NATIONAL

CENTER FOR
EARTHQUAKE
ENGINEERING
RESEARCH

· r·roce.eding of the-
NCEER Workshop on
-Evaluation of Liquefaction
Resistance of Soils
Edited by

_ T. Leslie Youd
Brigham Young University
Department of Civil and Environmental Engineering
Provo, Utah 84602-4081

Izzat M. Idriss
· University of California at Davis
Department of Civil Engineering
Davis, California 95616

Technical Report NCEER-97-0022

December 31, 1997 REPRODUCED BY,

U.S. Department or Commerce
~

National Technical lnfom1at1on Service

Sprmgfield, Virgm1a 22161

This workshop was conducted at the Inn at Temple Square, Salt Lake City, Utah, and was partially
supported by the Federal Highway Administration under contract number DTFH61-92-C-00112.
 NOTICE
This report was prepared by Brigham Young University as a result of
research sponsored by the National Center for Earthquake Engineering
Research (NCEER) through a contract from the Federal Highway Admin-
istration. Neither NCEER, associates of NCEER, its sponsors, Brigham
Young Uni\·ersity, nor any person acting on their behalf:

a. makes any warranty, express or implied, with respect to the use of

any information, apparatus, method, or process disclosed in this re-
port or that such use may not infringe upon privately owned rights;
or

b. assumes any liabilities of whatsoever kind with respect to the use of,
or the damage resulting from the use of, any information, apparatus,
method, or process disclosed in this report.

Any opinions, findings, and conclusions or recommendations expressL'd

in this publication are those of the author(s) and do not necessarily reflect
the views of NCEER or the Federal Highway Administration.
 5027 2 • 101
REPORT DOCUMENTATION 11. REPORT NO. Nl..1::.1:K-~/-UULl 2. 3. Reciciient's Accession No.

PAGE ·
.C.. Title and Subtitle 5. Report Date

Proceedings of the NCEER Workshop on Evaluation of December 31, 1997

Liquefaction Resistance of Soils 6.

8. Performing Organization Rept. No.

. es 1•1e
7. t'thol.'..5) You d an d lzzat M. Idriss, Editors

9 .\ •rig
rloal-\";f O~ization NtJie .and Add!ss
m oung nivers1 y
10. Projif./Task/Work Unit No.
112- -4.2
Department of Civil and Environmental Engineering
11
Provo, Utah 84602-4081 • 'Ft1fWS!f1 ec9fft'f'5c~ Number
0
·

(C) DTFH61-92-C-0012
(Gl

12... Span-soring Organization Name and Address 13. Type of .Repart & Period Covered
National Center for Earthquake Engineering Research
State University of New at Buffalo Technical report
Red Jacket Quadrangle u.
Buffalo, NY 14261
15. Supplementary Noles Th. k
. 1s wor shop was _conducted at the Inn at Temple Square, January 5-6, n 96,
Salt Lake City, Utah, and was partially supported by the Federal Highway Administration
under contract number DTFH61-92-C-00112.
Following disastrous earthquakes in Alaska and in Nii(cata, Japan in 1964,
16. Abstract <Limit: 200 words>
Seed and Idriss developed and published the basic simplified procedure or evaluating lique-
faction resistance of soils. The frocedure, which is largely empirical, evolved over decades.
until it has now established itsel as the standard practice in North America and through- •
out much of the world. The purpose of this 1996 workshop was to convene a group of
experts to review developments and gain consensus for further augmentations to the ·
procedure. The scope was limited to evaluation of liquefaction resistance. Post-liquefaction
phenomena, such as soil deformation and ground failure, although equally or more important
were beyond the scope of this workshop. The participants developed consensus recommen-
dations on the following topics: use of the standard and cone penetration tests for evalu-
ation of liquefaction resistance; use of shear wave velocity measurements for evaluation of
liquefaction resistance; use of the Becker penetration test for gravelly soils; magnitude
scaling factors; correction factors Kl and KaX; evaluation of seismic factors required for thi.
evaluation procedure. Probabilistic analysis and seismic energy considerations were also
reviewed. Seismic energy concepts were judged to be insufficiently developed to make
recommendations for engineering practice. Probabilistic methods have been used in some
risk analyses, but are still outside the mainstream of standard practice.

17. Document Analysis a. Oescrtpton

Earthquake engineering. Standard penetration tests. Cone penetration tests. Becker
penetration test. Gravel. Shear wave velocity. Magnitude scaling factors. Correction factors
Energy. Statistical analysis. Probabilistic analysis. Fines content. Soil plasticity.
Liquefaction res i sta nee. Liquefaction potential evaluation.· Simplified methods.
b. Identifiers/Open-Ended Terms

c. COSATI Field/Group

18. Availability Statement 19. Security Class (This Report) 21. NJ'2°4 Pages
Unclassified
Release unlimited 20. ~curt Cla,~/Tha Page) 22. Price
nc ass1 1e
I ~ - - &ltJC!I __ '7'2~ 1 D\ -:-- , .... , •. ,_.,___ ....... o ........,,. OPTIONAL FORM 272 (4-71'
 NATIONAL
CENTER FOR
EARTHQUAKE
ENGINEERING
RESEARCH
Headquartered at the Stare University of New Yo/Kat Buffalo

Proceedings of the
NCEER Workshop on
Evaluation of Liquefaction Resistance of Soils

held at
Inn at Temple Square
Salt Lake City, Utah
January 5-6, 1996

Edited by T. L. Youd 1 and I. M. Idriss 2

Publication Date: December 31, 1997

Technical Report NCEER-97-0022

NCEER Highway Project Task Number 112-D-4.2

FHWA Contract Number DTFH61-92-C-00112

1 Professor and Chair, Department of Civil and Environmental Engineering, Brigham Young
University
2 Professor, Department of Civil Engineering, University of California at Davis

NATIONAL CENTER FOR EARTHQUAKE ENGINEERING RESEARCH

State University of New York at Buffalo
Red Jacket Quadrangle, Buffalo, NY 14261
 PREFACE
Over the past twenty-five years, a procedure termed the "simplified procedure" has evolved for
evaluating the seismic liquefaction resistance of soils. This procedure has become the standard of
practice in the U.S. and Canada and throughout much of the rest of the world. The development of
the procedure has been incremental. Following the disastrous earthquakes in Alaska and in Niigata,
Japan, in 1964, Seed and Idriss (1971) developed and published the basic "simplified procedure." The
procedure has been revised and augmented periodically since that time, with landmark papers by Seed
(1976), Seed and Idriss (1982), and Seed et al. (1985). In 1985, the Committee on Earthquake
Engineering, National Research Council (NRC) organized a workshop with experts from the
profession at large to evaluate and update the procedure. That workshop was convened by Prof
Robert V. Whitman, the Massachusetts Institute of Technology, with thirty-six experts and observers
who thoroughly reviewed the state-of-knowledge and the state-of-the-art for assessing liquefaction
hazard. That workshop produced a report (NRC, 1985) that has become a widely used reference.
No additional general review or updating of procedures has been published since that time.

The purpose of the 1996 workshop, sponsored by the National Center for Earthquake Engineering
Research (NCEER), was to convene a group of21 experts to review recent developments and come
to consensus on further improvements and augmentations to the simplified procedure. Emphasis was
on developments that have been published in the ten-year period between the NRC and NCEER
workshop. To keep the workshop focused and the content tractable, the scope was purposely
restricted to a review of procedures for evaluating liquefaction resistance of soils as outlined in the
simplified procedure. Thus, the workshop was primarily concerned with evaluation of triggering of
liquefaction. Post liquefaction phenomena, such as soil deformation and ground failure, although
equally or more important, were omitted from discussion.

The simplified procedure was originally developed for assessment of liquefaction resistance of shallow
alluvial soils beneath level to gently sloping ground. Thus, by definition, valid application of the
procedure should be limited to these terrain conditions. Although the procedure has been applied by
some engineers to assess liquefaction hazard under steeply sloping terrain, constructed embankments,
or deep soil layers, such extrapolations are beyond the range of empirical data upon which the original
procedure was based. Such extrapolations should be made by experts with experience in such
applications. Thus, deliberations at the workshop were largely restricted to shallow deposits beneath
level or nearly level ground conditions. However, proposed revisions to the factors K,, and Ka,
respectively, used to correct the analyses for large static normal and shear stresses, respectively, were
discussed and consensus gained on Ka values to be used.

lll
 ACKNOWLEDGMENTS

Many hours of volunteer time were contributed by the participants in the workshop. This dedicated
effort is gratefully acknowledged. Reviews of the workshop proceedings were provided by Thomas
F. Blake, Fugro-West, Ventura, California, and Donald G. Anderson, CH2M-Hill, Bellevue,
Washington. Steven F. Noble and Sam Gilstrap, graduate students at Brigham Young University,
assisted with the workshop and preparation of the proceedings.

V
 Table of Contents

Reports

Summary Report ............................................................................................................ I

By Workshop Participants

Cyclic Liquefaction and Its Evaluation Based on the SPT and CPT .............................. 41
By Peter K. Robertson and Catherine E. Wride

Liquefaction Resistance Based on Shear Wave Velocity ............................................... 89

By Ronald D. Andrus and Kenneth H. Stakoe

Application of the Becker Penetration Test for Evaluating .......................................... 129

the Liquefaction Potential of Gravelly Soils
By Leslie F. Harder, Jr.

Magnitude Scaling Factors .......................................................................................... 149

By T. Leslie Youd and Steven K. Noble

Application ofK,,. and Ka Correction Factors .............................................................. 167

By Leslie F. Harder, Jr. and Ross Boulanger

Seismic Factors for Use in Evaluating Liquefaction Resistance .................................... 191

By T. Leslie Youd

Liquefaction Criteria Based on Statistical and Probabilistic Analyses .......................... 201

By T. Leslie Youd and Steven K. Noble

Liquefaction Criteria Based on Energy Content of Seismograms ................................. 217

By T. Leslie Youd, Robert E. Kayen and James K. Mitchell

Cyclic Liquefaction Based on the Cone Penetrometer Test .......................................... 225

By Richard S. Olsen

Appendices

A List of Participants ................................................................................................ A-1

B. Agenda ............................................................................................................ B-1
C. Definitions ofTerms ............................................................................................... C-1

vii
 Summary Report
Workshop Participants

Chair: T. Leslie Youd, Brigham Young University, Provo, UT; Co-chair: Izzat M. Idriss,
University of California at Davis, Davis, CA; Ronald D. Andrus, National Institute for Standards
and Technology, Gaithersburg, MD; Ignacio Arango, Bechtel Corp., Oakland, CA; Gonzalo
Castro, GEI Consultants, Inc., Winchester, MA; John T. Christian, Engineering Consultant,
Boston, MA; Ricardo Dobry, Rensselaer Polytechnic Institute, Troy, NY; W.D. Liam Finn,
University of British Columbia, Vancouver, BC; Leslie F. Harder, Jr., California Department of
Water Resources, Sacramento, CA; Mary Ellen Hynes,US Army Engineers Waterways Experiment
Station, Vicksburg, MS; Kenji Ishihara, Science University of Tokyo, Tokyo, Japan; Joseph P.
Koester, US Army Engineers Waterways Experiment Station, Vicksburg, MS; Sam S.C. Liao,
Parsons Brinckerhoff, Boston, MA; William F. Marcuson, III, US Army Engineers Waterways
Experiment Station, Vicksburg, MS; Geoffrey R. Martin, University of Southern California, Los
Angeles, CA; James K. Mitchell, Virginia Tech, Blacksburg, VA; Yoshiharu Moriwaki,
Woodward-Clyde Consultants, Santa Ana, CA; Maurice S. Power, Geomatrix Consultants, San
Francisco, CA; Peter K. Robertson, University of Alberta, Edmonton, Alberta; Raymond B. Seed,
University of California, Berkeley, CA; Kenneth H. Stokoe, II, University of Texas, Austin, TX

Abstract

Over the past twenty-five years, a procedure, termed the "simplified procedure," has evolved for
evaluating liquefaction resistance of soils. This procedure has become the standard of practice in
North America and throughout much of the world. Following disastrous earthquakes in Alaska and
in Niigata, Japan in 1964, Professors H.B. Seed and I.M. Idriss developed and published the basic
"simplified procedure." The procedure, which is largely empirical, evolved over the decades
primarily through summary papers by H.B. Seed and his colleagues. In 1985, Professor Robert V.
Whitman convened a workshop on behalf of the National Research Council (NRC) in which thirty-
six experts reviewed the state-of-knowledge and the state-of-the-art for assessing liquefaction hazard.
No general review or update of the simplified procedures has occurred since that time. The purpose
of the 1996 workshop, sponsored by the National Center for Earthquake Engineering Research
(NCEER), was to convene a group of experts to review developments and gain consensus for further
augmentations to the procedure. To keep the workshop focused and the content tractable, the scope
was limited to evaluation of liquefaction resistance. Post-liquefaction phenomena, such as soil
deformation and ground failure, although equally or more important, were beyond the scope of this
workshop. The participants developed consensus recommendations on the following topics: (1) use
of the standard and cone penetration tests for evaluation ofliquefaction resistance, (2) use of shear
wave velocity measurements for evaluation of liquefaction resistance, (3) use of the Becker
penetration test for gravelly soils, (4) magnitude scaling factors, (5) correction factors I<., and K 0 ,
and (6) evaluation of seismic factors required for the evaluation procedure. Probabilistic analysis
and seismic energy considerations were also reviewed. Seismic energy concepts were judged to be
insufficiently developed to make recommendations for engineering practice. Probabilistic methods
have been used in some risk analyses, but are still outside the mainstream of standard practice.
 Introduction

Over the past twenty-five years, a procedure, termed the "simplified procedure," has evolved for
evaluating the seismic liquefaction resistance of soils. This procedure has become the standard of
practice in North America and throughout much of the world. Following disastrous earthquakes in
Alaska and Niigata, Japan in 1964, Seed and Idriss (1971) developed and published the basic
"simplified procedure." The procedure has been corrected and augmented periodically since that
time with landmark papers by Seed (1979), Seed and Idriss (1982), and Seed et al. (1985). In 1985,
Professor Robert V. \Vhitman from the Massachusetts Institute of Technology convened a workshop
on behalf of the National Research Council (NRC) in which thirty-six experts and observers
thoroughly reviewed the state-of-knowledge and the state-of-the-art for assessing liquefaction
hazard. That workshop produced a report (NRC, 1985) that has become a widely used standard and
reference for liquefaction hazard assessment. No general review or update of the simplified
procedures has occurred since that time.

The purpose of the 1996 workshop, sponsored by the National Center for Earthquake Engineering
Research (NCEER), was to convene a group of experts to review recent developments and gain
consensus on further corrections and augmentations to the procedure. Emphasis was placed on new
developments since the NRC review. To keep the workshop focused and the content tractable, the
scope was limited to procedures for evaluating liquefaction resistance of soils under level to gently
sloping ground. In this context, liquefaction refers to the phenomena of seismic generation of large
pore-water pressures and consequent severe softening of granular soils. Post-liquefaction
phenomena, such as soil deformation and ground failure, although equally or more important than
triggering, were beyond the scope of the workshop.

The simplified procedure was developed from evaluations of field observations and field and
laboratory test data. Field evidence of liquefaction generally consisted of observed sand boils,
ground fissures or lateral spreads. Data were collected mostly from sites on level to gently sloping
terrain underlain by Holocene alluvial or fluvial sediment at shallow depths (less than 15 m). The
original procedure was verified for and is applicable only to these site conditions. The primary focus
of the workshop was to review and update procedures for evaluating soil liquefaction resistance for
these general site conditions. Limited attention was given to liquefaction resistance evaluation for
sediment layers at greater depths (high overburden pressures) and beneath steeply sloping terrain or
embankments.

Cyclic Stress Ratio (CSR) and Cyclic Resistance Ratio (CRR)

Calculation or estimation of two variables is required for evaluation of liquefaction resistance of

soils. These variables are the seismic demand placed on a soil layer, expressed in terms of cyclic
stress ratio (CSR), and the capacity of the soil to resist liquefaction, expressed in terms of cyclic
resistance ratio (CRR), hereafter referred to as liquefaction resistance or liquefaction resistance ratio.
CRR is a symbol proposed by Robertson and Wride that was endorsed by the workshop. Previously,
this factor had been called the cyclic stress ratio required to generate liquefaction, or the cyclic
strength ratio, and had been given different symbols by different writers. For example, Seed and
Harder (1990) used the symbol CSR£, Youd (1993) used the symbol CSRL, and Kramer (1996) used

2
the symbol CSRi, to denote this ratio. The workshop participants agreed that CRR conveys an
appropriate meaning and generates less confusion than the use of CSR with or without a subscript
to signify liquefaction resistance.

Seed and Idriss (1971) formulated the following equation for calculation of CSR:

(1)

where a,,,ax is the peak horizontal acceleration at ground surface generated by the earthquake, g is the
acceleration of gravity, ovo and o 'vo are total and effective vertical overburden stresses, respectively,
and rd is a stress reduction coefficient. The latter coefficient provides an approximated correction
for flexibility of the soil profile. The workshop participants recommend the following minor
modification to the procedure for calculation of CSR. For noncritical projects, the following
equations may be used to estimate average values of rd.

rd= 1.0 - 0.00765 Z for z s: 9.15 m (2a)

rd= 1.174 - 0.0267 Z for 9 .15 m < z s: 23 m (2b)
rd= 0.744 - 0.008 Z for 23 < z s'. 3 0 m (2c)
rd= 0.50 Forz> 30m (2d)

where z is depth below ground surface in meters. Parts a and b of this equation were proposed by
Liao and Whitman (1986b), part c was added by Robertson and Wride (this report), and part d was
suggested by William F. Marcuson (US Army Engineers, oral commun.) in post-workshop
discussions. Mean values ofrct calculated from Equation 2 are plotted on Figure I along with the
mean and range of values proposed by Seed and Idriss (1971). The workshop participants agreed
that for convenience in programming spreadsheets and other electronic aids, and to be consistent
with past practice, rct values determined from Equation 2 are suitable for use in routine engineering
practice. The user should understand, however, and take into account that rd values calculated from
Equations 2 or 3 give only the mean value from a range of possible rd values and that the range of
rd values increases with depth. Thus the certainty with which CSR can be calculated decreases with
depth when mean rd values are used to simplify calculations. In addition to the uncertainty in rd, the
simplified procedure is not well verified for depths greater than about 15 m, as indicated on Figure 1.
Thus the user should understand that results developed from the simplified procedure are quite
uncertain at depths greater than 15 m.

As an alternative to Equation 2, Thomas F. Blake (Fugro-West, Inc., Ventura, Calif., written

commun.) approximated the mean curve plotted on Figure 1 by the following equation:

(1.000 - 0.4113z 05 + 0.04052z + 0.001753z1. 5) (3)

05
(1.000 - 0.4177z + 0.05729z - 0.006205z 1.s + 0.001210z 2)

where z is depth beneath ground surface in meters. Equation 3 yields essentially the same values for
rd as Equations 2a-d, but is easier to program for many applications and may be used in routine
engineering practice.

3
 ('max)d
('max)r
O0 0.1 0.2 0.3 0.4 05 0.6 0.7 0.8 0.9 1.0

3 (10)
Average values
6(20)
Mean values of ra calculated
from equation 2
--s
¢;
'-'
9(30)

12(40)
..d
--
0.
ll)
15 (50)

Cl 18 (60)

21 (70)

24(80) -: region . : ..:

27 (90) . .. .
....
30(100)
. . . ...
FIGURE 1 rd Versus Depth Curves Developed by Seed and Idriss (1971)
with Added Mean Value Lines from Equation 2

The primary focus of the workshop was to improve procedures for evaluating liquefaction resistance
of soils, CRR. A plausible method for evaluating CRR is retrieving undisturbed soil specimens from
field sites and testing those specimens in the laboratory using cyclic tests to model seismic loading
conditions. Unfortunately, specimens of granular soils retrieved with typical drilling and sampling
techniques are generally too disturbed to yield meaningful laboratory tests results. Only through use
of specialized sampling techniques, such as ground freezing, can sufficiently undisturbed specimens
be obtained. The cost of such procedures is generally prohibitive for all but the most critical
projects. To avoid the difficulties associated with undisturbed sampling and testing, field tests have
become the state-of-the-practice for routine liquefaction investigations.

Several field tests have gained common usage for evaluation of liquefaction resistance, including
the cone penetration test (CPT), the standard penetration test (SPT), shear-wave velocity
measurements (V.), and the Becker penetration test (BPT). These tests were discussed at the
workshop along with associated criteria for evaluating liquefaction resistance. Possible
improvements to the state-of-the-art were reviewed and consensus recommendations developed for
engineering practice. A conscientious attempt was made to correlate liquefaction resistance criteria
from the various field tests to provide generally consistent results no matter which test is employed.
Thus the choice of test should depend on availability of equipment, site conditions, cost, and
preference. Primary advantages and disadvantages of each test are listed in Table 1.

4
 Table I. Comparison of Advantages and Disadvantages of Various Field Tests
for Assessment of Liquefaction Resistance

Feature Test Type

I I SPT I CPT I v~ IBPT I

Number oftest Abundant Abundant Limited Sparse
measurements at
liquefaction sites
Type of stress-strain Partially Drained, Smail strain Partially
behavior influencing drained, large strain drained, large
test large strain strain
Quality control and Poor to good Very good Good Poor
repeatability
Detection of variability Good Very good Fair Fair
of soil deposits
Soil types in which test Non-gravel Non-gravel All Primarily gravel
is recommended
Test provides sample Yes No No No
of soil
Test measures index Index Index Engineering Index
or engineering property property

Standard Penetration Test (SPT)

Criteria for evaluation of liquefaction resistance based on standard penetration test (SPT) blow
counts have been rather robust over the years. Those criteria are largely embodied in the CSR versus
(N 1) 60 plot reproduced in Figure 2. That plot shows calculated CSR and (N1) 60 data from sites where
liquefaction effects were or were not observed following past earthquakes along with CRR curves
separating data indicative ofliquefaction from data indicative ofnonliquefaction for various fines
contents. The CRR curve for a fines content less than five percent is the basic penetration criterion
for the simplified procedure and is referred to hereafter as the "simplified base curve." The CRR
curves in Figure 2 are valid only for magnitude 7.5 earthquakes.

Clean Sand Base Curve

Several changes to the SPT criteria were endorsed by workshop participants. The first change is to
curve the trajectory of the simplified base curve at low (N 1) 60 to a projected CRR intercept of about
0.05 (Figure 2). This adjustment reshapes the base curve to achieve consistency with CRR curves

5
 0.6r------,----,.,,...--r-------,-------.--------,
37
.29 251:l

Percent Fines = 35 15

O.Si-------+------+'-----t---.1---------1-------1
I
I
I
I
I
I
I
I
-0
0.4
0...... I
I
> I
~ I
I CRR curves for 5,15, and
31
• 35 percent fines, respectively
·-.0... 0.3
I
I
ce I
20 I
~ .12 I
en
en .so+ I
ll)
I
....
1-,
I
I
1Zl .21 /18.
·--
(.)

( .)
;>-.
60• •80 .20
.,,.s0+1 ®
20
1
10
U 0.2

Marginal No
Liquefaction Liquefaction Liquefaction
Pan - American data ■ 1:1
Japanese data • Q 0
Chinese data .,,. A
0 [__ _l,;;;;;;;;;;;;;:=;;;;;;;;;;;.!J....____L_ _ _--1.._ _ _-1.._ _-=--__J

0 10 20 30 40 50
Corrected Blow Count, (N1) 60
Figure 2 Simplified Base Curve Recommended for Calculation of CRR from SPT
Data along with Empirical Liquefaction Data (modified From Seed et al., 1985)

developed from CPT data and probabilistic analyses by Liao et al. (1988) and Youd and Noble
(Statistical and Probabilistic Analyses, this report). Seed and Idriss (1982) originally projected that
curve through the origin, but there were few data to constrain the curve in the lower part of the plot.
A better fit to the present empirical data is to bow the lower end of the base curve as indicated in
Figure 2.

6
Thomas F. Blake (Fugro-West, Inc., Ventura, Calif., written commun.) approximated the simplified
base curve plotted on Figure 2 by the following equation:
a + ex + ex 2 + gx 3 (4)
CRR 7 5 = -------~--
. 1 + bx + dx 2 + ft 3 + hx 4

where CRR7 _5 is the cyclic resistance ratio for magnitude 7.5 earthquakes; x = (N 1) 60 ; a= 0.048; b
= -0.1248; c = -0.004721; d = 0.009578; e = 0.0006136; f= -0.0003285; g = -l.673E-05; and h =
3.714E-06. This equation is valid for (N 1) 60 less than 30 and may be used in spreadsheets and other
analytical techniques to approximate the simplified base curve for engineering calculations.
Robertson and Wride (this report) indicate that Equation 4 is not applicable for (N 1) 60 less than three,
but the general consensus of workshop participants is that the curve defined by Equation 4 should
be extended to intersect the intercept at a CRR value of about 0.05.

Correlations for Fines Content and Soil Plasticity

Another change was the quantification of the fines content correction to better fit the empirical data
and to support computations with spreadsheets and other electronic computational aids. In the
original development, Seed et al. (1985) found that for a given (N 1) 60 , CRR increases with increased
fines content. It is not clear, however, whether the CRR increase is because of greater liquefaction
resistance or smaller penetration resistance as a consequence of the general increase of
compressibility and decrease of permeability with increased fines content. Based on the empirical
data available, Seed et al. developed CRR curves for various fines contents as shown on Figure 2.

After a lengthy review by the workshop participants, consensus was gained that the correction for
fines content should be a function of penetration resistance as well as fines content. The participants
also agreed that other grain characteristics, such as soil plasticity may affect liquefaction resistance;
hence any correlation based solely on penetration resistance and fines content should be used with
engineering judgement and caution. The following equations, developed by I.M. Idriss with
assistance from R.B. Seed are recommended for correcting standard penetration resistance
determined for silty sands to an equivalent clean sand penetration resistance:

(5)

where a, and p are coefficients determined from the following equations:

o:= 0 forFC ~ 5% (6a)
o: = exp[l.76 - (190/FC 2)] for 5% < FC < 35% (6b)
a= 5.0 for FC.: 35% (6c)

p= 1.0 for FC:,; 5% (7a)

p= [0.99 + (FCu/1000)] for 5% < FC < 35% (7b)
p= 1.2 forFC;;.: 35% (7c)

where FC is the fines content measured from laboratory gradation tests on retrieved soil samples.

7
These equations may be used for routine liquefaction resistance calculations. Back calculation of
CRR curves as a function of fines content and (Ni)60 for magnitude 7.5 earthquakes using Equations
5-7 yield curves that are essentially identical to the curves plotted on Figure 2.

Several workshop participants suggested that liquefaction resistance should also increase with soil
plasticity. Agreement could not be reached, however, on formulation of a correction for plasticity.
There is very little empirical data from which such a correction could be developed. Nevertheless,
some practitioners have been increasing CRR by about 10 percent for soils with fines and plasticity
indices greater than 15 percent. This increase seemed appropriate to several participants, but
consensus was not attained. The participants did agree, however, that plasticity should be measured
as part of liquefaction investigations with the goal of better defining the influence of plasticity on
liquefaction resistance.

Although not endorsed by workshop participants, Robertson and Wride ( this report) reviewed fines
content data as part of their workshop assignment to review liquefaction resistance criteria based on
SPT measurements. They suggest correcting the calculated (N1) 60 to an equivalent (N 1) 600 , using a
correction factor, K,, which is solely a factor of fines content as noted below:

(8a)
where
K. = 1 + [(0.75/30)(FC -5)] (8b)

This recommendation is for soils with nonplastic fines (PI s;_ 5 percent). For soil with plastic fines,
the correction factor, K,, would likely be larger, but the available empirical data are insufficient at
present to define a plasticity adjustment. For fines contents less than about 15 percent, the CRR
curves are not greatly different than the curves of Seed et al. (1985). However, for fines contents
greater than 15 percent, the Robertson and Wride CRR curves are significantly less conservative and
plot to the left of the curves of Seed and others. Although there are little empirical data to control
the positioning of the curves for fines contents greater than 15 percent and (N 1\ 0 greater than 10, the
general consensus of workshop participants was that the CRR curves should not be shifted to a less
conservative position, as proposed by Robertson and Wride, without additional supporting data.

Other Corrections

In addition to grain characteristics, several other factors affect SPT results. One of the more
important of these factors is the energy delivered to the SPT sampler. An energy ratio, ER, of 60%
has generally been accepted as the reference value. The ER delivered by a particular SPT setup
depends primarily on the type of hammer and anvil in the drilling system and on the method of
hammer release. Approximate correction factors (CE= ER/60%) to modify the SPT results to a 60%
energy ratio for various types of hammers and anvils are listed in Table 2. Because of variations in
drilling and testing equipment and differences in procedures used, a rather wide range in the energy
correction factor, CE, has been observed as noted in the table. Even when procedures are carefully
monitored to conform to established standards, such as ASTM D-1686, considerable variation in CE
may occur because of minor variations in equipment and procedures. Even within a given borehole,
variations in energy ratio between hammer blows or between tests typically may vary by as much

8
 Table 2. Corrections to SPT (Modified from Skempton, 1986)
as Listed by Robertson and Wride (this report)

Factor Equipment Variable Term Correction

Overburden Pressure CN (P ala' VO)05

CN '.5: 2
Energy ratio Donut Hammer CE 0.5 to 1.0
Safety Hammer 0.7 to 1.2
Automatic-Trip Donut- 0.8 to 1.3
Type Hammer

Borehole diameter 65 mm to 115 mm CB 1.0

150 mm 1.05
200mm 1.15

Rod length 3mto4m CR 0.75

4mto6m 0.85
6m to 10 m 0.95
10to30m 1.0
>30m <1.0

Sampling method Standard sampler Cs 1.0

Sampler without liners I.I to 1.3

as ten percent. Thus, the recommended practice is to measure the energy ratio frequently at each site
where the SPT is used. Where measurements can not be made, careful observation and notation of
the equipment and procedures is required to estimate a CE value for use in liquefaction resistance
calculations. Use of good-quality testing equipment and carefully controlled testing procedures
conforming to ASTM D-1686 will generally yield more consistent energy ratios and ¼ values from
the upper parts of the ranges listed in Table 2.

Additional correction factors are required for rod lengths less than 10 m and greater than 30 m,
borehole diameters outside the recommended interval (65 mm to 125 mm), and sampling tubes
without liners. Ranges of correction values for each of these variables are listed in Table 2. Careful
documentation of drilling equipment and procedures, including measurement of ER, is required to
select the most appropriate values for these correction factors. Even so, some uncertainty remains
in the actual factors that should apply for any field operation.

Because the SPT N-value also varies with effective overburden stress, an overburden stress
correction factor is also applied. This factor has commonly been calculated from the following
equation (Liao and Whitman, 1968a):

CN = (P ala' vo)0·5 (9)

9
where CN is a factor to correct measured penetration resistance for overburden pressure and Pa equals
100 kPa or approximately one atmosphere of pressure in the same units used for o'va· The effective
overburden pressure, o' vo, applied in this equation should be the overburden pressure that was
effective at the time the SPT test was conducted. Even though the ground water level may have
changed and a different water table level applied in the calculation of CSR (Equation I), the
correction of blow count requires use of the effective pressures that were effective at the time of
drilling and testing.

The SPT N-value corrected for each of the above variables is given by the following equation:

(10)

where Nm is the measured standard penetration resistance, CE is the correction for hammer energy
ratio (ER), C8 is a correction factor for borehole diameter, CR is the correction factor for rod length,
and Cs is the correction for samplers with or without liners. Suggested ranges of values for each of
these correction factors are listed in Table 2. Selection of appropriate factors from within these
ranges requires specific information on equipment and drilling procedures and engineering
judgement. The engineer should become familiar with details of the SPT procedure to avoid or at
least minimize possible errors associated with SPT testing and to gain expertise in selecting
appropriate correction factors.

A final change recommended by workshop participants is the use of revised magnitude scaling
factors rather than the original Seed and Idriss (1982) factors to adjust CRR7 .5 to CRR for other
earthquake magnitudes. Magnitude scaling factors are addressed later in this report.

Cone Penetration Test (CPT)

The workshop participants were unable to reach consensus on CPT criteria for evaluating
liquefaction resistance. Robertson and Wride (this report) developed the techniques presented below
with input from workshop attendees. Robertson and Wride verified these criteria against SPT and
other data from sites they had investigated. T.L. Youd and his students compared liquefaction
resistances calculated from CPT criteria against field performance at nineteen sites where surface
effects of liquefaction were or were not observed. The CPT criteria yielded apparently correct
prediction of liquefaction or nonliquefaction with greater than 90 percent reliability. Youd and his
students also compared liquefaction resistances from CPT criteria with results from SPT criteria at
50 sites with parallel CPT soundings and SPT borings, with a conclusion that the CPT criteria listed
below yield consistent and reasonably conservative results. G.R. Martin (oral commun., February
1998) and several colleagues from southern California also compared results developed from parallel
CPT soundings and SPT boreholes. They determined that liquefaction resistances estimated from
the CPT procedure are on average slightly smaller, and thus more conservative, than liquefaction
resistances developed from the parallel SPT tests. The above investigators endorse the CPT criteria
listed below, but strongly recommend that at least one parallel borehole near a CPT sounding be
drilled at each site to verify soil types and liquefaction resistances estimated from the CPT. l.M.
Idriss, on the other hand, reviewed the CPT criteria and concluded that inadequate development and
verification has been made to presently recommend these criteria to the geotechnical profession. In

10
particular, Professor Idriss indicated that the correction for grain characteristics using~ needs further
consideration and verification. R.S. Olsen reviewed the CPT criteria listed below and concluded that
the criteria are incorrectly developed and formulated. He recommends the criteria he has developed
and presents in a paper submitted to the workshop (Olsen, this report).

A primary advantage of the CPT is that a nearly continuous profile of penetration resistance is
developed for stratigraphic interpretation. The CPT results are generally more consistent and
repeatable than results from other penetration tests listed in Table 1. The continuous profile also
allows a more detailed interpretation of soil layers and soil types than the other tools listed in the
Table. This stratigraphic capability makes the CPT particularly advantageous for reconnaissance
investigations. In addition, CPT data can be used to estimate liquefaction resistance of penetrated
soil layers. Thus the CPT can be used to develop preliminary soil and liquefaction resistance profiles
for site investigations. These preliminary profiles should then be verified by other techniques, such
as drilling and SPT testing.

In recent years, increased field performance data have become available at liquefaction sites
investigated with CPT (Robertson and Wride, this report). These data have facilitated the
development of CPT-based liquefaction resistance correlations. These correlations allow direct
calculation of CRR, rather than through conversion of CPT measurements to equivalent SPT blow
counts and then applying SPT criteria, a technique that was commonly applied in the past.

Figure 3 shows a chart developed by Robertson and Wride (this report) for determining cyclic
resistance ratio (CRR75 ) for clean sands (fines content, FC s: 5%) from CPT data. The chart, which
is valid only for magnitude 7.5 earthquakes, shows calculated CRR plotted as a function of corrected
and normalized CPT resistance, q01 N, from sites where liquefaction effects were or were not observed
following past earthquakes. A CRR curve separates regions of the plot with data indicative of
liquefaction from regions indicative of nonliquefaction. Dashed curves showing approximate cyclic
shear strain potential, y ,, as a function of q01 N are drawn on Figure 3 to emphasize that cyclic shear
strain and ground deformation potential of liquefied soils decreases as penetration resistance
mcreases.

The CRR curve in Figure 3 is approximated by the following simplified equation:

If(qc!N)c, < 50 CRR75 = 0.833[(qclN)jl000J + 0.05 (1 la)

(11 b)

where (q 0 1N\, is the clean sand cone penetration resistance normalized to 100 kPa (approximately
one atmosphere of pressure).

Normalization of Cone Penetration Resistance

Although cone penetration resistance is commonly corrected only for overburden stress, resulting
in the term q 01 , truly normalized (i.e., dimensionless) cone penetration resistance corrected for
overburden stress (q 01 N) is given by:

11
 0.6
IM=7.5I 0.25 < Dso (mm) < 2.0
FC(%)<5
0.5

CRRCurve

0
M
-~ •
P::: 0.3 INo Liquefaction!
r.f.l
r.f.l
a
t
r:/')
A
A
fr
~
0.2 • 0 0

-
.....u
u
>-,
0
a
0~

U 0.1 1::,0
Field Performance Liq. NoLiq.
NCEER (1996) Stark & Olson (1995) • 0
Workshop Suzuki et al. (1995b) A a
0 -l-----+__:_--1---==:::::i:==:::i::::::::=:::::i:===:-1
0 50 100 150 200 250 300
Corrected CPT Tip Resistance, qc1N

Figure 3 Curve Recommended for Calculation of CRR from CPT Data along
with Empirical Liquefaction Data (After Robertson and Wride, this report)

(12)
where
(13)

C0 is a normalizing factor for cone penetration resistance, Pa is 100 kPa or approximately one
atmosphere of pressure in the same units used for a' vo, and q 0 is field cone penetration resistance
measured at the tip. A maximum C0 value of2.0 is generally applied to CPT data at shallow depths.
The value of the exponent, n, is dependent on grain characteristics of the soil and ranges from 0.5
for clean sands to 1.0 for clays (Olsen, this report). Selection of the value for use in liquefaction
resistance calculations is discussed in the following paragraphs.

The CPT friction ratio (sleeve resistance, f,, divided by cone tip resistance, qc) generally increases
with increasing fines content and soil plasticity. Robertson and Wride (this report) suggest that
appropriate grain characteristics, such as approximate soil type and a rough estimate of fines content,
termed apparent fines content herein, can be estimated directly from CPT data for sandy soils.
Relationships recommended by Robertson and Wride are reproduced in Figures 4 and 5. The
boundaries between soil types 2 through 7 on Figure 4 can be approximated as concentric circles
(Jeffries and Davies, 1993). The radius of each circle, referred to as the soil behavior type index, 10 ,
is calculated from the following equation:

12
 Cl
aS 100
u
C
.....
('j

·-
,;J'J
,;J'J
a.)
~
a.)
C
0
u
"O

·--s
a.)
N 10
( 'j

l-<
0
z

1
0.1 1 10

Normalized Friction Ratio, F = - -fs-O - x 100%

qt - vo

L Sensitive, fine grained 6. Sands - clean sand to silty sand

2. Organic soils - peats 7. Gravelly sand to dense sand
3. Clays - silty clay to clay 8. Very stiff sand to clayey sand*
4. Silt mixtures - clayey silt to silty clay 9. Very stiff, fine grained*
5. Sand mixtures - silty sand to sandy silt
*Heavily overconsilidated or cemented
Figure 4 CPT-Based Soil Behavior Type Chart Proposed by Robertson (1990)

le= [(3.47 - log Q)2 + (1.22 + Log F)2]°- 5 (14)

where
Q = [(qc - Ov 0 )/P 0 ][(P/a'v0 )"] (15)
and
F = [f/(qc - av◊>] x 100% (16)

13
 4.--------------------------~
Zone 2: organic soils - peats (Pl:,,2I"l•l_
u ._, 1.astic .fines • • • •
>-<
~
><: Zone 3: silty clay to clay - - • --
}Iig~lYJ • •

~ ---
~ 3
~&c:: "F-------------=--'•~•---:-J-~-~-~-~-~-~-~-~-~-~-~-~-::-:-:::::e~la=ti=o=n~
Zone 4: clayey silt to silty clay•
t---------....,..-.___;__..:.__ _ _ _ Recommen
ded General Corr

~
1-.. " ,.
,,. .. • - Zone 5: silty sand to sandy_sUt •• - -
- • • • •., •fi.nes (Pl< S%)
::l 2 ..1::---------=-0!!:::=---------.,,.....--'-!!..!:·-=-Non-p1asuC
0
·~~
.,
---
-------

~
a.)
o::i
---
--- - - - Zone 6: clean sand to silty sand

-·5 1 N.B. The correlation is approximate

r/J since le depends on other factors such Zone 7: gravelly sand to dense sand
as plasticity, mineralogy, sensitivity
and stress history.
0 -t----'----t------'---+-----'---+-----'---+------L------1
0 10 20 30 40 50
Apparent Fines Content, FC (%)
Figure 5 CPT Soil Behavior Type Index, l Versus Apparent Fines Content for
0
,

Normally Consolidated Soils (After Robertson and Wride, this report)

The soil behavior chart in Figure 4 was developed using an exponent, n, of 1.0, which is the
appropriate value for clayey-type soils. For clean sands, however, an exponent value of 0.5 is more
appropriate, and a value intermediate between 0.5 and 1.0 would be appropriate for silts and silty
sands. Robertson and Wride recommend the following procedure for selecting an exponent and
calculating the soil behavior type index, t

The first step is to differentiate soil types characterized as clays from soil types characterized as
sands and silts using Figure 4. This differentiation is performed by assuming an exponent, n, of 1.0
(characteristic of clays) and performing the following calculations. For clays, the dimensionless
normalized CPT penetration resistance, Q, is defined as:

(17)

If the calculated I0 calculated with an exponent of 1.0 is greater than 2.6, the soil is classed as clayey
and is considered too clay-rich to liquefy. Samples should be taken and tested, however, to confirm
the soil type and liquefaction resistance. Criteria, such as the Chinese criteria, might be applied to
confirm that the soil is nonliquefiable. The so-called Chinese criteria, as defined by Seed and Idriss
(1982), stipulate that liquefaction can only occur if all three of the following conditions are met:

( 1) The clay content (particles smaller than 5 µ) is less than 15 percent, by weight.
(2) The liquid limit is less than 35% percent.
(3) The natural moisture content is less than 0.9 times the liquid limit.

14
If the calculated Ic is less than 2.6, the soil is most likely granular in nature and Q should be
recalculated using an exponent, n, of 0.5. For this calculation, CQ should also be calculated with an
exponent, n, of 0.5 (Equation 13), and qc,N (calculated from Equation 12) substituted for Q in
Equation 14. Ic should then be recalculated using Equation 14. If the recalculated J.o is less than 2.6,
the soil can be classed as nonplastic and granular, and this Ic can be used to estimate liquefaction
resistance as noted below. If the recalculated Ic is greater than 2.6, however, the soil is likely to be
very silty and possibly plastic. In this instance, 4,N should be recalculated from Equation 12 using
an intermediate exponent, n, of 0. 7 in Equation 13 and Ic recalculated from Equation 14 using the
recalculated value for qc,N· This intermediate Ic is then used to calculate liquefaction resistance. In
this instance, a soil sample should be retrieved and tested to verify the soil type and whether the soil
is liquefiable by other criteria, such as the Chinese criteria.

Because the relationship between Ic and soil type is rather approximate, the consensus of the
workshop was that all soils characterized by an Ic of2.4 or greater should be sampled and tested to
confirm the soil type and to test the liquefiability with other criteria. Also, soil layers characterized
by an Ic greater than 2.6, but with a normalized friction ratio, F, less than 1.0 percent (Region 1 of
Figure 4) can be very sensitive, and hence should also be sampled and tested. Although perhaps not
technically liquefiable according to the Chinese criteria, such sensitive soils may suffer severe
softening and even strength loss under earthquake loading conditions.

Calculation of Clean Sand Equivalent Normalized Cone Penetration Resistance, (qc 1N)c,

To correct the normalized penetration resistance, (qc,N), of sands with fines to an equivalent clean
sand value,( qcJN)c,, for use in the calculation of liquefaction resistance, CRR, the following
relationships are applied:

(18)

where the CPT correction factor for grain characteristics, K" is defined by the following equations
(Robertson and Wride, this report):

Foric '.> 1.64 Kc= 1.0 (19a)

Foric > 1.64 Kc= -0.403 I/+ 5.581 I/- 21.63 I/+ 33.75 Ic - 17.88 (19b)

Although the measured fines content could be substituted for the apparent fines content in Figure 5
to determine an I0 such a substitution will likely yield erroneous results and should not be done.
As noted above, Ic is a function of plasticity and other factors as well as fines content. Thus when
using CPT data, Ic must be calculated from Equation 14 rather than estimated from the measured
fines content.

The Kc versus Ic curve defined by Equations 19a and 19b is plotted on Figure 6. For J.o greater than
2.6, the curve is shown as a dashed line, indicating that the soils are most likely too clay rich or
plastic to liquefy.

15
 o le= 2.6
~ 5,----------.------r----..:.....--r-----r---~
I

~ Gravelly Sands Sands Sand •

u 4.5 Mixtures 1

<S I
I

....§
~
4
I
I

as I

g 3.5 I

(.)
(.)

....·.;::
00
;..,
il)
3 4 3 2
K = - 0.403~ + 5.58 llc - 21.63Ic + 33. 75Ic - 17 .88

u 2.5 Silt Clays

~
Mixture

-5 2
....
~
ell
5b 1.5
E-<
0 l+----+----+---'---+-::_---,f-l---f-.1.---4----+-.J..._-~
0 0.5 1 1.5 2 2.5 3 3.5 4
Soil Behaviour Type Index, le
Figure 6 Grain-Characteristic Correction Factor, 1', for Determination of Clean-
Sand Equivalent CPT Resistance (After Robertson and Wride, this report)

With an appropriate Ic and Kc, Equations 11 and 19 can be used to calculate CRR75 . To adjust CRR
to magnitudes smaller or larger than 7.5, the calculated CRR75 is multiplied by an appropriate
magnitude scaling factor. The same magnitude scaling factors are used with CPT data as with SPT
or shear wave velocity data. Magnitude scaling factors are discussed in a later section of this report.

Although approached by a somewhat different route, the procedure for calculation of liquefaction
resistance, CRR, given above is generally consistent and will generally give compatible results with
the procedure proposed by Olsen (this report) for most level to gently sloping site conditions. As
noted by Olsen, almost any CPT normalization technique, such as the procedure noted above, will
give results consistent with his normalization procedure for shallow soil layers. For deep sites (a\ 0
> 150 kPa or depths greater than about 15 m), significant differences in results may develop between
the two procedures. Those depths are deeper than most documented occurrences of liquefaction at
natural sites and thus are deeper than the verified depth for the simplified procedure.

Correction of Cone Penetration Resistance for Thin Soil Layers

Theoretical as well as laboratory studies indicate that cone resistance is influenced by softer or stiffer
soil layers above or below the cone tip. As a result, the CPT will not usually measure the full
penetration resistance in thin sand layers sandwiched between layers of softer soils. The distance
to which cone tip resistance is influenced by an approaching interface increases with stiffness of the

16
 (a) Hide
20 20 40 60 80
\ :
\:
\
:\
\
\

(b)

Deposit D

t ------.... -.. -- ....-.. --.---.. ----··. ---..-.-.-.--.--.--.---.

--- '
Layer A \
I
--------------------------·----··-··J ___ _
I
_ _ _ _ __.I
I

Deposit B

Figure 7 Thin-Layer Correction Factor, KH, for Determination of Equivalent Thick-

Layer CPT Resistance (After Robertson and Fear, 1995)

17
stiff layer. In soft clays or loose sands, the distance of influence can be as small as 2 to 3 cone
diameters. In stiff clays or dense sands, the distance of influence may be as large as. 20 cone
diameters. (The diameter of the standard IO cm 2 cone is 36 mm.) Thus care should be taken when
interpreting cone resistance of sand layers sandwiched between silt or clay layers with lower
penetration resistances. Based on a simplified elastic solution, Vreugdenhil et al. (1994) developed
a procedure for estimating the full cone penetration resistance of thin stiff layers contained within
softer strata. Based on this model, Robertson and Fear (1995) suggest a correction factor for cone
resistance, KH, as a function oflayer thickness as shown in Figure 7. The correction applies only to
thin stiff layers embedded within thick soft layers. Because the corrections have a reasonable trend,
but appear rather large, Robertson and Fear (1995) recommend conservative corrections
corresponding to q,A/q, 8 = 2 as shown on Figure 7. The equation for evaluating the correction factor,
KH, is

KH = 0.5 [(H/1,000) - 1.45]2 + 1.0 (20)

where His the thickness of the interbedded layer in mm, and qcA and q,8 are cone resistances of the
stiff and soft layers, respectively.

