Professional Documents
Culture Documents
Proposal of Liquefaction Potential Assessment Procedure Using Real Earthquake Loading
Proposal of Liquefaction Potential Assessment Procedure Using Real Earthquake Loading
···································································································································································································································
Abstract
In this paper, the application of the energy-based excess pore pressure generation model using damage potential to the assessment
of liquefaction potential is examined through experimental and analytical investigations. For a more realistic description of the
dynamic responses of saturated sands, the model parameters of the proposed model were modified based on the relative density.
Dynamic undrained triaxial tests were performed for sand with different relative densities. Based on the test results, new equations
for the model parameters with the relative density were addressed. Further, a liquefaction potential assessment procedure using the
model based on the maximum cumulative excess pore pressure ratio is also proposed. Factors of safety calculated from the stress-
based method and the proposed method were compared through examples with different soils and earthquake conditions.
Keywords: liquefaction, dissipated energy, excess pore pressure, damage potential, cumulative absolute velocity, earthquake
loading, liquefaction potential assessment
···································································································································································································································
1. Introduction estimated CSR value, to the cyclic shear stress equivalent for a
given earthquake-induced loading. This approach is most popular
The excess pore pressure (EPP) is an important factor in- due to its simplicity and suitability to laboratory cyclic triaxial
fluencing the behavior of saturated sands under seismic loading tests. It should be noticed, however, that the stress-based method
conditions. When a dynamic force such as an earthquake is is based on the equivalent stress concept and does not reflect
applied to saturated sands, the EPP builds up continuously with a various seismic factors such as the amplitude, duration, and
concomitant decrease in soil strength, and the sands are shape of earthquake motions.
eventually liquefied. For assessment of the liquefaction potential During the past three decades, several energy-based liquefac-
or susceptibility of sands, various researchers have investigated tion potential assessment approaches have been proposed and
soil liquefaction both in the laboratory and in the field using also employed in practice to evaluate the liquefaction potential
methods such as SPT and energy measurement (Seed et al., 1975; and susceptibility of sands. These approaches have established a
Green, 2001). It is generally understood that the liquefaction foundation for the development of energy-based EPP generation
phenomenon is essential in the analysis and design of structures models (Nemat-Nasser and Shokooh, 1979; Law et al., 1990;
and foundations involving saturated sands subjected to dynamic Figueroa et al., 1994; Green et al., 2000; Davis and Berrill,
loading. 2001). These energy-based models define EPP in terms of the
Several methods have been developed to determine the dissipated energy produced by induced stresses and strains based
liquefaction potential of soil deposits. The stress-based method on certain model parameters that are empirically obtained from
using the equivalent shear stress concept proposed by Seed et al. laboratory test results. While energy-based liquefaction potential
(1975) is the most common method utilized by engineers for assessment procedures based on energy-based EPP generation
determining the liquefaction potential. According to this concept, models have been successfully verified for many dynamic
earthquake-induced loadings are simplified as equivalent cyclic geotechnical problems, these have been primarily for sinusoidal
shear stresses. The liquefaction resistance of a soil, on the other loading cases. When realistic earthquake loadings are involved,
hand, is defined by the cyclic stress ratio (CSR) at an equivalent the dynamic responses may differ from those for sinusoidal
number of loading cycles corresponding to a given earthquake loadings, and thus, correlations between the EPP and dissipated
magnitude. The factor of safety is then obtained as a ratio of the energy for sinusoidal loading conditions may no longer be
cyclic resistance ratio (CRR), which is obtained from the applicable (Ishihara and Yasuda, 1972; Kim et al., 2005; Park et
*Member, Professor, Department of Civil & Environmental Engineering, University of Yonsei, Seoul, Korea (E-mail: geotech@yonsei.ac.kr)
**Ph.D. Candidate, Department of Civil & Environmental Engineering, University of Yonsei, Seoul, Korea (Corresponding Author, E-mail: uscake@ yon-
sei.ac.kr)
al., 2007). In most energy-based models, a relationship between function of the number of loading cycles for a given earthquake
the dissipated energy, initial soil condition, and model parameters magnitude. When the capacity is determined in a laboratory,
can be derived by setting the EPP ratio (ER) to 1.0. Model liquefaction prediction is made by performing cyclic triaxial tests
parameters, which are implemented in the majority of these for more than three different stress ratios. In this method, the
models, should be obtained by curve fitting of experimental data. CSR of a soil (i.e., capacity) and the maximum or equivalent
Several laboratory sinusoidal test results are required before CRR likely to be induced in the soil during an earthquake (i.e.,
implementing a model (Baziar and Jafarian, 2007). Although demand) are estimated and the liquefaction potential of the
correlations have been proposed for the model parameters in deposit is expressed in terms of the FS against liquefaction,
some of these models, their accuracy is not satisfactory for all which is given by Eq. (1).
