject to JSTOR Terms and Conditions
 of Liquefaction
_Evaluation Potential_

0 10 20 30 40 50
Corrected Blow Count, ' V
(NT60
)

=
Figure 1. Cyclic resistance ratio for magnitude 7.5 (modified
after Seed et al., 3. Stratigraphy and measured
1985)_ Figure (Ni)60 profiles

thatthecurves inFigure 1 representthe liquefactionsusceptibility 24 56

under a magnitude 7.5 earthquake. it is often written as
MSF = \02 /
M2 (3)
Therefore, =
Where M magnitude of earthquake.
CRRT5.
IfCSR and CRR7 5 are calculated, then the factorof safety(FS) If themagnitude of the earthquake is 7.5, thenMSF is 1. If the
is less than 7.5, then MSF becomes larger than 1.
against liquefactioncan be determinedby thefollowingequation: magnitude
Since the FS is highly dependent on themagnitude of the
it is crucial that the magnitude be reliably estimated.
FS = (CRR751CSR )MSF (2)
earthquake,
In Korea, M =
6.5 is assumed in all design codes without a proper
evaluation (e.g. Moct, 1997; Moct, 2005).
where MSF ismagnitude scaling factor.
based on empirical
Various values of MSF have been proposed
SITE DESCRIPTION
data. Discussion on the differences between the proposed MSF is
The sites selected in this study are Pohang and Kwangyang,
out of the scope of thispaper. Korean design code (Moct, 2005) which are both located at southern coast of Korea (Figure 2). The
proposes use ofMSF proposed by Idriss (1995), which is defined
stratigraphy, (Nj)60 profiles of selected sites are shown in Figure 3.
as follows:
Pohang site is approximately 30 m in thickness, and is composed
of coarse sand, sandy gravel, sand, and silty clay. Kwangyang site,
composed of sand and silty clay, is 22 m in thickness. The shear
wave velocity Vs profiles were not available at the site. Therefore,
PGA contourwith retrun
period 1000years the following empirical equation proposed by Imai (1982) was
used to estimate the Vs:

F=97.0(AO6r (4)

The shear wave velocity of the bedrock was assumed to be 760

m/s.

SEISMIC HAZARD EVALUATION

The seismic hazard is most often evaluated via probabilistic
seismic hazard analysis (PSHA). Probabilistic seismic hazard
analysis (PSHA), firstdeveloped by Cornell (1968), accounts for
the uncertainties in the size, location, rate of recurrence of

earthquakes, and the variation of resulting ground motion

characteristics in estimation of the seismic hazard. PSHA

computes the mean annual rate of exceedance (or return period) of

a ground motion parameter at a particular site from the aggregate
risk of potential earthquakes of many different magnitudes

occurring at a range of source-site distances. The output of a

Longitude PSHA is amplitudes of ground motion parameters at a particular
site for a selected annual probability of exceedance (e.g. Figure 2).

Figure 2. Location of sites and seismic hazard map of Korean

=
Peninsula for return period 1000 years
earthquake_

Journal of Coastal Research, Special Issue 56, 2009

1036

This content downloaded from 128.111.121.42 on Sat, 31 Oct 2015 13:52:55 UTC
All use subject to JSTOR Terms and Conditions
 et
_Park al._

0.2 I_l_l_I_I_l_i_I
0 2 4 6 8 10 12 14
Time (sec)

Figure 6. Ground motion recorded during Ofunato earthquake

The groundmotion time history is needed to perform the site

and to determine
response analysis adepth in equation (1). Four
recorded motions that match the representative earthquake
scenarios (determined by deaggregation) most closely are selected
from PEER database (http://peer.berkeley.edu/), and are shown in