Shear Wave Velocity

During the past decade, several simplified procedures have been proposed for the use of field
measurements of small-strain shear wave velocity, Vs, to assess liquefaction resistance of granular
soils (Stokoe et al., 1988; Tokimatsu et al., 1991; Robertson et al., 1992; Kayen et al., 1992; Andrus,
1994; Lodge, 1994). The use of Vs as a field index of liquefaction resistance is justified because
both Vs and CRR are similarly influenced by void ratio, effective confining stresses, stress history,
and geologic age. The advantages of using Vs include the following: (1) Vs can be accurately
measured in situ using a number of techniques such as crosshole and downhole seismic tests, the
seismic cone penetration test, or spectral analysis of surface waves; (2) Vs measurements are
possible in soils that are difficult to penetrate with CPT and SPT or to extract undisturbed samples,
such as gravelly soils, and at sites where borings or soundings may not be permitted; (3)
measurements can be performed in small laboratory specimens, allowing direct comparisons between
measured laboratory and field behavior; and (4) V5 is directly related to small-strain shear modulus,
a parameter required in analytical procedures for estimating dynamic soil response at small and
intermediate shear strains.

Two significant limitations of using V5 in liquefaction hazard evaluations are that (1) seismic wave
velocity measurements are made at small strains, whereas liquefaction is a large strain phenomenon;
and (2) seismic testing does not provide samples for classification of soils and identification of
nonliquefiable soft clay-rich soils. To compensate for the latter limitation, a limited number of
borings should be drilled and samples taken to identify nonliquefiable clay-rich soils that might
classify as liquefiable by V5 criteria and also to identify weakly cemented soils that might be
liquefiable but classify as nonliquefiable because of their characteristically high Vs values.

18
 0.8 .-----.,--,,.---..--,---.,--,,-~--.-~
Data Based on: Data Based on: • Soil may
Mw= 7.3 to 7.8 earthquakes Mw= 7.3 to 7.8 earthquakes be weak!
-0 - Average value ofVs1 - -0 Average value ofVs1 cemente
,-< a max for the larger of two
\...,J
';: 0.6
horizontal components
-Solid - liquefaction --+-----1
6
.._, 0.6
a max for the larger of two
horizontal components
with
carbonate
> Solid - liquefaction ••
~ Open - no liquefaction
Robertson et al. _
~ ....
Open - no Iiquefactioir.. ••
0
·.:j
C':l
I
Liquefaction
,Jtf'
(1992)
0
4~
I
Liquefaction o
• 0 0

~ 1
0.4 t----+---+-+----1---l ~ 0.41----+---w-n+--+---1--------1
U'J
U'J I
• -·1
t
CZ) o ,s, I
-

-
.(.)....,
( .)
>,
: ... I No
0.2 1-----+-----r-+-- Liquefaction -
. •rI
••• 0

u / 0
/ -
Lodge (199 )
I I I I
0·00~~-10-0----20~0----30....o__......_____.400
100 200 300 400
Stress-Corrected Shear Wave Stress-Corrected Shear Wave
Velocity, Vs 1 , mis Velocity, Vs1, mis
(a) Mw = 7.5 (b)Mw =7

Figure 8 CSR Charts Based on Corrected Shear-Wave Velocities Suggested by

(a) Robertson et al. (1992), and (b) Kayen et al. (1992) and Lodge (1994)
(After Andrus and Stokoe, this report)

Criteria for Evaluating Liquefaction Resistance

Robertson et al. (1992) proposed a stress-based liquefaction assessment procedure using field
performance data from sites in the Imperial Valley, California. These investigators normalized Vs
by:

(21)

where Pa is a reference stress of I 00 kPa, approximately atmospheric pressure, and a' vo is effective
overburden pressure in kPa. Robertson et al. chose to modify V5 in terms of a' vo to follow the
traditional procedures for modifying standard and cone penetration test resistances. The liquefaction
resistance bound (CRR curve) determined by these investigators for magnitude 7.5 earthquakes is
plotted on Figure 8a along with data calculated from several field sites where liquefaction did or did
not occur. The cyclic stress ratios were calculated using estimates of a,,..2 x for the larger of two
horizontal components of ground acceleration that would have occurred at the site in the absence of
liquefaction.

Subsequent liquefaction resistance boundaries proposed by Kayen et al. (1992) and Lodge (1994)
for magnitude 7 earthquakes are shown on Figure 8b. These curves are based on field performance

19
 o.s.-------r------r------r----~---~---~
Boundaries Defined by:
CRR = a(Vs1/100) 2 + b[l/(Vs1c-Vs1)-lNs1c] Mw = 7.5
a= 0.03, b = 0.9
Vs1c = 220 mis for sands and gravels with PC::;; 5%
Vs 1c = 210 mis for sands and gravels with PC"' 20%
Vs1c = 200 mis for sands and silts with PC~ 35%
Chart Based on:
0.6 Average values ofVs1& a max
Uncemented, Holocene-age soils
-0 Average fines content (PC)
6....... FC~35 20.::;;5
;a,-
~ II I
III
·-p::.0....
c,:l
III
0.4
"'"'
<!)
.....
l-<
v.i

·--(.)

( .)
>-. Liquefaction
u

No
Liquefaction

0
--
0.0 .....--=-=-..::.-___._ _ ____._ _ ____._ _ ____,JL.,.__ _ _,L__ _--1
100 200 300
Stress-Corrected Shear Wave Velocity, Vs1 , mis

Figure 9 Curves Recommended by Workshop for Calculation of CRR

from Corrected Shear Wave Velocity (After Andrus and Stokoe, this report)

data from the 1989 Loma Prieta earthquake. With few exceptions, the liquefaction case histories are
bounded by the relationships proposed by these suggested bounds. The relationship proposed by
Lodge (1994) provides a conservative lower boundary for liquefaction case histories with V51 less
than about 200 mis. The relationship by Robertson et al. (1992) is the least conservative of the three.

20
Professor Ricardo Dobry suggested a relationship between cyclic resistance ratio and Vs 1 for constant
average cyclic shear strain, Yav, of the form:

(22)

where y av is constant average shear strain. This formula supports a CRR bound passing through the
origin and provides a rational approach for extrapolating beyond the limits of the available field
performance data, at least for lower values ofVs 1 (Vs, ~ 125 mis).

For higher values ofVs 1, Andrus and Stakoe (this report) reason that the CRR bound should become
asymptotic to some limiting Vs 1 value. This limit is caused by the tendency of dense granular soils
to exhibit dilative behavior at large strains. Thus Equation 22 is modified to:

(23)

where Vs1c is the critical value ofV 51 , which separates contractive and dilative behavior, and a and
b are curve fitting parameters.

Using the relationship between Vs, and CRR expressed by Equation 23, Andrus and Stokoe drew
curves to separate data from sites where liquefaction effects were and were not observed. Best fit
values for the constants a and b were 0.03 and 0.9, respectively, for magnitude 7.5 earthquakes.
Andrus and Stakoe also determined the following best-fit values for V510 :

Vs,c = 220 mis for sands and gravels with fines contents less than 5%
Vs,c = 210 m/s for sands and gravels with fines contents of about 20 %
Vsic= 200 mis for sands and gravels with fines contents greater than 35%

Figure 9 presents CRR boundaries recommended by Andrus and Stakoe for magnitude 7.5
earthquakes and uncemented Holocene-age soils with various fines contents. Although these
boundaries pass through the origin, natural alluvial sandy soils with shallow water tables rarely have
corrected shear wave velocities less than 100 mis, even near ground surface. For a V 5I of 100 mis
and a magnitude 7.5 earthquake, the calculated CRR is 0.03. This minimal CRR value is generally
consistent with intercept CRR values for the CPT and SPT procedures.

Equation 23 can be scaled to other magnitude values through use of magnitude scaling factors.
These factors are discussed in a later section of this paper.

Becker Penetration Tests

Liquefaction resistance of non-gravelly soils has been evaluated primarily through CPT and SPT,
with occasional Vs measurements. CPT and SPT measurements, however, are not generally reliable
in gravelly soils. Large gravel particles may interfere with the normal deformation of soil materials
around the penetrometer increasing penetration resistance. In an attempt to surmount these
difficulties, several investigators have employed large-diameter penetrometers. The Becker
penetration test (BPT) has become one of the more effective and widely used of these type of tools.

21
The BPT was developed in Canada in the late 1950s and consists of a 3-meter-long double-walled
casing driven into the ground with a double-acting diesel-driven pile hammer. The hammer impacts
are applied at the top of the casing and penetration is continuous. The Becker penetration resistance
is defined as the number of blows required to drive the casing through an increment of 300 mm.

The BPT has not been standardized and several different types of equipment and procedures have
been used. Also, only a few BPT blow counts have been measured at sites where liquefaction has
occurred. Thus the BPT is not correlated directly with liquefaction resistance, but is used to estimate
equivalent SPT blow counts though empirical correlation. The equivalent SPT blow count is then
used to estimate liquefaction resistance.

To provide uniformity, Harder and Seed (1986) recommend employment of newer AP-1000 drill rigs
equipped with supercharged diesel hammers, 168-mm O.D. casing, and a plugged bit. From several
sites where both BPT and SPT tests were conducted in parallel soundings, Harder and Seed (1986)
developed a preliminary correlation between Becker and standard penetration resistance (Figure
I 0a). Additional comparative data compiled since 1986 are plotted on Figure I Ob. The original
Harder and Seed correlation curve (solid line) is drawn on Figure 10b along with dashed curves
representing 20% over- and under-predictions of SPT blow counts. These plots indicate that SPT
blow counts can be roughly estimated from BPT measurements.

A major source of variation in BPT blow counts is deviations in hammer energy. Rather than
measuring hammer energy directly, Harder and Seed (1986) monitored bounce-chamber pressures
and found that uniform combustion conditions (e.g., full throttle with a supercharger) correlated
rather well with variations in Becker blow count. From this information, Harder and Seed (this
report) developed an energy correction procedure based on measured bounce-chamber pressure.

Direct measurement of transmitted hammer energy could provide a more theoretically rigorous
correction for Becker hammer efficiency. Sy and Campanella (1994) and Sy et al. (1995)
instrumented a small length of Becker casing with strain gages and accelerometers in an attempt to
measure transferred energy. They analyzed the recorded data with a pile-driving analyzer to
determine strain, force, acceleration, and velocity. The transferred energy was determined by time
integration of force times velocity. They were able to verify many of the variations in hammer
energy previously identified by Harder and Seed (1986), including effects of variable throttle settings
and energy transmission efficiencies of various drill rigs. However, they were not able to reduce the
scatter or uncertainty in converting BPT blow counts to SPT blow counts. Because the Sy and
Campanella procedure requires considerably more effort than monitoring of bounce-chamber
pressure without producing greatly improved results, the workshop participants agreed that the
bounce-chamber technique appears adequate for routine practice.

Friction along the driven casing also influences penetration resistance. Harder and Seed (1986) did
not evaluate the effect of casing friction; hence the correlation in Figure 1Ob intrinsically
incorporates casing friction. Casing friction, however, remains a concern for depths greater than 30
m and for measurement of penetration resistance in soft soils underlying thick deposits of dense soil.
Either of these circumstances could lead to greater casing friction than is intrinsically incorporated
in the Seed and Harder correlation.

22
 'g'
e.8
~
80r------r----...------r----T-----.----..---~
is (a) ♦
0 ♦
:0
'-'
0 601-----+----1-----+----+,.,::;_--l-----l-----l
'D
z,._J'
s::
6u 401-----l----1---.:::......~----1--------l----i-----l
is
.......
0 ·~~~~
CO (silt & sand)
t
r/J.
201--._..,._--+f-="-=--+------,1------+------I ■ Thermalito test site
(sand)
] ♦ San Diego test site
.., (silty sand & sand)
u
Q)

g 20 40 60 80 100 120 140

u Corrected Becker Blowcount, NBc (blows/foot)

,-.,_
.....
0
e.8 80
...__ (b)
c:l'l ♦
is
0
.......
.D
,_, 60
0 II ¢
'D
z,._J' ◊
Test Sites
s:: A Salinas test site
=
u
0 40 l------ll------,l-;....-7'~~a..,,-~.....::::.....i.---l ■ Thermalito test site
♦
San Diego test site
is0 □ Jackson Lake site A
....... □ Jackson Lake site H
i:o
■ Squamish FMC test site
t
r/J.
20 ~;;...a:=---+----+----+----1 11 MacDonalds farm site

◊ Duncan Darn toe site

"'O ◊ Duncan Darn crest site
.....
Q)
u ¢ Annacis north pier
~0 0
0 20 40 60 80 100 120 140
u
Corrected Becker Blowcount, NBc (blows/foot)

Figure 10 Correlation Between Corrected Becker Penetration Resistance, NBC,

and Corrected SPT Resistance, N 60, (a) Harder and Seed (1986); (b) Data from
Additional Sites (After Harder, this report)

23
The following procedures are recommended for routine practice: (1) The BPT should be conducted
with newer AP-1000 drill rigs equipped with supercharged diesel hammers used to drive plugged
168-mm O.D. casing. (2) Bounce-chamber pressures should be used to adjust measured BPT blow
counts to Nbc to account for variations in diesel hammer combustion efficiency. For most routine
applications, correlations developed by Harder and Seed ( 1986) may be used for these adjustments.
(3) The influence of casing friction is intrinsically accounted for in the Harder and Seed BPT-SPT
correlation. This correlation, however, has not been verified and should not be used for depths
greater than 30 meters or for sites with thick dense deposits overlying loose sands or gravels. For
these conditions, mudded boreholes may be needed to reduce casing friction, or sophisticated wave-
equation analyses may be applied to quantify frictional effects.

Magnitude Scaling Factors

In developing the simplified procedure, Seed and Idriss (1982) compiled a sizable data base from
sites where liquefaction did or did not occur during earthquakes with magnitudes near 7 .5. Analyses
were made of these data to calculate cyclic stress ratios (CSR) and (N 1) 60 values for various sites
where surface effects ofliquefaction were or were not observed. Results from clean sand sites (fines
content~ 5 percent) were plotted on a CSR versus (N 1) 60 plot. An updated version of that plot (Seed
et al., 1985) is reproduced in Figure 2. A deterministic curve was drawn through the plot to separate
regions with data indicative of liquefaction (solid symbols) from regions with data indicative of
nonliquefaction (open symbols). Where there was a mixture of data, the curve was conservatively
placed to ensure that data indicative of liquefaction plot above or to the left of the bounding curve.
This curve, termed the simplified base curve or CRR75 curve, is relatively well constrained by
empirical data between CSR of 0.08 and 0.35 and is logically extrapolated to higher and lower values
beyond that range. As shown in Figure 2, the workshop participants recommend bowing the lower
part of the simplified base curve to intersect the ordinate of the plot at a CRR of about 0.05.

To adjust the simplified base curve to magnitudes smaller or larger than 7.5, Seed and Idriss (1982)
introduced correction factors called "magnitude scaling factors." These factors are used to scale the
simplified base curve upward or downward on the CSR versus (N 1) 60 plot. Conversely, magnitude
weighting factors, which are the inverse of magnitude scaling factors, may be applied to correct CSR
for magnitude. Either correcting CRR via magnitude scaling factors, or correcting CSR via
magnitude weighting factors, leads to the same final result. Because the original papers by Seed and
Idriss were written in terms of magnitude scaling factors, the use of magnitude scaling factors is
continued in this report.

To illustrate the influence of magnitude scaling factors on calculated hazard, the equation for factor
of safety (FS) against liquefaction can be written in terms of CRR, CSR, and MSF as follows:

FS = (CRR7 /CSR)MSF (24)

where CRR75 is the cyclic resistance ratio determined for magnitude 7.5 earthquakes using Figure 2
or Equation 4 for SPT data, Figure 3 or Equation 11 for CPT data, or Figure 9 or Equation 23 for Vs 1
data. Equation 24 demonstrates that the factor of safety against development of liquefaction at a site
is directly proportional to the magnitude scaling factor selected.

24
 1.0.---------,,-----,--r---..------,
::t ::t
('<) -
\0 ID' r--' 00'
0.8 11 11 11 11
~ ~ ~ ~
.....0
~
0:.-:
<Zl
<Zl
~
0.6

l ll
......
'(/)

-u
.....u
0.4

1 itt 1--I
l lT
u
>. 0.2

089
o...________..,___ ____,_--'----'--------1
1 6 10 15 26 100
Number of Cycles to Cause ru = 100% and± 5% Strain
Figure 11 Representative Relationship Between CSR and Number of Cycles
To Cause Liquefaction and (After Seed and Idriss, 1982)

Seed and Idriss (1982) Scaling Factors

Because of the limited empirical data available in the 1970s, Seed and Idriss (1982) were unable to
narrowly constrain bounds between liquefaction and nonliquefaction regions on CRR plots for
magnitudes other than 7 .5. Consequently, they based their scaling factors on representative loading
cycles and laboratory test results. From a study of strong-motion accelerograms, the number of
representative loading cycles generated by an earthquake was correlated with earthquake magnitude.
For example, magnitude 7.5 earthquakes were characterized by 15 loading cycles, whereas,
magnitude 8.5 earthquakes were characterized by 26 loading cycles, and magnitude 6.5 earthquakes
by 10 loading cycles. Second, laboratory tests were conducted to measure the number of loading
cycles required to generate liquefaction and five percent cyclic strain. Laboratory tests were
conducted using a variety of clean sands, void ratios, and ambient stress conditions. From these
tests, a single representative curve was developed that relates cyclic stress ratio to the number of
loading cycles required to generate liquefaction (Figure 11). By dividing CSR values from this curve
for various numbers of cycles, representative of various earthquake magnitudes, by the CSR for 15
cycles (magnitude 7.5), the initial set of magnitude scaling factors was derived. These scaling factors
are listed in Column 2 of Table 3 and are plotted on Figure 12. These magnitude scaling factors have
been routinely applied in engineering practice since their introduction in 1982.

Idriss Scaling Factors

In preparing his H.B. Seed memorial lecture, I.M. Idriss reevaluated the data that he and the late
Professor Seed had used to calculate the original (1982) magnitude scaling factors. In so doing,
Idriss re-plotted the data on a log-log plot and found that the data plotted as a straight line. He
further noted that one outlier point had strongly influenced the original analysis, causing the original

25
 Table 3. Magnitude Scaling Factor Values Defined by Various Investigators
(from Youd and Noble, Magnitude Scaling Factors, this report)

Mag- Seed and Idriss Ambraseys Arango Andrus and Youd and Noble
nitude, Idriss ( 1988) ( 1996) Stokoe (this report)
M (I 982) (in press) PL<20% PL<32% PL<50%
(I) (2) (3) (4) (5) (6) (7) (8) (9) ( 10)

5.5 1.43 2.20 2.86 3.00 2.20 2.8 2.86 3.42 4.44

6.0 1.32 1.76 2.20 2.00 1.65 2.1 1.93 2.35 2.92

6.5 1.19 1.44 1.69 1.60 1.40 1.6 1.34 1.66 1.99

7.0 1.08 1.19 1.30 1.25 1.10 1.25 1.00 1.20 1.39

7.5 1.00 1.00 1.00 1.00 1.00 1.00 1.00

8.0 0.94 0.84 0.67 0.75 0.85 0.8? 0.73?

8.5 0.89 0.72 0.44 0.65? 0.56?

4.5
-+- Seed and Idriss, (1982)
4 -+------'s,---=+--~-----....., ------ Idriss

3.5
x Ambraseys ( 1985)
Workshop ◊ Arango (1996)
3 ♦ Arango ( 1996)
--+- Andrus and Stakoe
2.5 • Youd and Noble, PL<20%
1:!,. Youd and Noble, PL<32%
2 • Youd and Noble, PL<50%

1.5

0.5

0
5.0 6.0 7.0 8.0 9.0
Earthquake Magnitude, Mw

Figure 12 Magnitude Scaling Factors Derived by Various Investigators

(After Youd and Noble, Magnitude Scaling Factors, this report)

26
plot to be nonlinear and characterized by unduly low values for magnitudes less than 7.5. Based on
this reevaluation, Idriss defined a new set of magnitude scaling factors. These factors are listed in
Column 3 of Table 3, plotted on Figure 12, and are defined by the following equation:

MSF = I O2.24;M2.56 (25)

Idriss recommends these revised scaling factors for use in engineering practice in place of the
original factors.

The revised scaling factors are significantly larger than the original scaling factors for magnitudes
less than 7.5 and somewhat smaller than the original factors for magnitudes greater than 7.5.
Relative to the original scaling factors, the revised factors lead to a reduced calculated liquefaction
hazard for magnitudes less than 7.5 and increased calculated hazard for magnitudes greater than 7.5.

Ambraseys Scaling Factors

Field performance data collected since the 197Os for magnitudes less than 7.5 indicate that the
original Seed and Idriss (1982) scaling factors may be overly conservative. For example, Ambraseys
(1988) analyzed liquefaction data compiled through the mid-198Os and plotted calculated cyclic
stress ratios for sites that did or did not liquefy on CSR versus (N 1) 60 plots. From these plots,
Ambraseys developed empirical exponential equations that define CRR as a function of (N 1) 60 and
moment magnitude, Mw. By holding the value of (N1) 60 constant in the equations and taking the ratio
of CRR determined for various magnitudes of earthquakes to the CRR for a magnitude 7 .5
earthquakes, Ambraseys derived the magnitude scaling factors listed in Column 4 of Table 3. These
factors are also plotted on Figure 12. For magnitudes less than 7.5, the MSF suggested by
Ambraseys are significantly greater than both the original factors developed by Seed and Idriss
(Column 2, Table 3) and the revised factors by Idriss (Column 3). Because they are based on
observational data, these factors have validity for estimating liquefaction hazard; however, they have
not been widely used in engineering practice. Conversely, for magnitudes greater than 7.5,
Ambraseys factors are significantly lower than the original (Seed and Idriss, 1982) and Idriss's
revised scaling factors. Because there are little data to constrain Ambraseys' scaling factors for
magnitudes greater than 7.5, these factors are uncertain, are likely overly conservative, and are not
recommended for engineering practice.

Arango Scaling Factors

Arango (1996) developed two sets of magnitude scaling factors. The first set (Column 5, Table 3)
is based on farthest observed liquefaction effects from the seismic energy source, the estimated
average peak accelerations at those distant sites, and the absorbed seismic energy required to cause
liquefaction. The second set (Column 6, Table 3) was developed from energy concepts and the
relationship derived by Seed and Idriss (1982) between numbers of significant stress cycles and
earthquake magnitude. The MSF listed in Column 5 are similar in value (within about 10%) to the
MSF of Ambraseys (Column 4), and the MSF listed in Column 6 are similar in value (within 6%)
to the revised MSF proposed by Idriss (Column 3).

27
Andrus and Stokoe Scaling Factors

From their studies of liquefaction resistance as a function of shear wave velocity, V Andrus and
5
,

Stakoe (this report) developed Equation 23 for calculating CRR from V, for magnitude 7.5
earthquakes. Using this equation, Andrus and Stokoe drew curves on graphs with plotted values of
CSR as a function ofV 51 from sites where surface effects ofliquefaction were or were not observed.
Graphs were plotted for sites shaken by magnitude 6, 6.5, 7, and 7.5 earthquakes. The positions of
the CRR curves were visually adjusted on each graph until a best fit bound was obtained. Magnitude
scaling factors were then estimated by taking the ratio ofCRR for a given magnitude to the CRR for
magnitude 7.5 earthquakes. These MSF were then fitted to the following exponential function

MSF = (Mj7.5)"3 3 (26)

Values for magnitudes less than 6 and greater than 7.5 were extrapolated from this equation. MSF
values from this analysis are listed in Column 7, Table 3, and plotted on Figure 12. For magnitudes
less than 7.5, the MSF proposed by Andrus and Stokoe are rather close in value (within about 5
percent) to the MSF proposed by Ambraseys. For magnitudes greater than 7.5, the Andrus and
Stakoe MSF are slightly smaller than the revised MSF proposed by Idriss.

Youd and Noble Scaling Factors

Youd and Noble (Magnitude Scaling Factors, this report) used a logistic analysis to analyze case
history data from sites where effects of liquefaction were or were not reported following past
earthquakes. This analysis yielded the following probabilistic equation:

Legit (PL)= ln(P/(1-PL)) = -7.633 + 2.256 Mw - 0.258 (N 1\ocs + 3.095 In CRR (27)

where PL is the probability that liquefaction occurred, I-PL is the probability that liquefaction did not
occur, and (N 1)60c, is the corrected blow count, including the correction for fines content. Youd and
Noble recommend direct application of this equation to calculate the CRR for a given probability
ofliquefaction occurrence. In lieu of direct application, Youd and Noble define MSF for use with
the simplified procedure. These MSF were developed by rotating the simplified base curve to near
tangency with the probabilistic curves for PL of 50%, 32%, and 20% and various earthquake
magnitudes. These MSF are defined as the ratio of the ordinate of the rotated base curve at the point
of near tangency to the ordinate of the unrotated simplified base curve at the same (N 1)6ocs· Because
the rotated simplified base curves lie entirely below the given probability curve, CRR calculated with
these MSF are characterized by smaller probability of liquefaction occurrence than the associated
probabilistic curves. Thus the MSF listed in Columns 8, 9, and 10 (Table 3), are denoted by
PL<50%, PL<32%, and PL<20%, respectively. Because the derived MSF are less than 1.0, Youd and
Noble do not recommend use ofMSF for PL<32% and PL<20% for earthquakes with magnitudes
greater than 7 .0. Equations for defining the Youd and Noble MSF are listed below:

Probability, PL< 20% MSF = 103.s1;M4.53 ForM <7 (28)

Probability, PL< 32% MSF = l03.74JM433 ForM<7 (29)
Probability, PL< 50% MSF = 104.21JM48I ForM < 7.75 (30)

28
Recommendations for Engineering Practice

The workshop participants reviewed the MSF listed in Table 3 and all but one (S.S.C. Liao) agree
that the original factors were too conservative and that an increase is warranted for engineering
practice for magnitudes less than 7.5. Rather than recommending a single set of factors, the
workshop participants suggest a range of MSF with the engineer allowed to choose factors from
within that range requisite with the conservatism required for the given application. For magnitudes
less than 7.5, the lower bound for the recommended range is the revised set of magnitude scaling
factors proposed by Idriss (Column 3, Table 3, or Equation 23). The upper bound for the suggested
range is the MSF proposed by Andrus and Stokoe (Column 7, Table 3, or Equation 26). The upper
bound values are consistent with MSF suggested by Ambraseys, Arango, and Youd and Noble for
PL <20% (generally within about 10 percent).

For magnitudes greater than 7.5, the factors recommended by Idriss (Column 3, Table 3;
Equation 25) should be used for engineering practice. Above magnitude 7.5, these factors are
smaller than the original Seed and Idriss (1982) factors, and hence application of the new factors
leads to increased calculated liquefaction hazard compared to the original factors. The reasoning for
this recommendation is that the original factors by Seed and Idriss (1982) may not have been
sufficiently conservative for magnitudes greater than 7.5. There are insufficient case history data for
earthquakes with magnitudes greater than 8 to support use of the lower MSF values listed in Table
3. These lower values were generally extrapolated from smaller magnitude earthquakes. Thus, these
more conservative MSF are not recommended for engineering practice.

Corrections for High Overburden Pressures, Static Shear Stresses,

and Age of Deposit

The correction factors K0 and Ka were developed by Seed (1983) to adjust cyclic resistance ratios
(CRR) to static overburden and shear stresses larger than those embodied in the development of the
simplified procedure. As noted, the simplified procedure is only valid for level to gently sloping
sites (low static shear stress) and depths less than about 15 m (low overburden pressures). The K 0
correction factor extends cyclic ratios to high overburden pressures, while the Ka correction factor
allows extension of the simplified procedure to more steeply sloping ground conditions. Because
there are virtually no case histories available to help define these correction factors, the results from
laboratory test programs have been used to develop corrections for engineering practice.

K0 Correction Factor

Cyclically loaded, isotropically consolidated triaxial compression tests show that while liquefaction
resistance of a soil increases with increasing confining pressure, the resistance, as measured by the
cyclic stress ratio, is a nonlinear function that decreases with increased normal stress. To incorporate
the nonlinear effect of decreasing cyclic stress ratio with increasing confining pressure, Seed (1983)
recommended incorporation of a correction factor, K0 , for overburden pressures greater than 100
kPa. This factor allows correction of results obtained from the simplified procedure to overburden
pressures that are greater than those generally extant in the observational data base from which the
procedure was derived. Because of the lack of case history data, extrapolation of the simplified

29
 1.2
-~
I>
1 - .
n

t;i- ---
'
······· ······· ······
······· ······· ··X
0.8 • ~
: .: : .= .= : : :ii!'••· ··:::,.
~-~~~-,□-
..
.......... ::::::. :.~
Ka '---
_ r--....
I'- -
• A
- - -
- - - - - - "<:Jean Sand"
0.6 "• Sardis
Sardis Darn Shell
Darn Foundation See :-:-:--
□ Sac. River Sand dalld :--_
■ Monterey O Sand l!arder (1990--
0.4 ◊Reid Bedford Sand
♦ New Jersey Backfill
llE Tailings Sand Dr=60
0.2 ~ Tailings Sand Dr=65
:£: Ottowa Sand Dr=60 I1.0 TSF = 96 kPa I
*Ottowa Sand Dr=70 '
1 2 3 4 5 6 7 8 9 10
Effective Confining Pressure (tsf)
Figure 13 K0 values determined by various investigators (After Seed and Harder, 1990)

1.2

1
"-.. r--.....
Ka
0.8

0.6
........_
-
0.4

0.2
I1.0 TSF = 96 kPa I
I I I

1 2 3 4 5 6 7 8 9 10
Effective Confining Pressure (tsf)
Figure 14. Minimum Values for K,, Recommended for Clean and Silty Sands
and Gravels (After Harder and Boulanger, this report)

procedure to depths greater than 15 m using K0 factors yields results, such as factors of safety, that
are less certain than at shallower depths.

The K0 values developed by Seed (1983) were obtained by normalizing cyclic resistance ratios of
isotropically consolidated cyclic triaxial compression tests to CRR values associated with an
effective confining pressure of 100 kPa. For confining pressures greater than 100 kPa, the K.,
correction factor is less than one and decreases with increasing pressure. The original analyses by
Seed (1983) yielded a band of suggested Ka factors that decreased approximately linearly with
effective overburden pressure from a value of 1.0 at 100 kPa to values ranging from about 0.40 to
0.65 at 800 kPa (Figure 13). Seed and Harder (1990) analyzed additional data and suggested

30
generally lower values that are defined by a single concave curve with a K., value of 0.44 at an
effective confining pressure of 800 kPa. Vaid et al. (1985) and Vaid and Thomas (1994) performed
constant-volume cyclic simple shear tests on clean sands and derived smaller decreases in K.,. From
tests on mine-tailing, Ottawa, and Frazer Delta sands (Figure 13), several investigators (Byrne and
Harder, 1991; Pillai and Byrne, I 994; Arango, 1996) calculated values for K., ranging from about
0.75 to 0.90 for effective overburden pressures of 1,000 kPa to minimal values of 0.67 for effective
overburden pressures as great as 2,600 kPa. These analyses indicate that lower relative densities
generally produced higher K., values. The various analyses confirm the considerable variability in
derived K0 values, and that the factors developed by Seed and Harder were overly conservative.

Based on the above discussion and a review oftest results presented by Harder and Boulanger (this
report), the workshop participants gained consensus that the Seed and Harder K0 values were too
conservative and that an increase is recommended for general engineering practice. Based on this
review, the workshop recommended K0 values represented by the curve in Figure 14 as minimal
values for engineering practice for both clean and silty sands and for gravels.

K" Correction Factor for Sloping Ground

Sloping ground induces static shear within the body of a soil mass before the onset of earthquake
shaking. The relative magnitude of the static shear, ',i, on the horizontal plane can be assessed by
normalizing it with respect to the effective vertical stress, a' vo· The resulting parameter is called the
alpha ratio, where a= ,,la'vo· For level ground conditions, the alpha ratio is zero. Early researchers
suggested that the presence of a static shear stress always improved the cyclic resistance of a soil
because higher cyclic shear stresses were required to cause stress reversal. This conclusion is true
for dense soils under relatively low confining pressures. However, loose soils and some soils under
high confining pressures have lower liquefaction resistance under the influence of initial static shear
stresses than in the absence of these stresses. This behavior is due to the potential strain softening
nature of very loose soils.

To incorporate the effect of static shear stresses on liquefaction resistance of soils, Seed (1983)
recommended use of a correction factor, K". This factor is used to correct results obtained from the
simplified procedure for level ground to sloping ground sites with constant static shear stress.

For the workshop, Harder and Boulanger (this report) reviewed past publications, tests, and analyses
relative to K". They concluded that the wide ranges in potential K., values developed by past
investigators indicate a lack of consensus and a need for continued research and field verification
of the effects of static shear stress on liquefaction resistance. Different rates of pore pressure
generation and different limiting values for excess pore pressure at different locations within a slope
make liquefaction analyses for sloping ground conditions an extremely complicated endeavor.

The workshop participants agreed that the evaluation of liquefaction resistance beneath sloping
ground or embankments (slopes greater than about six percent) is not well understood and that such
evaluations are beyond routine application of the simplified procedure. Although curves relating K"
to a have been published (Harder and Boulanger, this report), the participants concluded that general
recommendations for use ofK" by the engineering profession is not advisable at this time.

31
Influence of Age of Deposit

Several investigators have shown that liquefaction resistance of soils increases with age. For
example, Seed (1979) observed significant increases in liquefaction resistance with age of
reconstituted sand specimens tested in the laboratory. Cyclically loaded tests were conducted on
freshly reconstituted sand specimens and on similar sand specimens at periods ranging up to one
hundred days. Increases of as much as 25 percent in cyclic resistance ratio were noted between the
freshly constituted and the 100-day-old specimens. Youd and Hoose (1977) and Youd and Perkins
(1978) note that liquefaction resistance increases markedly with geologic age. Sediments deposited
within the past few hundred years are generally much more susceptible to liquefaction than older
Holocene sediments; Pleistocene sediments are even more resistant; and Pre-Pleistocene sediments
are essentially insusceptible to liquefaction. Although qualitative increases in liquefaction resistance
have been well documented, insufficient quantitative data have been assembled from which
correction factors for age can be defined.

The age, and concomitantly the liquefaction resistance, of naturally sedimented deposits generally
increases with depth. In natural soils, this increase may partially or wholly counteract the influence
of the Ka factor which generates an apparent decrease in liquefaction resistance with depth. In the
absence of quantitative correction factors for age, engineering judgement is required in assessing
liquefaction resistance of sediments older than a few hundred years. In some instances where deeper
sediments have been dated as more than a few thousand years old, knowledgeable engineers have
ignored the Ka factor as partial compensation for unquantifiable, but known increases in liquefaction
resistance with age. For man-made structures, such as thick fills and embankment dams, ageing
effects are generally minimal and should be ignored in calculating liquefaction resistance.

Seismic Factors

Application of the simplified procedure for evaluating liquefaction resistance requires estimates of
earthquake magnitude and peak horizontal ground acceleration. In the procedure, these factors
characterize duration and intensity of ground shaking, respectively. The workshop addressed the
following questions with respect to selection of magnitude and peak acceleration.

Earthquake Magnitude

Records from past earthquakes indicate that the relationship between duration and magnitude is
rather uncertain and that factors other than magnitude influence duration. For example, unilateral
faulting, in which rupture begins at one end of the fault and propagates to the other, usually produces
longer shaking duration for a given magnitude than bilateral faulting, in which slip begins near the
midpoint on the fault and propagates in both directions. Duration also generally increases with
distance from the seismic energy source and may vary with site conditions and with bedrock
topography (basin effects). The workshop addressed the following questions with respect to the use
of magnitude as an index for shaking duration, and developed the following consensus answers.

Question: Should correlations or correction factors be developed to adjust duration of shaking to

account for the influence of earthquake source mechanism and other factors?

32
Answer: Faulting characteristics and variations in shaking duration are difficult to predict in advance
of an earthquake event. The influence of distance is generally of secondary importance within the
range of distances to which potentially damaging effects of liquefaction commonly develop. Basin
effects are not yet sufficiently predictable to be adequately accounted for in engineering practice.
Thus workshop participants recommend continued use of conservative relationship between
magnitude and duration embodied in the simplified procedure for routine evaluation ofliquefaction
resistance.

Question: An important difference between eastern US earthquakes and western US earthquakes

is that eastern ground motions are generally richer in high frequency energy and thus could generate
more significant stress cycles and equivalently longer durations than western earthquakes of the same
magnitude. Should a correction be made to account for higher frequencies of ground motions
generated by eastern US earthquakes?

Answer: The high-frequency motions of eastern earthquakes are generally limited to rock sites.
High-frequency motions attenuate or are damped out rather quickly as they propagate through soil
layers. This filtering action reduces the high-frequency energy at soil sites and should reduce
differences in numbers of significant loading cycles between eastern and western earthquakes.
Because liquefaction occurs only within soil strata, duration differences on soil sites between eastern
and western earthquakes are not likely to be great. Without more instrumentally recorded data from
which differences in ground motion characteristics can be quantified, there is little basis for the
development of additional correction factors for eastern localities.

Another difference between eastern and western US earthquakes is that strong ground motions
generally propagate to greater distances in the east than in the west. By applying present state-of-the-
art procedures for estimating peak ground acceleration at eastern sites, differences in ground motion
propagation between western and eastern earthquakes are properly accounted for.

Question: Which magnitude scale should be used by engineers in selecting a magnitude for use in
liquefaction resistance analyses?

Answer: Seismologists commonly calculate earthquake magnitudes using five different scales: (1)
local or Richter magnitude, ML; (2) surface-wave magnitude, Ms; (3) short-period body-wave
magnitude, m 0; (4) long-period body-wave magnitude, 111s; and (5) moment magnitude, M,_,. Moment
magnitude is the scale most commonly used for engineering applications and is the scale preferred
for calculation ofliquefaction resistance. As shown on Figure 15, magnitudes from other scales may
be substituted directly'for M,., within the following limits: ML< 6, m 8 < 7.5, and 6 <Ms< 8. mb, a
scale commonly applied in the eastern US, may be used for magnitudes between 5 and 6, provided
such magnitudes are corrected to Mw using the curves plotted in Figure 15 (Idriss, 1985).

Peak Acceleration

In the simplified procedure, peak horizontal acceleration (¾lax> is used to characterize the intensity
of ground shaking. To provide guidance for estimation of ¾lax, the workshop addressed the
following questions.

33
 9

~ ----
---
----- Ms

d: ~-
8 MJMA-
----
----
~---- ------ ------
.,.-
,......._-- mB

~ .,.- ML
7 .,
#/2/;,zi ,., ~------
/2
:)/ .,,/
i...---- ---·- -----
mb

I/✓¼-~- 0 ,.,,
/4, /

✓- r✓--:--J
',,f /
A
A
/ IScale I Magnitude I
AA I
~
A
I ML Local or Richter
4
✓f
~yAA
~// Ms Surface wave

I
3 / mb Short-period body wave

I/
mB Long-period body wave

MJMA Japanese Meteorological Agency

3 4 5 6 7 8 9 10
Moment Magnitude, Mw
Figure 15 Relationship between Moment, Mw, and Other Magnitude Scales
(After Heaton et al., 1982)

Question: What procedures are preferred for estimating amax at potentially liquefiable sites?

Answer: The following three methods, in order of preference, may be used for estimating 8max:

(1) The preferred method for estimating ~ax at a site is through application of empirical
correlations for attenuation of ~ax as a function of earthquake magnitude, distance from the
seismic energy source, and local site conditions. Several correlations have been developed
for estimating ~ax for sites on bedrock or stiff to moderately stiff soils. Preliminary
attenuation relationships have also been developed for soft soil sites (Idriss, 1991 ). Selection
ofan attenuation relationship should be based on factors such as region of the country, type
of faulting, site condition, etc.

(2) For soft sites and other soil profiles that are not compatible with available attenuation
relationships, ~ax may be estimated from local site response analyses. Computer programs
such as SHAKE, DESRA, etc., may be used for these calculations. Input ground motions

34
 in the form of recorded accelerograms are preferable to synthetic records. Accelerograms
derived from white noise should be avoided. A suite of plausible earthquake records should
be used in the analysis, including as many records as feasible from earthquakes with similar
magnitudes, source distances, etc.

(3) The third and least desirable method for estimating peak ground acceleration is through
amplification ratios, such as those developed by Idriss (1990; 1991 ), Seed et al.(l 994), and
BSSC (1994). These factors use a multiplier or ratio by which bedrock outcrop motions are
amplified to estimate motions at ground surface. Because amplification ratios are influenced
by strain level, earthquake magnitude, and perhaps frequency content, caution and
considerable engineering judgment are required in the application of these relationships.

Question: Which peak acceleration should be used? (a) the largest horizontal acceleration recorded
on a three-component accelerogram; (b) the geometric mean (square root of the product) of the two
maximum horizontal components; or (c) a vectorial combination of horizontal accelerations.

Answer: According to I.M. Idriss (oral communication at workshop), where recorded motions were
available, the larger of the two horizontal peak components of acceleration were used in the original
development of the simplified procedure. Where recorded values were not available, which was the
circumstance for most sites in the data base, peak acceleration values were estimated from
attenuation relationships based on the geometric mean of the two orthogonal peak horizontal
accelerations. In nearly all instances where recorded motions were used, the peaks from the two
horizontal records were approximately equal. Thus where a single peak was used that peak and the
geometric mean of the two peaks were about the same value. Based on this information, the
workshop participants concurred that use of the geometric mean is more consistent with the
derivation of the procedure and is preferred for use in engineering practice. However, use of the
larger of the two orthogonal peak accelerations would be conservative and is allowable. Vectorial
accelerations are seldom calculated and should not be used. Peak vertical accelerations are ignored
for calculation of liquefaction resistance.

Question: Liquefaction usually develops at soil sites where ground motion amplification may occur
and where sediments may soften as excess pore pressures develop. How should investigators
account for these factors in estimating peak acceleration?

Answer: The procedure recommended by the workshop is to calculate or estimate a peak

acceleration that incorporates the influence of site amplification, but neglects the influence of excess
pore-water pressure. Simply stated, the peak acceleration to be used in liquefaction resistance
evaluations is the peak horizontal acceleration that would have occurred at ground surface at the site
in the absence of increased pore-water pressure or the occurrence of liquefaction.

Question: Should high-frequency spikes (periods less than 0.1 sec) in acceleration records be
considered or ignored?

Answer: In general, short-duration, high-frequency acceleration spikes should be ignored for

liquefaction resistance evaluations. By using attenuation relationships for estimation of peak

35
acceleration, as noted above, high frequency spikes are essentially ignored because few high-
frequency peaks are incorporated in data bases from which attenuation relationships have been
derived. Similarly, ground response analyses programs such as SHAKE and DESRA generally
attenuate or filter out high-frequency spikes, reducing their influence. Where amplification ratios
are used engineering judgment should be used to determine which bedrock accelerations should be
amplified.

Energy-Based Criteria and Probabilistic Analyses

The workshop considered two additional topics: liquefaction resistance criteria based on seismic
energy passing through a liquefiable layer (Youd et al., this report) and probabilistic analyses of case
history data (Youd and Noble, Statistical and Probabilistic Analyses, this report). Although risk
analyses for several localities and facilities have been made using probabilistic criteria, the workshop
attendees agreed that probabilistic procedures are still outside the mainstream of standard practice.
Similarly, energy-based criteria need further development before recommendations can be made for
general practice. The workshop participants did agree that research and development should
continue on both of these potentially useful procedures.

Conclusions

The participants in the NCEER workshop reviewed the state-of-the-art for evaluation ofliquefaction
resistance and proposed several augmentations to that procedure that have been developed over the
past ten years. Specific conclusions, including recommended procedures and equations, are listed
within each section of this summary paper. General consensus recommendations from the workshop
are as follows:

I. Four field tests are recommended for general use in evaluating liquefaction resistance--the
cone penetration test (CPT), the standard penetration test (SPT), measurement of shear-wave
velocity (V.), and for gravelly sites, the Becker penetration test (BPT). The workshop
reviewed and revised criteria for each test to incorporate recent developments and to
maximize compatibilities between liquefaction resistances determined via the various tests.
Each field test has its advantages and limitations. The CPT provides the most detailed soil
stratigraphy and provides a preliminary estimate of liquefaction resistance. The SPT has
been used more widely and provides disturbed soil samples from which fines content and
other grain characteristics can be determined. v. measurements provide fundamental
information for evaluation of small-strain constitutive relations and can be applied at gravelly
sites where CPT and SPT may not be reliable. The BPT test has been used primarily at
gravelly sites and requires use of rough correlations between BPT and SPT. In many
instances, two or more test procedures should be applied to assure that both adequate
definition of soil stratigraphy and a consistent evaluation of liquefaction resistance is
attained.

2. The magnitude scaling factors originally derived by Seed and Idriss (1982) have proven to
be very conservative for earthquake magnitudes less than 7.5. The consensus of the
workshop was that a range of scaling factors should be recommended for engineering

36
 practice, with the lower end of the range being the revised MSF recommended by Idriss
(Column 3, Table 3), and the upper end of the range being the MSF suggested by Andrus and
Stakoe (Column 7, Table 3). These MSF are defined by Equations 25 and 26, respectively.
For magnitudes greater than 7.5, the revised factors by Idriss (Column 3, Table 3) should be
used. The latter factors are significantly more conservative than the original Seed and Idriss
(1982) factors, but the consensus was that these more conservative factors should be applied.

3. The K 0 factors suggested by Seed and Harder (1990) are too conservative for recommended
use in general engineering practice. The workshop participants recommend the K0 values
represented by the curve in Figure 14 as minimal values for engineering practice for clean
and silty sands and for gravels.

4. The workshop participants agreed that the evaluation of liquefaction resistance beneath
sloping ground or embankments (slopes greater than about six percent) is not well
understood at this time and that such evaluations are beyond the applicability of the
simplified procedure. Special expertise is required for evaluation of liquefaction resistance
beneath sloping ground.

5. Moment magnitude, Mw, should be used as an estimate of earthquake size for liquefaction
resistance calculations. No general corrections are recommended to adjust earthquake
magnitude to account for differences in duration due to source mechanism or geographic
region (eastern versus western US earthquakes).

6. The peak acceleration, Gtmax, recommended for calculation of cyclic stress ratio, CSR, is the
amax that would have occurred at the site in the absence of pore pressure increases or
liquefaction generated by the earthquake. Application of attenuation relationships
compatible with conditions at a given site is the preferred procedure for estimating Gtmax·
Where site conditions are incompatible with existing attenuation relationships, site-specific
response calculations, using programs such as SHAKE or DESRA, should be used. The least
preferable technique is application of amplification factors.

References

Andrus, R.D., 1994, "In Situ Characterization of Gravelly Soils That Liquefied in the 1983 Borah
Peak Earthquake," Ph.D. Dissertation, University of Texas at Austin, p. 533.

Ambraseys, N.N., 1988, "Engineering Seismology," Earthquake Engineering and Structural

Dynamics, Vol. 17, p. 1-105.

Arango, I., 1996, "Magnitude Scaling Factors for Soil Liquefaction Evaluations," Journal of
Geotechnical Engineering, ASCE, Vol. 122, No. 11, p. 929-936.

BSSC, 1994, NEHRP Recommended Provisions for Seismic Regulations for New Buildings, Part
2--Commentary, Federal Emergency Management Agency, 1994 Edition, Building Seismic
Safety Council, Washington, D.C., 335 p.

37
Byrne, P.M. and Harder, L.F ., jr., 1991, "Terzaghi Dam, Review of Deficiency Investigation, Report
No. 3," prepared for BC Hydro, Vancouver, British Columbia.