soils and loading conditions. Furthermore, many of these models For energy-based liquefaction potential assessment, the demand
lack a specified procedure to determine the model parameters. imposed on the soil is quantified in terms of seismic energies
Recently, an energy-based model to predict the EPP under real proposed by Gutenberg and Richter (1965), Arias (1970), and
earthquake loading conditions has been proposed using the Alkhatib (1994). The capacity represents a measure of liquefaction
dissipated energy (wd) and damage potential [such as the cumu- resistance in the energy-based EPP generation models. The
lative absolute velocity (CAV)] as a measure of the severity of liquefaction potential is predicted where the ratio of capacity to
earthquake motions (Park et al., 2007). This model has been demand is less than or equal to unity.
successfully verified for dynamic triaxial tests considering various
seismic factors such as the amplitude, duration, and shape of earth- 3. Energy-based EPP Generation Model
quake motions. However, no energy-based liquefaction potential
assessment procedure using real earthquake motions has yet 3.1 Energy-based EPP Generation Model using CAV
been presented. The energy-based EPP generation model proposed by Park et
This paper proposes a new procedure for liquefaction potential al. (2007) is based on a scalar damage mechanics approach. The
assessment based on the energy-based model proposed by Park concept postulates that a soil in the undamaged state undergoes
et al. (2007). This method has been developed to evaluate the internal microstructural changes under external force conditions.
liquefaction potential without requiring ground response analysis, These microstructural changes cause reorientation and relative
several triaxial tests, or determination of dynamic properties such motion of the particles of the soil. As a result, the state of the soil
as the shear modulus. In addition, undrained triaxial tests under is transformed continuously from the undamaged to a damaged
earthquake loading conditions have been performed for sands state. The EPP response of the soil at a certain loading condition
with different relative densities to investigate their influence on is then determined from the degree of the CAV and dissipated
the model parameters proposed by Park et al. (2007). Based on energy (wd), which is defined as the area bound by the stress–
the test results, for correlations of model parameters with full strain hysteresis loops, whereupon the EPP is quantified by the
consideration of the relative density, model parameters consi- normalized cumulative EPP ratio (NCER) as follows:
dering the variation in the relative density are proposed. A
NCER = 1 – exp ( –Ar NCW z ) r
(2)
liquefaction potential assessment procedure based on the model
proposed by Park et al. (2007) is also proposed. The proposed where NCER = CER/CERmax, NCW = CW/CWmax, and Ar and
assessment procedure is discussed and compared through Zr are the model parameters. The model parameters Ar and Zr in
examples for different soils and earthquakes conditions. Eq. (2) represent the scale and overall shape of the NCER–NCW
curve, respectively. CER, CERmax, CW, and CWmax are functions
2. Current Liqufaction Potential Assessment Pro- expressed in terms of the EPP and wd as follows:
cedure t
CER = ³ [ ( 1 – dCAV ) ⋅ ER ] dt (3)
With regard to the design practice for soil structures and 0
capacity
CWmax = ³ wd ( t ) dt (6)
FS = --------------------- (1) 0
demand
where t denotes time, dCAV = CAV/CAVmax, ER is the EPP ratio
In the conventional stress-based procedure (Seed and Idriss, (i.e., ER = EPP/σ'c), wd is the dissipated energy represented by
1971), the demand of a soil varies depending on the magnitude the area of the hysteresis loop of the stress–strain curves, and tmax
of the input earthquake because the CSR is determined as a denotes the time until the end of the loading cycle. CAV and
CAVmax are the area under the absolute accelerogram of an 3.3 Determination of Initial Liquefaction
earthquake excitation and at the end of loading cycle, respectively, According to Park et al. (2007), the key procedure for the
and are expressed as follows: evaluation of ER or EPP generation using Eq. (9) is the identi-
t fication of initial liquefaction. Initial liquefaction occurs at a
CAV = ³ a( t ) dt (7) certain level of applied stress at which wd reaches a limit value,
0
causing EPP to equal σ'c. This, in turn, indicates that initial lique-
tmax faction can be identified from wd and cumulative stress, such as
CAVmax = ³ a ( t ) dt (8) from the normalized cumulative stress (NCS)–normalized cu-
0
mulative dissipated energy (NCW) relationship. Similar to Eq.