Figure 5. The peak ground acceleration (PGA) were, scaled to

match the design guideline (Moct, =
Figure 4. Representative earthquake scenarios determined by 1997) of Korea (PGA
=
0.154g and 0.22g for earthquakes with return periods 1000 and
deaggregation_
2400 years, respectively). In Korea, the ground motions used in
the seismic design are not selected based on the predicted
To determine the earthquake scenario that has the highest
representative earthquake scenario, but arbitrarily selected. Shown
contribution to the seismic hazard, needs to be
deaggregation inFigure 6 is one of themost widely used groundmotion inKorea.
performed (Kramer, 1996). Deaggregation is simply the inverse It is the recorded motion =
during Ofunato earthquake (M 7.4,
process of PS HA, and calculates the contribution to hazard of all
1978), but is shortenedfromoriginal recording.
individual earthquake scenarios. Although deaggregation is very

important in characterizing the seismic hazard, surprisingly it has

not yet been performed in Korea. Without performing LIQUEFACTION POTENTIAL EVALUATION
deaggregation, Korean design codes assume that representative VIA SIMPLIFED METHOD
magnitude is 6.5. The liquefaction potentials at selected sites were evaluated
In thispaper, deaggregationwas performedboth inPohang and using the simplified approach described previously. To evaluate

Kwangyang using the seismicitydata used in generation of the the effect of earthquake magnitudes, two procedures were used.

probabilistic seismic hazard maps in Korea. The results of The first procedure, termed "This study", uses the representative

deaggregation are shown in Figure 4, which displays the earthquake scenarios determinedby deaggregation (Figure 4) and
contribution to hazardpossibleof earthquake scenarios corresponding ground motions (Figure 5). The second procedure,
represented in terms of magnitudes and distances. Note that the termed "Design Method", uses M=6.5 recommended in the

representative earthquake scenarios for an earthquake with return Korean design guideline and theOfunato groundmotion (Figure
periods 1000 and 2400 years are M=5.1/R(hypocentral 6). Total-stress nonlinear site response analyses were performed
distance)=15km and M=5.5/R=15km for Pohang, and for both procedures using one-dimensional code DEEPSOIL
M=4.9/R=15km and M=5.3m/R=15km for Kwangyang. Note that (Hashash and Park, 2001) to obtain peak acceleration profile
the magnitudes of the
representative earthquake scenario are within the soil column. The constitutive model incorporated in

significantly smaller than M=6.5 currently used in design DEEPSOIL is the modified hyperbolic model developed by
guidelines in Korea. Itmeans that the current design codes highly Matasovic (1993), which defines the backbone curve as follows:
overestimate the representative earthquake magnitude.

i
(5)
iiii i I r- r

1+ /? In

where r = shear stress, y

=
shear strain, Gmo
=
initial shear
= =
modulus, rmo shear strength, yr reference shear strain, p & s =
material constants that adjust the shape of the backbone curve.
The parameters selected for the constitutive model are listed in

Table 1:Parametersfor themodified hyperbolicmodel

P
Sandy Gravel 1.5
0.7
Time (sec) Time (sec)
Sand 1.30.7
Figure 5. Selected groundmotion timehistories thatmatch the
scenario determined Clay 0.9 0.7
earthquake by deaggregation_

Journal of Coastal Research, Special Issue 56, 2009

1037

This content downloaded from 128.111.121.42 on Sat, 31 Oct 2015 13:52:55 UTC
All use subject to JSTOR Terms and Conditions
 of Liquefaction
_Evaluation Potential_

LIQUEFACTION POTENTIAL EVALUATION

Strain(%)
0.0001 0.001 0.01 0.1
1 VIA ADVANCED METHOD
To validate whether the magnitude adjustment using the
magnitude predicted from deaggregation results in an improved
estimate of the liquefaction susceptibility, the liquefaction
potential was evaluated with an advanced method an
using
effective-stress nonlinear site response analysis. The program used
was DEEPSOIL (Hashash and Park, 2001) and the input
motions shown in Figure 5 were used. Two excess pore pressure
generation models are incorporated in the effective-stress analysis
code for sands and clays, respectively. The models are given in

equations (6) and (7), respectively.

p-f'F'N\y-ytvp)
(6)
Strain(%) \+
f-F.N\r-rJ
7. Dynamic curves used in site response
Figure analyses_ u = ANMr'-'J + BNMr'-rJ + CN-(r'-'J +D (7)