Harder, L.F., Jr., and Seed, H.B., 1986, "Determination of Penetration Resistance for Coarse-
Grained Soils using the Becker Hammer Drill," Earthquake Engineering Research Center,
Report No. UCB/EERC-86/06, University of California, Berkeley.

Heaton, T.H., Tajima, F., and Mori, A.W., 1982, "Estimating Ground Motions Using Recorded
Accelerograms," unpublished report by Dames and Moore to Exxon Production Res. Co.,
Houston, Texas.

Idriss, I.M., 1985, "Evaluating Seismic Risk in Engineering Practice," Proceedings, 11th Int. Conf.
on Solid Mechanics and Foundation Engineering, San Francisco, Vol. 1, p. 255-320.

Idriss, I.M., 1990, "Response of Soft Soil Sites During Earthquakes," Proceedings, H. Bolton Seed
Memorial Symposium, Vol. 2, BiTech Publishers, Ltd, Vancouver, B.C. Canada, p. 273-290.

Idriss, I.M., 1991, "Earthquake Ground Motions at Soft Soil Sites," Proceedings, 2nd Int. Conf. on
Recent Advances in Geotechnical Earthquake Engineering and Soil Dynamics, Vol. 3, p.
2265-2271.

Kayen, R.E., Mitchell, J.K., Seed, R.B., Lodge, A., Nishio, S., and Coutinho, R., 1992, "Evaluation
of SPT-, CPT-, and Shear Wave-Based Methods for Liquefaction Potential Assessment
Using Loma Prieta Data," Proceedings, 4th Japan-US Workshop on Earthquake Resistant
Design of Lifeline Facilities and Countermeasures for Soil Liquefaction, Honolulu, Hawaii,
NCEER, Buffalo, NY, Technical Report NCEER-92-0019, Vol. 1, p. 177-204.

Kramer, S.L., 1996, Geotechnical Earthquake Engineering, Prentice Hall, Upper Saddle River, NJ,
653 p.

Liao, S.S.C. and Whitman, R. V ., 1986a, "Overburden Correction Factors for SPT in Sand," Journal
ofGeotechnical Engineering, Vol.112, No. 3, p. 373-377.

Liao, S.S.C., and Whitman, R.V., 1986b, Catalogue of A Liquefaction and Non-Liquefaction
Occurrences During Earthquakes, Research Report, Dept. Of Civil Engineering, M.I.T.,
Cambridge, MA.

Liao, S.S.C., Veneziano, D., and Whitman, R.V., 1988, "Regression Models for Evaluating
Liquefaction Probability," Journ. ofGeotechnical Engineering, Vol. 114, No. 4, p. 389-411.

Lodge, A.L., 1994, "Shear Wave Velocity Measurements for Subsurface Characterization," Ph.D.
Dissertation, University of California at Berkeley.

National Research Council (NRC), 1985, Liquefaction of Soils During Earthquakes, National
Academy Press, Washington, D.C.

38
Robertson, P.K., and Campanella, R.G., 1985, "Liquefaction Potential of Sands Using the Cone
Penetration Test," Journal ofthe Geotechnical Division, ASCE, Vol. 111, No. 3, p. 298-307.

Robertson, P.K., 1990, Soil Classification Using CPT, Canadian Geotechnical Journal, Vol 27, No.
1, p. 151-158.

Robertson P.K., and Fear, C.E., 1995, "Liquefaction of Sands and Its Evaluation," Proceedings, 1st
Int. Conf. on Earthquake Geotechnical Engineering, Keynote Lecture, Tokyo, Japan.

Robertson, P.K., Woeler, D.J., and Finn, W.D.L, 1992, "Seismic Cone Penetration Test for
Evaluating Liquefaction Potential Under Cyclic Loading," Canadian Geotechnical Journal,
Vol. 29, p. 686-695.

Seed, H.B., 1979, "Soil Liquefaction and Cyclic Mobility Evaluation for Level Ground During
Earthquakes," Journal ofGeotechnical Engineering, ASCE, Vol. 105, No. GT2, p. 201-255.

Seed, H.B., 1983, "Earthquake Resistant Design of Earth Dams," Symposium on Seismic Design of
Earth Dams and Caverns, ASCE, New York, p. 41-64.

Seed, H.B., and Idriss, J.M., 1971, "Simplified Procedure for Evaluating Soil Liquefaction
Potential," Journal of the Soil Mechanics and Foundations Division, ASCE, Vol. 97, No.
SM9, p. 1249-1273.

Seed, H.B., and Idriss, J.M., 1982, "Ground Motions and Soil Liquefaction During Earthquakes,"
Earthquake Engineering Research Institute Monograph.

Seed, H.B., Tokimatsu, K., and Harder, L.F., Jr., 1984, "The Influence of SPT Procedures in
Evaluating Soil Liquefaction Resistance," Report, UCBIEERC-84-15, University of
California at Berkeley, Berkeley, California.

Seed, H.B., Tokimatsu, K., Harder, L.F., and Chung, R.M., 1985, "The Influence ofSPT Procedures
in Soil Liquefaction Resistance Evaluations," Journal ofGeotechnical Engineering, ASCE,
Vol. 111, No. 12, p. 1425-1445.

Seed, R.B., Dickenson, S.E., Rau, G.A., White, R.K., and Mok, C.M., 1994, "Site Effects on Strong
Shaking and Seismic Risk: Recent Developments and Their Impact on Seismic Design Codes
and Practice," Proceedings, Structural Congress II, ASCE, Vol. 1, p. 573-578.

Seed, R.B., and Harder, L.F., Jr., 1990, "SPT-Based Analysis of Cyclic Pore Pressure Generation
and Undrained Residual Strength,"Proceedings, H. Bolton Seed Memorial Symposium, May,
1990. BiTech Publishers, Ltd., p. 351-376.

Skempton, A.K., 1986, "Standard Penetration Test Procedures and the Effects in Sands of
Overburden Pressure, Relative Density, Particle Size, Aging and Overconsolidation,"
Geotechnique, Vol 36, No. 3, p. 425-447.

39
Stark, T.D. and Olson, S.M., 1995, "Liquefaction Resistance Using CPT and Field Case Histories,"
Journal ofGeotechnical Engineering, ASCE, Vol 121, No. 12, p. 856-869.

Stokoe, K.H., II, Roesset, J.M., Bierschwale, J.G., and Aoouad, M., 1988, "Liquefaction Potential
of Sands From Shear Wave Velocity," Proceedings, 9th World Conf. on Earthquake
Engineering, Tokyo, Japan, Vol III, p. 213-218.

Suzuki, Y., Tokimatsu, K., Koyamada, K., Taya, Y., and Kubota, Y., 1995, "Field Correlation of
Soil Liquefaction Based on CPT Data," Proceedings of the International Symposium on
Cone Penetration Testing, CPT'95, Link.oping, Sweden, Vol. 2, p. 583-588.

Sy, A. and Campanella, R.G., 1994, "Becker and Standard Penetration Tests (BPT-SPT)
Correlations with Consideration of Casing Friction," Canadian Geotechnical Journal, Vol.
31, p. 343-356.

Sy, A., Campanella, R.G., and Stewart, R.A., 1995, "BPT-SPT Correlations for Evaluation of
Liquefaction Resistance in Gravelly Soils," Proceedings of Specialty Session on Dynamic
Properties of Gravelly Soil, ASCE Annual Convention, San Diego, California,

Tokimatsu, K., Kuwayama, S., and Tamura, S., 1991, "Liquefaction Potential Evaluation Based on
Rayleigh Wave Investigation and Its Comparison with Field Behavior," Proceedings, 2"d Int.
Conf. on Recent Advances in Geotechnical Earthquake Engineering and Soil Dynamics, held
in St. Louis, Missouri, S. Prakash, Ed., University of Missouri-Rolla, Vol. 1, p. 357-364.

Vaid, Y.P., Chem, Jing C., and Tumi, Hadi, 1985, "Confining Pressure, Grain Angularity, and
Liquefaction," Journal of Geotechnical Engineering, ASCE, Vol. 111, No. 10, October.

Vaid, Y.P., and Thomas, J., 1994, "Post Liquefaction Behaviour of Sand," Proceedings of the 13th
Int. Conf. on Soil Mechanics and foundation Engineering, New Delhi, India.

Vreugdenhil, R., Davis, R. and Berrill, J., 1994, "Interpretation of Cone Penetration Results in
Multilayered Soils," Int. Journal for Numerical Methods in Geomechanics, Vol. 18, p. 585-
599.

Youd, T.L., 1993, Liquefaction-Induced Lateral Spread Displacement, US Navy, NCEL Technical
Note N-1862, 44p.

Youd, T.L., and Hoose, S.N., 1977, "Liquefaction Susceptibility and Geologic Setting,"
Proceedings, 6th World Conference on Earthquake Engineering, Prentice-Hall, Inc., Vol 3,
pp. 2189-2194.

Youd, T.L., and Perkins, D.M., 1978, "Mapping of Liquefaction-Induced Ground Failure Potential,"
Journal of the Geotechnical Engineering Division, ASCE, Vol. 104, No. GT4, pp. 433-446.

40
 Cyclic Liquefaction and its Evaluation
based on the SPT and CPT

P.K. Robertson and C.E. (Fear) Wride

Geotechnical Group, University of Alberta

Edmonton, Alberta, Canada

Abstract

Soil liquefaction is a major concern for structures constructed with or on sandy soils. This paper
describes the phenomena of soil liquefaction, provides suitable definitions, and provides an
update on methods to evaluate cyclic liquefaction using primarily the Standard Penetration Test
and the Cone Penetration Test (CPT). A new method is described to estimate grain
characteristics directly from the CPT and to incorporate this into one of the methods for
evaluating resistance to cyclic loading. A method is also described for correcting the results of
the CPT in thin layers. A worked example is also provided. This paper is the final submission
from the authors to the proceedings of the 1996 NCEER workshop on soil liquefaction; a similar
version has been submitted for review to the Canadian Geotechnical Journal.

41
 Introduction

Soil liquefaction is a major concern for structures constructed with or on saturated sandy soils.
The phenomenon of soil liquefaction has been recognized for many years. Terzaghi and Peck
(1948) referred to 'spontaneous liquefaction' to describe the sudden loss of strength of very loose
sands that caused flow slides due to a slight disturbance. Mogami and Kubo (1953) also used the
term liquefaction to describe a similar phenomenon observed during earthquakes. The Niigata
earthquake in 1964 is certainly the event that focused world attention on the phenomenon of soil
liquefaction. Since 1964, much work has been carried out to explain and understand soil
liquefaction. The progress of work on soil liquefaction has been described in detail in a series of
state-of-the-art papers, such as Yoshimi et al. (1977), Seed (1979), Finn (1981), Ishihara (1993),
and Robertson and Fear (1995). The major earthquakes of Niigata in 1964 and Kobe in 1995
have illustrated the significance and extent of damage caused by soil liquefaction. Liquefaction
was the cause of much of the damage to the port facilities in Kobe in 1995. Soil liquefaction is
also a major design problem for large sand structures, such as mine tailings impoundments and
earth dams.

The state-of-the-art paper by Robertson and Fear (1995) provided a detailed description of soil
liquefaction and its evaluation. In January 1996, the National Center for Earthquake Engineering
(NCEER) in the U.S.A. arranged a workshop (chaired by T.L. Youd) in Salt Lake City, Utah, to
discuss recent advances in the evaluation of cyclic liquefaction. This paper is the authors' final
presentation to the proceedings of that workshop; a similar version has been submitted for review
to the Canadian Geotechnical Journal. The objective of this paper is to provide an update on the
evaluation of cyclic liquefaction using primarily the Standard Penetration Test (SPT) and the
Cone Penetration Test (CPT). Several phenomena are described as soil liquefaction. In an effort
to clarify the different phenomena, the mechanisms will be described and a set of definitions for
soil liquefaction will be presented. Recent advances in the evaluation of cyclic liquefaction using
laboratory testing are also briefly described.

Liquefaction definitions

Before describing methods to evaluate liquefaction potential, it is important to first define the
terms used to explain the phenomena of soil liquefaction. Figure 1 shows the results from
undrained triaxial compression tests on Toyoura sand presented by Ishihara (1993). These results
present a clear picture of sand behaviour in undrained shear, since they show results at the same
void ratio, but at different effective confining stresses. The results are presented in the form of
deviator stress, q, versus axial strain and stress paths in q versus mean normal effective stress, p'.
Very loose sand ( density index = 16%), shows a marked strain softening response during
undrained shear. The shear stress reaches a peak then strain softens, eventually reaching an
ultimate condition referred to as critical or steady state. In this report, this ultimate condition will
be referred to as 'ultimate state' (US), as recommended by Poorooshasb and Consoli (1991).

42
 "t,
C.
~ 0.15 (a)

15'I
15 j
II
c- 0.10 J
J

.. .. I I
1
, I
• I ' I ' •
0 0 s 10 15" ;;o 25 30
Axial s!rain , f:1 (¼)

0 J
§: 0,15 ( b) ioyoura sand I
t> =Cl916, Dr= 16 '/. J
I
6I J
Ste<Jdy !
ti state 7
II 0.10
c- 7
j
.,.,., I
.. J
;;; 0.05'- ..j
c
C i
~ j
C, I
-I
0 0.01 0.02 0-06 0.1
Effective confining stress, p'=(Oi'•2u{)/.3 (MPo)

Figure 1 Undrained behaviour ofToyoura sand (after Ishihara, 1993).

The stress path during strain softening appears to follow a 'collapse surface', as suggested by
Sladen et al. (1985). However, at a lower confining stress, sand at the same void ratio shows a
strain hardening response before reaching ultimate state. For the same sand at a higher density,
a similar behaviour is seen, except that the ultimate state condition is reached at a higher stress
level (Ishihara, 1993). For dense sand, the response is predominately strain hardening since the
ultimate state strength is very large. This confirms the basic behaviour suggested by Castro
(1969) and that embodied in critical state soil mechanics (Roscoe et al., 1958). Been et al. (1991)
showed that steady state and critical state are the same condition and in e-p' space are
independent of the stress path followed to reach this ultimate state. The steady state or critical
state represents an ultimate state that can be represented in e-p'-q space, where p' is the mean
normal effective stress, q is the deviator stress and e is the void ratio.

Figure 2 shows a swnmary of the behaviour of a granular soil loaded in undrained triaxial
compression. In e-p' space, a soil with an initial void ratio higher than the ultimate state line
(USL) will strain soften (SS) at large strains, whereas a soil with an initial void ratio lower than

43
the USL will strain harden (SH) at large strains. It is possible to have a soil with an initial void
ratio higher than but close to the USL. For this soil state, the response can show limited strain
softening (LSS) to a quasi-steady state (QSS) (Ishihara, 1993), but eventually, at large strains, the
response strain hardens to the ultimate state. For some sands, very large strains are required to
reach the ultimate state, and in some cases, conventional triaxial equipment may not reach these
large strains (Ea > 20%) ( see Figure 1).

VoidI300 Notaiion:

• SS: Strain sofu:ning ,esponsc

SH: Strain hardening ,..ponsc
LSS: Limiud sttain sofu:ning response
QsT: StatiogmituiolllllsbearstreSs
Su: Ultimate undrained shear
Stffllgth .
US: Ul!imate Swe

Mt'111 cffcctiYe StreSS, p'

Shear q
•= q
Collapse surface
Strain sonening (SS)

;su
p' 1, she:ir s::::ll"!'l

q
q
Limited strain softening (LSS)

p' y

q q

"(
p'

Figure2 Schematic of undrained monotonic behaviour of sand in triaxial compression

(after Robertson, 1994).

If a soil slope or structure, such as an earth dam or tailings dam, is composed entirely of a strain
softening soil and the in-situ gravitational shear stresses are larger than the ultimate state strength
(i.e. a relatively steep slope consisting of very loose sand), a catastrophic collapse and flow slide
can occur if the soil is triggered to strain soften. Either cyclic or monotonic undrained loading
can trigger the collapse. Sasitharan et al. (1994) have shown that certain types of drained
monotonic loading (e.g. a slow rise in groundwater level) can trigger undrained collapse.

44
If a soil structure is composed entirely of strain hardening soil, lllldrained collapse can generally
not occur llllless the soil can become looser due to pore water redistribution. If a soil structure is
composed of both strain softening (SS) and strain hardening (SH) zones and the SS soil is
triggered to strain soften, a collapse and slide will occur only if, after stress redistribution due to
softening of the SS soil, the SH soil cannot support the gravitational shear stresses. A flow slide
will occur only if a kinematically admissible mechanism can develop. In general, a kinematically
admissible mechanism cannot form under level grolllld conditions in the absence of driving
loads. The trigger mechanism for a catastrophic flow slide can be cyclic, such as earthquake
loading, or monotonic, such as a rise in grolllldwater level or a rapid undrained loading.

During cyclic undrained loading (e.g. earthquake loading), almost all saturated cohesionless soils
develop positive pore pressures due to the contractive response of the soil at small strains. If
there is shear stress reversal, the effective stress state can progress to the point of essentially zero
effective stress (see Figure 3). For shear stress reversal to occur during earthquake loading,
ground conditions must be generally level or gently sloping; however, shear stress reversal can
occur in steeply sloping grolllld if the slope is of limited height (Pando and Robertson, 1995).
When a soil element reaches the condition of essentially zero effective stress, the soil has very
little stiffness and large deformations can occur during cyclic loading. However, when cyclic
loading stops, the deformations essentially stop, except for those due to local pore pressure
redistribution. If there is no shear stress reversal, such as in steeply sloping ground subjected to
moderate cyclic loading, the stress state may not reach zero effective stress. As a result, only
cyclic mobility with limited deformations will occur, provided that the initial void ratio of the
sand is below the USL and the large strain response is dilative (i.e. the material is not susceptible
to a catastrophic flow slide). However, shear stress reversal in the level ground area beyond the
toe of a slope may lead to overall failure of the slope due to softening of the soil in the toe region.
Shear stress, q Shear stress, q

Peak strength
envelope
Cyclic
Liquefaction

Shear stress reversal - - Zero effective stress

yclic
• efaction
Sh ear Stram
y
5% Strain

No. of cycles, N

Figure3 Schematic of undrained cyclic behaviour of sand illustrating cyclic liquefaction

(after Robertson, 1994).

45
Based on the above descriptions of soil behaviour in undrained shear and following the work by
Robertson (1994) and Robertson and Fear (1995), the following definitions are suggested:

Flow liquefaction

• Applies to strain softening soils only.

• Requires a strain softening response in undrained loading resulting in constant shear stress and
effective stress, as illustrated in Figure 2.
• Requires in-situ shear stresses greater than the ultimate or minimum undrained shear strength.
• Either monotonic or cyclic loading can trigger flow liquefaction
• For failure of a soil structure to occur, such as a slope, a sufficient volume of material must
strain soften. The resulting failure can be a slide or a flow depending on the material
characteristics and ground geometry. The resulting movements are due to internal causes and
can occur after the trigger mechanism occurs.
• Can occur in any metastable saturated soil, such as very loose granular deposits, very sensitive
clays, and loess (silt) deposits.

Cyclic softening

• Applies to both strain softening and strain hardening soils.

• Two terms can be used: cyclic liquefaction and cyclic mobility.

Cyclic liquefaction

• Requires undrained cyclic loading during which shear stress reversal occurs or zero shear stress
can develop (i.e. occurs when in-situ static shear stresses are low compared to cyclic shear
stresses), as illustrated in Figure 3.
• Requires sufficient undrained cyclic loading to allow effective stresses to reach essentially zero.
• At the point of zero effective stress no shear stress exists. When shear stress is applied, pore
pressure drops as the material tends to dilate, but a very soft initial stress strain response can
develop resulting in large deformations.
• Deformations during cyclic loading can accumulate to large values, but generally stabilize
when cyclic loading stops. The resulting movements are due to external causes and occur only
during the cyclic loading.
• Can occur in almost all saturated sands provided that the cyclic loading is sufficiently large in
magnitude and duration.
• Clayey soils can experience cyclic liquefaction but deformations are generally small due to the
cohesive strength at zero effective stress. Rate effects (creep) often control deformations in
cohesive soils.

Cyclic mobility

• Requires undrained cyclic loading during which shear stresses are always greater than zero;
i.e. no shear stress reversal develops.

46
• Zero effective stress will not develop.
• Deformations during cyclic loading will stabilize, unless the soil is very loose and flow
liquefaction is triggered. The resulting movements are due to external causes and occur only
during the cyclic loading.
• Can occur in almost any saturated sand provided that the cyclic loading is sufficiently large in
magnitude and duration, but no shear stress reversal occurs.
• Cohesive soils can experience cyclic mobility, but rate effects (creep) usually control
deformations.

Note that strain softening soils can also experience cyclic softening (cyclic liquefaction or cyclic
mobility) depending on the ground geometry.

Material Characterization

Strain Softening Strain Hardening

Behavior Behavior

Monotonic/Cyclic Size and duration

Trigger of cyclic loading

Gravitational stresses > Shear stress No shear stress

Undrained shear strength reversal reversal

Contained Uncontained j
Deformation Deformation

Small
Deformations

Deformations can continue Deformations essentially

after the trigger event stop after cyclic loading

Figure 4 Suggested flow chart for evaluation of soil liquefaction (after Robertson, 1994).

Figure 4 presents a suggested flow chart (after Robertson, 1994) for the evaluation of liquefaction
according to the above definitions. The first step is to evaluate the material characteristics in
terms of a strain softening or strain hardening response. If the soil is strain softening, flow
liquefaction is possible if the soil can be triggered to collapse and if the gravitational shear

47
stresses are larger than the ultimate or minimum strength. The trigger mechanism can be either
monotonic or cyclic. Whether a slope or soil structure will fail and slide will depend on the
amount of strain softening soil relative to strain hardening soil within the structure, the brittleness
of the strain softening soil and the geometry of the ground. The resulting deformations of a soil
structure with both strain softening and strain hardening soils will depend on many factors, such
as distribution of soils, ground geometry, amount and type of trigger mechanism, brittleness of
the strain softening soil and drainage conditions. Soils that are only temporarily strain-softening
(i.e. experience a minimum strength before dilating to US) are not as dangerous as very loose
soils that can strain-soften directly to ultimate state. Examples of flow liquefaction failures are
Fort Peck Dam (Casagrande, 1965), Aberfan flowslide (Bishop, 1973), Zealand flowslide
(Koppejan et al., 1948), and the Stava tailings dam. In general, flow liquefaction failures are not
common; however, when they occur, they take place rapidly with little warning and are usually
catastrophic. Hence, the design against flow liquefaction should be carried out cautiously.

If the soil is strain hardening, flow liquefaction will generally not occur. However, cyclic
softening can occur due to cyclic undrained loading, such as earthquake loading. The amount
and extent of deformations during cyclic loading will depend on the density of the soil, the
magnitude and duration of the cyclic loading and the extent to which shear stress reversal occurs.
If extensive shear stress reversal occurs, it is possible for the effective stresses to reach zero and,
hence, cyclic liquefaction can take place. When the condition of essentially zero effective stress
is achieved, large deformations can result. If cyclic loading continues, deformations can
progressively increase. If shear stress reversal does not take place, it is generally not possible to
reach the condition of zero effective stress and deformations will be smaller; i.e. cyclic mobility
will occur. Examples of cyclic softening were common in the major earthquakes in Niigata in
1964 and Kobe in 1995 and manifested in the form of sand boils, damaged lifelines (pipelines,
etc.), lateral spreads, slumping of small embankments, settlements, and ground surface cracks.
If cyclic liquefaction occurs and drainage paths are restricted due to overlying less permeable
layers, the sand immediately beneath the less permeable soil can loosen due to pore water
redistribution, resulting in possible subsequent flow liquefaction, given the right geometry.

Both flow liquefaction and cyclic liquefaction can cause very large deformations. Hence, it can
be very difficult to clearly identify the correct phenomenon based on observed deformations
following earthquake loading. Earthquake-induced flow liquefaction movements tend to occur
after the cyclic loading ceases due to the progressive nature of the load redistribution. However,
if the soil is sufficiently loose and the static shear stresses are sufficiently large, the earthquake
loading may trigger essentially 'spontaneous liquefaction' within the first few cycles of loading.
Also, if the soil is sufficiently loose, the ultimate undrained strength may be close to zero with an
associated effective confining stress very close to zero (Ishihara, 1993). Cyclic liquefaction
movements, on the other hand, tend to occur during the cyclic loading since it is the inertial
forces that drive the phenomenon. The post earthquake diagnosis can be further complicated by
the possibility of pore water redistribution after the cyclic loading resulting in a change in soil
density and possibly the subsequent triggering of flow liquefaction. Identifying the type of
phenomenon after earthquake loading is difficult and, ideally, requires instrumentation during
and after cyclic loading together with comprehensive site characterization.

48
The most common form of soil liquefaction observed in the field has been cyclic softening due to
earthquake loading. Much of the existing research work on soil liquefaction has been related to
cyclic softening, primarily cyclic liquefaction. Cyclic liquefaction applies to level or gently
sloping ground where shear stress reversal occurs during earthquake loading. This paper is
concerned primarily with cyclic liquefaction due to earthquake loading.

Cyclic resistance based on laboratory testing

Much of the early work related to earthquake-induced soil liquefaction resulted from laboratory
testing of reconstituted samples subjected to cyclic loading by means of cyclic triaxial, cyclic
simple shear, or cyclic torsional tests. The outcome of these studies generally confirmed that the
resistance to cyclic loading is influenced primarily by the state of the soil (i.e. void ratio, effective
confining stresses, and soil structure) and the intensity and duration of the cyclic loading
(i.e. cyclic shear stress and number of cycles), as well as the grain characteristics of the soil. Soil
structure incorporates features such as fabric, age and cementation. Grain characteristics
incorporate features such as grain size distribution, grain shape, and mineralogy.

Resistance to cyclic loading is usually represented in terms of a cyclic stress ratio or cyclic
resistance ratio (CRR). For cyclic simple shear tests, CRR is taken as the ratio of the cyclic shear
stress to cause cyclic liquefaction to the initial vertical effective stress; i.e. (CRR)ss = 'tcyJa'vo•
For cyclic triaxial tests, CRR is taken as the ratio of the maximum cyclic shear stress to cause
cyclic liquefaction to the initial effective confining stress; i.e. (CRR)tx = crctJ2cr'3c• The two tests
impose different loading conditions and the CRR values are not equivalent. Cyclic simple shear
tests are generally considered to be better than cyclic triaxial tests at closely representing
earthquake loading for level ground conditions. However, experience has shown that the (CRR)ss
can be estimated quite well from (CRR)tx, and correction factors have been developed (Ishihara,
1993). The CRR is typically taken at about 15 cycles of uniform loading to represent an
equivalent earthquake loading of Magnitude (M) 7.5; i.e. CRR7.s.

The CRR for any other size earthquake can be estimated using the following equation:

CRR = (CRR1.s)(MSF) (I)

where:
MSF = magnitude scaling factor (recommended values are provided in the report by Youd et al.
(1997), which summarizes the results of the 1996 NCEER Workshop).

It is common practice to define the point of 'liquefaction' in a cyclic laboratory test as the time at
which the sample achieves a strain level of either 5% double-amplitude axial strain in a cyclic
triaxial test or 3 to 4% double-amplitude shear strain in a cyclic simple shear test. For loose sand

49
samples subjected to shear stress reversal, this often occurs close to the point at which the
effective confining stress is essentially zero and deformations develop rapidly; hence, the
definition is the same as that for cyclic liquefaction (see Figure 3). However, for denser sand
samples, the 5% double-amplitude strain criteria can occur before sufficient pore pressure has
developed to take the sample to the state of essentially zero effective stress. Hence, the criteria
for liquefaction typically applied to laboratory results may well be unduly conservative, since
deformations may actually be progressing rather slowly.

While void ratio (relative density) has been recognized as a dominant factor influencing the CRR
of sands, studies by Ladd (1974), Mulilis et al. (1977), and Tatsuoka et al. (1986) have clearly
shown that sample preparation (i.e. soil fabric) also plays an important role. This is consistent
with the results of monotonic tests at small to intermediate strain levels. Hence, if results are to
be directly applied with any confidence, it is important to conduct cyclic laboratory tests on
reconstituted samples with a structure similar to that in-situ. Unfortunately, it is very difficult to
determine the in-situ fabric of natural sands below the water table. As a result, there is often
some uncertainty in the evaluation of CRR based on laboratory testing of reconstituted samples.
Tokimatsu and Hosaka (1986) suggested that either the small strain shear modulus or shear wave
velocity measurements could be used to improve the value of laboratory testing on reconstituted
samples of sand.

Based on the above observations, there has been increasing interest in testing high quality
undisturbed samples of sandy soils under conditions representative of those in-situ.
Yoshimi et al. (1989) showed that aging and fabric had a significant influence on the CRR of
clean sand from Niigata, as shown in Figure 5.

0.7

'o In-situ frozen samples

l:5 0.6 from Niigala site
N
'( 0-o'= 78 kPa
0.5
13
0 0.4
e 0.3
1/1
1/1 Freshly deposited
...a, sand Dr=56¼
in 0.2

u
u
u>, 0.1

0
~l
10
I"

- 100 1000
Number of cycles lo 5¼ DA axial strain

Figure 5 Comparison between triax.ial cyclic resistance ratio (CRR) for a clean Niigata
sand based on recently deposited and aged samples (after Yoshimi et al., 1989).

50
Yoshimi et al. (1994) also showed that sand samples obtained using conventional high quality
fixed piston samplers produced different CRR values than undisturbed samples obtained using
in-situ ground freezing, as summarized in Figure 6. Dense sand samples showed a decrease in
CRR and loose sand samples showed an increase in CRR when obtained using a piston sampler,
as compared to the results of testing in-situ frozen samples. The difference in CRR became more
pronounced as the density of the sand increased.

1·0

b. Double-tube core barrel

I
o Hydraulic piston sampler {
I
~
O·B I
u• I
t I
~
I
"
t.
"'
·~o 0·6
C\l
..,._
~
C)
t.)
C

"'
to
-~ 0·4
C:
.Q
ti
"':,
ai
O"
:.J
0·2

10 20 30 40
Normalized N value (N,}so

Figure6 Comparison between triaxial cyclic resistance ratio (CRR) and SPT (N1)6o
values for clean sands based on tube samples and undisturbed in-situ frozen
samples (after Yoshimi et al., 1994).

The relationship, shown in Figure 6, (after Yoshimi et al., 1994) between (CRR)ix to cause 5%
double-amplitude axial strain after 15 cycles and corrected SPT N value at 100 k:Pa effective
overburden stress and 60% energy ((N1)60) obtained from adjacent soundings was based on
undisturbed samples of sand obtained using ground freezing. It would appear that dense sand
with a normalized SPT (N1)60 between 30 and 40 has a (CRR)ix less than 1.0. This is in conflict
with field observation (Seed et al., 1985) and is almost certainly associated with the definition of
'liquefaction' based on a limiting double-amplitude axial strain of 5%. As explained earlier,
dense sand samples can progressively develop 5% double-amplitude axial strain but may not

51
have achieved the condition of rapid deformation associated with essentially zero effective
confining stress. Hence, it is important to clearly define the onset of 'liquefaction'. In general,
for design purposes, cyclic liquefaction is the point at which the soil experiences large
uncontrolled deformations.

Although the results shown in Figure 6 apply to a range of sands from Japan, it is likely that
changes in grain characteristics will influence the correlation between CRR and SPT (N1)60-
Based on the same laboratory test results on undisturbed in-situ frozen sand samples as those
shown in Figure 6 plus one additional site, Suzuki et al. (1995a) suggested a correlation between
CRR and corrected Cone Penetration Test (CPT) penetration resistance. To account for the
variation due to differences in grain characteristics, Suzuki et al. (1995a) suggested
a modification to the CPT corrected to incorporate the minimum void ratio (emm), as a measure of
the grain characteristics, as follows:

(2)

where qtN is the cone tip resistance corrected for overburden stress and minimum void ratio; Pa is
a reference pressure, usually equal to 100 kPa; cr'vo is the vertical effective stress and
f(emm) = (2.17 - emin/ / ( 1+ emm), is a factor to account for differences in grain characteristics.
The resulting correlation is shown in Figure 7.

1.0~,, 1 1
1 • • ·, 1 • • • ·,
DA=5'7c, ::--;:;=15 cycles
0 cmin<0.70
t:>° 0.8 D O.ill~ c min<0.80
c::: c.cmin~0.80
t:>"

Figure 7 Correlation between triaxial cyclic resistance ratio (CRR) and modified
normalized cone penetration resistance for a wide range of sands (after Suzuki
et al., 1995a).

52
Suzuki et al. (1995a) suggested that the correlation in Figure 7 would be applicable to a wide
range of sandy soils (i.e. sandy soils with various grain distributions, grain shapes and/or
mineralogy). It is interesting to note that the modification using f(emin) accounts for a correction
to the traditional normalized cone penetration resistance by a factor of 0.65 to 0.96 when emin
varies from typical values of0.6 to 0.8. However, the incorporation ofemm into the correlation is
cumbersome, difficult to apply, and appears to have a relatively small influence for most sands.

When a soil is fine-grained or contains some amount of fines, some cohesion or adhesion can
develop between the fine particles making the soil more resistant at essentially zero effective
confining stress. Consequently, a greater resistance to cyclic liquefaction is generally exhibited
by sandy soils containing some fines. However, this tendency depends on the nature of the fines
contained in the sand (Ishihara, 1993). Laboratory testing has shown that one of the most
important index properties influencing CRR is the plasticity index of the fines contained in the
sand (Ishihara and Koseki, 1989). Figure 8 shows the results of cyclic triaxial tests versus
plasticity index (Ip) for a variety of sandy soils (Ishihara, 1993) and illustrates that the (CRR)rx
appears to increase with increasing plasticity index.

0.5
..--.
""o
~ 0.4
[I

1:3
[I
~
It) • [I

en ' Q.3 • •
•5 (/) •
3~ • • •
Cu
u ~0.22
,Qo all
• •
- v ...
e~ 0
• a: Bentonite based
111 ·- o: Kaolinite
X! o
.l::
C
0.1 ( er;= so kPa ) .t. :
L
cam
(/) ... e = 0,53~1-60 v: Tailings
u iii •: Undisturbed
u <i. tailings
Gd o~.___..____.....____.____._____..___.c..-,
0 10 20 30 40 50 60
Plasticity index , Ip

Figure 8 Effect of soil plasticity on cyclic resistance ratio (CRR) of fines-containing

sands (after Ishihara, 1993).

Studies in China (Wang, 1979) suggest that the potential for cyclic liquefaction in silts and clays
is controlled by grain size, liquid limit, and water content. The interpretation of this criterion as
given by Marcuson et al. (1990) and shown in Figure 9 can be useful; however, it is important to
note that it is based on limited data. Figure 9 suggests that when a soil has a liquid limit less than
35% combined with a water content greater than 90% of the liquid limit, it is unclear if the soil

53
can experience cyclic liquefaction and that the soil should be tested to clarify the expected
response to undrained cyclic loading.

CRITERIA OF LIQUEFACTION ASSESSMENT Or CLAYEY SOILS FROM

STUDIES IN CHINA !SEED ET f.J... 1973 AND WJ.J-lG 19791

PERCENT FINEn THAN 0.OO!i MM

UOU10 LIMIT

WATEr\ CONTENT >0.9 x UOUIO LIMIT

GRAPHICAL RE?RESENTATION OF CRITERIA:

100

~
"'
L)
C:
w
II,.

_,
.J
150
1:·
_,z ~,-.

!_, 0

Figure 9 Liquefaction criteria for silts and clays (after Marcuson et al., 1990).

For high risk projects where the evaluation of the potential for soil liquefaction due to earthquake
loading is very important, consideration should be given to a limited amount of appropriate
laboratory testing on high quality undisturbed samples. Recently, in-situ ground :freezing has
been used to obtain undisturbed samples of sandy soils (Yoshimi et al., 1978; Yoshimi et al.,
1989; Yoshimi et al., 1994; Sego et al., 1994; Hofmann et al. 1995; and Hofmann, 1997). Cyclic
simple shear tests are generally the most appropriate tests although cyclic triaxial tests can also
give reasonable results.

Cyclic resistance based on field testing

Standard penetration test (SPT)

The above comments have shown that testing high quality undisturbed samples will give better
results than testing poor quality samples. However, obtaining high quality undisturbed samples
of saturated sandy soils is very difficult and expensive and can only be carried out for large

54
projects for which the consequences of liquefaction may result in large costs. Therefore, there
will always be a need for simple, economic procedures for estimating the CRR of sandy soils.
Currently, the most popular simple method for estimating CRR makes use of the penetration
resistance from the Standard Penetration Test (SPT) although, more recently, the
Cone Penetration Test (CPT) has become very popular due to its greater repeatability and the
continuous nature of its profile.

The late Professor H.B. Seed and his co-workers developed a comprehensive approach to
estimate the potential for cyclic softening due to earthquake loading. The approach requires an
estimate of the cyclic stress ratio (CSR) profile caused by a design earthquake. This is usually
done based on a probability of occurrence for a given earthquake. A site specific seismicity
analysis can be carried out to determine the design CSR profile with depth. A simplified method
to estimate CSR was also developed by Seed and Idriss (1971) based on the maximum ground
surface acceleration (amax) at the site. This simplified approach can be summarized as follows:

CSR= ~•v = 0.6s(amax)( ~vo )rd (3)

crvo g crvo

where •av is the average cyclic shear stress; amax is the maximum horizontal acceleration at the
ground surface; g = 9.8lm/s2 is the acceleration due to gravity; O'vo and cr'vo are the total and
effective vertical overburden stresses, respectively; and rd is a stress reduction factor which is
dependent on depth. The factor rd can be estimating using the following tri-linear function,
which provides a good fit to the average of the suggested range in ra originally proposed by Seed
and Idriss (1971):

ra = 1.0 - 0.00765 z ifz < 9.15 m (4)

= 1.174 - 0.0267 z if z = 9.15 to 23 m
= 0.744 - 0.008 z ifz =23to30m
= 0.5 ifz >30m

where z is the depth in metres. The first two formulae in Equation 4 (i.e. for depths less
than 23 m) were recommended by Liao and Whitman (1986b ). The third formula has been added
here to provide a better match with the average of the range in rd suggested by Seed and Idriss
(1971) at depths between 23 m and 30 m. The fourth formula has been added as a conservative
cutoff at large depths. These formulae are approximate, at best, and represent only average
values since rct shows considerable variation with increasing depth (Seed and Idriss, 1971). The
CSR profile from the earthquake can be compared to the estimated CRR profile for the soil
deposit, adjusted to the same magnitude using Equation 1. At any depth, if CSR is greater than
CRR, cyclic softening (liquefaction) is possible. This approach is the most commonly used
technique in most parts of the world for estimating soil liquefaction due to earthquake loading.

55
Tue approach based on the SPT has many problems, primarily due to the inconsistent nature of
the SPT. Tue main factors affecting the SPT have been reviewed (e.g. Seed et al., 1985;
Skempton, 1986; Robertson et al., 1983) and are summarized in Table 1. It is highly
recommended that the engineer become familiar with the details of the SPT in order to avoid or
at least minimize some of the major factors.

Table 1. Factors affecting the SPT (after Kulhawy and Mayne, 1990).

Cause Effects Influence on SPT

Nvalue
Inadequate cleaning of hole SPT is not made in original in-situ Increases
soil, and therefore soil may become
trapped in sampler and be compressed
as sampler is driven, reducing recovery
Failure to maintain adequate Bottom of borehole may become quick Decreases
head of water in borehole
Careless measure of hammer Hammer energy varies (generally Increases
drop variations cluster on low side)
Hammer weight inaccurate Hammer energy varies (driller supplies Increases or
wei!!ht; variations of 5-7 % common) decreases
Hammer strikes drill rod Hammer energy reduced Increases
collar eccentrically
Lack of hammer free fall Hammer energy reduced Increases
because of ungreased
sheaves, new stiff rope on
weight, more than two turns
on cathead, incomplete
release of rope each drop
Sampler driven above Sampler driven in disturbed, Increases greatly
bottom of casing artificially densified soil
Careless blow count Inaccurate results Increases or
decreases
Use of non-standard sampler Corrections with standard sampler Increases or
invalid decreases
Coarse gravel or cobbles in Sampler becomes clogged or impeded Increases
soil
Use of bent drill rods Inhibited transfer of energy of sampler Increases .

One of the single most important factors affecting SPT results is the energy delivered to the SPT
sampler. This is normally expressed in terms of the rod energy ratio (ER). An energy ratio of
60% has generally been accepted as the reference value, which represents the approximate
historical average SPT energy. Tue value of ER (%) delivered by a particular SPT set-up

56
 depends primarily on the type of hammer/anvil system and the method of hammer release.
Values of the correction factor to modify the SPT results to 60% energy (ER./60) can vary from
0.3 to 1.6 corresponding to field values of ER of20% to 100%. Additional correction factors are
also required for rod lengths less than 10 m, borehole diameters outside the recommended
interval (65 - 125 mm) and samplers without internal liners.

Since the SPT N value also varies with the effective overburden stress level, an overburden stress
correction factor is usually also applied to provide a consistent reference (i.e. (N 1) 60). The SPT
N value corrected for overburden stress, rod length, borehole diameter and sampling method is
given by:

(5)

5
where N is the measured SPT blowcount; CN = (Pafcr'v0 )°- (with a restriction that CN :5: 2) is a
correction for effective overburden stress; Pa is a reference pressure of 100 kPa; cr'vo is the
vertical effective stress; CE= ER./60% is a correction to account for rod energy; ER is the actual
energy ratio, in percent; CB is a correction for borehole diameter; CR is a correction for rod
length; and Cs is a correction for the sampling method. The correction for overburden stress (CN)
is the same as that proposed by Liao and Whitman (1986a), except that, as noted above, a
maximum value of CN = 2 should be applied for SPT values at shallow depths. Correction
factors for energy ratio, borehole diameter, rod length and sampling method were suggested by
Skempton (1986) and are given in Table 2.

Table 2. Corrections to the SPT (modified from Skempton, 1986).

Factor Equipment Variable Term Correction

5
Overburden Pressure CN (Pa/cr'vo)°- but :5: 2
Energy Ratio Donut Hammer CE 0.5 to 1.0
Safety Hammer 0.7 to 1.2
Automatic Hammer 0.8 to 1.5
Borehole diameter 65 to 115 mm (2.5 to 4.5 in) CB 1.0
150 mm (6 in) 1.05
200mm (8 in) 1.15
Rod length 3 to 4 m (10 to 13 ft) CR 0.75
4 to 6 m (13 to 20 ft) 0.85
6 to 10 m (20 to 30 ft) 0.95
10 to 30 m (30 to 100 ft) 1.0
>30 m (>100 ft) <1.0
Sampling method Standard sampler Cs 1.0
Sampler without liner 1.1 to 1.3

57
The ranges of correction values for ER (CE) are given only as a guide. Actual values of CE can
vary significantly from the global averages. It is recommended that ER be measured during the
actual site investigation to improve the level of reliability of the SPT.

A summary of the recommended procedure for performing the SPT is given in Table 3.
The borehole should be made using mud rotary techniques using a side or upward discharge bit.
Hollow stem auger techniques are not recommended in saturated sands and silts unless extreme
care is taken, since disturbance and heave in the hole is common. There is a need for care when
cleaning out the bottom of the borehole to avoid disturbance. The borehole should not exceed
115 mm (4.5 inches) in diameter, since the associated stress relief can reduce the measured N
value in some sands. The energy delivered to the SPT sampler can also be very low for an SPT
above a depth of about 10 m (30 ft.) due to rapid reflection of the compression wave in the rod.
The energy reaching the sampler can also become reduced for an SPT below a depth of about
30 m (100 ft) due to energy losses and the large mass of the drill rods. If the SPT sampler has
been designed to hold a liner, it is important to ensure that a liner is installed, since a correction
of up to 30% may apply if a liner is not used (Schmertmann, 1979).

Table 3. Recommended SPT procedure (after Seed et al., 1985).

SPT Set-up Recommended Procedure

Borehole size 66 mm < Diameter < 115 mm

Borehole support Casing for full length and/or drilling mud

Drilling Wash boring; side discharge bit

Rotary boring; side or upward discharge bit
Clean bottom of borehole•
Sampler Standard 51 mm O.D. + 1 mm
35 mml.D. + 1 mm
> 457 mm lemrth
Penetration Resistance Record number of blows for each 150 mm
N = number of blows from 150 to 450 mm penetration

• Maximum soil heave within casing < 70 mm

Seed (1979) developed a method to estimate the CRR for a sand under level ground conditions
based on the SPT. This method was based on extensive field performance data for Holocene
sands from essentially level ground sites which either had or had not experienced cyclic softening
(liquefaction) due to earthquake loading. Liquefaction was assumed to have occurred based on
the presence of observable surface features such as sand boils and ground cracks. A summary of
the SPT based method to estimate CRR for clean sand is shown in Figure 10 (Seed et al., 1985).

58
 Liquefccticn with
7j -2:i•;. • ro~~ • 3~•

0.5 Ne L:~sefcc:,~r.

•
-"l
'<)v
o;,'
•
•
•

1
0.2
•

.
o_
"
,?
0

0 0

0.1

10 20 ~o 40 50
(N1)60

Figure 10 Cyclic resistance ratio (CRR) for clean sands under level ground conditions
based on SPT and field performance data (after Seed et al., 1985).

Other SPT based methods have been developed (Tatsuoka et al., 1980; Shibata, 1981; Tokimatsu
and Yoshimi, 1983; Kokusho et al., 1983; Ishihara, 1993; Fear and McRoberts, 1995), but the
correlation by Seed et al. (1985) appears to maintain the most popularity, especially in North
America. The correlation shown in Figure 10 includes the estimated limiting shear strain values
associated with 'liquefaction'. Seed recognized that dense sands ((N1)6o>15) generally
experienced less deformation for a given cyclic loading (i.e. they experienced cyclic mobility)
than loose sands (which experienced cyclic liquefaction). Hence, the definition of 'liquefaction'
became flexible in that dense sand would not develop very large strains (i.e. would not reach the
condition of essentially zero effective confming stress). This is supported by laboratory test
results (Yoshimi et al., 1994) and field observations (Bartlett and Youd, 1995).

59
Based on discussions at the 1996 NCEER Workshop regarding the statistical analysis of the
liquefaction database and physical considerations discussed in Liao et al. (1988) and the concept
of a threshold strain (Dobry et al., 1982), the Seed et al. (1985) SPT curve was slightly modified
to avoid the extrapolation to zero CRR at zero penetration resistance. The modified clean sand
SPT curve is shown in Figure 11 and has an intercept of CRR.=0.05. Occurrence of liquefaction
is based on level ground observations of surface manifestations of cyclic liquefaction. For loose
sand (i.e. (N1) 60 < 15) this could involve large deformations resulting from a condition of
essentially zero effective stress being reached. For denser sand (i.e. (N 1)6o > 15), this could
involve the development of large pore pressures, but the effective stress may not fully reduce to
zero and deformations may not be as large as in loose sands. Hence, the consequences of
'liquefaction' will vary depending on the soil density as well as the size and duration ofloading.

Ne Llquefccticn

o.~

0.2 C

0.1

Q,___ _....____________________ ~

0 10 20 ~o 50
{N1)60
Figure 11 Recommended cyclic resistance ratio (CRR) for clean sands under level ground
conditions based on SPT.

60
The recommended SPT clean sand curve can be approximated by the following equation
proposed by Blake (1996):

a+ ex+ ex 2 + gx 3
CRR1.s= - - - - - - - - - - - (6)
1 + bx + dx 2 + fx: 3 + hx 4

where: x = (N 1)6ocs is the clean sand equivalent SPT blowcount corrected for overburden; and
a= 4.844E-02, b = -1.248E-0l, c = - 4.721E-03, d = 9.578E-03, e = 6.136E-04, f= - 3.285E-04,
g = - l.673E-05, and h = 3.714E-06 are all constants.