where t denotes time, and a(t) is the time-dependent ground (2), NCS–NCW curves can be defined by:
acceleration.
NCS = 1 – exp ( –Aw NCW z ) w
(17)
In this model, the relationship between the wd and ER to
trigger initial liquefaction is developed as follows: where Aw and Zw are the model parameters. The values of the
parameters Aw and Zw can be estimated from the first-order
wd · wd ·
ER = K ⋅ § ---------------
© 1 – dCW¹
- / K ⋅ § ---------------
© 1 – dCW¹
-
liq
(9) polynomial regression line that is obtained by taking the natural
logarithm of both sides in Eq. (17) as follows:
where K is obtained from the derivative of Eq. (2) as follows:
ln[ – ln{ ( 1 – NCS ) } ] = lnAw + Zw lnNCW (18)
dNCER
K = ------------------- = ( Ar ⋅ Zr ) ⋅ NCW Z – 1 ⋅ exp ( –Ar ⋅ NCWZ )
r r
(10) As NCS and NCW are known from the input loading motion
dNCW
and the constitutive relationship, the regression line and values of
From Eq. (9), the general equation under random loading Aw and Zw can be obtained without dynamic tests.
conditions with a given earthquake waveform can be rewritten For the identification of initial liquefaction using the NCS–
as: NCW curves, the minimum curvature method proposed by Park
(1997) is adopted. Based on the best curve fitting procedure
wd · § CERmax·
ER = K ⋅ § -----------------
- ⋅ ------------------ (11) using Eq. (17), the minimum curvature (Rp) on the NCS–NCW
© 1 – dCAV¹ © CWmax ¹
curve can be found as the minimum value of R given by the
where K = dNCER/dNCW. In Eq. (11), to determine ER from following equation of curvature:
the input stress time history, the CERmax value must be identified;
( NCS )''
this will be further discussed. Rp = min ( R ) = min ------------------------------------ (19)
( 1 + ( NCS )' 2 )3 / 2
3.2 Determination of Model Parameters Ar and Zr where Rp denotes the minimum curvature; R, the curvature of the
In the model proposed by Park et al. (2007), the model NCS–NCW function of Eq. (17); min( ), a function that repre-
parameters, Ar and Zr, are related to the CAV. In order to establish sents the minimum value; and (NCS)' and (NCS)", the 1st and
a correlation between Ar, Zr, and CAV, the following normalized 2nd derivatives of the NCS–NCW function, respectively.