Table 1 and the dynamic curves are shown in Figure 7. The u = =

where normalized pore water pressure after N cycles, ytvp
parameters for sandy gravel and sand were selected to match threshold strain under which excess does not
pore pressure
upper bound and mean curves developed by Seed and Idriss and A&B&C&D&r = curve
develop,/& p&F&s fitting
(1970), respectively,while parametersforclaywere selected toPI parameters for sands
and clays, respectively.
= 15 curves
developed byVucetic andDobry (1991). Note that both models are functions of shear strain amplitude
The FS profiles calculated by both procedures are shown in and number of cycles. The sand model, is developed
equation (6),
Figure 7. For Pohang site, pronounced discrepancies between the
by Dobry et al (1985), while the clay model, equation (7), is
calculated FS are observed. Since the magnitude
profiles developed byMatasovic (1993).
calculated by deaggregation is lower than M=6.5 recommended in of excess pore pressure results in decrease in effective
Build-up
the design guideline, the calculated FS was significantly larger. results in lower stiffness of the soil. The excess pore
stress, which
=
For return period 1000 year earthquake, while M=6.5 to be linked with
assuming pressure model needs the constitutive model,
resulted in FS < 1 at selected depths, applying newly selected for In
equation (5), hydro-mechanical coupled analysis.
magnitude resulted in significantlyhigher FS. Adjusting the DEEPSOIL, the following shear stress-strain-fluid model is
based on deaggregation > 1 even for return
magnitude
=
analysis, FS incorporated,which was originally developed by Matasovic
period 2400 year earthquake. This case demonstrated
that it is
(1993):
extremely important that the magnitude be accurately predicted
and applied for magnitude adjustment, such that liquefaction

susceptibility is not over-conservatively predicted and unnecessary xG xy

soil improvement techniques are employed at port facilities. The (8)
yjl-u x
same trend was observed at Kwangyang site. Lower magnitude
estimated by deaggregation resulted in significant increase in the
calculated FS profile, and use of M=6.5 cause severe x r
underestimation of the FS. (l-(?7)
where v = material constant.
Table 2 lists the pore pressure model parameters selected in this

study. The parameters for sandy gravel were representative of the

Factorof Safety Factor of Safety
1 0.1 10 0.1 1 10 Treasure Island sand layer, while the parameters for the sandy
,-,-. 0 ,-,-, are of the Owi Island sand
gravel representative layer
(Matasovic, 1993). Clay parameters are selected based on

20 I_I_I I_I_

?a? This Study,ReturnPeriod= 1000year

This Study,ReturnPeriod= 2400 year
? ? =
Design Method, ReturnPeriod 1000year 0 0.2 0.4 0.6 0.8 1
* =
Design Method, ReturnPeriod 2400 year Strain(%)

Figure 9. Calculated normalized excess pore pressure when

Figure 8. Factor of safetycalculated by thesimplifiedmethod applying five uniform
cycles_

Journal of Coastal Research, Special Issue 56, 2009

1038

This content downloaded from 128.111.121.42 on Sat, 31 Oct 2015 13:52:55 UTC
All use subject to JSTOR Terms and Conditions
 et
_Park al_

Table 2: Parameters for the excess pore water pressure models

Normalized excess pore pressure Normalized excess pore pressure
Parameters for sand model 0 0.2 0.4 0.6 0.8 10 0.2 0.4 0.6 0.8 1
0 ~l-1-1-1
P F s ytup v
_/ Pohang
5
Sand^ 1.0 1.0 0.7 1.0 0.020 3.8
gravel 10
Sand 1.0 1.1 2.6 1.7 0.020 3.8 ?
? 15
Parameters for clay model
Q
20
s r A B C D ytup
25
Clay 0.075 0.495 7.64 -14.72 6.38 0.69 0.1
30
-
ReturnPeriod= 1000year
laboratorytestof normallyconsolidated clays (Matasovic, 1993).
-
9 shows the calculated normalized excess pore pressure ReturnPeriod= 2400 year
Figure
when five cycles of shear strain are applied. Note that sand results
in highest excess followed 10. evaluation via advanced
pore pressure generation, by the sandy Figure Liquefaction potential
The method
gravel. clay shows very limited pore pressure development.
Figure 10 shows the calculated normalized excess pore water
pressure profiles calculated by the effective-stress site response
At both Pohang and Kwangyang sites, return
analyses.
= 1000 year earthquakemotions resulted in negligible
using
ACKNOWLEDGEMENTS
period This research was the research from the
excess supported by grant
pore pressure ratio generation. The calculated pore water
POSCO E&C. The authorsgratefullyacknowledge thesupport.
ratio was less than 0.05 at all depths, and demonstrated that