The field observation data used to compile the curves in Figure 10 and Figure 11 are apparently
based on the following:

Holocene age, clean sand deposits

Level or gently sloping ground
Magnitude M = 7.5 earthquakes
Depth range from 1 to 14 m ( 3 to 47 ft)
(90% is for depths< 10 m, (32 ft))
Representative average SPT N values for the layer that was considered
to have experienced cyclic liquefaction.

Hence, caution should be exercised when extrapolating the correlation to conditions outside of
the above range. An important feature to recognize is that the correlation appears to be based on
average values for the inferred liquefied layers. However, the correlation is often applied to all
measured SPT values, which include low values below the average. Hence, the correlation can
be conservative in variable deposits where a small part of the SPT data could indicate possible
liquefaction.

Seed et al. (1985) showed that for a given CRR, a sand with fines has a lower SPT (N1)60 value
and, based on this observation, developed the correlation further to include the influence of fines
content, as shown in Figure 12. The correlation showed that, for the same CRR, the penetration
resistance in silty sands was smaller. This is most likely due to the greater compressibility and
decreased permeability of silty sands, which reduces penetration resistance and moves the
penetration process toward an undrained penetration, respectively. Robertson and Fear (1995)
recommended an average blowcount correction, which was dependent on fines content, but not
on penetration resistance.

Although the original correlations shown in Figure 12 are based on fines content, it is clear that
the CRR of a soil is a :function of many factors, including type of fines; e.g. plasticity (Figure 8
and Figure 9), other grain characteristics (mineralogy, grain shape, etc), and fines content.
Hence, any correction should be applied with caution.

61
 06
... J7
,,
Percenr f.nes ='35I IS sS
I
I I
0.
I
l
I
I
'
I
I
I
.. ,o
04

!m'..
a.,'
....
0.3
.,o. "' .-i

FiNE:S CONTENT~ 5 %
Modified Oi1n~r Cede P:c~Cf Cc:cy c:::r.t~r =~~•.1 ®

F-:n•Atrltr1ccn d:.ra
Jc;,onnt dare
.
L.ia,;r.:,!lcn

•.
ffb'Qncr

..
No
:..~a.~:r::,, LiQ.;e-t:c"a,
:::
0
Chine~ QQIQ 6

IO 40

Figure 12 Relationship between cyclic resistance ratio (CRR) and SPT for sands and silty
sands based on field performance data (after Seed et al., 1985).

Based on extensive discussions by the NCEER Workshop participants, it is recommended that a

correction be applied to the measured (N1)6o to obtain the equivalent clean sand (N 1)6ocs using the
following equation:

(7)

where Ks is a correction factor that is a function of fines content and plasticity of the fines.
Based on observed field performance, the suggested correction factor Ks is shown in Figure 13.
This basic recommendation is based on soils with non-plastic fines (PI::;; 5%) for which Ks has
a maximum value of 1.75 at a fines content of 35%. For soils with more plastic fines (PI> 5%),
the correction factor is likely larger. However, the data are limited and contain much uncertainty
(see Figure 8). For soils with a fines content greater than 35%, other criteria such as that shown
in Figure 9 should be applied.

62
 2 ......- - - - - - - - - - - - - - - - - - - - - - - - - .

K,= 1.75
Recommended Relationship
forPI:,;5%

0 5 10 15 20 25 30 35
Fines content, FC (%}

Figure 13 Recommended grain characteristic correction to obtain clean sand equivalent

SPT penetration resistance in sandy soils with non-plastic fines.

The recommended procedure is to determine the fines content and plasticity of the fines and
apply a correction to the measured SPT (N1)60 value using Figure 13 and Equation 7. The clean
sand equivalent penetration resistance, (N 1)6ocs, can then be combined with the clean sand base
curve shown in Figure 11 and approximated by Equation 6 to estimate CRR1.s- However, for
high fines content soils or soils with highly plastic fines, the criteria shown in Figure 9 should
also be applied. Figure 14 shows the resulting equivalent CRR curves for fines contents of 15%
and 35% for sandy soils with non-plastic fines (Pl :s; 5%).

Cone penetration test (CPT)

Due to the inherent difficulties and poor repeatability associated with the SPT, several
correlations have been proposed to estimate CRR for clean sands and silty sands using corrected
CPT penetration resistance (e.g. Robertson and Campanella, 1985; Seed and de Alba, 1986;
Olsen, 1988; Olsen and Malone, 1988; Shibata and Teparaska, 1988; Mitchell and Tseng, 1990;
Olsen and Koester, 1995; Suzuki et al., 1995a & 1995b; Stark and Olson, 1995; Robertson and
Fear, 1995).

63
 0.6 - , - - - - - - , - - - - - . . . - - - - ~ - - - - - . - - - ~

(%)=35 15 ::; 5
0.5 + - - - - + - - - - + - - - - - - - ! f - - - - - - i - - - - - - t
I I
I I
I I

PI :55%

0;-----+-----i---------f-------1
0 10 20 30 40 50

Figure 14 SPT base curves for various fines contents for non-plastic fines (Pl:,; 5%).

Although cone penetration resistance is often just corrected for overburden stress (resulting in the
tenn 4c1), truly normalized (i.e. dimensionless) cone penetration resistance corrected for
overburden stress (qcIN) can be given by:

(8)

where 4c is the measured cone tip penetration resistance; CQ = (P Jcr'vot is a correction for
overburden stress; n = exponent, typically equal to 0.5; Pa is a reference pressure in the same units
as cr'vo (i.e. Pa=lOO kPa if cr'vo is in kPa); and Pa2 is a reference pressure in the same units as
qc (i.e. Pa2=0.l MPa if qc is in MPa). A maximum value of CQ=2 is generally applied to CPT
data at shallow depths. The normalized cone penetration resistance, 4cIN, is dimensionless.

64
Robertson and Campanella (1985) developed a chart for estimating CRR from corrected CPT
penetration resistance based on the Seed et al. (1985) SPT chart and SPT-CPT conversions.
Other similar CPT-based charts were also developed by Seed and de Alba (1986), Shibata and
Teparaska (1988), and Mitchell and Tseng (1990). A comparison between three of these CPT
charts is shown in Figure 15.

Cleon
sands-
0.5 - 0.325

Silly
sanas

0.26 ~
u
0

0.195 ~
!
C
.s
.:!)
"'
~
i::::
- 0.13 i-~
u
~
~
loo
...
~

0-1- - - :Shibata-Teparaska - 0.065

-<
(1988)
-=Robertson-Campanella
(1985)
-•-:Seed-Alba (1986)
0 ,__....,__.,_____.__..__---L.,_.,___,__ _.___,__-'-J
10 20
Corrected CPT tip resistance, qcl (MPa)

Figure 15 Comparison between three CPT based charts for estimating cyclic resistance
ratio (CRR) for clean sands (after Ishihara, 1993).

In recent years, there has been an increase in available field performance data, especially for the
CPT (Ishihara, 1993; Kayen et al., 1992; Stark and Olson, 1995; Suzuki et al., 1995b).
The recent field performance data have shown that the existing CPT-based correlations to
estimate CRR are generally good for clean sands. The recent field performance data show that
the correlation between CRR and qc1N by Robertson and Campanella (1985) for clean sands
provides a reasonable estimate of CRR. Based on discussions at the 1996 NCEER Workshop,

65
the curve by Robertson and Campanella (1985) has been adjusted slightly at the lower end, in
order to be more consistent with the SPT curve. The resulting recommended CPT correlation for
clean sand is shown in Figure 16.

0.6 - - - - - - - - - - - - - - - - - - - - ,
0.25 < Dso (mm) < 2.0
FC (%) :::;5
0.5 Ye .,,zo¾ ""1 o¾ .,,3%
I I
I I
• I I
I I
• ..., • I
... -:-t ,.
• j{Lique,actiqn I INo liquefaction j
M I
o.a.l I•

f . . .a.• .A.•._l·pl/ &

~
• .._JI. I o 0
0
cs 0
..... • • •, ~0
r .vt;. /"Ji,. LY)IY)
o 0
~

t::A ....-,"l'T"iO:--,.,-=-N),----,--,----:,..,.........,.........,
•
.
liq. No Llq.
0
~

0 50 100 150 200 250 300

Corrected CPT tip resistance, qc1N

Figure 16 Recommended cyclic resistance ratio (CRR) for clean sands under level ground
conditions based on CPT.

Included in Figure 16 are suggested curves of limiting shear strain, similar to those suggested by
Seed et al. (1985) for the SPT. Occurrence of liquefaction is based on level ground observations
of surface manifestations of cyclic liquefaction. For loose sand (i.e. qcJN < 75 ) this could involve
large deformations resulting from a condition of essentially zero effective stress being reached.
For denser sand (i.e. qc1N > 75 ) this could involve the development of large pore pressures, but
the effective stress may not fully reduce to zero and deformations may not be as large as in loose
sands. Hence, the consequences of 'liquefaction' will vary depending on the soil density as well
as the size and duration of loading. An approximate equation for the clean sand CPT curve
shown in Figure 16 is given later in this paper (Equation 14).

The field observation data used to compile the curve in Figure 16 are apparently based on the
following conditions, similar in nature to those for the SPT based data:

66
 Holocene age, clean sand deposits
Level or gently sloping grolllld
Magnitude M = 7.5 earthquakes
Depth range from 1 to 15 m (3 to 45 ft)
(84% is for depths< 10 m (30 ft))
Representative average CPT qc values for the layer that was considered
to have experienced cyclic liquefaction.

As for the SPT-based approach, caution should be exercised when extrapolating the CPT
correlation to conditions outside of the above range. An important feature to recognize is that the
correlation appears to be based on average values for the inferred liquefied layers. However, the
correlation is often applied to all measured CPT values, which include low values below the
average. Therefore, the correlation can be conservative in variable deposits where a small part of
the CPT data could indicate possible liquefaction. Although some of the recorded case histories
show liquefaction below the suggested curve in Figure 16, the data are based on average values
and, hence, the authors consider the suggested curve to be consistent with field observations.
The CPT curve in Figure 16 is also consistent with the SPT curve shown in Figure 11.

Based on data from 180 sites, Stark and Olson (1995) also developed a set of correlations
between CRR and qc 1 for various sandy soils based on fines content and mean grain size, as
shown in Figure 17.

0.6

0.
Dso (mm)
F.C. ..,, V I,
SLlY SAND-SAl'IJY
< 0.10
I I I
SU.lY SAND
0.10 ~ Dso Cmml ~ 0.25
5 < F.C. (f,) < 35
~ I I
i:i: I I I
u
'-' I
0.4 I CLEAN SAND
0
:;:I
m
f 025 < D50 Cmrnl < 20
{ F.C. ('l>) i 5
~
0.3 I
...
Cl)
I
=
.5
.,
·;;;
Cl)
i:i: 0.2
.!:!
u;,...
u 0.1
IM= 7.5 I
0 0 50 100 150 200 250 300
Corrected CPT tip resistance, qc1N

Figure 17 Summary of variation of cyclic resistance ratio (CRR) with fines content based
on CPT field performance data (after Stark and Olson, 1995).

67
The CPT combined database is now larger than the original SPT-based database proposed by
Seed et al. (1985). It is important to note that the simplified approach based on either the SPT or
the CPT has many wicertainties. The correlations are empirical and there is some wicertainty
over the degree of conservatism in the correlations as a result of the methods used to select
representative values of penetration resistance within the layers assumed to have liquefied (Fear
and McRoberts, 1995). A detailed review of the CPT data, similar to those carried out by Liao
and Whitman (1986b) and Fear and McRoberts (1995) on SPT data, would be required to
investigate the degree of conservatism contained in Figure 16 and Figure 17. The correlations are
also sensitive to the amowit and plasticity of the fines within the sand.

One reason for the continued use of the SPT has been the need to obtain a soil sample to
determine the fines content of the soil. However, this has been offset by the poor repeatability of
SPT data. With the increasing interest in the CPT due to its greater repeatability, several
researchers (e.g. Robertson and Campanella, 1985; Olsen, 1988; Olsen and Malone, 1988; Olsen
and Koester, 1995; Suzuki et al., 1995a & 1995b; Stark and Olson, 1995; Robertson and Fear,
1995) have developed a variety of approaches for evaluating cyclic liquefaction potential using
CPT results. It is now possible to estimate grain characteristics such as apparent fines content
and grain size from CPT data and incorporate this directly into the evaluation of liquefaction
potential. The following is a modification and update of the CPT approach suggested by
Robertson and Fear (1995).

As for the SPT, for the same CRR, the CPT penetration resistance in silty sands is smaller due to
the greater compressibility and decreased permeability of silty sands. Robertson and Fear (1995)
recommended an average correction, which was dependent on apparent fines content, but not on
penetration resistance. Similar to the SPT, it is possible to correct the CPT penetration resistance
based on grain characteristics, such as fmes content, plasticity, etc. The proposed equation to
obtain the equivalent clean sand normalized CPT penetration resistance, (qclN)cs, is a function of
both the measured penetration resistance, qc!N, and the grain characteristics of the soil, as
follows:

{qclN)cs = Kc qc!N (9)

where Kc is a correction factor that is a function of the grain characteristics of the soil, as
described later in this paper.

Grain characteristics from the CPT

In recent years, charts have been developed to estimate soil type from CPT data (Olsen and
Malone, 1988; Olsen and Koester, 1995; Robertson and Campanella, 1988; Robertson, 1990).
Experience has shown that the CPT friction ratio (ratio of the CPT sleeve friction to the cone tip
resistance) increases with increasing fines content and soil plasticity. Hence, grain characteristics

68
such as apparent fines content of sandy soils can be estimated directly from CPT data using any
of these soil behaviour charts, such as that by Robertson (1990) shown in Figure 18. As a result,
the measured penetration resistance can be corrected to an equivalent clean sand value.
The addition of pore pressure data can also provide valuable additional guidance in estimating
fines content. Robertson et al. (1992) suggested a method for estimating fines content based on
the rate of pore pressure dissipation (t50) during a pause in the CPT.

1000

7 8
9

0
...
w
0 100
z
~
(I)
6
ti}
w
a:
w
z
0
(.)
0
w 10
N
:J
<
:e
a:
0
:z 1
2
1
0.1 10
NORMALIZED
f :s
FRICTION RATIO, F- x100%
qt - <lvo

I. Sensitive, fine grained 6. Sands - clean sand to silty sand

2. Organic soils - peats 7. Gravelly sand to dense sand
3. Clays - silty clay to clay 8. Very stiff sand to clayey sand*
4. Silt Mixtures - clayey silt to silty clay 9. Very stiff, fine grained*
5. Sand Mixtures - silty sand to sandy silt
* Heavily overconsolidated or cemented

Figure 18 Normalized CPT soil behaviour type chart, as proposed by Robertson (1990).

69
Based on extensive field data and experience, it is possible to estimate grain characteristics
directly from CPT results using the soil behaviour type chart shown in Figure 18. The booodaries
between soil behaviour type Zones 2 to 7 can be approximated as concentric circles (Jefferies and
Davies, 1993). The radius of each circle can then be used as a soil behaviour type index. Based
on the CPT chart developed by Robertson (1990), the soil behaviour type index, L:, can be
defined as follows:

2 2 ~s
L: = [ (3.47 - log Q) + (log F + 1.22)] (10)

where Q=(qc-crvo)( ~•
Pa2 cr vo
)n is the normalized CPT penetration resistance, dimensionless;

n = exponent, typically equal to 1.0; F = [fs /(qc - crv0 )] x 100% is the normalized friction ratio, in
percent; fs is the CPT sleeve friction stress; O'vo and cr'vo are the total and effective overburden
stresses, respectively; Pa is a reference pressure in the same units as cr'vo (i.e. P a=l 00 kPa if cr'vo is
in kPa); and Pa2 is a reference pressure in the same units as qc and O'vo (i.e. Pa2=0.l MPa if qc and
O'vo are in MPa).

The soil behaviour type chart by Robertson (1990) uses a normalized cone penetration resistance
(Q) based on a simple linear stress exponent of n = 1.0 (see above), whereas the chart
recommended here for estimating CRR (see Figure 16) is essentially based on a normalized cone
penetration resistance (qc!N) based on a stress exponent n = 0.5 (see Equation 8). Olsen and
Malone (1988) correctly suggested a normalization where the stress exponent (n) varies from
around 0.5 in sands to 1.0 in clays. However, this normalization for soil type is somewhat
complex and iterative.

The Robertson (1990) procedure using n=l.0 is recommended for soil classification in clay type
soils when le > 2.6. However, in sandy soils when le ::;; 2.6, it is recommended that data being
plotted on the Robertson (1990) chart be modified by using n=0.5. Hence, the recommended
procedure is to first use n = 1.0 to calculate Q and, therefore, an initial value of le for CPT data.
If le > 2.6, the data should be plotted directly on the Robertson (1990) chart (and assume
qc1N = Q). However, if le ::;; 2.6, the exponent to calculate Q should be changed to n = 0.5
(i.e. essentially calculate 4cIN using Equation 8 since O'vo <<" qc) and L: should be recalculated
based on qc1N and F. If the recalculated le remains less than 2.6, the data should be plotted on the
Robertson (1990) chart using 4cIN based on n = 0.5. If, however, L: iterates above and below a
value of 2.6, depending which value of n is used, a value of n = 0. 75 should be selected to
calculate 4cIN (using Equation 8) and plot data on the Robertson (1990) chart. Note that if the
in-situ effective overburden stresses are in the order of 50 kPa to 150 kPa, the choice of
normalization has little effect on the calculated normalized penetration resistance.

The boundaries of soil behaviour type are given in terms of the index, le, as shown in Table 4.

70
 Table 4. Boundaries of soil behaviour type (after Robertson, 1990).

Soil Behaviour Type Index, le Zone Soil Behaviour Type (see Figure 18)

le< 1.31 7 Gravelly sand to dense sand

1.31 <le< 2.05 6 Sands: clean sand to silty sand

2.05 < le < 2.60 5 Sand Mixtures: silty sand to sandy silt

2.60 < le < 2.95 4 Silt Mixtures: clayey silt to silty clay

2.95 <le< 3.60 3 Clays: silty clay to clay

le> 3.60 2 Organic soils: peats

The soil behaviour type index does not apply to Zones 1, 8 or 9. Along the normally
consolidated region in Figure 18, soil behaviour type index increases with increasing apparent
fines content and soil plasticity, and the following simplified relationship is suggested:

ifle < 1.26 Apparent fines content, FC (%) = 0 (I la)

if 1.26 ~le~ 3.5 Apparent fines content, FC (%) = 1.75 L: 3·25 - 3.7 (llb)
iflc > 3.5 Apparent fines content, FC (%) = 100 (I le)

The range of potential correlations is illustrated in Figure 19, which shows the variation of soil
behaviour type index (le) with apparent fines content and the effect of the degree of plasticity of
the fines. The recommended relationship given in Equation 11 is also shown in Figure 19. Note
that this equation is slightly modified from the original work by Robertson and Fear (1995) in
order to increase the prediction of apparent FC for a given value of le.

The proposed correlation between CPT soil behaviour index (le) and apparent fines content is
approximate, since the CPT responds to many other factors affecting soil behaviour, such as
soil plasticity, mineralogy, sensitivity and stress history. However, for small projects, the above
correlation provides a useful guide. Caution must be taken in applying Equation 11 to sands that
plot in the region defined by 1.64 < L: < 2.36 and F < 0.5% in Figure 18, so as not to confuse very
loose clean sands with denser sands containing fines. In this zone, it is suggested that the
apparent fmes content is set equal to 5%, such that no correction will be applied to the measured
CPT tip resistance when the CPT data plot in this zone. To evaluate the correlation shown in
Figure 19 it is important to show the complete soil profile (CPT and samples; e.g. see Figure 24),
since comparing soil samples with an adjacent CPT at the same elevation can be misleading due
to soil stratigraphic changes and soil heterogeneity.

71
 4-r-----------------------------------,
Zone 2: organic soils - peats
- --
-. - - -----
Zone 3: silty clay to clay
- -, - - ---
Zone 4: clayey silt to silty ~lay • • 1ecouunen
•• -
, , •
,
, ,.
,. ,.
--- ---
--- --- -- - -- Zone 6: clean sand to silty sand

N.B. The correlation is approximate since I,

Zone 7: gravelly sand to dense sand
depends on other factors such as plasticity,
mineralogy, sensitivity and stress history.

0--t-----'-----+-----'-----+-----'----+---...._--+---....__ ___,

0 10 20 30 40 50
Apparent Fines Content, FC (%)

Figure 19 Variation of CPT soil behaviour type index (le) with apparent fines content in
or close to the normally consolidated zone of the soil behaviour chart by
Robertson (1990).

Based on the above method for estimating grain characteristics directly from the CPT using the
soil behaviour index (le), the recommended relationship between L: and the correction factor Kc is
shown in Figure 20 and given by the following equations:

ifL:::;; 1.64 Kc= 1.0 (12a)

ifL: >1.64 Kc= - 0.403 I/+ 5.581 L:3 - 21.63 L:2 + 33.75 L:-17.88 (12b)

The proposed correction factor, Kc, is approximate since the CPT responds to many factors, such
as soil plasticity, fines content, mineralogy, soil sensitivity and stress history. However, for small
projects or for initial screening on larger projects, the above correlation provides a useful guide.
Caution must be taken in applying the relationship to sands that plot in the region defined by
1.64 < L: < 2.36 and F::;; 0.5% so as not to confuse very loose clean sands with sands containing
fines. In this zone, it is suggested that the correction factor Kc be set to a value of 1.0
(i.e. assume that the sand is a clean sand).

72
 I,= 2.6
5
I
Gravelly Sands Sands Sand
I
::l 4.5 Mixtures

.: •
...
Q
(,>
I
,
~ 4
,
.::...= ,
.
(,>
QI

Q
3.5 I

(,>

.:!
.... 3 4 3 2
·c"' K, =- 0.4031c + 5.5811c - 21.631c + 33. 751c - 17.88
...
QI

.
(,>
= 2.5 Silt Clays

-==
(,>
Mixtures

·;=
. 2
t)J)

~
,:i.
u 1.5

1
0 0.5 1 1.5 2 2.5 3 3.5 4
Soil Behaviour Type Index, le

Figure 20 Recommended grain characteristic correction to obtain clean sand equivalent

CPT penetration resistance in sandy soils.

Note that the relationship between the recommended correction factor, Kc, and soil behaviour
type index, le, is shown dashed beyond an le of 2.6, which corresponds to an approximate
apparent fines content of 35%. Soils with le > 2.6 fall into the clayey silt, silty clay and clay
regions of the CPT soil behaviour chart (i.e. Zones 3 and 4). When the CPT indicates soils in
these regions (le > 2.6), samples should be obtained and evaluated using the criteria shown in
Figure 9. It is reasonable to assume, in general, that soils with le> 2.6 are non-liquefiable and
that the correction Kc could be large. Soils that fall in the lower left region of the CPT soil
behaviour chart (Figure 18), defined by le> 2.6 and F:s; 1.0%, can be very sensitive and, hence,
possibly susceptible to both cyclic and/or flow liquefaction. Soils in this region should be
evaluated using criteria such as that shown in Figure 9 combined with additional testing.

Figure 21 shows the resulting equivalent CRR curves for le values of 1.64, 2.07 and 2.59 which
represent approximate apparent fines contents of 5%, 15% and 35%, respectively.

73
 Apparent FC = 35% 15% S:5%
le = 2.59 2.07 1.64

0.5

~ 0.4
U
._,
0
.:::
~
i:r::
.
Qi
0.3
....=
~

.la
"'
~
.:! 0.2
;:;
>,
u

0.1

0 50 100 200 250 300

Figure 21 CPT base curves for various values of soil behaviour index, le (corresponding
to various apparent fines contents, as indicated).

Influence of thin layers

A problem associated with the interpretation of penetration tests in interbedded soils occurs when
thin sand layers are embedded in softer deposits. Theoretical as well as laboratory studies show
that the cone resistance is influenced by the soil ahead of and behind the penetrating cone. The
cone will start to sense a change in soil type before it reaches the new soil and will continue to
sense the original soil even when it has entered a new soil. As a result, the CPT will not always
measure the correct mechanical properties in thinly interbedded soils. The distance over which
the cone tip senses an interface increases with increasing soil stiffness. In soft soils, the diameter
of the sphere of influence can be as small as 2 to 3 cone diameters, whereas, in stiff soils, the
sphere of influence can be up to 20 cone diameters. Hence, the cone resistance can fully respond
(i.e. reach full value within the layer) in thin soft layers better than in thin stiff layers. Care
should, therefore, be taken when interpreting cone resistance in thin sand layers located within
soft clay or silt deposits. Based on a simplified elastic solution, Vreugdenhil et al. (1994) have

74
provided some insight as to how to correct cone data in thin layers. Vreugdenhil et al. (1994)
have shown that the error in the measured cone resistance within thin stiff layers is a function of
the thickness of the layer as well as the stiffness of the layer relative to that of the surrounding
softer soil. The relative stiffness of the layers is reflected by the change in cone resistance from
the soft surrounding soil to the stiff soil in the layer (qcA/qcs). Vreugdenhil et al. ( 1994) validated
the model with laboratory and field data.

Hid,
(a)
o 20
® m 80
2 -.-----.-""T""--,---,----,------,--,-----,---,

~
=
.:
s .."
."".,=
.."t 1.5
...
u..."
...,,.,
..l
.5
...-=
Relationship :

0 1000 2000 3000

Layer Thickness, H (mm) for a 10 cm' cone (i.e. d, =35.7 mm)

4c
(b)

4cB
Deposit B

Layer A - \

I
I
------------------------------------... r----
1
- - - - - - - q.* ---------,
4cB
Deposit B

Figure 22 Suggested correction (Kn) to CPT penetration resistance in thin sand layers
(based on results by Vreugdenhil et al., 1994) (after Robertson and Fear, 1995).

75
Based on this work, Robertson and Fear (1995) suggested a correction factor for cone resistance
(originally termed Kc by Robertson and Fear (1995), but now termed K8 in order to prevent
confusion with the correction outlined above for obtaining clean sand equivalent penetration
resistance) as a function of layer thickness (H), as shown in Figure 22. The corrections apply
only to thin sand layers embedded in thick fine-grained layers. The corrections appear to have
a reasonable trend, but are rather large. Therefore, Robertson and Fear (1995) recommended
a conservative correction (corresponding to (qcAlqc8 ) = 2) as is shown in Figure 22 and given by
the following expression:

2
Hid )
KH = 0.5 ( 28 c -1.45 + 1.0 (13)

where His the layer thickness, in mm, for H < 40.6 de; qcA and qc8 are the tip resistances in the
layer and in the soil surrounding the layer, respectively; and de is the cone diameter, in mm
(e.g. for a 10 cm2 cone, de=35. 7 mm).

Thin sand layers embedded in soft clay deposits are often incorrectly classified as silty sands
based on the CPT soil behaviour type charts. Hence, a slightly improved classification can be
achieved if the cone resistance is first corrected for layer thickness before applying the
classification charts.

Cyclic resistance from the CPT

In an earlier section, a method was suggested for estimating apparent fines content directly from
CPT results, using Equation 11. Following the traditional SPT approach, the estimated apparent
fines content could be used to estimate the correction necessary to obtain the clean sand
equivalent penetration resistance. However, since other grain characteristics also influence the
measured CPT penetration resistance, it is recommended that the necessary correction be
estimated from the soil behaviour type index, as described above. Hence, Equations 9, 10 and 12
can be combined to estimate the equivalent clean sand normalized penetration resistance, (qciN)cs,
directly from the measured CPT data. Then, using the equivalent clean sand normalized
penetration resistance (qcJN)cs, the CRR (for M = 7.5) can be estimated using the following
simplified equation (which approximates the clean sand curve recommended in Figure 16):

3
if 50 ::;; (qc1N)cs < 160 CRR = 93 ((qclN )cs) + 0.08 (14a)
1000

if (qc1N)cs < 50 CRR = 0.833 ((qc!N )cs)+ 0.05

1000
(14b)

In summary, Equations 9 to 12 and 14 (and Equation 13 if thin layers are present) can be
combined to provide an integrated method for evaluating the cyclic resistance (M=7.5) of

76
 I Liquefaction Analysis
I
I I
Flow Liquefaction
I Cyclic Softening

I
I
Laboratory In-Situ
Testing Testing

I
I I I
SPT CPT V,

,················••.. I
Thin layer
correction
...
q,
•
( tip resistance)
•f,
(sleeve friction)

~ I I
l................·..·· In-situ stresses (o-,., o-,.')
I
! !
Q F
= ( q,-0-.,.)/0-.,.' = [f,!(q,-o-,.)lxlOO%

.• •
Pa
_ qc
qc!N-(-p-)(--,)
o.s
. 1,>2.6
.I
a2 O' VO .
'
I,<2.6
I I,= [(3.47-logQ)2+(1ogF+l.22)'] 0 '
I .I
qdN = Q
I
P,=IOOkPa if o-,.' is in kPa I I,= [(3 .47-logq,IN)2+(1ogF+ 1.22)']°"5
P., =0.1 MPa if q, is in MPa I
I,<2.6
! I,>2.6
qc!N = ( -1£_
p )( O' , ~ l'"
a2 VO
I,= [(3.47-logq,lN)'+(logF+ 1.22)'] 0-'

if], 5 1.64
if I.,> 1.64
K,= 1.0
•
K, = • 0.403 1:+ 5.581 1/- 21.631/ + 33. 75 I, - 17.88
•
ifl• .: 2.6 evaluate using other criteria; likely non-Iiquefiable ifF> 1% as well
BUT,if 1.64<1,<2.36 and F< 0.5%, setK, = 1.0

.I
.I
•
(qc!Nla = (K.)(qclN)
I
if50 5 (q,IN)cs < 160 CRR75
•
= 93 ((qc!N )cs ) 3 + 0.08
1000
if(q,IN)cs< 50 CRR 0.833 /qc!N )cs) + 0.05
,., = 1000
Note: if l.>2.6, evaluate using other criteria; likely non-Iiquefiable ifF> I% as well

Figure 23 Flowchart illustrating the application of the integrated CPT method of

evaluating cyclic resistance ratio (CRR) in sandy soils.

77
saturated sandy soils based on the CPT. The CPT-based method is an alternative to the SPT or
shear wave velocity CVs) based in-situ methods; however, using more than one method is useful
in providing independent evaluations of liquefaction potential. The integrated CPT method is
summarized in Figure 23 in the form of a flowchart. The flowchart clearly shows the
step-by-step process involved in using the proposed integrated method based on the CPT for
evaluating CRR and indicates the recommended equations for each step of the process.

An example of this proposed modified CPT-based method is shown in Figure 24 for the Moss
Landing site that suffered cyclic liquefaction during the 1989 Loma Prieta earthquake in
California (Boulanger et al., 1995 and 1997). The measured cone resistance is normalized and
corrected for overburden stress to qc1N and F and the soil behaviour type index (le) is calculated.
The final continuous profile of CRR at N=15 cycles (M = 7.5) is calculated from the equivalent
clean sand values of qc1N (i.e. (qcJN)cs = K, qc!N) and Equation 14. Included in Figure 24 are
measured fines content values obtained from adjacent SPT samples. A good comparison is seen
between the estimated apparent fines contents and the measured fines contents. Note that for
L: > 2.6 (i.e. FC > 35%; see Equation 11), the soil is considered to be non-liquefiable; however,
this should be checked using other criteria (e.g. Marcuson et al., 1990; see Figure 9). The
estimated zones of soil that are predicted to experience cyclic liquefaction are very similar to
those observed and reported by Boulanger et al. (1995 and 1997).

Olsen (1988), Olsen and Koester (1995) and Suzuki et al. (1995b) have also suggested integrated
methods to estimate the CRR of sandy soils directly from CPT results with the correlations
presented in the form of soil behaviour charts. The Olsen and Koester (1995) method is
based on SPT-CPT conversions plus some laboratory based CRR data. The method by
Suzuki et al. (1995b) is based on limited field observations. The methods by Olsen and Koester
(1995) and Suzuki et al. (1995b) are shown in Figure 25. The Olsen and Koester (1995) method
uses a variable normalization technique, which requires an iterative process to determine the
normalization. The method by Suzuki et al. (1995b) uses the qc1N normalization suggested in this
paper (Equation 8 with n=0.5). The Olsen and Koester (1995) method is very sensitive to small
variations in measured friction ratio and the user is not able to adjust the correlations based on
site specific experience. The friction sleeve measurement for the CPT can vary somewhat
depending on specific CPT equipment and on tolerance details between the cone and the sleeve
and hence, can be subject to some uncertainty. The method proposed in this paper is based on
field observations and is essentially similar to those of Olsen and Koester ( 1995) and Suzuki et
al. (1995b); however, the method described here is slightly more conservative and the process has
been broken down into its individual components.

Built into each of the CPT methods for estimating CRR is the step of correcting the measured
cone tip resistance to a clean sand equivalent value. It is the size of this correction that results in
the largest differences between predicted values of CRR from the various methods. Figure 26
provides an approximate comparison between the methods by Stark and Olson (1995),
Olsen and Koester (1995) and Suzuki et al. (1995b), in terms of Kc from soil behaviour type
index, L:. Also superimposed on Figure 26 is the recommended relationship, based on the
equations given in this paper. The comparisons are approximate because the different authors

78
 'I.JN F(°/o) Apparent FC (%) I(,, (q.,,,.J. am,,
0 SO 100 ISO 200 0 2 4 4 0 20 40 60 80 100 0 3 6 9 12 15 0 SO 100 150 200 0 0.2 0.4 0.6
0+---,.--.--..---, o+-,;,-+--+--,1 0 t--t--+--+---+--1 0

2 2 2 2 2 -~ .
4 4 4 • 4
,.
4 4

•
,.-
~

6 6 6 6 6 6

g g g g 1,s 8 .ac=
,s 8 ,S8 ,s 8 ii
g ~
g ~
8
g
"' "'
-.J
\0 10 10 10 10 10 10
'

12 12 12 12 12 12

)
14 14 14 Dob represeot FC or 14 14 14
iuuqilr! from adjacent
C
SPI'OCBIO

16 1 0 ~ - - - - -0 1 16 16 16 16 16
F=-0.5'% Gitofflc=2.6

Figure 24 Application of the integrated CPT method for estimating cyclic resistance ratio (CRR) to the State Beach site
at Moss Landing, which suffered cyclic liquefaction during the 1989 Loma Prieta earthquake in California.
 1000

101X)+---~-
(a) : P ,~ L~Il
, q
• cl~
•q • I
c ~, lc•Mli] ~1.
·-- •
r;;..~;,7 . ~, 100 !
,;;.

•
j • •
,i r • '" ;=0,25
~ !'----~~,_,_~~-~....,JO
j 100
• Liquefaction
1000

e o No Liquefaction
i
.!I 10:
j '
?i
~ ... 100 ,5,
01 0.15 Ol O.l__.
CPT pr.dd.r IICIIIT~IH4 lqiaefacllon
1•mbl'IG• cyd'c.trna n,lio (CSR]
0-1-,--....::;;:,· ;-;·rr~:0--r·-+,c-,,..,,-i,-c,-,l+o.o
FndloR RaUo (")
•
,,,.,:=0.15
!'----~~,_,_-~-~....... 10
0.1 I 10
Friction Ratio, R (%)
1

Figure 25 Comparison of estimating CRR from the CPT (a) CRR=0.05 to 0.3, after Olsen
and Koester (1995); (b) CRR=0.15 and CRR=0.25, after Suzuki et al. (1995b).

5
-.-Olsen & Koester (1995): CRR=0.05; (qclN)cs=64
-<>-Olsen & Koester (1995): CRR=0.30; (qclN)cs=200
4,5 ....,._ Suzuki et al.(1995): CRR=0.15; (qclN)cs=l07
-tr-Suzuki et al.(1995): CRR=0.25; (qc1N)cs=l60
'it,..
---stark&Olson (1995): CRR=0.08; (qclN)cs=53
,.;-
4 --o-Stark &Olson (1995); CRR=0.50; (qclN)cs=l63
~
.,!l
- Recommendation of this paper
=
"'"" 3.5 N.B. if 1.64 <I,< 2.36 and F < 0.5%, then K,, = 1.0
!:
8
:i.. 3
Silt Clays
:i Sands

....."
Gravelly Sands Midurea
"" 2.5
"
......
.!:
f
2

u
1.5

0 0.5 1.5 2 2.5 3 3.5 4

Soil Behaviour Type Index, I,

Figure 26 Approximate comparison of various methods for correcting CPT tip resistance
to clean sand equivalent values, based on soil behaviour type.

80
use different normalizations when correcting CPT tip resistance for plotting data on a soil
classification chart. However, for effective overburden stresses in the order of 100 kPa c~ 1 tsf),
all of the normalization methods should give similar values of normalized penetration resistance
for the same value of measured penetration resistance. The boundaries between different soil
behaviour types are also indicated on Figure 26, as a guide to the type of soil in which corrections
of certain magnitudes are suggested by the various methods.

The methods by Suzuki et al. (1995b) and Olsen and Koester (1995) appear to be consistent with
each other, indicating that Kc increases with increasing CRR. Very large corrections result
especially in soils of high le. The method by Stark and Olson (1995) indicates that Kc decreases
with increasing CRR for a given le and does not fit in with the trend of the combined Suzuki
et al. (1995b) and Olsen and Koester (1995) lines. The magnitudes of the corrections suggested
by Stark and Olson (1995) are generally smaller than the other methods.

Figure 26 indicates that the recommended correction is generally more conservative than the
corrections proposed by the other authors. The other methods generally predict higher values of
Kc and suggest that corrections should be applied beginning at lower values of le, particularly for
higher values of CRR. Note that the recommended relationship between Kc and le is shown
dashed beyond le =2.6, which corresponds to an apparent FC of 35% (Equation 11 ). This shows
that the integrated CPT method for evaluating CRR, as outlined here, does not apply to soils that
would be classified as clayey silt, silty clay or clay. As explained earlier, when interpretation of
the CPT indicates that these types of soils are present, samples should be obtained and evaluated
using other criteria, such as that given in Figure 9 (Marcuson et al., 1990). It is logical that in
non-liquefiable clay soils, the equivalent correction factor, Kc, could be very large for le >2.6.

Recommendations

For low risk, small-scale projects, the potential for cyclic liquefaction can be estimated using
penetration tests such as the SPT or CPT. The CPT is generally more repeatable than the SPT
and is the preferred test, where possible. The CPT provides continuous profiles of penetration
resistance, which are useful for identifying soil stratigraphy and for providing continuous profiles
of estimated cyclic resistance ratio (CRR). When using the SPT, care should be taken to ensure
that the test is carried out according to the above recommended procedures. Corrections are
required for both the SPT and CPT for grain characteristics, such as fines content and plasticity.
For the CPT, these corrections are best expressed as a function of soil behaviour type index, Ic,
which is affected by a variety of grain characteristics.

For medium to high-risk projects, the SPT and CPT can be useful for providing a preliminary
estimate of liquefaction potential in sandy soils. For higher risk projects, it is important to
measure the rod energy during each SPT. For higher risk projects, it is also preferred practice to
drill sufficient boreholes adjacent to CPT soundings to verify various soil types encountered and
to perform index testing on disturbed samples. A procedure has been described to correct the

81
measured cone resistance for grain characteristics based on the CPT soil behaviour type index, L:.
The corrections are approximate, since the CPT responds to many factors affecting soil
behaviour. Expressing the corrections in terms of soil behaviour index is the preferred method of
incorporating the effects of various grain characteristics, in addition to fines content. When
possible, it is recommended that the corrections be evaluated and modified to suit a specific site
and project. However, for small-scale low risk projects and in the initial screening process for
higher risk projects, the suggested general corrections provide a useful guide. A guide has also
been given for correcting CPT results in thin sand layers embedded in softer fine-grained
deposits. The SPT and CPT are generally limited to sandy soils with limited gravel contents. In
soils with high gravel contents, penetration may be limited.

A summary of the CPT method is shown in Figure 27, which identifies the zones in which soils
are susceptible to cyclic liquefaction (primarily Zone A). In general, soils with le > 2.6 and
F > 1.0% (Zone B) are likely non-liquefiable. Soils that plot in the lower left portion of the chart
(Zone C; le> 2.6 and F < 1.0%) may be susceptible to cyclic and/or flow liquefaction due to the
sensitive nature of these soils. Soils in this region should be evaluated using other criteria.
1000 ~ - - - - - - - ~ - - - - - - - ~

Q
...
....=
.:: 100
.
:'.l
g
.
,:::

1C.

s:u
.,,. 10
•!:!
oi

J C
1
0.1 1 10
Normalized CPT friction ratio, F
Zone A: Cyclic liquefaction possible - depends on size and duration of cyclic loading.
Zone B: Liquefaction unlikely - check other criteria.
Zone C: Flow/cyclic liquefaction possible - depends on soil plasticity and sensitivity as well as size
and duration of cyclic loading.
Figure 27 Summary ofliquefaction potential on soil behaviour chart by Robertson (1990).

82
Caution should be exercised when extrapolating the suggested SPT and CPT correlations to
conditions outside of the range from which the field performance data were obtained.
An important feature to recognize is that the correlations appear to be based on average values for
the inferred liquefied layers. However, the correlations are often applied to all measured SPT
and/or CPT values, which include low values below the average for a given sand deposit. Hence,
the correlations could be conservative in variable stratified deposits where a small part of the
penetration data could indicate possible liquefaction.

Finally, as mentioned earlier, it is clearly useful to evaluate CRR using more than one method.
The seismic CPT provides a useful technique for independently evaluating liquefaction potential,
since it measures both the usual CPT parameters and shear wave velocities within the same
borehole. The CPT provides detailed profiles of cone tip resistance, but the penetration
resistance is sensitive to grain characteristics, such as fines content and soil mineralogy and,
hence, corrections are required. The seismic part of the CPT provides a shear wave velocity
profile typically averaged over l m intervals and, therefore, contains less detail than the cone tip
resistance profile. However, shear wave velocity is less influenced by grain characteristics and
few or no corrections are required (Robertson et al, 1992; Andrus and Stokoe, 1996). Shear wave
velocity should be measured with care in order to provide the most accurate results possible since
the estimated CRR is sensitive to small changes in shear wave velocity. There should be
consistency in the liquefaction evaluation using either method. If the two methods provide
different predictions of CRR profiles, samples should be obtained to evaluate the grain
characteristics of the soil.

Acknowledgements

The authors appreciate the contributions of the members of the 1996 NCEER Workshop on
Evaluation of Liquefaction Resistance (T.L. Youd, Chair) that was held in Salt Lake City, Utah.
Particular appreciation is extended to the following individuals: W.D.L. Finn, I.M. Idriss,
J. Koester, S. Liao, W.F. Marcuson III, J.K. Mitchell, R. Olsen, R. Seed, and T.L. Youd.

References

Andrus, R.D., and Stokoe, K.H. 1996. Guidelines for evaluation of liquefaction resistance using
shear wave velocity. Submission to the 1996 NCEER Workshop on Evaluation of
Liquefaction Resistance (T.L. Youd, Chair), Salt Lake City, Utah.
Bartlett, S.F., and Youd, T.L. 1995. Empirical prediction of liquefaction induced lateral spread.
Journal ofGeotechnical Engineering, ASCE, 121(3): 249-261.
Been, K., Jefferies, M. G., and Hachey, J. 1991. The critical state of sands. Geotechnique,
41(3): 365-381.

83
Bishop, A.W. 1973. The stability of tips and spoil heaps. Quarterly Journal of Engineering.
Geology, 6: 335-376.
Blake, T.F. 1996. Personal communication from T.L. Youd, 1996.
Boulanger, R.W., Idriss, I.M., and Mejia, L.H. 1995. Investigation and evaluation of liquefaction
related ground displacements at Moss Landing during the 1989 Loma Prieta earthquake.
Center for Geotechnical Modeling, Department of Civil and Environmental Engineering,
College of Engineering, University of California at Davis, Report No. UCD/CGM-95/02.
Boulanger, R.W., Mejia, L.H., and Idriss, I.M. 1997. Liquefaction at Moss Landing during
Loma Prieta earthquake, Journal of Geotechnical and Geoenvironmental Engineering,
123(5): 453-467.
Casagrande, A. 1965. The role of the "calculated risk" in earthwork and foundation engineering.
The Terzaghi Lecture, Journal of Soil Mechanics and Foundation Division, ASCE,
91(4): 1-40.
Castro, G. 1969. Liquefaction of sands. Harvard Soil Mechanics Series No. 81, Harvard
University, Cambridge, MA.
Dobry, R., Ladd, R.S., Yokel, F.Y., Chung, R.M., and Powell, D. 1982. Prediction of pore water
pressure buildup and liquefaction of sands during earthquakes by the cyclic strain method.
NBS Building Science Series 138, U.S. Department of Commerce, National Bureau of
Standards, Gaithersburg, MD, 152p.
Fear, C.E., and McRoberts, E.C. 1995. Reconsideration of initiation of liquefaction in sandy
soils. Journal of Geotechnical Engineering, ASCE, 121(3): 249-261.
Finn, W.D.L. 1981. Liquefaction potential: developments since 1976. Proceedings,
1st International Conference on Recent Advances in Geotechnical Earthquake Engineering and
Soil Dynamics, St. Louis, Vol. 2, pp. 655-681.
Hofrnann, B.A., Sego, D.C., and Robertson, P.K. 1995. Undisturbed sampling of a deep loose
sand deposit using ground freezing. Proceedings of the 47th Canadian Geotechnical
Conference, Halifax, Nova Scotia, pp. 287-296.
Hofrnann, B.A., 1997. In-situ ground freezing to obtain undisturbed samples of loose sand for
liquefaction assessment, Ph.D. Thesis, University of Alberta, Edmonton, Alberta, Canada.
Ishihara, K. 1993. Liquefaction and flow failure during earthquakes. The 33rd Rankine Lecture.
Geotechnique 43(3): 351-415.
Ishihara, K., and Koseki, J. 1989. Cyclic shear strength of fines-containing sands. Earthquake
Geotechnical Engineering. Proceedings, Discussion Session on Influence of Local Conditions
on Seismic Response, 12th International Conference on Soil Mechanics and Foundation
Engineering, Rio de Janeiro, pp. 101-106.
Jefferies, M.G., and Davies, M.P. 1993. Use of CPTu to estimate equivalent SPT N 60 • ASTM
Geotechnical Testing Journal, 16(4): 458-467.
Kayen, R.E., Mitchell, J.K., Lodge, A., Seed, R.B., Nishio, S., and Coutinho, R. 1992.
Evaluation of SPT-, CPT-, and shear wave-based methods for liquefaction potential
assessment using Loma Prieta data. Proceedings, 4th Japan-U.S. Workshop on Earthquake
Resistant Design of Lifeline Facilities and Countermeasures for Soil Liquefaction, Technical
Report NCEER-94-0019, Edited by M. Hamada and T.D. O-Rourke, Vol.I, pp. 177-204.