CAV (NCAVliq) at initial liquefaction is addressed:
CAVliq
4. Experimental Investigation of Dynamic Soil
NCAVliq = -------------------
- (12) Responses
tmax ⋅ amax
where CAVliq is the CAV at initial liquefaction; tmax, the time until 4.1 Cyclic and Static Triaxial Tests
the end of loading cycle; and amax, the peak acceleration. The In order to investigate the effects of the model parameters on
value of CAVliq is obtained as the CAV value corresponding to the model proposed by Park et al. (2007), dynamic undrained
the time when initial liquefaction occurs. The determination of triaxial tests with two more relative densities equal to 70 and 75
initial liquefaction is further described below. % of that of the sands were performed under earthquake
According to Park et al. (2007), through dynamic test results loading conditions. Four acceleration records were selected
for both sinusoidal and earthquake loading cases with relative from recent major earthquakes of magnitudes between 6.1 and
density (Dr) = 60%, the model parameters are given as: 7.0. The significant duration is the 5-95% RMS (root-mean-
square) duration (Abrahmson and Silva, 1996). The values of
Ar = 15.20 ⋅ NCAV liq – 0.87 for sinusoidal loading (13)
Zr = 0.34 ⋅ NCAVliq + 0.54 for sinusoidal loading (14)
Ar = 32.56 ⋅ NCAV liq + 2.45 for earthquake loading (15) Table 1. Basic Properties of Jumunjin Sand
Zr = 4.71 ⋅ NCAVliq + 0.68 for earthquake loading (16) γdmax γdmin D50
emax emin GS Cu* Cc**
(kN/m3) (kN/m3) (mm)
Once the initial liquefaction and NCAVliq are identified for a
given loading condition, the values of Ar and Zr can be obtained 15.7 13.6 0.719 0.625 2.63 0.52 1.35 1.14
* **
using Eqs. (13)–(16). C = coefficient of uniformity, C
u c = coefficient of curvature.
4.2 Test Results Fig. 2. Fig. 2 shows the test results for all earthquake loadings
The results of the earthquake loading tests are shown in Fig. 1. until initial liquefaction. The ER vs. wd plots appear to have
Fig. 1(a) shows the response time history of the deviatoric stress remarkably irregular shapes. As indicated in Fig. 2, the extent
(σd) and EPP for Ormond earthquake loading. As indicated in of scatter is fairly severe and no unique correlations appear to
Fig. 1(a), the EPP does not increase at the early stages, but be obtained. These results also indicate that the relationship
suddenly jumps when σd,p is reached, whereupon liquefaction between ER and wd is strongly dependent on the loading type.
sets in. Fig. 1(b) shows the hysteresis loops. Contrary to the This is consistent with previously reported results (Park et al.,
general sinusoidal loading test results, the loops for earthquake 2007).
loading tended to progressively increase until initial liquefaction.
However, after liquefaction, the loops became progressively 4.3 Effect of Model Parameters and Verification
flatter and their areas were near zero. The same phenomena were Fig. 3 shows the NCS and NCER versus NCW plots described
observed for the other earthquake loadings. previously. When compared with Fig. 2, the results in Fig. 3
To compare the liquefaction initiation procedure for earthquake show improved correlations and appear to be more suitable for
loadings, the relationships between wd and ER are plotted in quantifying the energy-based relationship between stress and ER
Table 3. Values of Model Parameters and Number of Loading Cycles at Initial Liquefaction for Verification Test Results
Earthquake (location and year) Nliq,p/Nliq,m Aw Zw Ar Zr Dr (%)
Taumaranui (NewZealand, 1973) 18/18 6.15 0.70 5.23 0.95 70
Synthetic earthquake loading 6/6 4.23 0.65 4.23 0.85 75
From the test results, it was found that Ar and Zr are variable 5. Liquefaction Potential Assessment Based on
and are closely related to the relative densities. To determine Ar Energy-based EPP Generation Model
and Zr with relative densities, the values of NCAVliq were
obtained and replotted in Fig. 5 as a function of 0.7⋅Dr⋅Ar and Zr. 5.1 Evaluation of CERmax
As shown in Fig. 5, it is observed that the correlation parameters In the model proposed Park et al. (2007), the expression for the
0.7⋅Dr⋅Ar and Zr increase with NCAVliq and NCAVliq⋅Dr-–4 values, CERmax value is defined as cumulative ER. It was also found that
respectively. Ar and Zr were found to be linearly dependent on the CERmax is the only unknown value for identifying the ER
NCAVliq with the relative density. Thus, new correlations for using the input stress time history. Fig. 7 shows the ER
earthquake loading with different Dr shown in Fig. 5 can be obtained from different CERmax values based on Eq. (11) and
approximated as: dynamic triaxial test results obtained for different loading
amplitudes. As can be seen in Fig. 7, the changes in the CERmax
Ar = ( 6.06 ⋅ NCAVliq + 0.51 ) ⁄ 0.7 D r
for earthquake loading value appear to be nearly identical to those of the ER. Once the
(20) CERmax value is obtained for a given loading condition, the ER
Zr = 0.91 ⋅ NCAVliq ⋅ Dr–4 + 0.46 for earthquake loading can be identified using Eq. (11). The results in Fig. 7 suggest
(21) that the CERmax value can be effectively used for the description
of ER development, and thus, as an index for the assessment of
In order to verify the effect of the model parameters, Ar and Zr, liquefaction potential.