liquefaction is unlikely to occur at both sites. However, when

using return period =

2400 year earthquake ground motions, LITERATURE CITED
normalized excess pore pressure reached 1.0, and liquefaction was Cornell, C.A., 1968. Engineering Seismic Risk Analysis, Bulletin

predicted to occur. of theSeismological Society ofAmerica, 58, 1583-1606

The results of the advanced method were compared to the Dobry, R.; Pierce, W.G.; Dyvik, R.; Thomas, G.E., and Ladd,

simplified analysis. At Pohang, FS > 1 using the simplified R.S., 1985. Pore Pressure Model for Cyclic Straining of Sand,
procedurewithmagnitude adjusted by deaggregation,while FS < New York:
Troy, Civil Engineering Department, Rensselaer
1 using theM=6.5 for an earthquakewith returnperiod of 1000 Polytechnic Institute, Research Report, 56p.
years. This showed that the procedure recommended Y.M.A. and 2001. Non-linear one
by the design Hashash, Park, D.,
code is highly conservative and that the proposed procedure is dimensional seismic ground motion propagation in the
more reasonable. At FS < 1 for both 62(1-3), 185
Kwangyang, simplified Mississippi embayment, Engineering Geology,
procedures. While the magnitude adjustment through 206
deaggregation significantly increased the predicted FS, it fell http ://peer.berkeley.edu/
- 1995.
below 1 at depth between 7 8m. Overall, the predicted Idriss, I.M., Seed Memorial Lecture, University of

liquefactionpotentialusing simplifiedmethod ismuch higher than California at Berkerley.

when using the advanced method. Imai, T. and Tonouchi, K., 1982. Correlation of N-Value with S
demonstrated that the magnitude is Wave Velocity and shear Modulus, Proceedings of 2nd
Comparisons adjustment
very important indeed, and the current design code highly Europe Symposium on Penetration Testing (Amsterdam,
overestimates the liquefaction potential. The magnitude Netherlands), pp. 67-72
adjustment through deaggregation significantly increased the Kramer, S.L., 1996. Geotechnical Earthquake Engineering,
accuracy of the evaluation procedure. Upper Saddle River,New Jersey:PrenticeHall, 653p.
Matasovic, N., 1993. Seismic Response of Composite
Soil Deposits, Los Angeles, California:
CONCLUSION Horizontally-layered
University of California at Los Angeles, Ph.D. thesis, 449p.
This paper evaluated the liquefactionsusceptibilitiesof two port
sites in southernKorea using the simplifiedmethod and the MOCT, 1997. Seismic Design Guideline (II), Seoul, Korea.

stress MOCT, 2005. Highway Bridge Design Specification, Seoul, Korea.

effective site response analysis. In application of the
Seed, H.B. and Idriss, I.M., 1970. Soil Moduli and Damping
simplifiedmethod, the representativemagnitude predicted by Factors for
= Dynamic Response Analyses, Berkeley,
deaggregation and magnitude 6.5 recommended in the design
California: Reaserch Center,
guideline were both applied. Computed factorof safety against
Earthquake Engineering
UniversityofCalifornia,Berkeley,Report EERC 70-10.
liquefaction indicated that the currentdesign guideline highly H.B. and 1971. Procedure for
to overestimation Seed, Idriss, I.M., Simplified
overestimates the liquefaction susceptibility due
of the representative The of the Evaluating Soil Liquefaction Potential, Journal of the Soil
earthquake magnitude. accuracy
Mechanics and Foundations Division, 107(9), 1249-1274
simplified evaluation procedure can be highly increased by and Chung,
Seed, H.B.; Tokimatsu, K.; Harder, L.F., R.M.,
properly adjusting the magnitude of the earthquake. Further 1985. Influence of SPT Procedures in Soil Liquefaction
comparisons with effective-stress nonlinear analysis indicated that
theapplication ofmagnitude predictedby deaggregationwas more
Resistance Evaluations, Journal of Geotechnical Engineering,

accurate than when =

6.5 recommended in the 111(12), 1425-1145
using magnitude on
Vucetic, M. and Dobry, R., 1991. Effect of Soil Plasticity
design code.
Cyclic Response, Journal of Geotechnical Engineering, ASCE,
117(1), 89-107

JournalofCoastal Research, Special Issue 56, 2009

1039

This content downloaded from 128.111.121.42 on Sat, 31 Oct 2015 13:52:55 UTC
All use subject to JSTOR Terms and Conditions