84
 Kokusho, T., Yoshida, Y., and Eashi, Y. 1983. Evaluation of seismic stability of dense sand
layer (Part 2) - Evaluation method by standard penetration test. Electric Power Central
Research Institute, Japan, Report 383026(in Japanese).
Koppejan, A.W., Van Wamelen, B.M., and Weinberg, L.J.H. 1948. Coastal landslides in the
Dutch province of Zealand. Proceedings, 2nd International Conference on Soil Mechanics and
Foundation Engineering, Rotterdam, Holland, pp. 89-96.
Kulhawy, F.H., and Mayne, P.W. 1990. Manual on Estimating Soil Properties for Foundation
Design. Cornell University, EL-6800, Research Project 1493-6, prepared for Electric Power
Research Institute.
Ladd, R.S. 1974. Specimen preparation and liquefaction of sands. Journal of ASCE,
ll0(GTIO): 1180-1184.
Liao, S.S.C., and Whitman, R.V. 1986a. Overburden correction factors for SPT in sand, Journal
of Geotechnical Engineering, ASCE, 112(3): 373-377.
Liao, S.S.C., and Whitman R.V. 1986b. A catalog of liquefaction and non-liquefaction
occurrences during earthquakes. Research Report, Department of Civil Eng., M.I.T.,
Cambridge, MA.
Liao, S.S.C., Veneziano, D., and Whitman, R.V. 1988. Regression models for evaluating
liquefaction probability. Journal of Geotechnical Engineering, ASCE, 114(4): 389-411.
Marcuson, III, W.F., Hynes, M.E., and Franklin, A.G. 1990. Evaluation and use of residual
strength in seismic safety analysis of embankments. Earthquake Spectra, 6(3): 529-572.
Mitchell, J.K., and Tseng, D.-J. 1990. Assessment of liquefaction potential by cone penetration
resistance. Proceedings, H. Bolton Seed Memorial Symposium, Berkeley, California,
Edited by J.M. Duncan, Vol. 2, pp. 335-350.
Mogami, and Kubo 1953. The behaviour of soil during vibration. Proceedings, 3rd International
Conference on Soil Mechanics and Foundation Engineering, Vol. 1, pp. 152-153.
Mulilis, J.P., Seed, H.B., Chan, C.K., Mitchell, J.K., and Arulanandan, K. 1977. Effects of
sample preparation on sand liquefaction. Journal of ASCE, 103(GT2): 99-108.
Olsen, R.S. 1988. Using the CPT for dynamic response characterization. Proceedings,
Earthquake Engineering and Soil Dynamics II Conference, ASCE.
Olsen R.S., and Koester J.P. 1995. Prediction of liquefaction resistance using the CPT.
Proceedings, International Symposium on Cone Penetration Testing, CPT'95. Linkoping,
Sweden, Vol. 2, pp. 251-256.
Olsen, R.S., and Malone, P.G. 1988. Soil classification and site characterization using the cone
penetrometer test. Penetration Testing 1988, ISOPT-1, Edited by De Ruiter, Balkema,
Rotterdam, Vol. 2, pp. 887-893.
Pando, M., and Robertson, P.K. 1995. Evaluation of shear stress reversal due to earthquake
loading for sloping ground. Proceedings, 48tll Canadian Geotechnical Conference, Vancouver,
B.C., Vol. 2, pp. 955-962.
Poorooshasb, H.B., and Consoli, N.C. 1991. The ultimate state. Proceedings, IX Pan-American
Conference, pp. 1083-1090.
Robertson, P.K. 1990. Soil classification using the CPT. Canadian Geotechnical Journal,
27(1): 151-158.

85
Robertson, P.K. 1994. Suggested terminology for liquefaction. Proceedings, 47th Canadian
Geotecbnical Conference, Halifax, Nova Scotia, pp. 277-286.
Robertson, P.K., and Campanella, R.G. 1985. Liquefaction potential of sands using the cone
penetration test. Journal of Geotechnical Division of ASCE, March 1985, 22(3): 298-307.
Robertson, P.K., and Campanella, R.G. 1988. Design manual for use of CPT and CPTu.
Pennsylvania Department of Transportation, (Penn Dot), 200 p.
Robertson, P.K., and Fear, C.E. 1995. Liquefaction of sands and its evaluation. Proceedings,
IS Tokyo '95, 1st International Conference on Earthquake Geotechnical Eng., Keynote Lecture.
Robertson, P.K., Campanella, R.G., and Wightman, A. 1983. SPT-CPT correlations. Journal. of
Geotecbnical Division of ASCE, 109: 1449-1459.
Robertson, P.K., Woeller, D.J., and Finn, W.D.L. 1992. Seismic cone penetration test for
evaluating liquefaction potential under cyclic loading. Canadian Geotechnical Journal,
29: 686-695.
Roscoe, K.H., Schofield, A.N., and Wroth, C.P. 1958. On the yielding of soils. Geotechnique,
8: 22-53.
Sasitharan, S., Robertson, P.K., Sego, D.C., and Morgenstern, N.R. 1994. State boundary surface
for very loose sand and its practical implications. Canadian Geotechnical Journal,
31(3): 321-334.
Schmertrnann, J.H. 1979. Statics of SPT. Journal of the Geotechnical Engineering Division,
ASCE, 105(GT5): 655-670.
Seed, H.B. 1979. Soil liquefaction and cyclic mobility evaluation for level ground during
earthquakes. Journal of the Geotechnical Engineering Division, ASCE, 105(GT2): 201-255.
Seed, H.B., and de Alba, P. 1986. Use of SPT and CPT tests for evaluating the liquefaction
resistance of sands. Use of In-situ Tests in Geotechnical Engineering, ASCE, Geotechnical
Special Publication, 6: 281-302.
Seed H.B., and Idriss, I.M. 1971. Simplified procedure for evaluating soil liquefaction potential.
Journal of the Soil Mechanics and Foundation Division, ASCE, 97(SM9): 1249-1273.
Seed, H.B., Tokimatsu, K., Harder, L.F., and Chung, R. 1985. Influence of SPT procedures in
soil liquefaction resistance evaluations. Journal of Geotechnical Engineering, ASCE
111(12): 1425-1445.
Sego, D.C., Robertson, P.K., Sasitharan, S., Kilpatrick, B.L., and Pillai, V.S. 1994. Ground
freezing and sampling of foundation soils at Duncan Dam. Canadian Geotechnical Journal,
31(6): 939-950.
Shibata, T. 1981. Relations between N-value and liquefaction potential of sand deposits.
Proceedings, 16th Annual Convention of Japanese Society of Soil Mechanics and Foundation
Engineering, pp. 621-624 (in Japanese).
Shibata, T., and Teparaska, W. 1988. Evaluation of liquefaction potentials of soils using cone
penetration tests. Soils and Foundations, 28(2): 49-60.
Skempton, A.W. 1986. Standard penetration test procedures and the effects in sands of
overburden pressure, relative density, particle size, aging and overconsolidation.
Geotechnique 36(3): 425-447.
Sladen, J.A., D'Hollander, R.D., and Krahn, J. 1985. The liquefaction of sands, a collapse
surface approach. Canadian Geotechnical Journal, 22: 564-578.

86
Stark, T.D., and Olson, S.M. 1995. Liquefaction resistance using CPT and field case histories.
Journal ofGeotechnical Engineering, ASCE 121(12), 856-869.
Suzuki, Y., Tokimatsu, K., Taye, Y., and Kubota, Y. 1995a. Correlation between CPT data and
dynamic properties of in-situ frozen samples. Proceedings, 3rd International Conference on
Recent Advances in Geotechnical Earthquake Engineering and Soil Dynamics, St. Louis,
U.S.A., Vol.I.
Suzuki, Y., Tokimatsu, K., Koyamada, K., Taya, Y., and Kubota, Y. 1995b. Field correlation of
soil liquefaction based on CPT data. Proceedings, International Symposium on Cone
Penetration Testing, CPT'95. Linkoping, Sweden, Vol. 2, pp. 583-588.
Tatsuoka, F., Iwasaki, T., Touida, K., Yasuda, S., Hirose, M., Imai, T., and Kon-no, M. 1980.
Standard penetration tests and soil liquefaction potential evaluation. Soils and Foundations,
20(4): 95-111.
Tatsuoka, F., Ochi, K., Fujii, S., and Okamoto, M. 1986. Cyclic undrained triaxial and torsional
shear strength of sands for different sample preparation methods. Soil and Foundations,
26(3): 23-41.
Terzaghi, K. and Peck, R.B. 1948. Soil mechanics in engineering practice. John Wiley and
Sons, Inc., 2nd edition.
Tokimatsu, K., and Hosaka, Y. 1986. Effects of sample disturbance on dynamic properties of
sand. Soils and Foundations, 26(1): 53-64.
Tokimatsu, K., and Yoshimi, Y. 1983. Empirical correlation of soil liquefaction based on SPT
N-value and fines content. Soils and Foundations, 23(4): 56-74.
Vreugdenhil, R., Davis, R., and Berrill, J. 1994. Interpretation of cone penetration results in
multilayered soils. International Journal for Numerical Methods in Geomechanics,
18: 585-599.
Wang, W. 1979. Some findings in soil liquefaction. Water Conservancy and Hydroelectric
Power Scientific Research Institute, Beijing, China.
Yoshimi, Y., Richart, F.E., Prakash, S., Balkan, D.D., and Ilyichev 1977. Soil dynamics and its
application to foundation engineering. Proceedings, 9th International Conference on Soil
Mechanics and Foundation Engineering, Tokyo, Vol. 2, pp. 605-650.
Yoshimi, Y., Hatanaka, M., and Oh-Oka, H. 1978. Undisturbed sampling of saturated sands by
:freezing. Soils and Foundations, 18(3 ): 105-111.
Yoshimi, Y., Tokimatsu, K., and Hosaka, Y. 1989. Evaluation of liquefaction resistance of clean
sands based on high-quality undisturbed samples. Soils and Foundations, 29(1 ): 93-104.
Yoshimi, Y., Tokimatsu, K., and Ohara, J. 1994. In situ liquefaction resistance of clean sands
over a wide density range. Geotechnique, 44(3): 479-494.
Youd, T.L., Idriss, I.M., Andrus, R.D., Arango, I., Castro, G., Christian, J.T., Dobry, R., Finn,
W.D.L., Harder, L.F., Hynes, M.E., Ishihara, K., Koester, J., Liao, S., Marcuson III, W.F.,
Martin, G.R., Mitchell, J.K., Moriwaki, Y., Power, M.S., Robertson, P.K., Seed, R., and
Stokoe, K.H. 1997. Summary report of the 1996 NCEER Workshop on Evaluation of
Liquefaction Resistance, Salt Lake City, Utah.

87
88
 Liquefaction Resistance Based on Shear Wave Velocity

By Ronald D. Andrusl and Kenneth H. Stokoe, IJ2

1Research Civil Engineer

National Institute of Standards and Technology

2Professor of Civil Engineering

The University of Texas at Austin

Abstract

This report reviews the current simplified procedures for evaluating the liquefaction resistance
of granular soil deposits using small-strain shear wave velocity. These procedures were
developed from analytical studies, laboratory studies, or very limited field performance data.
Their accuracy is evaluated through field performance data from 20 earthquakes and in situ
shear wave velocity measurements at over 50 different sites (124 test arrays) in soils ranging
from sandy gravel with cobbles to profiles including silty clay layers, resulting in a total of 193
liquefaction and non-liquefaction case histories. The current procedures correctly predict high
liquefaction potential at many sites where surface manifestations of liquefaction were observed.
Revisions and enhancements to the current procedures are proposed using the compiled case
history data. The recommended procedure follows the general format of the SPT- and CPT-
based procedures. Liquefaction potential boundaries are established by applying a modified
relationship between shear wave velocity and cyclic stress ratio for constant average cyclic shear
strain suggested by Dobry. These new boundaries, which are simply defined mathematically
and easy to implement, correctly predict moderate to high liquefaction potential for more than
95% of the liquefaction case histories. Additional case histories are needed of all types of soils
that have and have not liquefied during earthquakes, particularly from deeper deposits (depth >
8 m) and from denser soils (Vs > 200 mis) shaken by stronger ground motions (amax > 0.4 g), to
further validate the proposed procedures.

This report is a U.S. Government work and, as such, is in the public domain of the United States
of America.

89
 Introduction

During the past decade, several simplified procedures using small-strain shear wave velocity,
Vs, have been proposed for assessing the liquefaction resistance of granular soils (Stokoe et al.
1988b; Tokimatsu et al. 1991a; Robertson et al. 1992; Kayen et al. 1992; Andrus 1994; Lodge
1994). The use of Vs as an index of liquefaction resistance is justified since both Vs and
liquefaction resistance are influenced by many of the same factors (e.g. void ratio, effective
confining pressure, stress history, and geologic age).

The in situ Vs can be measured by a number of techniques such as the crosshole seismic test,
the Seismic Cone Penetration Test (SCPT), or the Spectral-Analysis-of-Surface-Wave (SAS\V)
test. The accuracy of these techniques can be sensitive to procedural details, soil conditions, and
interpretation methods. Some advantages of using Vs are:

• Measurements are possible in soils that are hard to sample, such as gravelly soils, and at
sites where borings or soundings may not be permitted, such as capped landfills;

• Measurements can be performed in small laboratory specimens, allowing direct

comparisons between measured laboratory and field behavior;

• Vs is directly related to small-strain shear modulus, Gmax, a parameter required in

analytical procedures for estimating dynamic shearing strain in soils; and

• For large earthquake magnitudes and long durations of shaking, the cyclic shear strain
needed for liquefaction decreases and approaches the threshold strain in sand(= 0.02%),
thus making it possible to conduct analytical evaluations ofliquefaction using Vs and Gmax
as basic parameters (Dobry et al. 1981; Seed et al. 1983).

Two limitations of using Vs to evaluate liquefaction resistance are: (1) Field seismic
measurements are made with small strains, whereas liquefaction is a large-strain phenomenon
(Roy et al. 1996). This limitation can be significant for cemented soils, since Vs is highly
sensitive to weak interparticle bonding which is eliminated at large strains. (2) Seismic testing
does not provide samples for classification of soils and identification of non-liquefiable soft
clay-rich soils. Non-liquefiable soils by the so-called Chinese criteria have clay contents
(particles smaller than 5 µm) greater than 15%, liquid limits greater than 35%, or moisture
contents less than 90% of the liquid limit (Seed and Idriss 1982). To compensate for these
limitations, a limited number of borings should be drilled and samples taken to identify weakly
cemented soils that might be liquefiable but classed as non-liquefiable by Vs criteria and also to
identify non-liquefiable clay-rich soils that otherwise might be classed as liquefiable.

90
The purpose of this report is to recommend guidelines for evaluating liquefaction resistance
using in situ measurements ofV S· To accomplish this purpose, current procedures are reviewed
and their accuracy is evaluated using Vs measurements at over 50 different sites (124 test
arrays) and field performance data from 20 earthquakes, resulting in a total of 193 liquefaction
and non-liquefaction case histories.

Earthquake and site characteristics used in the evaluations are summarized in Table I. In
Column 2 of Table 1, test array refers to the two boreholes used for crosshole measurements, the
borehole (or cone sounding) and source used for downhole measurements, or the line of
receivers used for SASW measurements. The occurrence of liquefaction is based on the
appearance of sand boils, ground cracks and fissures, or ground settlement. The shear wave
velocities used in the subsequent evaluations are either the average or minimum of values
reported by the investigator(s) for the most vulnerable layer at the test array. Shown in Fig. 1
are the relationships between shear wave velocity and depth. Some of the velocities are from
measurements made before the earthquake, and others are from measurements made following
the earthquake. The values of total vertical stress, crv, and effective vertical stress, cr'v, listed in
Columns 8 and 9 of Table 1 are averages for the depth range of the measurements, estimated
using total unit weights reported by the investigator(s). When no values are reported, total unit
weights of 17.3 kN/m 3 for soils above the water table and 18.9 kN/m 3 for soils below the water
table are assumed. The materials comprising the most vulnerable layer at all sites are Holocene
to latest Pleistocene age (< 15,000 years). The peak horizontal ground surface accelerations,
amax, used in subsequent evaluations are either the peak value for the larger of the x and y
ground motion records or the average of peak values for the x and y ground motion records that
would have occurred at the site in the absence of liquefaction. Values of a max are determined by
averaging estimates reported by the investigator(s) and estimates made as part of this study
using attenuation relationships developed from published ground surface acceleration data.

The proposed liquefaction assessment procedures can be divided into three general categories:
(I) procedures developed from analytical studies; (2) procedures developed from laboratory
studies; and (3) procedures developed from field performance studies.

Procedures Developed From Analytical Studies

Stokoe et al. (1988b) applied the cyclic strain approach developed by Dobry and his colleagues
(1982) in a parametric study of the liquefaction potential of sandy soils in the Imperial Valley,
California. In the cyclic strain approach, the peak cyclic shearing strain at which the cyclic pore
water pressure equals the confining pressure is used as the criterion for liquefaction occurrence.

Two generalized soil profiles were used in the parametric study. The first generalized soil
profile contained a shallow (~ 12 m) liquefiable sand layer. The three parameters of the sand
layer which were varied are: soil stiffness in terms ofV s (or small-strain shear modulus), depth,
and thickness. Depicted in Fig. 2a are three variations of the first generalized soil profile. The
second generalized soil profile is presented in Fig. 2b, and was simply a 61-m thick clay deposit
representative of a soil site in the Imperial Valley upon which strong-motion accelerographs

91
 Table 1 - Vs-based Liquefaction and Non-liquefaction Case Histories

Lique- Water Top of Layer Aver- Aver- Aver-

Measure- faction table layer thick- age age age Cyclic
Test rnent observed depth depth ness <Jv cr'v Soil Vs Vs1 amax stress
Site array tyre ? (rn) (rn) (m) (kPa) (kPa) type (mis) (mis) (g) ratio Reference
(I) (2) ( ) (4) (5) (6) (7) (8) (9) (10) (II) (12) (13) (14) (15)
(a) 1906 San Francisco, California Earthquake (Mw = 7.7)
Coyote Creek SRI crosshole yes 2.4 3.5 2.5 83.6 62.1 sand& 136 153 0.36 0.30 Youd and
RIR2 crosshole yes 2.4 3.5 2.5 75.4 58.2 gravel 154 177 0.36 0.29 Hoose
RIR3 crosshole yes 2.4 3.5 2.5 75.4 58.2 161 185 0.36 0.29 (1978);
R2R3 crosshole yes 2.4 3.5 2.5 75.4 58.2 173 198 0.36 0.29 Barrow
(1983);
Salinas River, SRI crosshole no 6.0 9.1 1.5 178.2 140.8 sandy silt 177 162 0.32 0.24 Bennett and
north RIR2 crosshole no 6.0 9.1 1.5 178.2 140.8 195 179 0.32 0.24 Tinsley
RIR3 crosshole no 6.0 9.1 1.5 178.2 140.8 200 184 0.32 0.24 (1995)
R2R3 crosshole no 6.0 9.1 1.5 178.2 140.8 199 183 0.32 0.24

Salinas River, SRI crosshole yes 6.0 6.5 4.5 142.2 123.5 sand & 131 124 0.32 0.22
south RlR2 crosshole yes 6.0 6.5 4.5 142.2 123.5 silWsand 149 141 0.32 0.22
RIR3 crosshole yes 6.0 6.5 4.5 142.2 123.5 158 150 0.32 0.22
R2R3 crosshole ves 6.0 6.5 4.5 142.2 123.5 168 159 0.32 0.22
(b) 1964 Nugata, Japan Earthquake (Mw = 7.5)
Niigata City Al SASW no 5.0 5.0 2.5 I 10.9 97.7 sand 163 164 0.16 0.1 I Tokimatsu et
Cl SASW yes 1.2 1.6 6.5 90.0 54.7 sand 115 136 0.16 0.16 al. (1991a)
C2 SASW es 1.2 1.2 4.8 67.8 44.5 sand 118 148 0.16 0.14
(c) 1975 Haicheng, PRC Earthquake (Mw = 7.1)
Paper Mill downhole yes 1.0 3.0 2.0 54.7 35.3 clayey 122 158 0.12 0.12 Arulanandan
silt et al. (1986)

Glass Fiber downhole yes 0.8 3.0 3.5 90.0 50.l sandy silt 98 117 0.12 0.14
to clayey
silt

Construction downhole yes 1.5 5.0 4,5 124,9 73.7 clayey 103 111 0.12 0.13
Building silt

Fishery & downhole yes 0.5 2.5 4.0 81.7 43.6 silty sand IOI 124 0.12 0.14
Shipbuilding to clayey
silt
Middle School downhole no 1.0 9.0 2.5 191.8 101.2 clayey 143 142 0.12 0.13
silt

Chemical downhole marginal 1.5 6,0 5,5 159.4 90.1 sand to 147 152 0.12 0.13
Fiber clayey
silt
(d) 1979 Impenal Valley, California Earthquake (Mw = 6.5)
Wildlife 1 crosshole no 1.5 2.5 4.3 83.8 53.9 silty sand 127 148 0.13 0.13 Bennett et al.
2 crosshole no 1.5 2.5 4,3 83,8 53.9 to sandy 124 145 0.13 0.13 (1981.
SASW no 1.5 2.5 4.3 91.8 57.8 silt 115 132 0.13 0.13 1984);
Sykora and
Radio Tower SASW yes 2.0 2.7 3.4 79.2 55.8 silty sand 90 104 0.21 0.18 Stokoe
to sandy (1982);
silt Youd and
Bennett
McKim SASW yes 1.4 1.4 3.5 54.3 38.1 silty sand 126 161 0.51 0.45 (1983);
Bierschwale
Vail Canal SASW no 2.7 2.7 2.8 70.4 58.4 sand to 101 116 0.12 0.10 and Stokoe
silty sand (1984);
Stokoe and
Kombloom SASW no 2.5 2.5 3.5 74.7 57.8 sandy silt 105 120 0.12 0.09 Nazarian
(1984);
Heber Road, SRI crosshole yes 2.0 2.0 3.3 63.0 48.0 silty sand 131 158 0.50 0.41 Dobry et al.
channel fill RlR2 crosshole yes 2.0 2.0 3,3 63.0 48.0 133 160 0.50 0.41 (1992)

Heber Road, SRI crosshole no 2.0 2.0 2.3 60.1 46.6 sand 164 200 0.50 0.40
ooint bar RIR2 crosshole no 2.0 2.0 2.3 60.1 46.6 173 210 0.50 0.40

92
 Table 1 (cont.) - Vs-based Liquefaction and Non-liquefaction Case Histories

(I) (2) (3) (4) I (5) I (6) I c1i I csi 1 (9) 1 (10) 1 01) 1 02) 1 (13) 1 (14) (15)
(e) 1980 Mid-Chiba, Japan Earthquake (Mw = 5.9)
Owi Island C2, upper downhole no 1.35 4.5 3.3 105.4 59.2 silty sand 155 178 0.08 0.09 Ishihara et
No. I C2, lower downhole no 1.35 13.0 3.6 251.6 120.2 195 186 0.08 0.08 al. (1981;
1987)
(f) 1981 Westmorland California Earthquake (Mw = 5 9)
Wildlife 1 crosshole yes 1.5 2.5 4.3 83.8 53.9 silty sand 127 148 0.27 0.26 Bennett et al.
2 crosshole yes 1.5 2.5 4.3 83.8 53.9 to sandy 124 145 0.27 0.26 (1981,
SASW yes 1.5 2.5 4.3 91.8 57.8 silt I 15 132 0.27 0.27 1984);
Sykora and
Radio Tower SASW yes 2.0 2.7 3.4 79.2 55.8 silty sand 90 104 0.20 0.18 Stakoe
to sandy (1982);
silt Youd and
Bennett
McKim SASW no 1.4 1.4 3.5 54.3 38.1 silty sand 126 161 0.06 0.05 (1983);
Bierschwale
Vail Canal SASW yes 2.7 2.7 2.8 70.4 58.4 sand to IOI 116 0.30 0.23 and Stakoe
silty sand (1984);
Stakoe and
Kombloom SASW yes 2.5 2.5 3.5 74.7 57.8 sandy silt 105 120 0.36 0.29 Nazarian
(1984);
Heber Road, SRI crosshole no 2.0 2.0 3.3 63.0 48.0 silty sand 131 158 0.02 O.Q2 Dobry ct al.
channel fill RIR2 crosshole no 2.0 2.0 3.3 63.0 48.0 133 160 0.02 0.02 (1992)

Heber Road, SRI crosshole no 2.0 2.0 2.3 60.1 46.6 sand 164 200 0.02 0.02
uoint bar RIR2 crosshole no 2.0 2.0 2.3 60.1 46.6 173 210 0.02 0.02
(g) 1983 Borah Peak Idaho Earthquake (Mw = 6 9)
Pence Ranch SAi SASW yes 1.7 1.8 1.9 57.2 46.2 gravelly 107 131 0.36 0.28 Andrus and
SA2 SASW yes 1.5 1.5 2.8 52.7 40.5 sand to 94 118 0.36 0.29 Youd
SA3 SASW yes 1.4 1.4 1.8 44.5 36.0 sandy 102 132 0.36 0.28 (1987);
SA4 SASW yes 1.8 1.8 2.8 62.1 49.4 gravel 109 131 0.36 0.28 Stakoe et al.
SAS SASW yes 1.5 1.5 1.9 60.5 45.6 122 151 0.36 0.29 (1988a);
SAA SASW yes 2.0 2.0 1.7 57.5 46.3 134 164 0.36 0.28 Andrus et al.
SAB SASW yes 1.5 1.5 1.7 38.8 32.9 128 170 0.36 0.26 (1992);
SAC SASW yes 1.5 1.5 1.9 38.4 32.4 107 142 0.36 0.27 Andrus
SAD SASW yes 1.5 1.5 1.7 39.4 33.8 131 173 0.36 0.26 (1994)
SAE SASW yes 1.7 1.7 1.5 43.3 38.3 122 155 0.36 0.26
XDXE crossholc yes 1.5 1.5 2.3 48.5 38.1 154 198 0.36 0.29

Goddard SA2 SASW yes 1.2 1.2 2.0 47.3 36.0 sandy 122 158 0.30 0.24
Ranch SA4 SASW yes 1.2 1.2 2.0 41.1 32.7 gravel 105 137 0.30 0.23

Andersen Bar XIX2 crosshole yes 0.8 0.8 2.4 40.6 28.7 sandy 106 146 0.29 0.26
SAi SASW yes 0.8 0.8 2.4 39.0 27.8 gravel 105 145 0.29 0.25

Larter Ranch X3X4 crosshole yes 0.8 2.2 1.3 59.9 39.0 silty 176 223 0.50 0.49
SAl,85 SASW yes 0.8 2.2 1.3 55.4 38.4 sandy 153 194 0.50 0.46
SAl,90 SASW yes 0.8 2.2 1.3 59.9 40.5 gravel 183 230 0.50 0.47

Whiskey WSla crosshole yes 0.8 1.8 2.2 59.1 38.2 sandy 181 230 0.50 0.49
Springs SAS SASW yes 0.8 1.8 2.2 45.6 31.7 silty 210 271 0.50 0.46
gravel

North Gravel SAi SASW no I.0 1.8 1.2 51.0 36.0 sandy 206 266 0.46 0.41
Bar SA2 SASW no 3.0 3.0 1.3 75.2 53.5 gravel 274 322 0.46 0.42

Mackay Dam, SA2 SASW no 2.3 2.3 2.7 57.4 silty

66.6 271 313 0.23 0.17
downstream sandy
toe ~ravel
(h) 1985 Chiba-lbarag1-Kenkyo, Japan Earthquake (Mw = 6.0)
Owi Island C2, upper downhole no 1.35 4.5 3.3 105.4 59.2 silty sand 155 178 0.06 0.07 Ishihara et
No. I C2. lower downhole no 1.35 13.0 3.6 251.6 120.2 195 186 0.06 0.06 al. (1987)
(i) 1/16/86 Taiwan Earthquake (Mw ..
=6 6- Event LSST4)
Shen et al.
LotungLSST L8L3 crosshole no 0.5 2.0 5.0 85.4 35.4 silty sand 146 190 0.22 0.33
Facility L8L4 crosshole no 0.5 2.0 5.0 85.4 35.4 to sandy 133 173 0.22 0.33 (1991); EPRI
L2L5L6 crosshole no 0.5 2.0 5.0 85.4 35.4 silt 127 166 0.22 0.33 (19()2)
L2L7 crosshole no 0.5 2.0 5.0 85.4 35.4 130 171 0.22 0.33
(j) 5/20/86 Taiwan Earthquake (Mw =6.6; Event LSST7)
Lotung LSST L8L3 cross hole no 0.5 2.0 5.0 85.4 35.4 silty sand 146 190 0.18 0.27 Shen et al.
Facility L8L4 cross hole no 0.5 2.0 5.0 85.4 35.4 to sandy 133 173 0.18 0.27 (1991); EPRI
L2L5L6 crossho)e no 0.5 2.0 5.0 85.4 35.4 silt 127 166 0.18 0.27 (1992)
L2L7 crosshole no 0.5 2.0 5.0 85.4 35.4 130 171 0.18 0.27

93
 Table 1 (cont.) - Vs-based Liquefaction and Non-liquefaction Case Histories

(1) (2) (3) (4) I (5) I(6) I

(7) (8) I (9) I
(10) I
(11) I I (12) I (13) I (14) I (15)
(k) 5/20/86 Taiwan Earthquake (Mw = 6.2; Event LSST8)
Lotung LSST L8L3 crosshole no 0.5 2.0 5.0 85.4 35.4 silty sand 146 190 0.04 0.06 Shen et al.
Facility L8L4 crosshole no 0.5 2.0 5.0 85.4 35.4 to sandy 133 173 0.04 0.06 (1991); EPRI
L2L5L6 crosshole no 0.5 2.0 5.0 85.4 35.4 silt 127 166 0.04 0.06 (1992)
L2L7 crosshole no 0.5 2.0 5.0 85.4 35.4 130 171 0,04 0.06
(1) 7/30/86 Taiwan Earthquake (Mw = 6.2; Event LSST12)
LotungLSST L8L3 crosshole no 0.5 2.0 5.0 85.4 35.4 silty sand 146 190 0.18 0.27 Shen et al.
Facility L8L4 crosshole no 0.5 2.0 5.0 85.4 35.4 to sandy 133 173 0.18 0.27 (1991); EPRI
L2L5L6 crosshole no 0.5 2.0 5.0 85.4 35.4 silt 127 166 0.18 0.27 (1992)
L2L7 crosshole no 0.5 2.0 5.0 85,4 35.4 130 171 0.18 0.27
(m) 7/30/86 Taiwan Earthquake (Mw = 6.2; Event LSST13)
Lotung LSST L8L3 crosshole no 0.5 2.0 5.0 85.4 35.4 silty sand 146 190 0.05 0.08 Shen et al.
Facility L8L4 crosshole no 0.5 2.0 5.0 85.4 35.4 to sandy 133 173 0.05 0.08 (1991); EPRI
L2L5L6 crosshole no 0.5 2.0 5.0 85.4 35.4 silt 127 166 0.05 0.08 (1992)
L2L7 crosshole no 0.5 2.0 5.0 85.4 35.4 130 171 0.05 0.08
(n) I 1/4/86 Taiwan Earthquake (Mw = 6.2·' Event LSST16)
LotungLSST L8L3 crosshole no 0.5 2.0 5.0 85.4 35.4 silty sand 146 190 0.16 0.24 Shen et al.
Facility L8L4 crosshole no 0.5 2.0 5.0 85.4 35.4 to sandy 133 173 0.16 0.24 (1991); EPRI
L2LSL6 crosshole no 0.5 2.0 5.0 85.4 35.4 silt 127 166 0.16 0.24 (1992)
L2L7 crosshole no 0.5 2.0 5.0 85.4 35.4 130 171 0.16 0.24
(o) 1987 Chiba-Toho-Oki, Japan Earthquake (Mw = 6.5)
Sunamachi downhole no 6.2 6.2 5.8 168.2 140.2 sand with 150 138 0.10 0.07 Ishihara et
silt to al. (1989)
silt sand
(p) 1987 Elmore Ranch, California Earthquake (Mw= 5.9)
Wildlife 1 cross hole no 1.5 2.5 4.3 83.8 53.9 silty sand 127 148 0.12 0.12 Bennett et al.
2 crosshole no 1.5 2.5 4.3 83.8 53.9 to sandy 124 145 0.12 0.12 (1981,
SASW no 1.5 2.5 4.3 91.8 57.8 silt 115 132 0.12 0.12 1984);
Sykora and
Radio Tower SASW no 2.0 2.7 3.4 79.2 55.8 silty sand 90 104 0.11 0.10 Stokoe
to sandy (1982);
silt Youd and
Bennett
McKim SASW no 1.4 1.4 3.5 54.3 38.l silty sand 126 161 0.06 0.05 (1983);
Bierschwale
Vail Canal SASW no 2.7 2.7 2.8 70.4 58.4 sand to 101 116 0.13 0.10 and Stokoe
silty sand (1984);
Stokoe and
Kombloom SASW no 2.5 2.5 3.5 74.7 57.8 sandy silt 105 120 0.24 0.19 Nazarian
(1984);
Heber Road, SRI crosshole no 2.0 2.0 3.3 63.0 48.0 silty sand 131 158 0.03 0.02 Dobry et al.
channel fill RIR2 crosshole no 2.0 2.0 3.3 63.0 48.0 133 160 0.03 0.02 (1992)

Heber Road, SRI crosshole no 2.0 2.0 2.3 60.1 46.6 sand 164 200 0.03 0.02
ooint bar RIR2 crosshole no 2.0 2.0 2.3 60.J 46.6 173 210 0.03 0.02
(q) 1987 Superstition Hills, California Earthquake (Mw = 6.5)
Wildlife I crosshole yes 1.5 2.5 4.3 83.8 53.9 silty sand 127 148 0.20 0.19 Bennett et al.
2 crosshole yes 1.5 2.5 4.3 83.8 53.9 to sandy 124 145 0.20 0.19 (1981,
SASW yes 1.5 2.5 4.3 91.8 57.8 silt 115 132 0.20 0.20 1984);
Sykora and
Radio Tower SASW no 2.0 2.7 3.4 79.2 55.8 silty sand 90 104 0.20 0.18 Stokoe
to sandy (1982);
silt Youd and
Bennett
McKim SASW no 1.4 1.4 3.5 54.3 38.1 silty sand 126 161 0.19 0.17 (1983);
Bierschwale
Vail Canal SASW no 2.7 2.7 2.8 70.4 58.4 sand to 101 116 0.20 0.15 and Stokoe
silty sand (1984);
Stokoe and
Kornbloom SASW no 2.5 2.5 3.5 74.7 57.8 sandy silt 105 120 0.21 0.17 Nazarian
(1984);
Heber Road, SR! crossho!e no 2.0 2.0 3.3 63.0 48.0 silty sand 131 158 0.18 0.15 Dobry et al.
channel fill RIR2 crosshole no 2.0 2.0 3.3 63.0 48.0 133 160 0.18 0.15 (1992)

Heber Road, SRI crosshole no 2.0 2.0 2.3 60.l 46.6 sand 164 200 0.18 0.15
ooint bar RIR2 crosshole no 2.0 2.0 2.3 60.1 46.6 173 210 0.18 0.15

94
 Table 1 (cont.) - Vs-based Liquefaction and Non-liquefaction Case Histories

(1) (2) (3) (4) 1 cs) 1 (6) 1 c1) 1 cs) 1 (9) 1 c10) 1 01) 1 02) 1 (13) 1 c14i I (15)
(r) 1989 Loma Prieta, California Earthquake (Mw = 7 OJ
Treasure X!X2 crosshole marginal 1.4 4.2 7.3 106.6 63.5 silty sand 130 145 0.14 0.15 Furhriman
Island, fire B2B3 crosshole marginal 1.4 4.5 7.7 148.7 83.7 to clayey 157 164 0.14 0.15 (1993);
station B!B4 crosshole marginal 1.4 4.5 7.7 147.2 83.0 silty sand 157 165 0.14 0.15 Andrus
B4B5 crosshole marginal 1.4 4.5 7.7 101.3 60.9 131 148 0.14 0.15 (1994);
B2B4 crosshole marginal 1.4 4.5 7.7 I 18.5 69.2 136 150 0.14 0.15 Redpath
SASW marginal 1.4 4.5 7.7 139.9 78.6 148 145 0.14 0.15 (1991);
downhole marginal 1.4 4.5 7.7 163.0 90.6 137 142 0.14 0.15 Gibbs et al.
downhole marginal 1.4 4.5 7.7 154.4 86.4 152 158 0.14 0.15 (1992);
SCPT marginal 1.4 4.5 7.7 146.3 82.5 146 154 0.14 0.15 Hryciw et al.
(1991);
Rollins et al.
(1994)

Treasure UM03 SCPT no 1.5 4.4 5.6 133.J 77.5 sand to 178 190 0.14 0.15 Hryciw
Island, UM05 SCPT yes 2.4 3.5 4.5 102.6 71.0 silty sand 163 178 0.15 0.14 (1991);
perimeter UM06 SCPT yes 1.4 2.0 4.0 75.4 48.8 154 185 0.14 0.14 Hryciw et al.
UM09 SCPT yes 2.7 2.7 3.7 82.1 63.9 143 160 0.15 0.12 (1991);
UM!! SCPT yes 1.4 4.0 3.0 101.2 61.2 160 181 0.14 0.15 Geomatrix
(1990)

Port of SRI crosshole yes 3.5 4.0 4.0 110.1 84.7 silt to 143 149 0.16 0.13 Stokoe et al.
Richmond RIR2 crosshole yes 3.5 4.0 4.0 !10.1 84.7 silty sand 135 140 0.16 0.13 (1992);
SASW yes 3.5 4.0 4.0 97.0 78.8 I 17 124 0.16 0.12 Mitchell et
POR2 SCPT yes 3.5 4.0 4.0 98.9 79.4 152 161 0.16 0.12 al. (1994)
POR3 SCPT yes 3.5 5.0 2.0 98.9 79.4 121 128 0.16 0.12
POR4 SCPT yes 3.5 5.0 2.0 98.9 79.4 138 147 0.16 0.12

Port of SRI crosshole no 3.5 3.5 5.0 104.4 82.0 silty to 148 155 0.16 0.12
Richmond, RIR2 crosshole no 3.5 3.5 5,0 104.4 82.0 silty sand 145 152 0.16 0.12
Hall Ave. SASW no 3.5 3.5 5.0 109.2 84.3 133 139 0.16 0.12

Bay Bridge SRI crosshole yes 3.0 5.5 1.5 115.9 82.4 sand to 134 141 0.24 0.21
Toll Plaza R!R2 crosshole yes 3.0 5.5 1.5 !15.9 82.4 silty sand 134 141 0.24 0.21
SFOBBl SCPT yes 3.0 5.5 1.5 108.3 78.8 146 155 0.24 0.21
SFOBB2 SCPT yes 3.0 6.0 3.0 136.6 92.4 148 151 0.24 0.22

Port of SRI crosshole yes 3.0 5.5 2.5 121.6 85.8 sand 145 151 0.24 0.21
Oakland RIR2 crosshole yes 3.0 5.5 2.5 121.6 85.8 179 186 0.24 0.21
SASW yes 3.0 5.5 2.5 115.8 83.1 157 165 0.24 0.21
P007I SCPT yes 3.0 5.5 2.5 122.5 86.2 142 148 0.24 0.21
P0072 SCPT yes 3.0 5.5 2.5 122.5 86.2 145 150 0.24 0.21
P0073 SCPT yes 3.0 5.5 1.5 113.1 81.7 176 185 0.24 0.21

Bay Farm SRI crosshole no 3.6 3.6 2.8 87.1 75.2 sand 193 207 0.27 0.20
Island, dike RIR2 crosshole no 3.6 3.6 2.8 87.1 75.2 212 227 0.27 0.20
SASW no 3.6 3.6 2.8 91.9 77.0 204 219 0.27 0.20

Bay Farm SRI crosshole yes 3.0 3.0 1.7 69.9 60.9 sand 97 109 0.27 0.20
Island, So. RJR2 crosshole yes 3.0 3.0 1.7 69.9 60.9 116 13! 0.27 0.20
Loop Road SASW yes 3.0 3.0 1.7 67.0 59.6 125 143 0.27 0.19

Marina school downhole yes 2.7 2.7 1.6 61.9 54.4 sand to 153 177 0.15 0.11 Kayen et al.
District 2 SASW yes 2.9 2.9 7.1 117.0 82.2 silty sand 120 129 0.15 0.12 (1990);
3 SASW yes 2.9 2.9 7.1 !17.0 82.2 105 113 0.15 0.12 Tokimatsu et
4 SASW yes 2.9 2.9 2.1 69.9 59.6 120 137 0.15 0.11 al. (1991b)
5 SASW no 5.9 5.9 4.1 140.6 105.7 220 217 0.15 0.12

Coyote Creek SRI crossho1e no 2.4 3.5 2.5 83.6 62.1 sand& 136 153 0.19 0.16 Barrow
RIR2 crosshole no 2.4 3.5 2.5 75.4 58.2 gravel 154 177 0.19 0.16 (1983);
RIR3 crosshole no 2.4 3.5 2.5 75.4 58.2 161 185 0.19 0.16 Bennett
R2R3 crosshole no 2.4 3.5 2.5 75.4 58.2 173 198 0.19 0.16 (1995);
Bennett and
Salinas River, SRI crosshole no 6.0 9.1 1.5 178.2 140.8 silty sand 177 162 0.15 0.11 Tinsley
north R!R2 crosshole no 6.0 9.1 1.5 178.2 140.8 195 179 0.15 0.11 (1995)
R!R3 crosshole no 6.0 9.1 1.5 178.2 140.8 200 184 0.15 0.11
R2R3 crosshole no 6.0 9.1 1.5 178.2 140.8 199 183 0.15 0.11

Salinas River, SRI crosshole no 6.0 6.5 4.5 142.2 123.5 sand & 131 124 0.15 0.11
south R1R2 crosshole no 6.0 6.5 4.5 142.2 123.5 silty 149 141 0.15 0.11
RIR3 crosshole no 6.0 6.5 4.5 142.2 123.5 sand 158 150 0.15 0.11
R2R3 crosshole no 6.0 6.5 4.5 ]42.2 123.5 168 159 0.15 0.11

95
 Table 1 (cont.) - Vs-based Liquefaction and Non-liquefaction Case Histories

(I) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15)
Santa Cruz SC02 SCPT yes 0.6 1.3 2.6 48.1 28.7 sand to 116 160 0.42 0.44 Hryciw
SC03 SCPT yes 2.1 2.1 2.3 60.1 48.1 sandy silt 145 174 0.42 0.33 (1991)
SC04 SCPT no 1.8 1.8 2.2 51.0 41.0 126 158 0.42 0.33
SC05 SCPT no 2.8 3.0 1.6 67.7 57.8 135 155 0.42 0.31
SC!3 SCPT no 1.8 2.0 4.0 69.2 49.8 158 188 0.42 0.36
SC14 SCPT yes 1.2 1.4 1.6 41.0 30.5 126 170 0.42 0.37

Moss Landing, UC-15 SCPT yes 1.8 1.8 2.8 63.6 46.9 Sand 116 140 0.25 0.21 Boulanger et
State Beach UC-16 SCPT yes 2.3 2.3 7.1 l01.3 69.8 162 178 0.25 0.22 al. (1995);
Boulanger et
Moss Landing, UC-4 SCPT yes 1.8 2.1 1.5 54.2 42.4 Sand 130 161 0.25 0.20 al. (1997)
Sandholt Rd. UC-4 SCPT no 1.8 5.9 4.1 148.5 87.7 209 216 0.25 0.26
UC-6 SCPT marginal 1.7 3.0 4.3 85.6 59.5 171 196 0.25 0.22
Moss Landing, UC-12 SCPT yes 1.9 3.0 1.6 74.8 53.1 Silty sand 150 175 0.25 0.22
Harbor
Office

Moss Landing, UC-9 SCPT yes 1.2 2.6 1.4 60.3 39.6 Sand 143 180 0.25 0.24
Woodward
Marine
(s) 1993 Hokkaido-nansei-oki, Japan Earthquake (Mw = 8.3)
Pension House BHI downhole yes 1.0 1.0 2.5 45.5 33.4 sandy 79 105 0.19 0.16 Kokusho et
BH2 downhole marginal 0.7 3.7 4.8 I 22.9 70.0 gravel 144 159 0.19 0.21 al. (1995a.
1995b)
(t) 1995 Hyogo-ken Nanbu, Japan Earthquake (Mw = 6.9)
Port Island, 1991 downhole yes 2.4 2.4 12.6 160.8 98.8 sandy 197 202 0.50 0.43 Sato et al.
instrumented 1995 downhole yes 2.4 2.4 12.6 185.9 110.9 gravel 174 172 0.50 0.44 (I 996);
array with silt Shibata et
al. (1996);
SGK(TRC) downhole no 7.0 7.0 4.0 158.5 139.1 sand, silt 149 138 0.48 0.32 Sugito et al.
(1996)
TI(S (TPS) downhole yes 2.5 2.5 4.6 73.8 57.9 gravel, 135 157 0.20 0.15
sand, silt

KNK(KPS) downhole no 2.0 3.8 13.2 193.6 111.0 sand, silt 179 184 0.12 0.10
Test array refers to the two boreholes used for crosshole measurements, the borehole (or cone sounding) and source used for downhole
measurements. or the line of receivers used for SASW measurements.
Vs is shear wave velocity and Vs1 is shear wave velocity modified to an overburden pressure of 100 kPa using Vs1 = Vs (100 kPa / cr'v)0.25
(Robenson et al. 1992). Averages for the Treasure Island and Santa Cruz SCPT data are of the unfiltered data. One high
velocity measurement is omitted from the average for Santa Cruz test array SC04. Refracted wave velocities measured at 5.5 m
are omitted from the averages for Coyote Creek (test arrays RIR2, RIR3 and R2R3).
Average •max is the average of two peak ground surface accelerations obtained from the x and y ground motion records that would have
occurred at the site in the absence of liquefaction.
Mw is moment magnitude.
At Owi Island No. 1, Lotung LSST Faculity, Sunamachi, Wildlife (1987 earthquakes), and Port Island sites the assessment of liquefaction
or no liquefaction is supported by pore water pressure measurements.
At Larter Ranch and Whiskey Springs, soil may be weakly cemented by carbonate.
At Lotung LSST Facility, the artesian pressure is assumed to vary linearly from a pressure head of 8.1 m at a depth of 7 m to a pressure head
of 1.9 mat a depth of 2 m.
At Treasure Island Fire Station, Moss Landing Sandholt Road UC-6, and Pension House BH2 no sand boils or damaged observed,
although some liquefaction observed in adjacent areas. Thus, liquefaction behavior is listed as marginal for these sites.
At Moss Landing Sandholt Road UC-4 no lateral displacement occurred below 5.9 m based on slope inclinometer data.

96
 Shear Wave Velocity, Vs, mis Shear Wave Velocity, Vs, mis
0 100 200 300 400 0 100 200 300 400
0 ' ' ' 0 . ' '

- 0.
0 0~
- -
cc I> ~
-
~ ~I}~ +bl>

\i ~;o
n 1Da,, ~

5 5 ~
'
,1-T '!>
~ cit > I>
-
...
>;!~
cl>
o'
t>
>
- - ~cm\
oO O , <I
-
E E
c,/o
£10
C.
£Cl. 10 -f
u
Ql Q) 0 00 C
CJ CJ rPoc
.... Sim - -
o Treasure lsland'Jerimeter ~
+ Port of Richmon o Treasure Island, fire station
□ Bay Bridge Toll Plaza + Treasure Island, perimeter
15 >-- I> Port of Oakland 15 - □ Port of Richmond, Hall Ave.
<l Bay Farm Island, So. Loop Rd. I> Bay Farm Island, dike
• Marina District <l Marina District
- Note: An averaging filter has been
applied to the Treasure Island,
- - Note: An averaging filter has been
applied to the Treasure Island,
-
perimeter SCPT data perimeter SCPT data
I I I
20 ' ' ' ' 20 ' ' '
(a) Northern California Fill Sites Exhibiting (b) Northern California Fill Sites Not Exhibiting
Surface Manifestations of Liquefaction Surface Manifestations of Liquefaction

Shear Wave Velocity, Vs, mis Shear Wave Velocity, Vs, mis
0 100 200 300 400 0 100 200 300 400
0 ' ' 0 . ' '
Ii
- I),

~4' baJ!>
- - t.\ ~+,IJ· ++
-

<) &lo<:Jo
1- + -1:,.c•
5 5 -□ +
0 <l
D J
+ Al+
- <l - - D -
E <l
+ +++
E
CD 0
D
D
£C. 10 :510
Cl. 0
Q)
CJ
Q)
CJ
"'
- ~ - - . -
o Coyote Creek
+ Salinas River, south
c Santa Cruz
15 - I> Sandholdt Road 15 - ~
<l State Beach o Salinas River, north
• Harbor Office + Santa Cruz
z Woodward Marine _ o Sandholdt Road
- - -
Note: An averaging filter has been Note: An averaging filter has been
applied to the Santa Cruz SCPT data applied to the Santa Cruz SCPT data
20 ' ' ' ' 20 ' ' .
(c) Northern California Natural Soil Sites (d) Northern California Natural Soil Sites Not
Exhibiting Surface Manifestations of Liquefaction Exhibiting Surface Manifestations of Liquefaction

Fig. 1 - The Distribution of Shear Wave Velocity with Depth for the Most
Vulnerable Layer at the Sites Listed in Table 1.