for the evaluation of initial liquefaction and EPP generation,
additional dynamic triaxial tests using earthquake loading were 5.2 Determination of Capacity and Demand
performed. The test soil was Jumujin sand with Dr = 70 and 75 % From the energy-based excess pore pressure generation model
and σ'c = 100 kPa. Table 3 shows the input loadings, predicted proposed by Park et al. (2007), it was found that the mobilization
and measured number of cycles at initial liquefaction, and values of the ER or EPP under dynamic loading conditions is closely
of model parameters obtained from the proposed procedure. related to the CERmax value. From Fig. 7, the occurrence of
From Table 3, it is seen that the predicted values of the number liquefaction can be thought of as CERmax value (i.e., CERmax,liq)
of cycles at initial liquefaction (i.e., Nliq,p) show good agreement when initial liquefaction is first generated for given earthquake
with the measured values (i.e., Nliq,m). As shown in Fig. 6, the waveform; that is the critical resistance of the soil to resist the
predicted pore pressure build up shape also shows reasonable development of EPP. If exceeded, initial liquefaction may be
agreement with the measured results for Taumaranui earthquake induced leading to the development of EPP. CERmax,liq value is a
loading of Dr = 70%. Based on the analysis results, it is seen that unique value for given earthquake waveform. Thus, it represents
the overall procedure still remains valid in the model proposed the liquefaction capacity of soil and is comparable with cyclic
by Park et al. (2007). It is also found that the values of Ar and Zr resistance ratio (CRR) in stress based method. Simple rearrange-
are functions of the Dr. ment of Eq. (11) and substitution of the definition of ER = 1.0
Fig. 6 Measured and Predicted ER Curves for Dr = 70% Fig. 7. Proposed and Measured ER for σ'c = 100 kPa and Dr = 60%
where tmax is the time until the end of loading cycle, dCAV = CAV/
CAVmax, ER is the EPP ratio, and t denotes time.
Table 5. CERmax,tx and CERmax,liq for Different Earthquakes and Soil based method proposed by Seed and Idriss (1971) was selected
Conditions and used in the comparison. As stress-based method employs the
Name of earthquake peak acceleration at the target layer for liquefaction potential
Dr (%) σd,i (kPa) CERmax,tx CERmax,liq
(location and year) assessment, ground response analysis was required to obtain the
24 13.37 peak acceleration for the given earthquakes and soil conditions.
Cost Rica * However, the input deviatoric stresses for a given earthquake
60 40 22.20 22.20
(Costa Rica, 1991)
waveform with a magnitude at which the liquefaction is generated
60 22.93
can be determined through the proposed model and test results
24 13.74 (i.e., Table 4). The σd,i values of the earthquake loadings can be
Baja California
60 33* 17.18 17.18 directly used to obtain the peak acceleration from Newton’s
(Mexico, 1934)
43 19.08 second law for a given soil mass and transformed force. For
39 12.77
liquefaction potential assessment by the stress-based method, the
Parkfield *
peak acceleration calculated using the σd,i value was used.