97
 Shear Wave Velocity, Vs, mis Shear Wave Velocity, Vs, mis
0 100 200 300 400 o 100 200 300 400
0 ' ' ' 0 ' '

D
- - - C) -
+~ at<l 00

~
0 0
+ (I)
+
5 5
+ I:.
DCC

E
- -
E
- -

£10
C.
:S
C.
10
Q) Q)
0 0
- - - -

film
15 - oWildlife 15
+ Radio Tower
aMcKim
_ t> Vail Canal - -
<lKombloom - film
• Heber Road, channel fill Io Heber Road, point bar I
' I ' ' ' ' ' ' I '
20 20
(e) Imperial Valley, California Sandy Soil Sites (f) Imperial Valley, California Sandy Sites Not
Exhibiting Surface Manifestations of Liquefaction Exhibiting Surface Manifestations of Liquefaction

Shear Wave Velocity, Vs, mis Shear Wave Velocity, Vs, mis
0 100 200 300 400 0 100 200 300 400
0 ' ' ' ' 0 ' ' '

l ~~~~ - - C
-
-
0
C
ol~ t> <l
0
0
+
+
+ 0

5 5

- - - -
E
-5
C.
10
Ql
0
- - - -
film
o Pence Ranch
15 - + Goddard Ranch 15
a Andersen Bar
t> Larter Ranch •
<I Whiskey Springs•
- - - ~ -
•Soils may be weakly o North Gravel Bar
cemented with carbonate + Mackay Dam, downstream toe
I I I I I
20 ' ' ' ' 20 ' ' ' '
(g) Idaho Gravelly Soil Sites Exhibiting (h) Idaho Gravelly Soil Sites Not Exhibiting
Surface Manifestations of Liquefaction Surface Manifestations of Liquefaction

Fig. 1 (cont.) - The Distribution of Shear Wave Velocity with Depth for the Most
Vulnerable Layer at the Sites Listed in Table 1.

98
 Shear Wave Velocity, Vs, mis Shear Wave Velocity, Vs, mis
0 100 200 300 400 0 100 200 300 400
0 ' ' ' ' 0 ' ' ' '
:r
- ~ + 0
- - -
□□□ □
<JI 0
II

5
<J
,, 5
□ □□

• > <!,
. :r □□ "ij

- - :r
>
:r . I
-
E I>
-

+
£10
. I a.
Q)
0
+ :r
<J

~
- - -
o Niigata City, Jagan 11
+ Paper Mill, PR 11 ~
15 □ Glass Fiber, PRC 15 - +o Middle
Niigata City, Japan
School, PRC
I> Construction Building, PRC
<l Fishery & Shipbuilding, PRC □ Lotung LSST Facility, Taiwan
• Chemical Fiber, PRC I> SGK, Japan
~ :r Pension House, Japan
- ~ <J KNK, Japan -
11 Port Island, Japan • Owi Island No. 1, Japan
0TKS, Japan :r Sunamachi, Japan
20 ' 20 ' ' '

(i) Asia Soil Sites Exhibiting 0) Asia Soil Sites Not Exhibiting
Surface Manifestations of Liquefaction Surface Manifestations of Liquefaction

Fig. I (cont.) - The Distribution of Shear Wave Velocity with Depth for the Most
Vulnerable Layer at the Sites Listed in Table 1.

i
....,__...-1 -;-water
Clay

iJ~6{i
t&{} table

Clay Clay 61 m

(a) Sand Liquefaction Site (b) "Reference" Site

'/
l
Fig. 2 - Soil Model Used in the Parametric Study by Stokoe et al (1988b).

99
were placed. The variations in shear modulus and material damping ratio with shearing strain
assumed for the sand and clay layers were based on resonant column and cyclic triaxial tests on
specimens from the Imperial Valley (Ladd 1982; Turner and Stokoe 1982).

Most of the analyses were performed (Bierschwale and Stokoe 1984; Aouad 1986) with the
strong-motion acceleration time history which was recorded at the Salton Sea station during the
1981 Westmorland earthquake (moment magnitude, Mw = 5.9). This strong-motion record
exhibited a peak horizontal ground surface acceleration, amax, of 0.20 g and an equivalent
number of cycles, Ne, of about 10. Records of larger magnitude were fabricated by simply
multiplying the Salton Sea record by a pre-selected factor. Records with Ne of about 20 cycles
and 30 cycles were generated by doubling and tripling the strong-motion portion of the Salton
Sea record.

Stresses and strains within each soil profile were computed with program SHAKE (Schnabel et
al. 1972), an equivalent linear analysis. These calculations were repeated with either a larger or
smaller magnitude record until the estimated shearing strain within the liquefiable sand layer
equaled the cyclic strain required for initial liquefaction. Initial liquefaction was assumed to
occur at shearing strains of about 2%, 1% and 0.5% for 10 cycles, 20 cycles and 30 cycles of
loading, respectively, based on undrained, strain-controlled cyclic triaxial tests on two Imperial
Valley sands (Ladd 1982). The sand layer had been divided into 1.5-m thick sublayers, each
having the same stiffness. The computed strain within the bottom sublayer was always greater
than the computed strain in the other sublayers. Thus, criterion for initial liquefaction was first
satisfied in the bottom sublayer. Next, the scaled record that generated initial liquefaction was
applied at bedrock beneath the second profile, shown in Fig. 2b, to determine amax at the ground
surface of the non-liquefiable or "reference" soil site. These procedures were followed for each
set of parameters characterizing the liquefiable sand layer (Vs, depth, and thickness). A total of
46 velocity profiles was considered.

Since it seemed more likely engineers would estimate amax at the ground surface of non-
liquefiable soil sites than at liquefiable sites, Stokoe et al. (1988b) correlated Vs of the
liquefiable sand layer with amax estimated for a "reference" soil site at the candidate-site
location. The data from their parametric study are summarized in Figs. 3a, 3b and 3c for Ne of
10 cycles, 20 cycles and 30 cycles, respectively. As noted by Stokoe et al., the plotted data
exhibit the following general trends: (1) the higher the Vs, the less likely the site is to liquefy
for a given amax; (2) the greater the thickness of the liquefiable sand layer, the less likely the site
is to liquefy for a given Vs; and (3) the greater the depth to the bottom of the liquefiable sand
layer, the slightly more likely the site is to liquefy at a given VS· These findings suggest that
liquefaction potential is dependent on layer thickness and depth, and indicate that a separating
band (to allow for variations in thickness and depth) is more appropriate than a separating line to
distinguish between liquefaction and non-liquefaction.

Stokoe et al. (1988b) created liquefaction assessment charts by dividing Figs. 3a, 3b and 3c each
into three regions: the region left of the plotted data, the region of the plotted data, and the
region right of the plotted data. Liquefaction is predicted to not occur left of the plotted data

100
 200.----,----,---.,----,-----,---,

~
E No Liquefaction
UJ Liquefaction Likely
>_ 150 l--+---+~~~...,.,,:,..,,~~---1
~
u
0
No cii
Liquefaction >(I)

~.... 100 1---+--A-'-<1-'-..¥-1----+----1----1

Cll
(I)
..c: Sand
en 1---+---+---+--'"c = 20 cycles
No Drainage
Level Ground
50 '---..J....--'----'-'---....L.---"-----' 50'----'----'---'---....L.----L--'
0.0 0.2 0.4 0.6 0.0 0.2 0.4 0.6
amax at "Reference" Soil Site, g ~ax at "Reference" Soil Site, g

(a) N0 = 1o cycles (b) N0 = 20 cycles

200.----.----.---.----.----,----,
Chart Based on:
0 Sand
~ Liquefaction-+----+---Nc = 30 cycles
E No Drainage
ui Liquefaction Level Ground
>:_ Likely
150
~
·u
0

~ Liquefaction
~
Cll
S: 100 1----t-t.,-,.....ri- ~th m
I I Thick-
9 ill 46 ~
gi
..c:
en • • ... ♦ 3.0m
D 0 A 4.6 m
Ciil ~ 6.1 m
50'----'---"---'---
o.o 0.2 0.4 0.6
amax at "Reference" Soil Site, g

(c) N0 = 30 cycles

Fig. 3 - Comparison of Liquefaction Assessment Charts Proposed by Stokoe et al. (1988b)

with Relationship Between Vs ofLiquefiable Sand Layer and amax for 10, 20
and 30 Cycles of Shaking as Determined Analytically by
Bierschwale and Stokoe (1984) and Aouad (1986).

101
because the sand is too stiff to liquefy. Within the region of the plotted data, liquefaction would
likely occur, but depends on layer thickness and depth. Right of the plotted data, liquefaction is
predicted to occur.

To test the accuracy of these liquefaction assessment charts, field performance data for the
magnitude 5.9 to 6.6 earthquakes listed in Table 1 are plotted on the chart for Ne of 10 cycles
shown in Fig. 4a. The chart for Ne of 10 cycles is used since it was developed using a strong
motion record from the magnitude 5.9 Westmorland earthquake. The field performance data for
the magnitude 6.9 to 7.1 earthquakes are plotted on the chart for Ne of 15 cycles shown in Fig.
4b. For each case history, the shear wave velocity shown is the minimum measurement made
within the most vulnerable layer. The value of amax is for the larger of the x and y records of
ground acceleration that would have occurred at the site in the absence of liquefaction. With
several exceptions, the liquefaction (solid symbols) and non-liquefaction (open symbols) case
histories are distinctly separated by the likely liquefaction region. Marginal liquefaction (half
open symbols) is shown for the Chemical Fiber, Treasure Island Fire Station, and Sandholt Road
UC-6 sites. Liquefaction behavior predicted by the procedure by Stokoe et al. (1988b) is
nonconservative for lower levels of shaking ( amax < 0.3 g) and lower values of Vs (Vs < 180
mis). A similar conclusion was reached by Arulanandan et al. (1986) based on the six sites
shaken by the I 975 Haicheng earthquake listed in Table I.

40Q.---,---.-----.--,--..----,---~
Data Based on: Chart ased on: Data Based on- =h"-'-"""""""-"'!.h
Mw = 5.9-6.6 earthquakes Sand Mw = 6.9-7.1 earthquakes Sand
Minimum value of Vs T~ 3 m Minimum value of Vs T~3m
amax for the larger of two Ne = 10 cycles amax for the larger of two Ne = 15 cycles
1/) hortzontal components No drainage 1/) hortzontal components No drainage
E 300 Solid - liquefaction
Open - no liquefaction
Level round
E 300 Solid - liquefaction Level ground
Open - no liquefaction
en I .j I o
> No No
~ Liquefaction ~ Liquefaction 0
·5 ·5
0 0
0
~ 200t---;;-o,.
0 -+----+----+----I ~ 200t----:=+---"'---+--....,,.--+----I

I .
. 0

Liquefaction
• I
• •soil may be weakly
0 .____.__...___._....___..____.__..._____, o.___.__...__________ ..____.__..,___,
cemented with carbonate

0.0 0.2 0.6 0.8 0.0 0.2 0.4 0.6 0.8

amax, 9
(a) Ne = 10 cycles (b) Ne= 15 cycles

Fig. 4 - Comparison of Liquefaction Assessment Charts Proposed by Stokoe et al. (1988b)

Based on Vs and amax with Case Histories of Sites Shaken by Earthquakes with
Magnitude ofS.9 to 7.1 (after Stokoe et al.1988b; Andrus 1994).

102
While it has been suggested (Andrus 1994; after Robertson et al. 1992) that Vs be modified to a
reference overburden stress, this modification alone does not improve the distribution of the
performance data shown in Fig. 4. More work is needed to quantify the effects of layer
thickness and depth.

Procedures Developed from Laboratory Studies

Tokimatsu et al. (1991a) proposed a procedure for evaluating liquefaction resistance using the
stress approach developed by Seed and his colleagues (1971, 1983, and 1985) and results from
laboratory cyclic triax:ial tests on reconstituted sand specimens. In the stress approach, cyclic
loading is represented by the ratio of cyclic shear stress to initial vertical effective stress acting
on a horizontal plane, called cyclic stress ratio. The cyclic stress ratio, CSR, at a particular
depth in a level soil deposit can be expressed as (Seed and Idriss 1971 ):

CSR= 'tavl<J'v= 0.65 (amaxlg) (crv/cr'v) Td (I)

where 'tav is average cyclic shear stress generated by the earthquake, cr'v is initial effective
vertical (overburden) stress, O'v is total overburden stress, g is acceleration of gravity, and rd is a
shear stress reduction factor with a value less than 1.

Resistance to liquefaction in a soil deposit is represented by a cyclic stress ratio or cyclic

resistance ratio, CRR. Tokimatsu et al. (1991a) defined the cyclic resistance ratio for cyclic
triax:ial tests, CRRtx, as the ratio of cyclic deviator stress to initial effective confining stress,
<Jd/2cr' 0 , at the time the double-amplitude axial strain, DA, reaches 5%. Their correlations
between CRRtx at different number of cycles and stress corrected shear wave velocity, Vs I, are
shown in Fig. 5. They used the assumption that Vs is a function of the cube root of the mean
normal effective stress, cr'm, and corrected Vs by:

V SI'."' Vs (l/cr'm)0.33 (2)

where cr'm is in kgf/cm2 (1 kg£icm2 = 98.07 kPa). Tokimatsu et al. selected an exponent of 0.33
rather than 0.25, as determined by Hardin and Dmevich (1972), because it seemed that a slightly
better correlation could be obtained.

For converting CRRtx to an equivalent field cyclic resistance ratio, Tokimatsu et al. (1991a)
suggested the following expression (after Seed 1979):

CRR = 'tJ/cr'v = 1/3 (1 +2K0 ) re (CRRtx) (3)

where 'tJ is average cyclic shear stress resisting liquefaction, K0 is the coefficient of earth
pressure at rest, and re is a constant to account for the effects of multidirectional shaking with a
value between 0.9 and 1.0. As noted by Tokimatsu et al., any value of Ko between 0.5 and 1 can
be assumed for all practical purposes since the effects involved in Eqs. 2 and 3 almost cancel
each other out.

103
 1.5 . - - - - - - - - - - - - - - . - - - - - - - . - - - - - - - - ,

Number of Cycles 3 5101525

0 ._________.___ _ ___._ _ _ ___.__ _ _ ___,

100 150 200 250 300

Stress-Corrected Shear Wave Velocity, V 81 , m/s
(a) Clean Sand
1.5 . - - - - - - - , , - - - - - - - , - - - - - - , - - - - - - - - ,

Number of Cycles 3 5101525

~
LO
II
c3 1.01-----------+-+-r+-----+------1
(I)
(/)
::I
as
()

-
0

OL--------'-------'--------------'
100 150 200 250 300
Stress-Corrected Shear Wave Velocity, V81 , m/s
(b) Silty Sand

Fig. 5 - Liquefaction Assessment Charts Based on Vs1 Proposed by

Tokimatsu et al. (1991a).

104
The field performance data for 20 earthquakes are plotted in Fig. 6. The plotted data are based
on the procedure of T okimatsu et al. (1991 a) outlined above using minimum values of Vs 1 from
the most vulnerable layer and estimates of amax for the larger of the x and y records of ground
acceleration that would have occurred at the site in the absence of liquefaction. Included in Fig.
6 are the liquefaction potential boundaries by Tokimatsu et al. The boundaries are constructed
from the relationships shown in Fig. 5 using Eq. 3 and assuming K 0 of 0.6 and re of 0.95.
Liquefaction behavior predicted by these boundaries is nonconservative for Ne greater than
about 10 cycles and V SI greater than about 150 mis (see Figs. 6c and 6d).

Procedures Developed from Field Performance Studies

Robertson et al. (1992) proposed another stress-based liquefaction assessment procedure using
field performance data from primarily the Imperial Valley, California sites. They corrected Vs
by:

Vs1 =Vs (Pa/cr'v-P·25 (4)

where Pa is a reference stress, 100 kPa or approximately atmospheric pressure, and cr'v is in kPa.
Robertson et al. chose to correct Vs in terms of cr'v to follow the traditional procedures for
correcting standard and cone penetration resistances. It is implied by Eq. 4 that Ko equals 1,
since Vs is a function of mean effective stress (Hardin and Dmevich 1972). Their liquefaction
potential boundary for earthquakes with magnitude of 7.5 is shown in Fig. 7a. ·

Two subsequent liquefaction potential boundaries proposed by Kayen et al. (1992) and Lodge
(1994) for earthquakes with magnitude of about 7 are shown in Fig. 7b. These later curves are
based on field performance data from primarily the 1989 Loma Prieta, California earthquake.
Kayen et al. used field performance data from the Port of Richmond, Bay Bridge Toll Plaza,
Port of Oakland, and Bay Farm Island sites._ They assumed average values ofVs1 and amax for
the larger component of acceleration time histories recorded at neighboring seismograph
stations.

Lodge (1994) considered the same sites that Kayen et al. (1992) evaluated as well as several
additional sites that had been shaken by the Loma Prieta earthquake. The boundary by Lodge
was developed as follows. First, cyclic stress ratios for the entire soil profile at each site were
calculated using Eq. I and amax for the larger component of acceleration time histories recorded
at neighboring seismograph stations. Second, soil layers with a high and a low liquefaction
potential were identified with the simplified procedure of Seed et al. (1985) and SPT blow
counts. Soil layers where the modified blow count fell within 3 blows per 0.3 m of the SPT-
based liquefaction potential boundary of Seed et al. were eliminated due to uncertainties in the
correlation. Third, shear wave velocity measured by the SCPT and crosshole methods were
normalized using Eq. 4. Fourth, on a "meter by meter" basis values of Vs1 and cyclic stress
ratio were plotted for both layer types, those which were predicted liquefiable and those which
were predicted non-liquefiable. Finally, a curve was drawn to include all data for liquefiable
layers.

105
 0.8 .-----.---,--.......-....-----.--....-~
0.8 .Data
--- - -on:
Based - - - - - -Boundaries:
----~ Data Based on: Boundaries:
Mw = 5.9 to 6.2 earthquakes Ne= 6cycles Mw = 6.5 to 6.6 earthquakes Ne= 9cycles
Minimum value of Vs1 Ko= 0.6 Minimum value of Vs1 Ko=0.6
amax for the larger of two re= 0.95 amax for the larger of two re= 0.95
horizontal components horizontal components
> 0.6 Solid - liquefaction
-> O·6 Solid • liquefaction
----4------1

~ Open - no liquefaction
) Open - no !liquefaction

: I ; 0
Clean
Sand
'al Liquefaction 'al Liquefaction
a: 0.4 1-----+-----,...!lf-- a: 0.4 1 - - - - - + - - - - + l - 1 - - - - - 1 - - - - - - - 1
UJ UJ

~
UJ
~
ci5 ci5
.Q No .Q No
'g,
0 •oLiquefaction
0.2 ,___ ___._,_ _,__,__....__ _-4-_ _ __, g,
0
Liquefaction
0.2 1------+---'.+-+----+-----i

0
Explanation 0 Explanation
oo Silts o Silts
Ot Silty sands
0.0 ._...,,._ __..._....__...___,__---''-----' ot _
0.0 ,____.,_....,__...___...___,__ Silty sands
.___ _,
0 100 200 300 400 0 100 200 300 400
Stress-Corrected Shear Wave Stress-Corrected Shear Wave
Velocity, V51, mis Velocity, V51, mis
(a) Ne = 6 cycles (b) Ne= 9 cycles
0.8 ,----,.--,--,----,.--,--,----,.--, 0.8 I
Data Based on: Boundaries:
I
!;!ala Based on:
'
'
Bo!,!ndaries:
Mw = 6.9 to 7.1 earthquakes Ne = 12 cycles Mw = 7 .5 to 8.3 earthquakes Ne = 15 cycles
Minimum value of Vs1 Ko= 0.6 '"Minimum value of Vs1 Ko=0.6 -
amax for the larger of two re= 0.95 amax for the larger of two re= 0.95
horizontal components horizontal components
Solid - liquefaction .... > 0.6 ~Solid - liquefaction
Open - no liquefaction _...
. ~
Open • no liquefaction
I
-
I
. (IJ
.
-
.Q i.-_Clean
Sand
ai Liquefaction
a: 0.4
"'~"'
I,

...,
... S~ty
l • Sand
ci5 - J, No
-
.Q
g, A • I
Liq~efaetion
O 0.2 I

o• Silts
-
...
~1 D 0
Explanation
Silts
Ot Silty sands t Silty sands
•soil may be weakly O I Clean sands o • Clean sands
O.O ..c_e_m_e...
nt_ed_wi
....·1_h_ca_rbo..._na_1e....._..._A_•...;.G.....ra_v_el..,_I.....so~ils_,
0.0
. . A• Gravellv soils
0 100 200 300 400 0 100 200 300 400
Stress-Corrected Shear Wave Stess-Corrected Shear Wave
Velocity, V51, mis Velocity, Vs 1, mis
(c) Ne = 12 cycles (d) Ne= 15 cycles

Fig. 6 - Comparison of Liquefaction Assessment Charts Based on V s1 and CSR

Proposed by Tokimatsu et al. (1991a) with Case Histories
from 20 Earthquake

106
 0.8 ' I ' I '
0.8 ,----,---.--.,.---,.--,--.....----,.--,
Data Bsi!S!iQ Qn: Data Based on: •soil may
Mw = 7.5 to 7.7 earthquakes Mw"' 6.9 to 7.1 earthquakes be weakly
-Average value of Vs 1 - Average value of Vs, cemented
amax for the larger of two amax for the larger of two with
hortzontal components hortzontal components carbonate
> 0.6 -Solid - liquefaction Solid - liquefaction ---,-----1
Open - no liquefaction Open - no liquefaction .:•
)
ci
~
a: 0.4
r

I
Liquefaction
r
Robertson et al. _
(1992) •
• * •* •*
0 0

Cl)
Cl)
I
~ • ., /
ci5 I-
' -
0 ID'
.2
g, •••• , I No
, Liquefaction- Liquefaction
0 0.2 0
·I
- I/
/ 0
-
Lodge ( 1994)
' ' ' ' 0.0 ...._........._ ......._....._........_ _.__.,_________,
0.0
0 100 200 300 400 0 100 200 300 400
Stress-Corrected Shear Wave Stress-Corrected Shear Wave
Velocity, Vs1, mis Velocity, Vs 1, mis
(a) Mw= 7.5 {b) Mw= 7

Fig. 7 - Comparison of Liquefaction Assessment Charts Based on Vs1 and CSR

Proposed by (a) Robertson et al. (1992) and (b) Kayen et al. (1992) and
Lodge (1994) with Case Histories from Earthquakes with
Magnitude of6.9 to 7.7.

Field performance data from earthquakes with magnitude of 6.9 to 7.7 are also plotted in Fig. 7.
The plotted data are based on average values of Vs1 from the most vulnerable layer at the
investigated sites. The cyclic stress ratios are calculated using estimates of amax for the larger of
two horizontal components of ground acceleration that would have occurred at the site in the
absence of liquefaction. With a few exceptions, the liquefaction case histories are bounded by
the relationships by Kayen et al. (1992) and Lodge (1994). The relationship by Robertson et al.
(1992) is the least conservative of the three relationships.

Recommended Liquefaction Potential Boundaries Based on V s1 and CRR

After reviewing the proposed procedures outlined above, this workshop agreed that a careful
review of the case histories should be conducted. It was suggested that the recommended Vs-
based procedure follow the general format of the CPT- and SPT-based procedures.

107
The compiled case histories for magnitude 5.9 to 7.7 earthquakes are shown in Figs. 8, 9 and 10.
The plotted data have been separated into three categories: (1) sands and gravels with average
fines (particles smaller than 75 µm) content less than or equal to 5%, Fig. 8; (2) sands and
gravels with average fines content of 6% to 34%, Fig. 9; and (3) sands and silts with average
fines content greater than or equal to 35%, Fig. 10. Where possible, the fines content is noted
next to the data point corresponding to soils with over 5% fines. The data for the Larter Ranch
and Whiskey Springs sites are not shown, since the soils at these two sites may be weakly
cemented with carbonate. Following the recommendation of this workshop, the plotted data are
based on representative values of Vs 1 and arnax for the average of peak values for the x and y
ground acceleration time histories that would have occurred at the site in the absence of
liquefaction. Values of VSI are calculated using Eq. 4. Values of rd are estimated using the
relationship by Seed and Idriss (1971 ).

0.8 I
' '
0.8 ' I ' I ' '
Data Bas~d on: Ds!s1 Ba~ed 2n:
Mw = 6.9 to 7.0 earthquakes Mw = 7 .5 to 7. 7 earthquakes
-Average fines content:;; 5% - -Average fines content:;; 5% -
Average values of Vs1 & amax
Average values of V51 & amax
Uncemented, Holocene-age soils
Uncemented, Holocene-age soils _Solid - liquefaction
> 0.6 -Solid - liquefaction > 0.6
Open - no liquefaction

'
~> Open - no liquefaction
.,,co
c5 - I I
- "'
.Q - I I
-

~ Liquefaction I b.
1u Liquefaction I
a: 0.4 b. a: 0.4
ti)
I I (/)
(/)
I
(/)

......,
~ No ~ No
ci5
.2
- ♦ ♦ ♦ I
_.\•_. □
Liquefaction - ci5
.9
f,- ♦ ♦♦ Liquefaction
-

~
I
~ U 0.2
U 0.2 -

V
I

- V
C Explanation
f,- Explanation
a ■ Sands
~ ♦ Sands-gravels a• Sands
,..,
- --100
♦ Sands-gravels
0.0
0
- 100
'
200
1:,._. Gravels

300 400
0.0
0
I

200 300 400

Stress-Corrected.Shear Wave Stress-Corrected Shear Wave
Velocity, Vs 1,·mJs Velocity, Vs 1, mis
(a) Mw=7 (b) Mw=7.5

Fig. 8 - Comparison of Liquefaction Assessment Charts Based on Vs1 and CSR from
Analysis for this Report with Case Histories ofUncemented Soils with
Fines Content Less than or Equal to 5%.

108
 0.8
' Qn: ' ' . 0.8
I
Dall!' B,!i:;ed Data' Based
' on:• I

Mw = 5.9 to 6.2 earthquakes Mw = 6.5 to 6.6 earthquakes

-Average fines content= 10% to Z 1/o - ~Average fines content= 10% to 22% -
Average values of Vs, & amax Average values of Vs1 & !!max
Uncemented, Holocene-age soils Uncemented, Holocene-age soils
_> 0.6 _Solid - liquefaction _>0.6 _Solid • liquefaction
Open - no liquefaction Open - no liquefaction
~ -!:2
~CG ~~
- - ~
-
6
1ii I
Liquefaction
.Q
1ii Liquefaction
I 20,
I
I
cc cc •2
rn 0.4 <fl 0.4 10
<f) <fl
!!! !!!
u5 ,.. No u5 ,.. No

J Liquefaction
- Liquefaction
-
.Q .Q
~ 113 ~ J
!/
(.) 0.2 U 0.2

~
1W
Explanation Explanation
- 820 01 Silty sands
- 3
5
01 Silty sands
2°«) D Sands with silt D Sands with silt
/
0.0 --r ~ I Ji/ 13 Fines
0.0
.... ✓
'
13 Fines
0 100 200 300 400 0 100 200 300 400
Stress-Corrected Shear Wave Stress-Corrected Shear Wave
Velocity, Vs1, mis Velocity, Vs1, mis
(a) Mw= 6 (b) Mw= 6.5
0.8 '
0.8 . . ---
Data - . -Qn:
Based --------------. ' I
Data Based Qn:
I
'
Mw = 6.9 to 7.0 earthquakes
Average fines content= 6% to 34%
Mw = 7.7 earthquakes
-Average fines content= 14%
I -
Average values ofVs1 & amax Average values of Vs1 & amax
Uncemented, Holocene-age soils
Uncemented, Holocene-age soils
Solid - liquefaction -+-----i > 0.6 _Solid - liquefaction
Open - no liquefaction
Open - no liquefaction ~
I
~CG
- -
6 I
~ Liquefaction I
cc<fl 0.4
en I
No I!! No
Liquefaction u5 ~

Liquefaction -
,Q
g, 14....
U 0.2

O.O
Ot Silty sands
□ ■ Sands
with silt
•~Gravels with silt
.._______.,,........__________.__13-'F_in_e_s_______.
~

-✓
L) Explanation
1 Silty sands
13 Fines
0.0 '
0 100 200 300 400 0 100 200 300 400
Stress-Corrected Shear Wave Stress-Corrected Shear Wave
Velocity, Vs1, mis Velocity, v81 , mis
(c) Mw=7 (d) Mw=7.5
Fig. 9 - Comparison of Liquefaction Assessment Charts Based on Vs1 and CSR from
Analysis for this Report with Case Histories of Uncemented Soils with
Fines Content of6% to 34%.

109
 0.8 I ' I ' 0.8 ' ' .
' '
t;lsl!.i 12<1:.i!llQ Qn: Dm.i B<!§!ilQ Qn:
Mw = 5.9 to 6.2 earthquakes Mw = 6.5 to 6.6 earthquakes
,_Average fines content.: 35% - -Average fines content.: 35% -
Average values of Vs1 & amax Average values of Vs1 & amax
Uncemented, Holocene-age soils Uncemented, Holocene-age soils
,_Solid - liquefaction _> 0.6 -solid - liquefaction
Open • no liquefaction ~.,.(lj
Open • no liquefaction
I- - - -
I I
~
r5 I I
Liquefaction I a: Liquefaction I
(/} 0.4

.;I U) I
~

~•,s,
50(1[) I
- No - ci5 I-
No
05o Liquefaction _g so(![) Liquefaction -
g ., J
U 0.2

'½
1f
35

Explanation 0 37 Explanation
- 35 37 I- 75o
Silts
0 • 0 Silts
~ffi O• Silts-sands 0 • Silts-sands
.,, / 3510 50
, 58 Fines ...-
/
' , 58 Fines
0.0 ' 0.0
0 100 200 300 400 0 100 200 300 400
Stress-Corrected Shear Wave Stress-Corrected Shear Wave
Velocity, Vs 1, mis Velocity, v81 , mis
(a) Mw= 6 (b) Mw=6.5
0.8 • I I ' 0.8 ' ' ' I ' '
Data Based on: Data Based on:
Mw = 6. 9 to 7.1 earthquakes Mw = 7. 7 earthquakes
-Average fines content.: 35%
Average values of Vs1 & amax
- -Average fines content.: 35%
Average values ofVs1 & amax
-
Uncemented, Holocene-age soils Uncemented, Holocene-age soils
> 0.6 ,-Solid - flquefaction >0.6 ,-Solid - liquefaction
Open • no liquefaction
Open - no liquefaction
~
\ -
J, -
I
I-> I- I-

0
i Liquefaction
I
I
r5
~
I
Liquefaction I
a: 0.4 a: 0.4
(/}
(/}
I U)
U)

I!? No ~
ci5
.9
- nao
Liquefaction - ci5
.Q
- No
Liquefaction
-
g_ so(![)
~
580 GD

U 0.2 U 0.2

~ V
Explanation
I- 35 ~ o •Silts
I-
Explanation

0.0
0
-t"' ---
100 200
'
300
Stress-Corrected Shear Wave
O• Silts-sands
58 Fines
400
0.0
0 -- 100 200
I

300
Stress-Corrected Shear Wave
'
0 Silts
s8 Fines

400

Velocity, Vs 1, mis Velocity, Vs1, mis

(c) Mw =7 (d) Mw= 7.5
Fig. 10 - Comparison of Liquefaction Assessment Charts Based on Vs1 and CSR from
Analysis for this Report with Case Histories of Uncemented Soils with
Fines Content Greater than or Equal to 35%.

110
Dobry (1996) derived a relationship between cyclic stress ratio and VsI for constant average
cyclic strain, Yav, using the equations:

Yav = 'tavl(G}yav (5)

and

Gmax = P Vs 2 (6)

where (G)yav is shear modulus at Yav, Gmax is small-strain shear modulus, and p is mass density.
Combining Eqs. 5 and 6, and dividing both sides by <i'v leads to:

'tav/<i'v = (p/<i' v) Yav (G/Gmax)rav V s 2 (7)

If everything is done at a reference stress, Pa, then Vs = Vs I and a line of constant average
cyclic strain is of the form:

(8)

where f('Yav) = (p/P3 ) 'Yav (G/Gmax)yav• This formulation assumes the modulus reduction factor,
(G/Gmax)yav, is independent of confining pressure and pore water pressure buildup. Equation 8
is strong evidence for extending the liquefaction potential boundaries to the origin, and provides
a rational approach for establishing the boundaries at low values ofV SI (say V SI ::; 125 mis).

For higher values ofVs1, it seems reasonable that the boundary separating liquefiable and non-
liquefiable soils would become asymptotic to some limiting value of Vs I. This limit is caused
by the tendency of dense granular soils to exhibit dilative behavior at large strains. Thus, Eq. 8
is modified to:

CRR = -r1/cr'v = a (Vsi/100)2 + b [1/(Vsic - V s1)- lNs1cJ (9)

where Vs1c is the critical value ofVsI that separates contractive and dilative behavior, and "a"
and "b" ·are curve fitting parameters,

Using the relationship between CRR and V s1 expressed by Eq. 9, curves have been drawn to
separate the liquefaction and non-liquefaction case histories plotted in Figs. 8, 9 and 10. The
curves are drawn assuming a = 0.03 and b = 0.9 for earthquakes with magnitude of 7.5.
Depending on fines content (FC), the following values ofV Slc are also assumed:

V Sic= 220 mis for sands and gravels with FC s; 5% (10a)

Vs1c = 210 mis for sands and gravels with FC"" 20% (10b)
Vs1c = 200 mis for sands and silts with FC ~ 35% (10c)

For earthquakes with magnitude of 6, 6.5 and 7, scaling factors of 2.1, 1.6 and 1.25,
respectively, are applied to the curves for magnitude 7.5 earthquakes. The curves shown in
Figs. 8, 9 and 10 correctly predict more than 95% of the occurrences ofliquefaction.

111
The three liquefaction case histories that lie slightly below the boundary curves shown in Figs.
8a and 9c are for the Treasure Island UM06 and UMl l, and Marina District School sites. The
data point for Treasure Island UMl 1 (see Fig. 9c) would lie on the boundary for 7% fines
content, the average fines content of the most vulnerable layer for this site. In addition, the
Treasure Island sites are located along the perimeter of the island where liquefaction was
moderate during the 1989 Loma Prieta earthquake and where sloping ground may have been a
factor. The Marina District School site is located on the margin of mapped artificial fill and
liquefaction damage caused by the Loma Prieta earthquake. Hence, there are only two cases of
liquefaction that incorrectly lie outside the region of predicted liquefaction as defined by these
procedures, and they are cases of marginal to moderate liquefaction.

Figure 11 presents the recommended liquefaction potential boundaries for magnitude 7.5
earthquakes and uncemented Holocene-age soils with various fines content. Although these
boundaries pass through the origin, natural alluvial sandy soils with shallow water tables rarely
have stress corrected shear wave velocities less than 100 mis, as shown by the in situ
measurements presented in Figs. 8, 9 and 10. For a V s1 -value of 100 mis and a magnitude 7.5
earthquake, the calculated CRR is 0.03. This minimal CRR value is consistent with intercept
CRR values of 0.03 to 0.05 suggested by the CPT and SPT procedures. The recommended
boundary for uncemented soils with fines content::; 5% and earthquakes with magnitude of 7,
shown in Fig. 8a, is similar to the boundaries of Kayen et al. (1992) and Lodge ( 1994), shown in
Fig. 7b, at lower values ofVs1 (Vs1 < 200 mis).

Values of Vs1c between 200 mis and 220 mis are consistent with values determined using the
relationship between SPT blow count and shear wave velocity by Ohta and Goto (1976)
modified to blow count with theoretical free-fall energy of 60% (Seed et al. 1985). Assuming a
corrected blow count of30 and a depth of 10 m, approximate values ofVs1 range from 190 mis
for clays to 220 mis for sandy gravels of Holocene-age. More work is needed to further validate
and refine the values ofV Slc•
. ·-
The magnitudes scaling factors of 2.1, 1.6 and 1.25 for earthquakes with magnitude of 6, 6.5 and
7, respectively, compare well with SPT-based factors developed in recent years by several
investigators (Youd and Noble in press), as noted in Columns 3 through 7 of Table 2. They
form the upper bound of scaling factors recommended by this workshop (Section 1, workshop
report) for earthquakes with magnitude less than 7.5. The lower bound of the range of
recommended scaling factors is defined by the scaling factors developed by Idriss (1996), as
listed in Column 3 of Table 2.

The relationship between earthquake magnitude and magnitude scaling factor, MSF, can be
expressed by (modified from Idriss 1996):

MSF = (Mw/7.S)Il (11)

112
 0.8 r-----.,-----...,.....---...,.....----,-----,------.
Boundaries Defined by:
CRR = a(Vs 1/100) 2 + b[1/(Vs 10-Vs 1)-1Ns1cl Mw=7.5
a = 0.03, b = 0.9
Vs 1c = 220 mis for sands and gravels with FC:,; 5%
Vs 1c = 210 mis for sands and gravels with FC"" 20%
V s1c = 200 mis for sands and silts with FC ~ 35%
Chart Based on:
0.6 Average values of V51 & amax
Uncemented, Holocene-age soils
Average fines content (FC)

>
-t:) FC ~35 20 $5
>
"'(Tl
III
0 III
~
a:en 0.4 1------------------1-1------------1
en
III
-~
Cf)
.2
0>,
(.) Liquefaction

No
Liquefaction

0.0
0
----
i..---==---=-.1.----....1.----....1.----..1...----.1..---_,I
100 200
Stress-Corrected Shear Wave Velocity, V 81 , m/s
300

Fig. 11 - Recommended Liquefaction Assessment Chart Based on V s1 and CSR for

Magnitude 7.5 Earthquakes and Uncemented Soils of Holocene Age.

113
where "n" is a curve fitting parameter. The scaling factors developed by Prof. Idriss as listed in
Column 3 of Table 2 are defined by Eq. 11 with n = -2.56. For the scaling factors used to
construct the Vs-based liquefaction potential boundaries shown in Figs. 8, 9 and 10 (MSF = 2.1,
1.6, 1.25, and 1.0 for Mw = 6, 6.5, 7 and 7.5, respectively), the value of"n" is -3.3.

While only the scaling factors detennined by Idriss (1996) for earthquakes with magnitude
greater than 7.5 have been recommended by this workshop, the scaling factors determined using
n = -3.3 are slightly more conservative. For example, Eq. 11 with n = -3.3 provides scaling
factors of 0.81 and 0.66 for earthquakes with magnitude of 8 and 8.5, respectively. These
scaling factors are slightly less than the scaling factors of 0.84 and 0.72 for earthquakes with
magnitude of 8 and 8.5, respectively, determined by Prof. Idriss.

Using Eq. 11 with n = -3.3 and the boundary for uncemented clean sands and gravels shown in
Fig. 11, leads to the family of curves shown in Fig. 12. The curves shown in Fig. 12 imply that
liquefaction will never occur in any earthquake if V s1 exceeds 220 mis and the soils are
uncemented and of Holocene age.

In areas with cemented soils, local correlations between shear wave velocity and penetration
resistance should be developed to determine the effects of cementation. The boundaries shown
in Fig. 11 could then be modified by increasing the abscissas by some factor. For example,
measurements from the Larter Ranch and Whiskey Springs sites which liquefied during the
1983 Borah Peak, Idaho earthquake suggest a correction factor of about 1.3 to 1.4 (Andrus
1994) for those distal alluvial fan sediments.

Table 2. Magnitude Scaling Factors Obtained by Various Investigators

(modified from Youd and Noble in press).

Magnitude Scaling Factor (MSF)

Moment Seed and Idriss Ambraseys Youd and Arango This
Magnitude, Idriss (1996) (1988) Noble, (1996) Report
Mw (1982) p<32%
(in press)
(I) (2) (3) (4) (5) (6) (7) (8)
5.5 1.43 2.20 2.86 3.42 3.00 2.20 2.8*
6.0 1.32 1.76 2.20 2.35 2.00 1.65 2.1
6.5 1.19 1.44 1.69 1.66 1.60 1.40 1.6
7.0 1.08 1.19 1.30 1.20 1.25 1.10 1.25
7.5 1.00 1.00 1.00 1.00 1.00 1.0
8.0 0.94 0.84 0.67 0.75 0.85 0.8*
8.5 0.89 0.72 0.44 0.65*

*Extrapolated from scaling factors for Mw = 6, 6.5, 7 and 7.5 using MSF = (Mwl7.st3· 3.

114
 0.8 . . - - - - - - . - - - - - . . . - - - - . . . - - - - - , - - - - - - , . - - - - ,
Boundary for Mw = 7.5 Defined by: Scaling
CRR = a(Vs/100) 2 + b[1/(Vs 1c·V81 )-1N81 J Magnitude Factor
a= 0.03, b = 0.9, V810 = 220 m/s 5.5 2.8*
6.0 2.1
Chart Based on: 6.5 1.6
Average values of V81 & ~ax 7.0 1.25
Uncemented, Holocene-age clean sands 7.5 1.0
and gravels 8.0 0.8*
0.6 8.5 0.65*
*Extrapolated from other
scaling factors, (Mwf7.5)·3.3
>
~
Ji
r5
~
er:
fl)
0 . 4 1 - - - - - - - - - - 1 - - - - - - - -........~ 1 - - - - - - - - - I
fl)

-
~
en
.2
~
(.) Liquefaction

No
Liquefaction

100 200 300

Stress-Corrected Shear Wave Velocity, Vs 1 , rn/s

Fig. 12 - Recommended Chart Based on Vs1 and CSR for Evaluation of Liquefaction
Potential of Uncemented Clean Sands and Gravels of Holocene Age.

115
Recommended Liquefaction Potential Boundaries Based on Vs and amax

By combining Eqs. I, 4 and 9, a relationship based on Vs and amax is obtained in the form of:

(12)

where fi = cr'v/(0.65 crv rd) and f2 = (Palcr'v) 0-25 . Assuming (1) the water table is located
midway between the ground surface and the center of the most vulnerable layer and (2) the total
unit weight of soil is 17.3 kN/m 3 above the water table and 18.9 kN/m3 below the water table,
then f1 and f2 can be approximated by:

(13)

and

f2 = (7.3/z)025 (14)

where z is depth to center of the most vulnerable layer in meters. For noncritical projects, this
workshop suggests the following equations to estimate average values of rd (Liao and
Whitman):

rd= 1.0 - 0.00765 z forz::;;9.15m (15a)

rd= 1.174- 0.0267 z for 9.15 m < z ::;; 23 m (15b)

Equations 12 through 15 provide a simple relationship between Vs and amax that depends on
depth. A relationship that depends on depth agrees with the analytical study by Stokoe et al.
(1988b). For exarilple, the critical values ofV s shown in Fig. 3c at amax equal to 0.2 g and layer
thickness of 3.0 m are about 110 mis for a depth of 4.6 m and 170 mis for a depth 12.2 m.

Liquefaction potential boundaries defined by Eqs. 12 through 15 are shown in Figs. 13, 14 and
15. Also shown are the case history data. Liquefaction behavior predicted by these boundaries
is similar to behavior predicted by the boundaries based on V s1 and CRR. The three
liquefaction case histories that lie slightly below the boundaries shown in Figs. 13a and 14c are
the same three that lie slightly below the boundaries shown in Figs. 8a and 9c (Treasure Island
UM06 and UMl 1, and Marina District School sites). Thus, the procedure based on Vs, amax
and depth is a good approximation to the recommended procedure based on Vs 1 and CRR.

The application of Eqs. 12 through 15 should be limited to sites with characteristics similar to
the database (i.e., level ground, depth of most vulnerable layer less than 12 m, depth of water
table 0.5-7.6 m, and uncemented soils of Holocene age).

116
 0.8 . . - - - - - - - - - - - - . . . -......... 0.8 .Data
. - Based
- - -on:
----------
Data Based on:
Mw = 6.9 to 7.0 earthquakes Mw = 7.5 to 7.7 earthquakes
Average fines content:. 5% Average fines content :. 5%
Average values of Vs & amax Average values of Vs & amax
Uncemented, Holocene-age soils
Uncemented, Holocene-age soils Depth to center of most vulnerable layer
0.6 Depth to center of most vulnerable layer 0.6

Depth =4 Depth=4 710m

I 3.6
t,.
I I I
Cl
'x I I I
al 0.4 1------+-----,,-+-,--,---1-------l
E 4.8
al I ♦♦♦

No Liquefaction No
Liquefaction Liquefaction

Explanation
□■ Sands
o ♦ Sands-gravels a ■ Sands
ll.• Gravels I Sands-gravels
0.0 .._....
--_ -~ ......_ _ _....__._5.8
__ "--_--'
Depth ~
0.0 '---='----'--'-----'--L------'
5.8 Depth

0 100 200 300 400 0 100 200 300 400

Shear Wave Velocity, v8 , mis Shear Wave Velocity, Vs, mis
(a) Mw= 7 (b) Mw= 7.5

Fig.13 - Comparison of Liquefaction Assessment Charts Based on Vs and Average amax

from Analysis for this Report with Case Histories of Uncemented Soils with
Fines Content Less than or Equal to 5%.

117
 0.8 r-----r----,---,--,----,-----,---,---, 0.8 ,---,---,-----,---,---,--,-----,,---,
Data Based on: Data Based on:
Mw = 5.9 to 6.2 earthquakes Mw = 6.5 to 6.6 earthquakes
Average fines content= 10% to 2 %
Average values of Vs & amax
Uncemented, Holocene-age soils
J Average fines content = 10% to
Average values of Vs & amax
Uncemented, Holocene-age soils
1/o
1
0.6 Depth to center of most vulnerabl layer 0.6 Depth to center of most vulnerabl layer

Depth= 4 710 m
Depth =4 710 m 20. 3.2.. co 10, 3.2
I I I I 22, 3.6 I I

~ Liquefaction I II Ol
'x Liquefaction I I
~ 0.4 1------~1-,+-,-,---+----1 ~ 0.4
a, a,

13, 4.1
No No
Liquefaction Liquefaction

13, 4.1 Explanation Explanation

13, 4.1
/
/20, 3 -2 1D
820
' s. I
o• Silty sands
a Sands with silt
o•
Silty sands
,:;, Sands with silt
D
0.0 ~ "'22, :i.sllD _ __.___.___......,:__,__.
....,.:....i,._...,..__. 9910, s.2 13, 4.1 Fines, depth 0.0 .✓-
.__......,_._......__......__..._ ,a.
4.1 Fines, depth
......______ -'--'
0 100 200 300 400 0 100 200 300 400
Shear Wave Velocity, Vs, mis Shear Wave Velocity, Vs, mis
(a) Mw= 6 (b) Mw = 6.5

0.8 - - - - - - - - - - - -.....----, 0.8 ,-----,-.....--,.--,-----,---,--..,...---,

Data Based on: Data Based on:
Mw = 6.9 to 7.0 earthquakes Mw = 7.7 earthquakes
Average fines content= 6% to 34%
Average values of Vs & amax
Uncemented, Holocene-age soils
I
Average fines content= 14%
Average values of Vs & amax
Uncemented, Holocene-age soils
l
0.6 0.6 Depth to center of most vulnerabl layer
Depth to center of most vulnerable layer
Depth=4 710m Depth= 4 7 10 m
~ .,rnla.7
na.90 I I I I I
na, 2.6-4.0 I
tO 10 ~al 0.4 1------1----.--,,--,---+------t
I I I
Liquefaction

<12, 3.8~ II 15, JS

10. 5.o I
No
Liquefaction
a,E Liquefaction
....
14,8.8
No
Liquefaction

~.s2•t. f.s
A6,3.6
0.2 >----na, 4.8,..--+-,....----+-----s
14, 4.
na,6.4~ I Explanation
14-26, 7.8-ll.8 o• Silty sands Explanation
a ■ Sands with silt
/. tu. Gravels with silt 1 Silty sands
0.0 .,_;...:a._.....,_...__..___.__
_,., -,, 4. 1 Fines, depth
13,_____ __,___. 0.0 .,,,, 13,_
-----=-......__......__..._...__ Fines,
4.1 _ __ depth ~

0 100 200 300 400 0 100 200 300 400

Shear Wave Velocity, Vs, mis Shear Wave Velocity, Vs, m/s
(c) Mw=7 (d) Mw= 7.5

Fig.14- Comparison of Liquefaction Assessment Charts Based on Vs and Average amax

from Analysis for this Report with Case Histories of Uncemented Soils with
Fines Content of6% to 34%.

118
 0.8 ...----,---,---,.--...----,---,---,.---, 0.8 ,......--,---,--..--...----,.--.--~---.
Data Based on: Data Based on:
Mw = 5.9 to 6.2 earthquakes Mw =6.5 to 6.6 earthquakes
Average fines content.: 35% Average fines content .: 35%
Average values of Vs & amax Average values of Vs & amax
Uncemented, Holocene-age soils Uncemented, Holocene-age soils
0.6 0.6 Depth to center of most vulnerable layer
Depth to center of most vulnerable layer

Depth= 4 710 m Depth =4 710 m

I I II I I I I
~ Liquefaction I I .!::J I I/
Liquefaction
~ 0.4 ~ 0.4 1-----+---,1-,r,-'- - - + - - - - - f
al 75, 4.2
I I
al
No No
Liquefaction Liquefaction

Explanation Explanation
/. 50, 4.5 o • Silts o Silts
-,,:
~ ll!l
U[J
8 OI Silts-sands
30, 14.8
p Ot Silts-sands
.,.._
0.0 ,_...... _.__.____._---1._ _Fines,
58, 4.1 _..:.....__._
depth....
0.0
,., sa, 4.1 Fines, depth
0 100 200 300 400 0 100 200 300 400
Shear Wave Velocity, Vs, mis Shear Wave Velocity, Vs, mis.
(a) Mw=6 (b) Mw= 6.5

0.8 ..--.---...--..--.--.------, 0.8 ...----.--.......-...---.a.,.--,--.----,.-~

Data Based on: Data Based on:
Mw = 6.9 to 7 .1 earthquakes Mw = 7.7 earthquakes
Average fines content.: 35% Average fines content .: 35%
Average values of Vs & amax Average values of Vs & amax
Uncemented, Holocene-age soils Uncemented, Holocene-age soils
0.6 Depth to center of most vulnerable layer 0.6 Depth to center of most vulnerable layer

Depth= 4 7 10 m Depth= 4 7 10 m
I
I II I I I I
I I I
~m 0.4 Liquefaction O I I I
na, 3.a~....,.t.- - - + - - - - - t
~m 0.4 1------t----,---.t-,----t------1
Liquefaction

E 1 11 E
a:l <ti
No No
Liquefaction Liquefaction

42-00, 4.5-6.8 Explanation

12. 4.o o • Silts Explanation
.,-:: 0 I Silts-sands o Silts
--< ;::e: 58, 4.1 Fines, depth 58. 9.8 Fines, depth
0.0
0 100 200 300 400 100 200 300 400
Shear Wave Velocity, Vs, mis Shear Wave Velocity, Vs, mis
(c) Mw= 7 (d) Mw= 7.5

Fig.15 - Comparison of Liquefaction Assessment Charts Based on Vs and Average amax

from Analysis for this Report with Case Histories of Uncemented Soils with
Fines Content Greater than or Equal to 35%.

119
 Conclusions and Recommendations

This report summarizes liquefaction and non-liquefaction case histories from 20 earthquakes
and over 50 sites in soils ranging from sandy gravel with cobbles to profiles including silty clay
layers. The data are limited to relatively level ground sites with the following characteristics:
(1) depth of most vulnerable layer less than 12 m; (2) uncemented soils of Holocene age, with a
few exceptions; and (3) depth of water table between 0.5 m and 7.6 m.

The compiled case histories are used to evaluate current liquefaction assessment procedures
based on small-strain shear wave velocity. Most sites where surface manifestations of
liquefaction were observed are correctly predicted by the current procedures. However, the
boundaries by Stokoe et al. (1988b) are nonconservative at values ofV s less than about 180 mis.
The boundaries by Tokimatsu et al. (1991a) for earthquakes with greater than about 10 cycles of
loading are nonconservative at values of Vs1 greater than about 150 mis. The boundary by
Robertson et al. (1992) for earthquakes with magnitude of 7.5 is nonconservative at values of
V s1 less than about 200 mis. With few exceptions, the liquefaction case histories for
earthquakes with magnitude of 7 are bounded by the relationships by Kayen et al. (1992) and
Lodge (1994).

This workshop agreed that a careful review of the compiled case histories should be conducted.
It was suggested that the recommended Vs-based procedure follow the general format of the
CPT- and SPT-based procedures.

To develop the recommended liquefaction potential boundaries, the compiled case histories are
separated into three categories: (1) sands and gravels with average fines content less than or
equal to 5%; (2) sands and gravels with average fines content of 6% to 34%; and (3) sands and
silts with average fines content greater than or equal to 35%. The data for two sites are not
considered, since soils at these sites may be weakly cemented with carbonate. Representative
values of Vs 1 for the most vulnerable layer and average values of amax that would have occurred
at the site in the absence of liquefaction are used. Values of VSI are calculated using Eq. 4.
Values of rd are estimated using the relationship by Seed and Idriss ( 1971 ).

The recommended liquefaction potential boundaries are established by applying a modified

relationship between Vs1 and cyclic stress ratio for constant average cyclic shear strain
suggested by Dobry (1996). The relationship by Dobry provides strong evidence for extending
the boundaries to the origin. It is modified to become asymptotic to some limiting value ofVs1-
This limit is caused by the tendency of dense granular soils to exhibit dilative behavior at large
strains.

Figure 11 presents the recommended liquefaction potential boundaries for magnitude 7.5
earthquakes and uncemented Holocene-age soils. These boundaries are defined by Eq. 9 with a
= 0.03, b = 0.9, and V s1c = 200 mis to 220 mis depending on fines content.

120
Using scaling factors of 2.1, 1.6, 1.25 and 1.0 for earthquakes with magnitude of 6, 6.5, 7 and
7.5, respectively, provide boundaries that included more than 95% of the liquefaction case
histories. These magnitude scaling factors lie within the range of scaling factors recommended
by this workshop.

Caution should be exercised when applying the liquefaction potential boundaries to sites where
conditions are different from the database. More work is needed to further validate and refine
the values of Vs 1c• Additional well-documented case histories of all types of soil that have and
have not liquefied during earthquakes should be compiled, particularly from deeper deposits
(depth> 8 m) and from denser soils (Vs> 200 mis) shaken by stronger ground motions (amax >
0.4 g), to further validate these boundaries.

Liquefaction potential boundaries based on Vs, amax and depth defined by Eqs. 12 through 15
provide a good approximation to the recommended procedure based on V SI and CRR. These
simpler boundaries are suggested for initial site screening, and should be limited to sites with
characteristics similar to the database.

Two limitations of using shear wave velocity are its high sensitivity to weak interparticle
bonding, and the lack of a sample for identifying non-liquefiable clayey soils. Therefore, the
preferred practice is to drill sufficient boreholes and take samples to verify or develop local
correlations for soil types encountered, to identify non-liquefiable clay-rich soils, and to detect
liquefiable weakly cemented soils. A combination of techniques may provide the most cost-
effective approach for evaluating sites of large areal extent. In some cases, such as many
landfills where borings are not permitted, evaluation based on shear wave velocity may be the
only feasible approach.

Acknowledgments

The reporters thank the participants of this workshop for their constructive review, which
greatly enhanced the quality of this report. Workshop participants included T. L. Youd (Chair),
I. M. Idriss (Co-chair), I. Arango, G. Castro, J. T. Christian, R. Dobry, W. D. L. Finn, L. F.
Harder, Jr., M. E. Hynes, K. Ishihara, J. P. Koester, S. S. C. Liao, W. F. Marcuson, Ill, G. R.
Martin, J. K. Mitchell, Y. Moriwaki, M. S. Power, P. K. Robertson, and R. B. Seed. The review
comments and encouragement ofR. M. Chung are also greatly appreciated.

121
 References

Ambraseys, N. N. (1988). "Engineering Seismology," Earthguake Engineering and Structural

Dynamics, Vol.17,p.1-105.

Andrus, R. D., Stokoe, K. H., II, Bay, J. A., and Youd, T. L. (1992). "In Situ Vs of Gravelly
Soils Which Liquefied," Proceedings, Tenth World Conference on Earthquake Engineering,
held in Madrid, Spain, A.A. Balkema, Rotterdam, The Netherlands, pp. 1447-1452.

Andrus, R. D. (1994). "In Situ Characterization of Gravelly Soils That Liquefied in the 1983
Borah Peak Earthquake," Ph.D. Dissertatiol!, University of Texas at Austin, 533 p.

Andrus, R. D., and Youd, T. L. (1987). "Subsurface Investigation of a Liquefaction-Induced

Lateral Spread, Thousand Springs Valley, Idaho," Geotechnical Laboratory Miscellaneous Paper
GL-87-8, U.S. Anny Engineer Waterways Experiment Station, Vicksburg, MS, 131 p.

Aouad, M. F. (1986). "Analytical Investigations of the Relationship between Shear Wave

Velocity and the Liquefaction Potential of Sands," M.S. Report, Geotechnical Engineering
Center, Department of Civil Engineering, University of Texas at Austin, 87 p.

Arango, I. (1996). "Magnitude Scaling Factors for Soil Liquefaction Evaluations," Journal of
Geotechnical Engineering, ASCE, Vol. 122, No. 11, pp. 929-936.

Arulanandan, K., Yogachandran, C., Meegoda, N. J., Ying, L., and Zhauji, S. (1986).
"Comparison of the SPT, CPT, SV and Electrical Methods of Evaluating Earthquake Induced
Liquefaction Susceptibility in Ying Kou City During the Haicheng Earthquake," Proceedings,
Use ofln Situ Tests in Geotechnical Engineering, Geotechnical Special Publication No. 6, held
in Blacksburg, Virginia, S. P. Clemence, Ed., ASCE, New York, NY, pp. 389-415.

Barrow, B. L. (1983). "Field Investigation of Liquefaction Sites in Northern California,"

Geotechnical Engineering Thesis GT83-1, University of Texas at Austin, 213 p.

Bennett, M. J. (1995). Personal communication to R. D. Andrus.

Bennett, M. J., Youd, T. L., Harp, E. L., and Wieczorek, G. F. (1981). "Subsurface
Investigation of Liquefaction, Imperial Valley Earthquake, California, October 15, 1979," Open-
File Report 81-502, U.S. Geological Survey, 83 p.

Bennett, M. J., McLaughlin, P. V., Sarmiento, J. S., and Youd, T. L. (1984). "Geotechnical
Investigation of Liquefaction Sites, Imperial Valley, California," Open-File Report 84-252, U.S.
Geological Survey, I 03 p.

122
Bennett, M. J., and Tinsley, J. C. (1995). "Geotechnical Data from Surface and Subsurface
Samples Outside of and within Liquefaction-Related Ground Failures Caused by the October 17,
1989, Loma Prieta Earthquake, Santa Cruz and Monterey Counties, California," Open-File
Report 95-663, U.S. Geological Survey.

Bierschwale, J. G., and Stokoe, K. H., II (1984). "Analytical Evaluation of Liquefaction

Potential of Sands Subjected to the 1981 Westmorland Earthquake," Geotechnical Engineering
Report GR-84-15, University of Texas at Austin, 231 p.

Boulanger, R. W., Idriss, I. M., and Mejia, L. H. (1995). "Investigation and Evaluation of
Liquefaction Related Ground Displacements at Moss Landing During the 1989 Loma Prieta
Earthquake," Report No. UCD/CGM-95/02, University of California at Davis.

Boulanger, R. W., Mejia, L. H., and Idriss, I. M. (1997). "Liquefaction at Moss Landing During
Loma Prieta Earthquake, Journal of Geotechnical and Geoenviromental Engineering, ASCE,
11

Vol. 123, No. 5, pp. 453-467.

Dobry, R. (1996). Personal communication to R. D. Andrus.

Dobry, R., Stokoe, K. H., II, Ladd, R. S., and Youd, T. L. (1981). "Liquefaction Susceptibility
from S-Wave Velocity," Proc., In Situ Testing to Evaluate Liquefaction Susceptibility, ASCE
National Convention, held in St. Louis, MO.

Dobry, R., Baziar, M. H., O'Rourke, T. D., Roth, B. L., and Youd, T. L. (1992). "Liquefaction
and Ground Failure in the Imperial Valley, Southern California During the 1979, 1981 and 1987
Earthquakes," Case Studies of Liquefaction and Lifeline Performance During Past Earthquakes,
Technical Report NCEER-92-0002, T. O'Rourke and M. Hamada, Eds., National Center for
Earthquake Engineering Research, Buffalo, NY, Vol. 2.

Dobry, R., Ladd, R. S., Yokel, F. Y., Chung, R. M., and Powell, D. (1982). "Prediction of Pore
Water Pressure Buildup and Liquefaction of Sands During Earthquakes by the Cyclic Strain
Method," NBS Building Science Series 138, U.S. Department of Commerce, National Bureau of
Standards, Gaithersburg, MD, 152 p.

EPRI (1992). Lotung Large-Scale Seismic Test Strong Motion Records, EPRI NP-7496L,
Electric Power Research Institute, Palo Alto, California, Vols. 1-7.

Fuhriman, M. D. (1993). "Crosshole Seismic Tests at Two Northern California Sites Affected
by the 1989 Loma Prieta Earthquake," M.S. Thesis, University of Texas at Austin, 516 p.

Geomatrix (1990). "Results of Field Exploration and Laboratory Testing Program for Perimeter
Dike Stability Evaluation Naval Station Treasure Island San Francisco, California, Project No.
1539.05, Vol. 2.

123
 Gibbs, J. F., Fumal, T. E., Boore, D. M., and Joyner, W. B. (1992). "Seismic Velocities and
Geologic Logs from Borehole Measurements at Seven Strong-Motion Stations that Recorded the
Loma Prieta Earthquake," Open-File Report 92-287, U.S. Geological Survey, 139 p.

Hardin, B. 0., and Drnevich, V. P. (1972). "Shear Modulus and Damping in Soils: Design
Equations and Curves," Journal of the Soil Mechanics and Foundation Division, ASCE, New
York, NY, Vol. 98, SM7, pp. 667-692.

Hryciw, R. D. (1991). "Post Loma Prieta Earthquake CPT, DMT and Shear Wave Velocity
Investigations of Liquefaction Sites in Santa Cruz and on Treasure Island," Final Report to the
U.S. Geological Survey, Award No. 14-08-000I-Gl865, University ofMichigan, 68 p.

Hryciw, R. D., Rollins, K. M., Homolka, M., Shewbridge, S. E., and McHood, M. (1991). "Soil
Amplification at Treasure Island During the Loma Prieta Earthquake," Proceedings, Second
International Conference on Recent Advances in Geotechnical Earthquake Engineering and Soil
Dynamics, held in St. Louis, Missouri, S. Prakash, Ed., University of Missouri-Rolla, Vol. II,
pp. 1679-1685.

Idriss, I. M. (1996). Personal communication to T. L. Youd.

Ishihara, K., Shimizu, K., and Yamada, Y. (1981). "Pore Water Pressures Measured in Sand
· Deposits During an Earthquake," Soils and Foundations, Japanese Society of Soil Mechanics
and Foundation Engineering, Vol. 21, No. 4, pp. 85-100.

Ishihara, K., Anazawa, Y., and Kuwano, J. (1987). "Pore Water Pressures and Ground Motions
Monitored During the 1985 Chiba-Ibaragi Earthquake," Soils and Foundations, Japanese ~ociety
of Soil Mechanics and Foundation Engineering, Vol. 27, No. 3, pp. 13-30.

Ishihara, K., Muroi, T., and Towhata, I. (1989). "In-situ Pore Water Pressures and Ground
Motions Monitored During the 1985 Chiba-Ibaragi Earthquake," Soils and Foundations,
Japanese Society of Soil Mechanics and Foundation Engineering, Vol. 29, No. 4, pp. 75-90.

Kayen, R. E., Liu, H. -P., Fumal, T. E., Westerland, R. E., Warrick, R. E., Gibbs, J. F., and Lee,
H.J. (1990). "Engineering and Seismic Properties of the Soil Column at Winfield Scott School,
San Francisco," Effects of the Loma Prieta Earthquake on the Marina District San Francisco,
California, Open-file Report 90-253, U.S. Geological Survey, pp. 112-129.

Kayen, R. E., Mitchell, J. K., Seed, R. B., Lodge, A., Nishio, S., and Coutinho, R. (1992).
"Evaluation of SPT-, CPT-, and Shear Wave-Based Methods for Liquefaction Potential
Assessment Using Loma Prieta Data," Proceedings, Fourth Japan-U.S. Workshop on Earthquake
Resistant Design of Lifeline Facilities and Countermeasures for Soil Liquefaction, Technical
Report NCEER-92-0019, held in Honolulu, Hawaii, M. Hamada and T. D. O'Rourke, Eds.,
National Center for Earthquake Engineering Research, Buffalo, NY, Vol. 1, pp. 177-204.

124
Kokusho, T., Sato, K., and Matsumoto, M. (1995a). "Nonlinear Seismic Amplification of Soil
Ground During 1995 Hyogoken-Nanbu Earthquake," Proceedings, Fifth International
Conference on Seismic Zonation, held in Nice, France, Presses Academiqes, Vol. II, pp. 1603-
1610.

Kokusho, T., Tanaka, Y., Kudo, K., and Kawai, T. (1995b). "Liquefaction Case Study of
Volcanic Gravel Layer during 1993 Hokkaido-Nansei-Oki Earthquake," Proceedings, Third
International Conference on Recent Advances in Geotechnical Earthquake Engineering and Soil
Dynamics, held in St. Louis, Missouri, S. Prakash, Ed., University of Missouri-Rolla, Vol. I, pp.
235-242.

Kokusho, T., Yoshida, Y., and Tanaka, Y. (1995c). "Shear Wave Velocity in Gravelly Soils
with Different Particle Gradings," Proceedings, Static and Dynamic Properties of Gravelly Soils,
Geotechnical Special Publication No. 56, M. D. Evans and R. J. Fragaszy, Eds., ASCE, pp. 92-
106.

Liao, S.S. C., and Whitman, R. V. (1986). "A Catalogue of Liquefaction and Non-liquefaction
Occurrences During Earthquakes," Research Report, Dept. of Civil Engineering, M.I.T.,
Cambridge, MA.

Ladd, R. S. (1982). "Geotechnical Laboratory Testing Program for Study and Evaluation of
Liquefaction Ground Failure Using Stress and Strain Approaches: Heber Road Site, October 15,
1979 Imperial Valley Earthquake," Woodward-Clyde Consultants, Wayne, New Jersey,
February.

Lodge, A. L. (1994). "Shear Wave Velocity Measurements for Subsurface Characterization,"

Ph.D. Dissertatio11, University of California at Berkeley.

Mitchell, J. K., Lodge, A. L., Coutinho, R. Q., Kayen, R. E., Seed, R. B., Nishio, S., and Stakoe,
K. H., II (1994). "Insitu Test Results from Four Loma Prieta Earthquake Liquefaction Sites:
SPT, CPT, DMT, and Shear Wave Velocity," Report No. UCB/EERC-94/04, Earthquake
Engineering Research Center, University of California at Berkeley, 171 p.

Ohta, Y., and Goto, N. (1976). "Estimation of S-Wave Velocity in Terms of Characteristic
Indices of Soil," Butsuri-Tanko (Geophysical Exploration), Vol. 24, No. 4, pp. 34-41 (in
Japanese).

Redpath, B. B. (1991). "Seismic Velocity Logging in the San Francisco Bay Area," Report to
the Electric Power Research Institute, Palo Alto, California, 34 p.

Robertson, P. K., Woeller, D. J., and Finn, W. D. L. (1992). "Seismic Cone Penetration Test for
Evaluating Liquefaction Potential Under Cyclic Loading," Canadian Geotechnical Journal,
Ottawa, Canada, Vol. 29, pp. 686-695.

125
Rollins, K. M., McHood, M. D., Hryciw, R. D., Homolka, M., and Shewbridge, S. E. (1994).
"Ground Response on Treasure Island," The Loma Prieta, California, Earthquake of October 17,
1989--Strong Ground Motion, U.S. Geological Survey Professional Paper 1551-A, R. D.
Borcherdt, Ed., United States Government Printing Office, Washington, pp. A109-A121.

Roy, D., Campanella, R. G., Byrne, P. M., and Hughes, J. M. 0. (1996). "Strain Level and
Uncertainty of Liquefaction Related Index Tests," Proceedings, Uncertainty in the Geologic
Environment: From Theory to Practice, Geotechnical Special Publication No. 58, held in
Madison, Wisconsin, C. D. Shackelford, P. P. Nelson, and M. J. S. Roth, Eds., ASCE, Vol. 2,
pp. 1149-1162.

Sato, K., Kokusho, T., Matsumoto, M., and Yamada, E. (1996). ''Nonlinear Seismic Response
and Soil Property During Strong Motion," Soils and Foundations, Special Issue on the 1995
Hyogoken-Nambu Earthquake, Japanese Geotechnical Society, January, pp. 41-52.

Schnabel, P. B., Lysmer, J., and Seed, H. B. (1972). "SHAKE: A Computer Program for
Earthquake Response Analysis of Horizontally Layered Sites," Earthquake Engineering
Research Center Report No. UCB/EERC-72-12, University of California, Berkeley.

Seed, H. B. (1979). "Soil Liquefaction and Cyclic Mobility Evaluation for Level Ground
During Earthquakes," Journal of the Geotechnical Engineering Division, ASCE, New York, NY,
Vol. 105, GT2, pp. 201-255.

Seed, H. B., and Idriss, I. M. (1971). "Simplified Procedure for Evaluating Soil Liquefaction
Potential," Journal of the Soil Mechanics and Foundation Division, ASCE, New York, NY, Vol.
97,SM9,pp. 1249-1273.

Seed, H. B., and Idriss, I. M. (1982). Ground Motions and Soil Liquefaction During
Earthquakes, monograph series, Earthquake Engineering Research Institute, Berkeley,
California, 134 p.

Seed, H. B., Idriss, I. M., and Arango, I. (1983). "Evaluation of Liquefaction Potential Using
Field Performance Data," Journal of Geotechnical Engineering Divisio!!, ASCE, New York,
NY, Vol. 109, No. 3, pp. 458-482.

Seed, H. B., Tokimatsu, K., Harder, L. F., and Chung, R. M. (1985). "Influence of SPT
Procedures in Soil Liquefaction Resistance Evaluations," Journal of Geotechnical Engineering
Division, ASCE, Vol. 111, No. 12, pp. 1425-1445.

Shen, C. K., Li, X. S., and Wang, Z. (1991). "Pore Pressure Response During 1986 Lotung
Earthquakes," Proceedings, Second International Conference on Recent Advances in
Geotechnical Earthquake Engineering and Soil Dynamics, held in St. Louis, Missouri, S.
Prakash, Ed., University of Missouri-Rolla, Rolla, Mo., Vol. I, pp. 557-563.

126
Shibata, T., Oka, F., and Ozawa, Y. (1996). "Characteristics of Ground Deformation Due to
Liquefaction," Soils and Foundations, Special Issue on the 1995 Hyogoken-Narnbu Earthquake,
Japanese Geotechnical Society, January, pp. 65-79.

Stokoe, K. H., II, Andrus, R. D., Bay, J. A., Fuhriman, M. D., Lee, N. J., and Yang, Y. (1992).
"SASW and Crosshole Seismic Test Results from Sites that Did and Did not Liquefy During the
1989 Loma Prieta, California Earthquake," Geotechnical Engineering Center, Department of
Civil Engineering, University of Texas at Austin.

Stokoe, K. H., II, and Nazarian, S. (1985). "Use of Rayleigh Waves in Liquefaction Studies,"
Proceedings, Measurement and Use of Shear Wave Velocity for Evaluating Dynamic Soil
Properties, held in Denver, Colorado, R. D. Woods, Ed., ASCE, New York, NY, pp. 1-17.

Stokoe, K. H., II, Andrus, R. D., Rix, G. J., Sanchez-Salinero, I., Sheu, J. C., and Mok, Y. J.
(1988a). "Field Investigation of Gravelly Soils Which Did and Did Not Liquefy During the
1983 Borah Peak, Idaho, Earthquake," Geotechnical Engineering Center Report GR 87-1,
University of Texas at Austin, 206 p.

Stokoe, K. H., II, Roesset, J.M., Bierschwale, J. G., and Aouad, M. (1988b). "Liquefaction
Potential of Sands from Shear Wave Velocity," Proceedings, Ninth World Conference on
Earthquake Engineering, held in Tokyo, Japan, Vol. III, pp. 213-218.

Sugito, M., Sekiguchi, K., Oka, F., and Yashima, A. (1996). "Analysis of Borehole Array
Records from the South Hyogo Earthquake of Jan. 17, 1995," Proceedings of the International
Workshop on Site Response Subjected to Strong Earthquake Motions, held in Yokosuka, Japan
on January 15-17, Port and Harbor Research Institute, Yokosuka, Japan, Vol. 2, pp. 343-357.

Sykora, D. W., and Stokoe, K. H., (1982), "Seismic Investigation of Three Heber Road Sites
After the Oct. 15, 1979 Imperial Valley Earthquake," Geotechnical Engineering Report GR82-
24, University of Texas at Austin.

Tokimatsu, K., Kuwayama, S., and Tamura, S. (1991a). "Liquefaction Potential Evaluation
Based on Rayleigh Wave Investigation and Its Comparison with Field Behavior," Proceedings,
Second International Conference on Recent Advances in Geotechnical Earthquake Engineering
and Soil Dynamics, held in St. Louis, Missouri, S. Prakash, Ed., University of Missouri-Rolla,
Rolla, Mo., Vol. I, pp. 357-364.

Tokimatsu, K., Kuwayma, S., Abe, A., Nomura, S., and Tamura, S. (1991b). "Considerations to
Damage Patterns in the Marina District During Loma Prieta Earthquake Based on Rayleigh
Wave Investigation," Proceedings, Second International Conference on Recent Advances in
Geotechnical Earthquake Engineering and Soil Dynamics, held in St. Louis, Missouri, S.
Prakash, Ed., University of Missouri-Rolla, Vol. II, pp. 1649-1654.

127
Turner, E. and Stokoe, K. H., II (1992). "Static and Dynamic Properties of Clayey Soils
Subjected to the 1979 Imperial Valley Earthquake," Geotechnical Engineering Report GR82-26,
University of Texas at Austin, 208 p.

Youd, T. L., and Noble, S. K. (in press). "Magnitude Scaling Factors," Proceedings, NCEER
Workshop on Evaluation of Liguefaction Resistance, held in Salt Lake City, Utah, T. L. Youd
and I. M. Idriss, Eds., National Center for Earthquake Engineering Research, Buffalo, NY.

Youd, T. L., and Bennett, M. J. (1983). "Liquefaction Sites, Imperial Valley, California,"
Journal ofGeotechnical Engineering Division, ASCE, Vol. 109, No. 3, pp. 440-457.

Youd, T. L., and Hoose, S. N. (1978). "Historic Ground Failures in Northern California
Triggered by Earthquakes," U.S. Geological Survey Professional Paper 993, 177 p.

128
 Application of the Becker Penetration Test for Evaluating
the Liquefaction Potential of Gravelly Soils

Leslie F. Harder, Jr.

Chief, Division of Engineering

California Department of Water Resources
Sacramento, California 94236-000 I

Abstract

The Becker Penetration Test has seen increased use in North America as an in situ technique for
measuring the liquefaction potential of gravelly soils. This test involves the use of a large dynamic
penetrometer and incorporates elements of both pile driving and the Standard Penetration Test.
Despite its increasing use, it remains a non-standard test with many uncertainties involving driving
energy and casing friction. This paper describes the development of the test, its application, and
recommendation for its use in the assessment ofliquefaction potential.

129
 INTRODUCTION

The liquefaction potential of sandy and silty soils in situ is generally evaluated with the use of either
the Standard Penetration Test (SPT) or the Cone Penetrometer Test (CPT). However, soils with
substantial gravel contents cannot be reliably evaluated with these tools because the gravel particles
are large relative to the effective size of the penetrometers. The inevitable result is that these tools
will give erroneously high penetration resistances because the gravel particles disrupt the normal
displacement of the soil by the penetrometer. Further, the drive shoe of a SPT sampler may become
blocked by a gravel particle and driven as a solid penetrometer rather than as a hollow penetrometer.

The difficulties associated with drilling and testing gravelly soils with conventional penetrometers
has led to the use oflarger penetrometers. Large dynamic penetrometers have been used both in Italy
and Japan (Jamiolkowski, et al.; Tokimatsu, 1988). In North America, the Becker Penetration Test
has seen increased use for the evaluation of gravelly deposits. The Becker Penetration Test was
developed in Canada in the late 1950's and is now widely used for exploring the characteristics of
deposits containing gravel and cobble-size particles. The test consists of driving a double-walled
casing into the ground with a double-acting diesel pile hammer. The general approach is to count
the number of diesel hammer blows per 30 centimeters of penetration. Although similar in nature
to the SPT, the Becker Penetration Test is a continuous sounding method. Typically, a 3-meter
length of casing is driven into the ground and the blowcounts are recorded. After the initial casing
length is driven into the ground, another section of casing is threaded onto the end of the previous
casing, and driving is resumed. The process is continued until the desired depth is reached and the
casing is jacked out of the ground.

In recent years, the test has been used by several investigators to evaluate the equivalent relative
density and/or the liquefaction potential of gravelly soils. The general approach is to determine
Becker Penetration resistance in the field and then convert it into an equivalent SPT resistance to
evaluate future performance of the gravelly soil. Although the Becker Penetration Test has been
used successfully in this manner for many projects, it remains a non-standard test that is used with
a multitude of different types of equipment and procedures. The test continues to need improved
standardization and interpretation in the following main areas:

• Elimination of the use of open-bit Becker soundings for evaluation of penetration

resistance.

• Standardization of driving and penetration equipment.

• Determination/Interpretation of diesel hammer driving energy.

• Evaluation/Interpretation of the effect of casing friction on penetration resistance.

These subjects are described further in the following sections.

130
 OPEN-BIT BECKER SOUNDINGS

The original intent of the Becker apparatus was to rapidly obtain samples of gravelly material. This
is accomplished by driving with an open bit and using reverse air circulation in the annular space
between the double-walled casing. During driving, air is forced down the annulus of the casing
system to the drive bit. Soil particles entering the bit are then transported up the inner casing to the
surface by the air flow and are then collected in a cyclone for examination or testing (see Figure 1).

Company literature from Becker Drills, Inc. and others have recommended that reliable penetration
resistance should be obtained with a plugged drive bit, and that sampling should be obtained with
a separate sounding with an open-bit and reverse air circulation. However, this approach requires
additional soundings, and many firms have in the past opted to use only open-bit soundings to obtain
both penetration resistance and sampling information. This latter approach is likely to result in
excessively low ( overly conservative) values of penetration resistance. Harder and Seed (1986)
documented the fact that the air recirculation process commonly loosens and removes material
ahead of the bit, particularly in saturated sandy material. This then results in umeasonably low
blowcount data, even for very dense soils.

The study by Harder and Seed (1986) also recommended that penetration and sampling information
be obtained in separate plugged-bit and open-bit soundings. Although this recommendation
currently appears to be generally followed, some investigators still try to obtain both types of data
with open-bit soundings alone. In addition, some investigators will stop the driving of an open-bit
sounding and then complete SPT tests through the Becker casing within sandy layers that have been
encountered. As documented by Harder and Seed (1986), this also commonly results in
umeasonably low blowcount data because the soil ahead of the Becker bit has been disturbed by the
air recirculation sampling process performed earlier. The only way this SPT sampling approach
would work successfully is if the disturbed material ahead of the bit was removed by mud rotary
drilling performed through the Becker casing. This would probably require drilling and removing
material for a distance ofat least 2 to 3 meters ahead of the bit before performing the SPT test.

STANDARDIZATION OF DRIVING AND PENETRATION EQUIPMENT

Becker Penetration Tests continue to be performed with a variety of different types of driving and
penetration equipment:

• Drill rigs: Older HAV-180 or B-180 rigs

Newer AP-1000 rigs*

• Superchargers: Some diesel hammers have superchargers*

Some diesel hammers do not have superchargers

131
 • Casing sizes: 140-mm (5.5-inch) O.D.
168-mm (6.6-inch) O.D.*
229-mm (9.0-inch) O.D.

• Drive bit: Crowd-in open bit

Crowd-out open bit
Crowd-in plugged bit
Crowd-out plugged bit*

* Recommended by Harder and Seed (1986)

Local correlations can generally be made between Becker and SPT blowcounts with most of the
above equipment combinations, except with open bits as noted previously. However, different sets
of equipment types will result in different penetration resistances being determined for the same
deposit. The correlation between Becker and SPT blowcounts published by Harder and Seed (1986)
is intended for a plugged, crowd-out bit, 168-mm O.D. casing, and driven with an AP-1000 drill
rig. The original Harder and Seed (1986) correlation has been supplemented with data from other
projects and appears to remain the most practical correlation available (see Figure 2). It is also the
only correlation that has been used to evaluate actual gravel deposits which have liquefied during
earthquake shaking. Consequently, even with the use oflocal correlations it is desirable to use the
Harder and Seed (1986) correlation as a check on the information being gathered. Therefore, the
above recommended set of equipment types should be used essentially as a standard to perform most
investigations if the Harder and Seed (1986) correlation is to be used.

A second choice would be to use the same set of equipment, but with a HAV-180 drill rig. Several
comparison tests have been performed to show that this older style of drill rig is significantly more
efficient in allowing the hammer energy to be transmitted to the drive casing. This difference is
thought to result from the different manner by which the diesel hammer is mounted to the mast of
the different types of drill rigs. Comparison tests generally indicate that corrected Becker
blowcounts obtained using a HAV-180 drill rig should be multiplied by a 1.5 correction factor to be
converted into equivalent AP-1000 drill rig blowcounts. Several investigations have been
successfully conducted with HAV-180 drill rigs using this 1.5 correction factor.

DETERMINATION/EVALUATION OF DIESEL HAMMER DRIVING ENERGY

Constant energy conditions are not a feature of the double-acting diesel hammers used in the Becker
Penetration Test. One reason for this is that the energy is dependent upon combustion conditions.
Thus, anything that affects combustion, such as fuel quantity, fuel quality, air mixture and pressure
all have a significant effect on the energy produced. Combustion efficiency is also operator-
dependent because the operator controls a variable throttle which affects how much fuel is injected
for combustion. On some rigs, the operator also controls a rotary blower, or supercharger, which
adds additional air to the combustion cylinder during each stroke. This additional air is thought to

132
better scavenge the cylinder of burnt combustion gases and has been found to produce higher
energies.

To monitor the level of energy produced by the diesel hammer during driving, use has been made
of the bounce chamber pressure. For the ICE Model 180 diesel hammers used on the Becker drill
rigs, the top of the hammer is closed off to allow a smaller stroke and a faster driving rate. At the
top, air trapped in the compression cylinder and a connected bounce chamber acts as a spring. The
amount of potential energy within the ram at the top of its stroke can be estimated by measuring the
peak pressure induced in the bounce chamber.

The studies by Harder and Seed (1986) made use of the bounce chamber to monitor combustion
efficiencies. They did not attempt to directly measure the actual energy transferred to the casing
during driving. Rather, they found that constant combustion conditions (e.g. full throttle with a
supercharger) resulted in a relatively unique relationship between bounce chamber pressure and
Becker blowcount. This relationship took the form of a curve and was designated a constant
combustion rating curve. The Harder and Seed (1986) studies showed that different combustion
conditions (e.g. reduced fuel or no supercharger) resulted in different Becker blowcounts and
different constant combustion curves. Presented in schematic form in Figure 3 are typical results
obtained for different combustion efficiencies. In the upper plot, three combustion rating curves
representing three different combustion efficiencies are shown. With different combustion
conditions, the resulting blowcounts from tests performed in the same materials can be radically
different. Consequently, tests in the same material at a depth of 40 feet can give a Becker blowcount
of 14 when the hammer is operated at high combustion efficiency, but give blowcounts of26 and
50 at succeeding reductions of combustion energy.

To account for variable combustion effects, Harder and Seed (1986) adopted a standard constant
combustion rating curve designated Curve AA (see Figure 4). They also developed correction
curves for correcting data with low combustion efficiencies to this standard rating curve. Becker
penetration resistance corrected to the standard rating curve was designated as the corrected Becker
blowcount, NBc· To use the correction curves, it is simply necessary to locate each uncorrected test
result on the chart using both the uncorrected blowcount and the bounce chamber pressure, and then
follow the correction curves down to the standard rating curve AA, to obtain the corrected Becker
blowcount. For example, if the uncorrected Becker blowcount was 43 and it was obtained at sea
level with a bounce chamber pressure of 18 psig, then the corrected Becker blowcount would be 30
(see Figure 4). Harder and Seed (1986) developed their SPT-Becker correlation by comparing
corrected SPT blowcounts, N 60, in sandy and silty soils to the corrected Becker blowcount, NBc, in
adjacent borings and soundings (see Figure 2).

The standard constant combustion curve in Figure 4 is for sea level atmospheric pressures. To obtain
equivalent sea level bounce chamber pressures, it is necessary to increase the measured bounce
chamber pressures. Bounce chamber pressures measured at approximately 2,000 feet above sea level
must be increased typically about 1.5 to 2 psi. Bounce chamber pressures measured at approximately
6,000 feet above sea level must be increased typically about 4 to 6 psi.

133
Several researchers have advocated the use of instrumentation and wave equation techniques
developed for pile driving in an effort to better quantify the effects of variable hammer energy. Such
an approach is very attractive as it would theoretically capture variations in both hammer energy and
energy transmission through the hammer anvil to the casing. To date, the most notable of these
efforts are those by Sy and Campanella (1994, 1995). In their studies, Sy and Campanella employed
a small length of Becker casing fitted with strain gages and accelerometers together with a "Pile-
Driving Analyzer" to determine strain, force, acceleration, and velocity. The transferred energy is
determined by time integration of force times velocity. Measured blowcounts were corrected by Sy
and Campanella to a reference energy value equal to 30 percent of the theoretical rated energy of the
ICE 180 hammer. The energy corrected blowcount was designated as Nb30 .

In their studies, Sy and Campanella were able to verify many of the variations in hammer energy
previously identified by Harder and Seed (1986). These include the effect of variable throttle
settings and the different energy transmission efficiencies of the HAV-180 and AP- I 000 drill rigs.
However, they were unable to improve or reduce the scatter associated with correlating corrected
Becker and SPT blowcounts. Shown in Figure 5 is a comparison between Becker and SPT
blowcounts using data obtained from the Richmond and Annacis Test Sites (after Sy and
Campanella, 1994). In both cases the level of scatter is comparable. However, it should be noted
that, as observed by these researchers, the Harder and Seed (I 986) correlation developed for other
sites fit well through the middle of the data when applied with the bounce chamber pressure
correlation. This was true for SPT data obtained at depths extending down as far as 24 and 42
meters (Richmond and Annacis, respectively).

It is believed that such instrumentation and wave equation techniques will eventually be used with
established correlations on a routine basis. However, this approach currently has had limited
verification in the field and it requires significantly more time, cost, effort, and interpretation than
the bounce chamber technique. For most investigations, therefore, there appears to be little benefit
for the increased effort over the proper application of the Harder and Seed (1986) bounce pressure
approach. However, both approaches point to the critical need to evaluate the hammer energy in at
least some form during the evaluation of penetration resistance.

EVALUATION/INTERPRETATION OF THE EFFECT OF CASING FRICTION

The studies by Harder and Seed (1986) did not specifically address the effect of casing friction on
the determined Becker blowcount. However, it is an important consideration because, unlike the
SPT, the Becker penetration resistance is determined by both the tip resistance and any friction acting
on the casing along the entire depth for which testing is being performed. During the studies
performed by Harder and Seed, there was concern that casing friction would mask any loose, or low
blowcount, gravelly soil layers. However, the initial studies found that low blowcount sand and
gravel layers could be identified at depth by the Becker Penetration Test. Consequently, the initial
concerns were somewhat abated and casing friction was considered to have an almost negligible
impact for routine investigations to depths of about 30 meters.

134
During the last ten years, various investigations using the Harder and Seed approach have been able
to identify loose sand and/or gravel layers both at depth and below layers of denser material. This
has given continued support for the approach. However, casing friction still remains a concern in
special circumstances, particularly involving great depths and/or the investigation of very soft soils
underlying very thick deposits of very dense material. Recent investigations have identified some
important results involving Becker casing friction:

1. A certain amount of casing friction is built into the Harder and Seed approach. Recent
studies have shown that:

a. There are now several test sites (e.g. Seymour Falls Dam, Duncan Dam, and
Mormon Island Auxiliary Dam) where Becker Penetration Tests have been
performed through pre-drilled and cased borings. When a significant portion of
the depth is pre-drilled and cased, the cased Becker penetration resistance is
significantly less than the uncased penetration resistance. In addition, the cased
Becker penetration resistance used with the Harder and Seed approach generally
under predicts measured values of SPT resistance of sandy soils at depth.
Accordingly, the use of pre-drilled holes and casing to remove excessive friction
is overly conservative and counter productive.

b. Recent studies by Yan and Wightman (1992) have involved a mudded Becker
sounding where bentonite drill mud is injected out of the casing just above the
drive bit during driving. This technique also uses a modified drive bit that has
a larger diameter than the casing to further reduce friction. The resulting Becker
penetration resistance yields substantially lower blowcounts, commonly only 30
percent of the penetration resistance developed using conventional driving
techniques (see Figure 6). This type of modified equipment may have great
potential for sounding deep deposits. However, it would require a different
correlation than those developed by Harder and Seed (1986). It may also have
a depth limitation in that mudded Becker soundings may relieve too much of the
casing friction and lead to conservatively low blowcounts at depth. As shown
in the right plot in Figure 6, a traditional Becker sounding tends to give higher
ratios of Becker blowcounts to SPT blowcounts with depth, indicating that
casing friction may be significant. For the mudded Becker sounding, however,
the ratio of Becker blowcount to SPT blowcount is approximately constant until
about a 30-meter depth, and then it decreases significantly. It may be that the
unsupported weight of the casing lengths and the diesel hammer significantly
reduces penetration resistance when not supported by casing friction.

2. Several investigations have shown the ability of the Becker Penetration Test to identify
loose sandy and/or gravelly deposits at depth. However, there are three sites (Jackson
Lake Dam - Site A, McDonald's Farm, and Annacis) where the Becker Penetration
Test used with the Harder and Seed approach missed the presence of soft silt deposits
at depth. In all three cases the soft silt had SPT blowcounts of less than 5 and the silt

135
 layer was at a depth of between 10 and 45 meters underlying moderately dense sand.
Corrected Becker blowcounts were typically three to eight times higher at these
particular depth intervals.

Recent studies by Sy and Campanella (1994, 1995) have attempted to directly address the effect of
casing fiiction. In addition to correcting for energy content, Sy and Campanella perform CAPWAP
analyses to separately estimate casing fiiction, Rs, and tip resistance. CAPWAP is a computer
program which uses the force and velocity traces obtained with the Pile Driving Analyzer to estimate
loading conditions and resistances through a trial and error process. As described by Lum and Yan
(1994), this process can be time consuming and expensive. Consequently, CAPWAP analyses have
generally been performed only at three or four depths for any Becker sounding. The casing fiiction
determined at these depths is then interpolated and extrapolated for the rest of the sounding profile.
Sy and Campanella produced a correlation between Becker and SPT blowcounts which uses both
the energy corrected Becker blowcount and the estimated shaft resistance (see Figure 7). This
correlation was developed using CAPWAP analyses of data obtained at the Annacis Test Site.

The Sy and Campanella correlation has been compared to the Harder and Seed correlation at a few
projects. Figure 8 presents comparisons developed by Lum and Yan (1994) for three soundings
performed within gravelly soil at Keenlyside Dam. As may be observed in the figure, the two
correlations give similar equivalent SPT blowcounts for much of each profile. However, as noted
by Lum and Yan (1994), the Sy and Campanella correlation yields much more variability. In
addition, it may be seen that the Sy and Campanella correlation predicts extremely high penetration
resistance at some depth intervals. To provide support for their correlation, Sy and Campanella
compared equivalent Becker and actual SPT N 60 values for one of their correlation sites, Annacis.
As shown in Figure 9, the Sy and Campanella correlation compared reasonably well for the Annacis
site down to depths of about 42 meters, not surprising since this was the site used to develop the
correlation. Also shown in Figure 9 is a comparison for the same site, only using the Harder and
Seed approach to interpret the Becker soundings. As may be observed, the Harder and Seed
correlation produced predicted SPT blowcounts that correlated with measured SPT resistance just
as well as the Sy and Campanella correlation. The Sy and Campanella work published to date does
not appear to attempt to predict the presence of soft silt layers at depth.

As noted previously, the "mudded" casing approach may provide a way to eliminate the effects of
casing friction. In this approach, it may not be necessary to account for friction effects, either by
explicit or implicit means, as the casing friction is nearly eliminated (see Figure 6). The effects of
drill rig type and hammer energy might be determined with a few site specific calibrations between
bounce chamber pressure and a limited number of measurements of hammer energy delivered to the
casing. However, the time consuming and expensive CAPWAP analyses would not be necessary
as friction is essentially eliminated. In this way, a simple and inexpensive approach might be very
successful in determining equivalent SPT blowcounts at large depths beneath even very dense
material. More research and correlations need to be developed and published for practitioners to
evaluate the potential of this approach. One concern is that the weight of the casing and hammer will
result in conservatively low penetration resistance at depth in a mudded sounding due to the

136
elimination of casing friction. As shown in Figure 6, the ratio of mudded Becker blowcounts to SPT
blowcounts decreases with depth. This issue needs to be addressed in future developments of this
approach. Nevertheless, it is a very promising direction for future improvements in the Becker
Penetration Test.

RECOMMENDATIONS

The Becker Penetration Test remains a non-standard test with results subject to interpretation. Many
uncertainties will eventually be cleared up with the use of new equipment and approaches. However,
for current use, the following procedures are recommended:

I. The use of the Becker Penetration Test should be carried out with plugged-bit
soundings in order to avoid an overly conservative evaluation of subsurface deposits.

2. In order to avoid the use of several correction factors, it is recommended that the
Becker Penetration Test be performed with the following set of equipment:

• AP-1000 Drill Rig

• Plugged 168-mm O.D. Drive Bit and Casing

3. It is necessary to monitor the efficiency/performance of the diesel hammer during

driving. This can be done using the bounce chamber pressure with the Harder and
Seed method, or may be performed using more sophisticated instrumentation similar
to that used by Sy and Campanella. The Sy and Campanella approach provides insight
on the tip and casing friction elements of Becker penetration resistance. However, for
most investigations where depths are less than about 30 meters, the simpler Harder
and Seed approach is probably warranted because of the greater data base together with
its ease of implementation and lower cost.