70 87 28.37 28.37
(USA, 1966) From Table 6, it is seen that the FS values using the proposed
104 29.50 method are reasonably predicted for liquefaction initiation with
15 14.22 different soil conditions and earthquake loading levels. It is also
Loma Prieta seen that the results from the proposed method are reasonable
75 25* 23.69 23.69
(USA, 1989)
and consistent with the variations in the soil and earthquake
35 24.50
loading conditions. The FS values from the stress-based method,
*
stress at which liquefaction was first generated for a given earthquake on the other hand, appear to be unconservative compared with
waveform
those from the proposed method. The differences in FS between
Table 6. Summary of Liquefaction Assessment Results the stress-based and proposed method are more pronounced. This
result indicates that the stress-based method may overestimate
Name of earthquake Dr σd,i
(location and year) (%) (kPa)
FSproposed FSSeed and Idriss the FS for all soil and earthquake loading conditions.
24 1.66 6.37
Cost Rica
6. Conclusion
*
60 40 1.00 3.82
(Costa Rica, 1991)
60 0.97 2.54 When a dynamic force such as an earthquake is applied to
24 1.25 6.37 saturated sands, the EPP builds up continuously with a concom-
Baja California itant decrease in soil strength, and the sands are eventually liq-
60 33* 1.00 4.63
(Mexico, 1934) uefied. In this paper, a new and simple method for determining
43 0.90 3.55 liquefaction potential based on the energy-based EPP generation
39 2.22 10.52 model proposed by Park et al. (2007) was proposed and
Parkfield * evaluated through experimental and analytical investigations. In
70 87 1.00 4.72
(USA, 1966)
104 0.96 3.94
order to describe in a more realistic manner the effect of the
model parameters proposed by Park et al. (2007), as an ex-
15 1.67 12.18
perimental investigation, undrained dynamic triaxial tests with
Loma Prieta *
75 25 1.00 7.31 saturated sand samples were carried out under earthquake
(USA, 1989)
35 0.97 5.22 loading conditions with different relative densities. From the test
*
stress at which liquefaction was first generated for a given earthquake results, it was observed that the model parameters were closely
waveform related to the relative density.
Based on the test results, new equations for the model
parameters based on the relative density were presented. In order
determination of wd, the DSC model was adopted. The model to verify the proposed model for the evaluation of EPP genera-
parameters and initial liquefaction were then determined and tion using the new equations for the model parameters, additional
identified using Eqs. (18)-(21). Through Eq. (22), CERmax,liq was undrained dynamic triaxial tests were performed. From these test
calculated until the end of the loading cycles. For Costa Rica results, it was seen that the predicted EPP generation and initial
earthquake, the value of the CERmax,liq with the same number of liquefaction match the measured results for both sinusoidal and
cycles then equals to 22.20. The values of the capacity induced real earthquake loading cases.
by the other earthquake loading cases are given in Table 5. The liquefaction potential assessment procedure using the
Table 6 shows the FS values for different soils and earthquakes model based on the maximum cumulative EPP ratio, which
using the proposed assessment method. For the proposed assess- represents the accumulated EPP in a dynamic loading process,
ment method, the FS was defined as the ratio of CERmax,liq to was also proposed. The FS values calculated from the stress-
CERmax,tx. In order to compare FS values, the existing stress- based procedure proposed by Seed and Idriss (1971) and the
proposed method [using the model by Park et al. (2007)] were Green, R.A. (2001). Energy based Evaluation and Remediation of
compared using examples of different soil and earthquake Liquefiable Soils, PhD Dissertation, Virginia Polytechnic Institute
conditions. It was seen that the proposed procedure produces and State Univ, Blacksburg, Virginia.
Gutenberg, B. and Richter, C.F. (1956), “Earthquake magnitude, inten-
reasonable FS values that were consistent with the variations in
sity, energy and acceleration.” Bulletin of the Seismological Society
the soil and earthquake conditions. of America, Vol. 46, pp. 105-145.
Ishihara, K. and Yasuda, S. (1972). “Sand liquefaction due to irregular
References excitation.” Soils and foundations, Vol. 12, No. 4, pp. 65-77.
Kim, S.I., Park, K.B., Park, S.Y., Hwang, S.J., Lee, J.H., and Choi, J.S.
Abrahamson, N., and Silva, W. (1996). Empirical Ground Motion (2005). “Effects of irregular dynamic loads on soil liquefaction.”
Models, Draft Report, Brookhaven National Lab, Upton, New York. 16th International Conference Soil Mechanics and Geotechnical
AlKahatib, M. (1994). Liquefaction Assessment by Strain Energy Engineering, Osaka, Vol. 4, No. 4b, pp. 2673-2676.
Approach, Ph.D. Thesis (T. Kagawa, Advisor), Wayne State Ladd, R.S. (1978). “Preparing test specimens using under compaction.”
University, p. 212. Geotechnical Testing Journal, Vol. 1, No. 1, pp. 16-23.
Arias, A. (1970). “Measure of earthquake intensity.” Seismic Design for Law, K.T., Cao, Y.L., and He, G.N. (1990). “An energy approach for
Nuclear Power Plants, R.J., Hansen, ed, The MIT Press, Cambridge, assessing seismic liquefaction potential.” Canadian Geotechnical
Massachusetts. Journal, Vol. 27, pp. 320-329.
Baziar, M.H., and Jafarian, Y. (2007). “Assessment of liquefaction Nemat-Nasser, S. and Shokooh, A. (1979). “A unified approach to
triggering using strain energy concept and ANN model: capacity densification and liquefaction of cohesionless sand in cyclic shear-
energy.” Soil Dynamics and Earthquake Engineering, Vol. 27, pp. ing.” Canadian Geotechnical Journal, Vol. 16, pp. 659-678.
1056-1072. Park, I.J. (1997). Disturbed State Modeling for Dynamic and
COSMOS Virtual Data Center, http://db.cosmos-eq.org, the Consortium Liquefaction Analysis, PhD Dissertation, Dept of CEEM, Univ of
of Organizations for Strong-Motion Observation System 1999. Arizona, Tucson.
Davis, R.O. and Berrill, J.B. (2001). “Pore pressure and dissipated Park, K., Kim, S., Lee, J., and Park, I. (2007). “Energy-based evaluation
energy in earthquakes-field verification.” Journal of Geotechnical of excess pore pressure using damage potential.” International
and Geoenvironmental Engineering, ASCE, Vol. 127, No. 3, pp. Journal of Offshore and Polar Engineering, (accepted).
269-274. Seed, H.B. and Idriss, I.M. (1971). “Simplified procedure for evaluating
Desai, C.S. (2001). Mechanics of Materials and Interfaces-the soil liquefaction pontential.” Journal of the Soil Mechanics and
Disturbed State Concept, CRC Press, Boca Raton, Florida. Foundation Division, Vol. 97, No. SM9, pp. 1249-1273.
Figueroa, J., Saada, A.S., Liang, L., and Dahisaria, N.M. (1994). Seed, H.B., Idriss, I.M., Makdisi, F., and Banerjee, N. (1975).
“Evaluation of soil liquefaction by energy principles.” Journal of “Representation of irregular stress time histories by equivalent
Geotechnical Engineering, ASCE, Vol. 120, pp. 1554-1569. uniform stress series in liquefaction analysis.” Earthquake En-
Green, R.A., Mitchell, J.K., and Polito, C.P. (2000). “An energy based gineering Research Center, Univ. California, Berkeley No.EERC
excess pore pressure generation model for cohesionless soils.” Proc 75-29.
Developments in Theoretical Geomechanics - The John Booker Vasquez-Herrera, A. and Dobry, R. (1989). The Behavior of Undrained
Memorial Symp, D.W. Smith and J.P. Carter, eds, Sydney, New Contractive Sand and Its Effects on Seismic Liquefaction Flow
South Wales, Australia, AA Balkema Publishers, Rotterdam, pp. Failures of Earth Structures, Contract No. 86-003, Dept. of Army,
383-390. U.S. Army Corps of Engineers, Washington, D.C.