4. Casing friction will remain a concern until either different equipment or approaches
are developed to make the determined penetration resistance independent of casing
length. For most investigations, the Harder and Seed approach with friction effects
implicitly incorporated will probably be adequate. However, for depths greater than
30 meters and/or for sites with thick deposits of very dense material overlying much
looser material, more sophisticated approaches involving wave equation techniques
may be necessary. Unreasonably low penetration resistance will be determined with
the Harder and Seed approach if Becker soundings are performed through cased
boreholes to reduce friction effects.

5. The Sy and Campanella studies have shown that wave equation techniques were able
to confirm results previously produced by Harder and Seed (1986) using less
sophisticated methods. This research has shown that the wave equation approach may

137
 eventually produce better correlations for predicting the performance of gravelly
materials. Hopefully, such methods \'/ill accurately account for casing friction in all
cases, including conditions where soft silt layers are present at depth below much
denser sand and gravel deposits. However, the Sy and Campanella approach can
predict extremely high equivalent SPT blowcounts when the Becker blowcounts are
only moderately high. This trend needs to be examined further and confirmed \1./ith
comparisons of actual SPT blowcounts. The Sy and Campanella work, however,
shows that this approach is promising and that it should be pursued further.

6. Additional research into the use of mudded Becker soundings with and without wave
equation techniques should be performed with the goal of developing a sounding tool
that could be efficiently used for all conditions regardless of friction concerns. There
may be a concern that the use of the mudded technique relieves too much friction and
that excessively low (overly conservative) penetration resistance may be predicted at
large depths.

7. For all projects involving the Becker Penetration Test to investigate gravelly deposits,
it is recommended that a local correlation or check be performed either at the project
site or nearby. This local check would consist of performing Becker soundings in
sandy material near the depth of interest and to also perform high quality SPT tests in
the same layer. In this way, a check on the applicability of the Becker equipment and
correlations may be made.

8. Several investigations have indicated that the Becker Penetration Test may not detect
the presence of a very soft silt layer at depth. If such layers are thought to be present
and of concern for the project, it is recommended that other investigative techniques
(e.g. SPT) be carried out to explore and characterize such materials.

ACKNOWLEDGEMENTS

The investigations performed for the Harder and Seed (1986) studies were carried out under the
direction of the late H. Bolton Seed. Dr. Seed's leadership, review, and guidance during the phases
of these studies were instrumental in their successful completion. Many of the recent investigations
associated \1./ith the Becker Penetration Test were sponsored by the British Columbia Hydroelectric
Power Authority (BC Hydro). Permission to publish this data and sponsorship of these studies is
gratefully acknowledged. In this regard, the assistance of Mr. Al Imrie is particularly appreciated.

138
 REFERENCES

1. British Columbia Hydroelectric Power Authority (BC Hydro), Uncorrected Becker Data
obtained at FMC, McDonald Farm, and Duncan Dam Test Sites.

2. Harder, L. F. (1988) "Use of Penetration Tests to Determine the Cyclic Load Resistance of
Gravelly Soils," Dissertation submitted as partial satisfaction for the degree of Doctor of
Philosophy, University of California, Berkeley.

3. Harder, L. F. Jr. and Seed, H. Bolton (1986), "Determination of Penetration Resistance for
Coarse-Grained Soils using the Becker Hammer Drill," Earthquake Engineering Research
Center, Report No. UCB/EERC-86/06, University of California, Berkeley.

4. Lum, Ken K. Y. and Yan, Li (1994) "In-situ Measurements of Dynamic Soil Properties and
Liquefaction Resistances of Gravelly Soils at Keenleyside Dam," Proceedings of Specialty
Session on Ground Failures under Seismic Conditions, ASCE Annual Convention, Atlanta,
GA, October 9-13, 1994.

5. Sy, A. and Campanella, R.G. (1994) "Becker and standard penetration tests (BPT-SPT)
correlations with consideration of casing friction," Canadian Geotechnical Journal, Vol. 31,
pp. 343-356.

6. Sy, Alex, Campanella, R. G., and Stewart, Raymond A. (1995) "BPT-SPT Correlations for
Evaluation of Liquefaction Resistance in Gravelly Soils," Proceedings of Specialty Session
on Dynamic Properties of Gravelly Soil, ASCE Annual Convention, San Diego, California,
October, 1995.

7. Tokimatsu, Kohji (1988) "Penetration Tests for Dynamic Problems," Proceedings of the first
International Symposium on Penetration Testing, ISOPT-1, Orlando, March 20-24, 1988,
A.A. Balkema.

8. Yan, L. and Wightman, A. (1992) "A Testing Technique for Earthquake Liquefaction
Prediction in Gravelly Soils," Report by Foundex Explorations Ltd. and Klohn LeonoffLtd.
to National Research Council of Canada, Contract No. IRAP-M 40401W.

139
 DIESEL HAMMER

-. . ..
..
. . .. . .
. .. ..

• . " • • •, • • • • • " "

•
• • • " • II , II
.. 0

• .. • • " •, 0
,. • • • , •
• ,, • • ,
• •
• ,. •
0
, 0• •
•
• 0 • "
" • • , • • . • • "
0 0 •
• • • • , • • •
•
• ., • ,
. , •
, 0
,,
• • • • "
• ,,
0
..
• • •
• ,.
0
I
• • • ., •'
• , • •
0

• 0 • • • •• • • • , • 0

" • • •
~
• • • •
., • • • • • . • • 0
• .
• II
• "
,. 0

• • • • • • • 0
0

• • • • • , • • 0 , •
0
• • • • 0 • • • •• • "
"
." • • 0
0
0

Figure 1: Schematic Diagram of Becker Sampling Operation

140
 II()

0
.!:
'"
•
•
0
.0
0

z
,._..
z
:,
0
u 40
3:
0
.J
OJ
I-
a.
"'
a
"'u
I-
20
... SALINAS TEST SITE
tSILT a SANO)
"'a:er • TH ERMALITO TUT Sm:
l!A.NOl
0
u
., 100
• SAN DIEGO TEST SITE
(SILTY SJ.NO
120
a SAND)
140

CORRECTED BECKER BLOWCOUNT, Noc (blows/Ioctl

a. Original Harder and Seed (1986) Correlation

- 80 ,------,-------,------~----,-------,--------,------,
0 /
(!) ♦
z ,.
i--: /

z
:::> 60 1---------1-----+----+~----+.~----+------t-------j
/
/

0 0 ,;;,

0
s:g ◊ TEST SITES
.AsAUNAS TEST SITE
CO 40 f--------t----,J-i\----71hW--;;--'1,--°'-----t-------J ■ THEAMALITO TEST SfTE
f- ♦ SAN □ !EGO TEST SITE
a.. 0
(/) JACKSON LAKE SITE A

0 0 JACKSON LA.KE SITE H

w ■ SOUAMISH FMC TEST SITE

I- 20 l------.1.''H;;r,i'/f!n,,.....,,.----+----+-------+------,
0 O MACDONALDS FARM SITE
w ◊ DUNCAN DAM TOE SITE
a:
a: ◊ DUNCAN DAM CREST SITE
0 ◊ANNACIS NORTH PIER
0 0OC--"''--<><>---'-------'--------'-----'----.l.--------~
0 ~ ~ ~ 00 100 1~ 140
CORRECTED BECKER BLOWCOUNT, NBC (blows per foot)

b. Harder and Seed (1986) Correlation Supplemented

with Data from Additional Test Sites

Figure 2: Correlation Between Corrected Becker and SPT Blowcounts

(from Harder and Seed, 1986)

141
 100,-----,,-----,.---,---""T'"---r----r---.-----,r------r--~--.

z"'
..,:
z:::,
0
u
f5 10
..J
cc
0:
w
!.:
u
w
cc

I
2 4 6 8 10 12 14 16 18 20 22 24
BOUNCE CHAMBER PRESSURE {psigl

10

.; 20
~
:i::
I-
CL 30
Lu
0

40

50
0 10 20 30 40 50
BECKER BLOWCOUNT, N9

Figure 3: Effect of Diesel Hammer Combustion Efficiency on Becker Blowcount

(after Harder and Seed, 1986)

142
 1000

' \ I'\ [',. ~FICIENCIES•D

BLOWCOUNT CORRECTION QJRVES
FOR REOUCED COMBUSTION - r--

\ I'\
.......
I'\
' I' " k r-,.. r-,..
I'--
\
\ ' r--....
100

"'
\
"\
\
'''
....
... ... "' ~

-
I'-- r-,...

.....
r--,..
~
....
,I
A-
z• ..... /
'
---
\. 1--

' \ !\
"\ I'.. -"-
1--
z ' I'\ .......
' o"'
:,
0
u
~
"
I\ \ \
"\'
,.\ I'\ " "r-.... .......
~
I....,_
....... ,__ /,
V(,Y

""
/~~

..J
' \ '\ '\ "-,, ,> '
al
\, \ \ i'-.. r--.... ...... l1
~ '\
'""
0
a: ~ /.
I.LI
/4 ~
'\ '\ i'\' I'\ ""
X
u r.....
I.LI
\ [',. >1, d~
""' "
al

i\_ "\ [',. I'\. /.. ~

~ ~'
10
' ' ' .... .... f'",
,_r,· '"'''
' ' ... ' '- L.., ~
" ,... -, /2 \'

,,, ~~
"
V-c

A 1c.- NOTE• ALL BECKER BLOWCOUNTS FALLING ON OR

CORRECTED TO CALIBRATION CURVE ARE
DESIGNATED AS CORRECTED BECKER
BLOWCOUNT, Nae

'o 2 6 8 10 12 14 16 18 20 22 24 26 28
BOUNCE CHAMBER PRESSURE AT SEA LEVEL, BPsc (psigl

Figure 4: Correction Curves Adopted to Correct Becker Blowcounts to

Constant Combustion Curve Adopted for Correlation
(from Harder and Seed, 1986)

143
 100 100
-
• ANNACIS SITE - •a ANNACIS SITE
a RICHMOND SITE RICHMOND SITE
-
80 60
~
-
E -
")
ci
-
60 60
"- •
U)
~
-
-
0
:0
....,,
40 40
- .•
:;:
z - a •
- -ti
•"□
f-
0..
(/)
20 20
-
- A
□

•• • a
..,_,. ~ . • C

• •
- ......". '
-
0 I I I I I I I I I I I I 0 I I

0 20 40 60 80 100 0 20 40 60 80 100
BPT Nbc (blows/0.3m) BPT Nb 30 (blows/0.3m)

Figure 5: SPT N60 vs. Becker NBc and Nb3o for Annacis and Richmond Test Sties
(from Sy and Campanella, 1993)

144
 NbJ-O & N60 (blows/0.3m) Nb:io/Nso
0 50 100 150 200 0.0 1.0 2.0 3.0

DELTA SITE
Injection - - F8PT5 NbJ-0
Regu1ur - - - - 8PT1 0 Nb30
___.... SPT Nso

I
15
0- -
~::--0
~,
20 a:- - -b
' <Sl
I
,....._ -, I
E 25 -, ... G-

---,,
:r:
~
CL
.... -_-_- - -
~

w 30
0 --,- - - a,
,,
,.d

•'I
';O
,,
35
(

40 ' ----
~- - ---- __ -_-_.
--; -
45 ..:-_ :=,
---=- -
- -=- =: -------
.. 50 ,,

55-L---------------'

Figure 6: Effect of "Mudded" Becker Casing on Determined Becker Blowcounts

(after Sy and Campanella, 1994)

145
 80
,,-...
E
t")

0
"-..
[/)
60
3
0
.D
'-"
0
40
z"'
f-
Q_
(/)

20

0 20 40 60 80 100
BPT Nb 3o (blows/0.3m)

Figure 7: Computed BPT vs. SPT Correlation for Different BPT Casing
Shaft Resistance (from Sy and Campanella, 1994)

146
 EQUIVALENT SPT N.,. EOUIVAI..ENT SPT N,.. ECUIVALENT SPT N_.
0 2S 50 75 100 2S 50 75 100 0 2S 50 75 100
0-+-'-,-!-,L--'-~~~'-'--~~~~--;

-------------
BPT D-i91 - 3C:
BPT D-!91-2C:
--HARC€R
10 --HARacR ----- SY
BPT OH91-1C: ····· ST
- - HARDER
----- SY

20

:r:
l-
o..
w
o 30

40

50 ~ - - - - - - - - - - -

Figure 8: Comparison of the Equivalent SPT N60 Values from Becker Soundings
using Correlations Developed by Harder and Seed (1986) and
Sy and Campanella (1994) - (from Lum and Yan, 1994)

147
 SPT N60 (blows/0.3m)
0 50 100
0
ANNACIS NORTH PIER
-oH91-B1 BECKER N60

5 -- DH91-B2 BECKER N60

• DH-90 SPT N60

• DH-92 SPT N60

10

15

60
20
-.....
-
E
'--'25
I
::::.,

I
f-
0..
UJ
80

I-
Q_
0 •
w 100
0 30

•
35
____,__ 120

- ---------~~·--
40
~-~~~~- -------------
-------
45
Equivalent (BPT 81) 1601,_.J_....!...._L.__L--''--.U....-'---'----'--'
Equivalent (BPT 82) 0 20 40 60 80 100
50 • Measured (DH90/92)
BECKER AND SPT N60

a. Sy and Campanella b. Harder and Seed

Figure 9: Comparisons of Predicted and Actual SPT N60 Values at Annacis Test Site
using Harder and Seed (1986) and Sy and Campanella (1994) Methods

148
 Magnitude Scaling Factors

T. Leslie Youd

Professor of Civil Engineering

Brigham Young University
Provo, Utah, 84602-4081

Steven K. Noble

Project Engineer
DOWL Engineers, Inc.
Anchorage, Alaska, 99503-5999

Abstract

In developing the "simplified procedure" for evaluating liquefaction resistance, Seed and Idriss
(1982) compiled a sizable database from sites where liquefaction did or did not occur during
earthquakes with magnitudes near 7.5. From this database, these investigators defined a conservative
deterministic bound separating data indicative of liquefaction from data indicative of
nonliquefaction. They then scaled this bound to other earthquake magnitudes using magnitude
scaling factors (MSF). Because insufficient observational data were available for magnitudes other
than 7.5, Seed and Idriss ( 1982) analyzed recorded ground motions and laboratory test data to define
the original set of scaling factors. As more field performance data were collected, the observed
occurrences and nonoccurrences of liquefaction for smaller earthquakes (magnitudes less than 7)
indicated that the original Seed and Idriss (1982) MSF may have been overly conservative for
smaller magnitudes. In response to this apparent overconservatism, Idriss reevaluated the original
seismic and laboratory data, made corrections, linearized the data on a logarithmic plot, and
developed a revised set ofMSF. Ambraseys (1988), Arango (1996), Andrus and Stokoe (this report)
and Youd and Noble (herein) defined alternative sets ofMSF based on empirical field observations.
These MSF lie within a narrow range for Mw:,; 7.5. The workshop gained consensus that a range of
MSF values should be suggested for engineering practice. Practitioners could then select MSF from
within this range, depending on the degree of risk they or their clients are willing to accept for
various applications. The lower bound for that recommended range is the MSF proposed by Idriss
(Column 3, Table 1) and the upper bound for the range is the factors proposed by Andrus and Stokoe
(Column 7, Table 1). For magnitudes greater than 7.5, the revised factors ofldriss should be used.
These factors are smaller than the original Seed and Idriss (1982) factors, and hence lead to an
increase of calculated liquefaction hazard compared to the old factors. The workshop participants
agreed, however, that the original factors by Seed and Idriss ( 1982) may not have been adequately
conservative for large magnitude earthquakes.

149
 Introduction

In developing the "simplified procedure" for evaluating liquefaction resistance, Seed and Idriss
(1982) compiled a sizable database from sites where liquefaction did or did not occur during
earthquakes with magnitudes of near 7.5. Data from the following earthquakes were compiled: 1964
Niigata, Japan (M = 7.5), 1974 Haicheng, China (M = 7.3), 1976 Tangshan, China (M = 7.6), 1976
Guatemala (M = 7.6), 1977 western Argentina (M = 7.4), and 1978 Miyagiken-Oki, Japan (M = 7.4).
Cyclic stress ratios were calculated for sites where surface effects of liquefaction were or were not
observed, and results from Holocene clean sand (fines contents :;; 5 percent) were plotted on a cyclic
stress ratio versus (N 1) 60 plot. An updated version of that plot by Seed et al. (1984) is reproduced
in Figure 1. A deterministic CRR curve was drawn on the plot to separate regions with data
indicative ofliquefaction occurrence (solid symbols) from regions indicative ofnonoccurrence.

0.6~-------------------

0.51-----------+-----,l--,----1-------I
I
e'
~
Q I
:
~
...... 0.4

0
>

•
.
t----'..'---+-----+-----L.,'-1------1-------1

.: /
e

-~ • I Ii)
cd • /

. .. I
~ 0.31-----+-----+--.J¥,1..
-'--+----1------1

1 a

u
u
:.=
~ 0.21----.....
■
• •• •
•• 0
.--t/,-------------..
■ ..
•
/
I

..00
ti
a

_-1-------I

,J• FINES CONTENT :,;.5%

• e ri Chinese Building Code (clay content= 0) ®
0.1 .:. ,.
a~0 Maroinal No

7
Liquefaction Liquefaction Liquefaction
A Pan - American data ■ Iii El
Japanese data • o o
Chinese data " A
0
0 10 20 30 40 50
Corrected Blow Count, (N1) 60

Figure 1 Cyclic Stress Ratio Versus (N 1) 60 for Magnitude 7.5 Earthquakes with Simplified
Base Curve Drawn to Separate Data Indicative of Liquefaction from that Indicative
on Nonliquefaction (modified from Seed 1984)

150
In areas where a mixture of both types of data populated the chart, the curve was conservatively
positioned to assure that nearly all of the points from sites where liquefaction had occurred plotted
above the bounding curve. The position of the curve is well constrained by observational data
between cyclic stress ratios of 0.08 and 0.35, and was logically extrapolated to higher and lower
values beyond that range. This curve, hereafter called the "simplified base curve," provides the
fundamental SPT-based criterion for assessing liquefaction resistance of clean granular soils for
magnitude 7.5 earthquakes. As noted in the Summary Report, the workshop participants agreed that
the lower limb of the curve should be bowed to intersect the ordinate of the plot at a CRR of about
0.05.

To adjust the simplified base curve for magnitudes other than 7.5, correction factors called
"magnitude scaling factors (MSF)" were applied to scale the base curve upward or downward on a
cyclic resistance ratio (CRR) plot as required. Conversely, magnitude weighting factors (MWF),
which are the inverse of magnitude scaling factors (MWF = 1/MSF), may be applied to correct the
stress ratio generated by the earthquake (CSR) to incorporate the influence of magnitude. Correcting
either CRR via magnitude scaling factors or CSR via magnitude weighting factors leads to the same
final result. Because the original papers by Seed and Idriss were written in terms of magnitude
scaling factors, use of magnitude scaling factors will be continued through this report.

The influence of magnitude scaling factors on calculated liquefaction resistance is illustrated by the
equation for factor of safety (FS) against liquefaction:

FS = (CRR7./CSR)*MSF (1)

where CRR75 is the cyclic liquefaction resistance ratio for a magnitude 7.5 earthquake as determined
from the simplified base curve. This equation indicates that the factor of safety against liquefaction
is directly proportional to the selected magnitude scaling factor.

Original Seed and Idriss Scaling Factors

Because sufficient empirical data were not available in the 1970s to constrain bounds between
liquefaction and nonliquefaction regions on CSR plots for magnitudes other than 7.5, Seed and Idriss
(1982) developed magnitude scaling factors based on strong motion records and laboratory test data
as follows: From a study of strong motion accelerograms, Seed and Idriss correlated the number of
representative loading cycles generated by an earthquake with earthquake magnitude. For example,
a magnitude 7.5 earthquake was characterized by 15 loading cycles, whereas a magnitude 8.5
earthquake was characterized by 26 loading cycles, and a magnitude 6.5 earthquake by IO loading
cycles. Laboratory test data were then used to assess the number of loading cycles required to
generate liquefaction and 5 percent cyclic strain as a function of cyclic stress ratio. Tests were
conducted on several different clean sands at various void ratios or relative densities and ambient
stress conditions. From these tests, a single representative curve was developed that relates cyclic
stress ratio to the number of loading cycles required to generate liquefaction (Figure 2). Earthquake
magnitudes were then assigned to various points on the curve based on the equivalent number of
representative loading cycles. By dividing ordinates on the representative curve for various
earthquake magnitudes by the ordinate for magnitude 7.5, the original set of magnitude scaling

151
 1.0.----------r----ir---""T"""-----r-------,
:::!: :::!:
("'"l -
\0 \0' r--' '
00
0.8 II II 11 II
~ ~ ~ ~
.....6
~
~
00
00
~
0.6

l ll
l1 it
...., 0.4
C/.l

t
ir
-u
.....u
u
;,-. 0.2
089
o---------'-----''----'----'-------'
1 6 10 15 26 100
Number of Cycles to Cause ru = 100% and± 5% Strain
Figure 2 Representative Relationship between -ch: 1 and Number of Cycles
Required to Cause Liquefaction (modified from Seed and Idriss, 1982)

factors was derived (Seed and Idriss, 1982). These scaling factors are listed in Column 2 of Table 1,
and plotted on Figure 3. These scaling factors have been widely applied as the standard of practice
since their initial introduction into the simplified procedure.

Idriss Revised Magnitude Scaling Factors

In preparing his H.B. Seed memorial lecture, I.M. Idriss (written commun.) reevaluated the data he
and the late Prof. H.B. Seed had used to calculate the original (1982) magnitude scaling factors. In
so doing, Idriss re-plotted the data on a logarithmic plot and determined that the scaling factors
should plot as a straight line. He further noted that one outlier point had unduly influenced the
original analysis, causing the original plot to be nonlinear on the logarithmic plot and characterized
by much lower values than those on the revised linear plot. Based on this reevaluation, Idriss defined
the revised set of magnitude scaling factors listed in Column 3 of Table 1, and plotted on Figure 3.
Idriss recommends the revised scaling factors for use in engineering practice. These revised
magnitude scaling factors are defined by the following equation:

MSF = 173(MY256 (2)

or in terms of magnitude weighting factors (MWF):

MWF = M 256/l 73 (3)

The revised scaling factors are significantly larger than the original scaling factors for magnitudes
less than 7 and significantly smaller than the original factors for magnitudes greater than 8. Relative
to the original scaling factors, the revised factors lead to reduced calculated liquefaction hazard for

152
 Table 1. Magnitude Scaling Factor Values Defined by Various Investigators

Mw Seed and Idriss Ambraseys Arango Andrus and Stokoe Youd and Noble
Idriss revised ( 1988) (I 996) (this report) (herein)
( 1982) PL<20%PL<32% PL<50%
(I) (2) (3) (4) (5) (6) (7) (8) (9) (I 0)

5.5 1.43 2.20 2.86 3.00 2.8 2.86 3.45 4.44

6.0 1.32 1.76 2.20 2.00 1.65 2.1 1.93 2.35 2.92

6.5 I. I 9 1.44 1.69 1.6 1.34 1.65 1.99

7.0 1.08 I. I 9 1.30 1.25 1.25 0.96 1.19 1.39

7.5 1.00 1.00 1.00 1.00 1.00 1.0 0.70? 0.88? 1.00

8.0 0.94 0.84 0.67 0.75 0.8 0.73?

8.5 0.89 0.72 0.44 0.76 0.65 0.56?

Table 2. Equations for Calculation of Magnitude Scaling Factors

Idriss (written commun.) MSF = 10224/M2s6 = (Mj7.5y2s6 (3)

Andrus and Stokoe (this report) MSF = (Mj7.5Y 33 (7)
Youd and Noble (herein)
Probability, PL< 20%, MSF = I03·81 /M4·53 For M<7 (11)
Probability, PL< 32%, MSF = 10381 /M4 .4 2 For M<7 (12)
Probability, PL< 50%, MSF = 10421 /M4·81 For M<7.75 (13)

magnitudes less than 7.5, but increased calculated hazard for magnitudes greater than 7.5. Because
there have been few large earthquakes (M > 8.0) causing widespread liquefaction since the 1960s,
there is little high-quality field data with which to verify or constrain magnitude scaling factors for
earthquakes greater than magnitude 8. To be conservative, the revised factors should be used, even
though they may lead to reclassification of some areas as liquefiable that were previously considered
to be nonliquefiable.

Ambrayses Magnitude Scaling Factors

As more field performance data were collected, the observed occurrences and nonoccurrences of
liquefaction for smaller earthquakes (magnitudes less than 7) indicated that the original Seed and
Idriss (1982) scaling factors were overly conservative. Consequently, Ambraseys (1988) analyzed
liquefaction data compiled through the mid-1980s and plotted calculated cyclic stress ratios for sites
that did or did not liquefy on CSR versus (N 1) 60 plots. Ambraseys segregated the data into magnitude
ranges (6.0 to 6.6, 6.7 to 7.2, 7.3 to 7.5, and 7.6 to 8.2) and fit a deterministic bound through the data

153
 4.5
-+- Seed and Idriss, (1982)
4 - ; - - - - _ . , . . - = - 1 r - - . , , . . - - - - - - - ; - . . . . - - ! ------ Idriss

3.5
x Ambraseys (1985)
Workshop ◊ Arango (1996)
3 ♦ Arango (1996)
_...... Andrus and Stokoe
2.5 .& Youd and Noble, PL<20%
t. Youd and Noble, PL<32%
2 .& Youd and Noble, PL<50%

1.5

0.5

0
5.0 6.0 7.0 8.0 9.0
Earthquake Magnitude, Mw

Figure 3 Magnitude Scaling Factors Proposed by Various Investigators

on each plot using an exponential form for the bounding equation. The equation Ambraseys
developed using the compiled data (all magnitude ranges and all soil gradations) is:

(4)

where Q is equivalent to cyclic resistance ratio (CRR) as used in this report, M.v is moment
magnitude, and (N 1\ 0 is corrected standard penetration resistance. Ambraseys developed a second
equation using only data from earthquakes with magnitudes less than 7.5 and data from sites
underlain by clean sands (fines contents less than 5 percent):

(5)

The exponential form of these equations is consistent with the exponential form used with many
attenuation relationships for peak acceleration as a function of distance. One criticism raised at the
workshop was that the derived scaling factors are strongly influenced by the functional form of the
equation. The equation may be appropriate for attenuation of peak acceleration, but may not be
adequate for occurrence of liquefaction, which is a function of shaking duration or magnitude, as
well as peak acceleration.

By holding the value of (N 1) 60 constant in Equations 4 and 5 and taking the ratio of Q determined for
various magnitudes of earthquakes to the Q for magnitude 7.5 earthquakes, Ambraseys derived the

154
set of magnitude scaling factors listed in Column 4 of Table I and plotted on Figure 3. For
magnitudes less than 7.5, the scaling factors suggested by Ambraseys are significantly greater than
the original factors developed by Seed and Idriss (1982) and the revised factors by Idriss. Use of
these larger factors leads to greater calculated liquefaction resistances and less calculated hazard for
earthquakes with magnitudes less than 7.5. Conversely, for magnitudes greater than 7.5, use of
Ambraseys' factors leads to significantly less calculated liquefaction resistance and much greater
hazard than any of the prior factors.

Procedures used to develop the original factors by Seed and Idriss and the revised factors by Idriss
are considerably different from procedures used by Arnbraseys. Seed and Idriss used laboratory tests
and strong motion records to develop their scaling factors; whereas, Ambraseys used field
observations of liquefaction and nonliquefaction effects. The different scaling factors determined
by these different investigators may be partly due to differences in developmental procedures.

Arango Magnitude Scaling Factors

Arango (1996) developed two sets of magnitude scaling factors: The first set (Column 5, Table 1)
is based on the farthest observed liquefaction effects past earthquakes of various magnitudes and
estimated average peak accelerations at those distant sites. The second set (Column 6, Table 2) is
based on absorbed seismic energy required to generate liquefaction and the relationship derived by
Seed and Idriss (1982) between number of significant stress cycles and earthquake magnitude.
Arango's scaling factors listed in Column 5 generally have values similar (within 10%) to those
derived by Ambraseys (1988). The MSF listed in Column 6 are rather close in value (within 6%)
to the revised MSF ofldriss (Column 3).

Andrus and Stokoe Magnitude Scaling Factors

From their studies of liquefaction resistance as a function of shear wave velocity, Vs, Andrus and
Stokoe (this report) developed the following equation for calculating CRR from Vs for magnitude
7.5 earthquakes:

(6)

where V 51 is the corrected shear wave velocity; Vsic is acritical value of Vsi, which separates
contractive and dilative behavior; and a and b are curve-fitting parameters. Values of a and b
recommended by Andrus and Stokoe for magnitude 7.5 earthquakes are 0.03 and 0.9, respectively.
The value ofVsic depends on the fines content of the soil, but ranges from 220 mis for clean sands
to 200 mis for silty sands with fines contents greater than 35%.

Using Equation 6, Andrus and Stokoe drew curves on graphs with plotted values of CSR as a
function of calculated Vsi from sites where surface effects ofliquefaction were or were not observed.
Graphs were plotted for sites shaken by magnitude 6, 6.5, 7, and 7.5 earthquakes. The positions of
the curves were visually adjusted on each of these graphs until a best-fit bound was obtained,
separating data indicative ofliquefaction from data indicative of nonliquefaction. Magnitude scaling
factors were then estimated by taking the ratio of CRR ordinates for a given V51 from plots or

155
magnitude 6, 6.5 and 7 earthquakes to CRR at the same V 51 for magnitude 7.5 earthquakes. These
MSF values were then fitted to an exponential equation and values extrapolated for magnitudes less
than 6 and greater than 7.5. That equation is

MSF = (Mj7.5)" 3·3 (7)

MSF values calculated from this equation are listed in Column 7, Table 1, and plotted on Figure 3.
For magnitudes less than 7.5, the MSF proposed by Andrus and Stokoe are rather close in value to
the MSF proposed by Ambraseys. For magnitudes greater than 7.5, the Andrus and Stokoe MSF are
rather close in value to the MSF proposed by Idriss.

Youd and Noble Magnitude Scaling Factors

In a study for this workshop, the authors reanalyzed case history data using logistic regression (Youd
and Noble, Probabilistic Analysis, this report). The data analyzed were from sites for which
subsurface soil information is available and for which surface effects of liquefaction were or were
not observed. From that regression, Youd and Noble developed the following empirical equation:

where PL is the probability that liquefaction will occur, 1-PL is the probability that liquefaction will
not occur, Mw is moment magnitude, and (N 1\ 00, is the corrected blow count, including a correction
for fines content. This equation allows cyclic resistance ratios, CRR, to be calculated as a function
of(N 1) 600,, earthquake magnitude, and a given probability ofliquefaction occurrence. Figure 4 shows
CRR curves plotted from Equation 6 for magnitude 7.5 earthquakes and PL of2O%, 32%, and 50%,
respectively. For reference, the simplified base curve for clean sands is also plotted on that figure
as well as calculated CSR from sites in the case history database where effects of liquefaction were
or were not observed. The plotted data are from earthquakes with magnitudes between 7 .25 and
7.75. For (N 1) 60 between 3 and 30, the simplified base curve is enveloped between probabilistic
CRR curves with PL between 20% and 50%.

PL<50% Magnitude Scaling Factors

Although Youd and Noble (Statistical and Probabilistic Analyses, this report) recommend direct use
of Equation 8 for evaluating liquefaction resistance, an alternative method is to calculate magnitude
scaling factors using Equation 8, and then apply those factors with the simplified procedure. The
development of magnitude scaling factors for a probabilities ofliquefaction of5O% or less (PL:;;5O%)
is rather simple and straightforward. The following text and examples illustrate this procedure.
Setting PL equal to 50% in Equation 8, and solving for CRR yields the following equation:

CRRPL~50% = e(2.466 -0.7289M + 0.0834(Nl)60cs) (9)

Holding (N 1\0c, constant at any value and taking the ratio of CRR for any magnitude to the CRR for
magnitude 7.5 earthquakes yields the following equation for calculating magnitude scaling factors
with PL=5O%:

156
 M SFPL=50% = CRRICRR7.5 = e-o 7289M/e-o 7289(75) (10)

For example, solving this equation for magnitude 6.5 earthquakes yields a calculated MSFPL=so¾ of
2.07. Scaling factors for earthquake magnitudes ranging from 5.5 to 8.5 were calculated from this
equation and are plotted with a plus symbol on Figure 5. These values are very close in value to the
MSF based on a probability of 32% previously published by Loertscher and Youd (1994). Using
these scaling factors, one can calculate the 50% probability curve for any magnitude simply by
multiplying the CRR ordinates of the magnitude 7.5 curve by the appropriate MSF. For example,
the 50% probability curve for magnitude 6.5 earthquakes can be generated by multiplying the CRR75
for each (N 1) 6ocs by 2.07.

..
0.50

•
•• • 0
0.40

• •
~
(/.l 0
u
....0
0.30 •
.....
ell • ••
~
VJ
VJ
~
.....
;..
"'·• 0
0

...-
(/.l

-u.....
u
u
0.20 •••
;>-.
Magnitude, Mw, = 7.5 ± 0.25
• •o•
Legend
0
Liquefaction datum O
0.10
.. Non-liquefaction datum •
Youd and Noble, PL= 50%
0 Youd and Noble, PL= 32%
Youd and Noble, PL= 20%
- Simplified base curve
0.00 +----+---+---~---+--+----,f----+----1-----1----l
0 10 20 30 40 50
Corrected Blow Count, (N 1)60cs

Figure 4 Probabilistic CRR Curves for PL= 20%, 32% and 50% from Youd and Noble
(Probabilistic Analysis, this report) with Simplified Base Curve and Empirical
Liquefaction Data for Magnitude 7.25 to 7.75 earthquakes

157
 Explanation
5.0 • MSF from procedure
in Figure 7
4.0
- MSF from equations
t:I..
Cll in Table 2
~
;...," 3.0
.....u0
cl:
Of)

·--
C
ro
u
2.0
Cll
<l)
"O

·-
:::l
.....
C
Of)
ro
~

1.0

5.0 5.5 6.0 6.5 7.0 7.5 8.0

Earthquake Magnitude, Mw
Figure 5 Logarithmic Plot of Probabilistically Determined Magnitude Scaling
Factors Versus Earthquake Magnitude

The simplified base curve is nearly tangent with, but lies slightly below, the 50% probability curve
at an (N 1) 6acs of about 12, and lies below that curve for all other (N 1) 6acs up to an (NJ 60 c,of 30
(Figure 4). Thus multiplying the simplified base curve by MSFPL=so% calculated from Equation 8
yields scaled base curves that lie below the 50% probability curve for all earthquake magnitudes.
These scaled curves, however, would be nearly tangent with, but lie slightly below the 50%
probability curves at an (Ni)6acs of approximately 12. Scaling the simplified base curve in this manner
yields predicted CRR that are always characterized by a probability ofliquefaction less than 50%,
or PL<50%. For illustration, the scaled simplified base curve for magnitude 6.5 earthquakes is
plotted on Figure 6. Thus the PL=50% magnitude scaling factors can be used as scaling factors for
use with the simplified procedure and will give estimates of CRR that are characterized by
probabilities ofliquefaction occurrence ofless than 50%.

Rather than using the PL= 50% scaling factors directly, the authors followed the model of Idriss and
plotted the derived magnitude scaling factors on a log-log plot (Figure 5) and developed a straight-
line fit to those MSF. To develop the straight-line fit, magnitude scaling factors calculated from

158
 0.50

•
0.40
•
•
••
~
C/)
u 0.30
·-
6
.....ro
~
lfl
Magnitude, Mw, = 6.5 ± 0.25
•
lfl

.....~ Legend
C/)

·--
u
u
>-.
0.20 Liquefaction datum
Non-liquefaction datum
0
•e
u Youd and Noble, PL= 50%
Youd and Noble, PL= 32% llE
Youd and Noble, PL= 20%
0.10 Scaled Base Curves
MSF, PL <50% - -- --..
-
0~ MSF,PL< 32%
0 0 0
0 MSF, PL <20%
Scaled base curve, original MSF
0.00
0 10 20 30 40 50

Corrected Blow Count, (N1)60cs

.Figure 6 Probabilistic CRR Curves with Scaled Simplified Base Curve

for Magnitude 6.5 Earthquakes

equation 8 were plotted on Figure 5 as noted above for earthquakes with magnitudes between 5.5
and 8.5. The straight line was drawn conservatively (slightly below the plotted data for magnitudes
Jess than 7.5) to assure that all CRR calculated using these magnitude scaling factors indeed are
characterized by a probability of liquefaction occurrence of less than fifty percent. The straight line
approximation for MSF with PL <50% is defined by the following equation:

MSF = 10421/M4s1 (11)

159
This equation is also listed in Table 2, and MSF values calculated from this equation are listed in
Table 1, Column 10, and plotted on Figure 3. For clarity, these MSF are marked with the designator
PL<50%, indicating that CRR calculated using these MSF yielded values characterized by less than
50% probability of liquefaction occurrence. These MSF may be used for engineering applications
where up to 50% probability of liquefaction occurrence is acceptable, but are restricted to
earthquakes with magnitudes less than 7.75.

As noted by Youd and Noble (Statistical and Probabilistic Analyses, this report), the case history
data are sufficient to adequately constrain the probabilistic analysis for magnitudes between 5.5 and
7.75. Above magnitude 7.75, the data are too sparse and scattered to adequately constrain the
probabilistic analysis. Thus the validity of MSF determined by this procedure is questionable for
magnitudes greater than 7.75. Although MSF values for PL<50% are listed for reference for
magnitudes 8 and 8.5 in Table 1, Column 10, these values are marked with a question mark
indicating they are uncertain. MSF in this tabulation marked with a question marks are not
recommended for use in engineering practice.

PL<32% and PL <20% Magnitude Scaling Factors

Most engineering applications require a more conservative likelihood of liquefaction than a

probability of 50%, thus magnitude scaling factors are also derived for probabilities less than 32%
and less than 20%, respectively. Because the simplified base curve for magnitude 7.5 earthquakes
lies above both the 32% and 20% probability curves for magnitude 7 .5 earthquakes over part of its
trajectory, the technique used for PL<50% can not be used directly for these lower probability values.
That is, because the probabilistic CRR curves intersect the base curve, scaling these curves directly
would also produce curves that intersect and hence do not lie entirely below the given probabilistic
curve. The magnitude 7.5 simplified base curve, however, lies entirely below the 32% and 20 %
probability curves for magnitudes less than 7. For these magnitudes, scaling the simplified base
curve upward to near tangency with the probabilistic CRR curves is a pragmatic procedure for
defining magnitude scaling factors. In essence, this is the procedure that was used in developing the
PL <50% MSF. In that instance, the simplified base curve was effectively scaled upward to near
tangency with the 50% probabilistic curve for earthquake magnitudes less than 7.5.

The development of MSF using this technique is illustrated by the following example. Figure 7
shows the simplified base curve and the PL=32% probabilistic curve for magnitude 6.5 earthquakes.
The simplified base curve was rotated upward about the origin until the curves become tangent at
some point. In this instance, the point of tangency occurs at an (N1) 6 ocs of about 12, as shown by the
dashed curve. The CRR from the PL=32% curve for an (N 1) 6 ocs of 12 is 0.224. The CRR from the
simplified base curve for an (N 1) 60cs of 12 is 0.134. By taking the ratio of these two CRR, MSF are
defined for a probability ofliquefaction occurrence equal to less than 32% (PL~32%). For magnitude
6.5 earthquakes, that value is:

MSFM=6.5.PL<32% = 0.224/0.134 = 1.67

Similarly, upwardly scaling of the simplified base curve to a point of tangency with PL=32%
probability curves for other earthquake magnitudes, and then taking the ratio of the CRR at the point

160
 Probabilistic CRR I
curve-Mw = 6.5,
0.4 PL=32% ~
+-- Simplified base curve

---+-- Rotated simplified

base curve
0.3

Point near
tangency
0.2
MSF = CRR6_5 = 0.224 = 1.67
CRR7 _5 0.134

0.1

0 IO 20 30 40

Corrected Blow Count, (N 1) 60

Figure 7 Diagram Showing Calculation of PL=32% Magnitude Scaling Factor

of tangency to the CRR from the simplified base curve at the same (N1) 60c,, would create a set of
MSF with PLd2%. These MSF were calculated and are plotted on the logarithmic chart in Figure
5. A straight line was then drawn slightly below the PLs:32% MSF yielding the following equation:

(12)

Values of MSF listed in Column 9, Table I, were calculated from this equation. Because the
straight-line fit was drawn conservatively (slightly below) the plotted values, CRR calculated using
these MSF will always be characterized by PL<32%, and this designation is used as a descriptor for
these MSF. A simplified base curve scaled by these MSF for magnitude 6.5 earthquakes is plotted
on Figure 6. The MSF listed in Column 9, Table 1, and plotted on Figure 3 are valid for use in
engineering applications where up to 32% probability of liquefaction occurrence is acceptable.
Extrapolation of Equation 12 to a magnitude of 7.5 yields MSF less than 1.0. This lower value

161
occurs because Equation 12 scales the base curve to a point of near tangency with the probabilistic
curves. Because the PL=32% curve lies below the simplified base curve at magnitude 7.5, the scaling
factor has to be less than 1.0.

The procedure used to develop PL<20% MSF was the same as that used for PL<32% MSF, except
that the simplified base curve was scaled up to tangency with the 20% probability curves rather than
the 32% probability curves. The MSF determined by that procedure are plotted on Figure 5. A
slightly conservative straight line fitted to those calculated MSF is quantified by the following
equation:

(13)

MSF calculated from Equation 13 are listed in Column 8, Table 1, and plotted on Figure 3. These
MSF are valid for use in engineering applications where up to 20% probability of liquefaction
occurrence is acceptable. Again, probabilistic curves for Mw equal to or less than 7 lie below the
simplified base curve. As with MSF for PL<32%, extrapolation of Equation 13 to magnitude of7
or less yields MSF that are less than 1.0.

Magnitude Scaling Factors For Engineering Practice

Magnitudes Equal to or Less than 7.5

Magnitude scaling factors from the various investigators noted herein are plotted on Figure 3. This
plot shows considerable consistency between values developed by the various investigators, each of
whom used distinctly different methodologies to develop their scaling factors. This consistency adds
confidence that the derived MSF are generally correct and appropriate for use in engineering
practice. For magnitudes less than 7.5, the MSF suggested by Ambraseys, Arango, and Andrus and
Stokoe ( Columns 4 through 7, Table I) lie within a narrow grouping. The PL<20% MSF of Youd
and Noble (Column 8, Table I), also lie within this grouping for magnitudes less than 6.5. The MSF
recommended by Idriss (Column 3, Table 1) generally lie below and are about 20% smaller than the
average values within the above grouping. The PL<32% MSF of Youd and Noble (Column 9,
Table 1) lie above and are roughly 20% larger than the values within the group for magnitudes less
than 6.5. The PL<50% MSF developed by Youd and Noble (Column 10, Table 1) are significantly
higher than all of the other factors. The original MSF of Seed and Idriss (Column 2, Table 1) plot
well below the MSF of all the more recent investigators and appear to be overly conservative. As
noted by Idriss, some errors and extra-onservative interpretations influenced the original derivation.
Thus the original Seed and Idriss ( 1982) scaling factors should no longer be used in engineering
practice, but should be superseded by more recently generated MSF as recommended below.

After evaluation of the various proposed MSF, the workshop gained consensus (with S.S.C. Liao
dissenting) that a range of values should be suggested for engineering practice for earthquakes with
magnitudes less than 7.75. This range is marked by a stippled pattern on Figure 3. Practitioners
could then select MSF from within that range, depending on the degree of risk they or their clients
are willing to accept for a given application. The lower bound for this recommended range is the
MSF proposed by Idriss (Column 3, Table 1) and the upper bound of the range is the MSF proposed

162
by Andrus and Stokoe (Colwnn 7, Table 1). Based on the probabilistic analysis of Youd and Noble,
the upper values in this range are characterized by a probability ofliquefaction occurrence of20%
or less. The MSF proposed by Idriss, which form the lower bound of the range, are characterized
by a probability of liquefaction much less than 20%. These MSF are valid for use with liquefaction
resistance criteria based on SPT, CPT, shear-wave velocity, or Becker penetration resistance.

Magnitudes greater than 7.5

For magnitudes greater than 7.75, the MSF developed by Idriss (Column 3, Table 1; and Equation 3)
are up to 20% smaller than the original Seed and Idriss MSF, but are still greater than those
suggested by other investigators as listed in Table 1. The values by Ambraseys, Andrus and Stokoe,
and Youd and Noble (Columns 4 and 7-10, respectively, Table 1), are extrapolated from smaller
magnitude earthquakes and thus are uncertain for magnitudes greater than 7.75. The MSF by Arango
are derived from distance and energy relationships, but still generally yield values much smaller than
the Idriss values. The workshop participants agreed that there is insufficient evidence and
verification to recommend MSF lower than those proposed by Idriss for use in engineering practice.
Even so, the MSF proposed by Idriss still lead to greater calculated liquefaction hazard than was
predicted by the original Seed and Idriss MSF, which have been the standard of practice. Thus some
areas that have been determined to be safe against liquefaction using the original MSF may be
predicted to liquefy using the Idriss MSF. Use ofMSF proposed by all of the other investigators
would lead to an even greater calculated hazard, but as noted above, there is little verification for this
greater hazard for earthquakes with magnitudes greater than 7. 75.

The MSF by Idriss for magnitudes greater than 7. 75 were derived from a reanalysis of the original
Seed and Idriss (1982) data and thus not extrapolated from smaller magnitude earthquakes. The case
history data provide some verification if the MSF proposed by Idriss for magnitudes greater than
7. 75. For example, Figure 8 shows a plot ofliquefaction data from the compiled case histories for
earthquakes with magnitudes between 7.75 and 8.25. Scaled simplified base curves, using the
various MSF for magnitude 8 earthquakes listed in Table 1, are also plotted on this chart. The case
history data are rather sparse on this plot and nearly all of the data are from three pre-1925
earthquakes, primarily the 1991 Mino-Owari, Japan (M = 7.8), 1906 San Francisco, California
(Mw = 7.9), and 1923 Kwanto, Japan (Mw = 7.9). Also, there is insufficient case history data over
much of the trajectories to adequately constrain the positioning of the curves. The data do indicate,
however, that the simplified base curve scaled by the Idriss scaling factor for a magnitude of 8
generates a slightly conservative bound for the plotted data. The data is even more sparse for larger
magnitude earthquakes. Little data have been collected for magnitude 8.25 to 8.75 earthquakes, and
essentially the only data from earthquakes with magnitudes greater than 8. 75 are from sites observed
to have liquefied during the 1964 Alaska earthquake (Mw = 9.2). These data are insufficient to
constrain empirical analyses of magnitude scaling factors for these larger magnitudes.

Although these data are insufficient to fully verify that the MSF by Idriss are conservative, they also
do not provide evidence that the Idriss factors are unconservative. The data are insufficient to justify
MSF smaller than those proposed by Idriss, as is indicated by nearly all of the factors in columns 4-
10 of Table 1. Thus the consensus of workshop participants is that the MSF by Idriss are the best
estimates available for large earthquakes for use in engineering practice.

163
 0.50

0.40

•
~
Cl'.l
u 0.30
0
0
.......
..... •
• •
~
C's$

r/:)
00

....
Cl)
.....
Cl'.l
u 0.20
0
•
•
0 •
0
0
.• Magnitude, Mw, = 8.0 ± 0.25
.......
u>-, • •• Legend
u Liquefaction datum 0
• • Non-liquefaction datum •
Scaled base curves
0.10
Idriss (revised)
Seed and Idriss (original)
Ambraseys (1988) - - -
Arango (1996) - - - - -
Andrus and Stokoe - - - -
0.00
0 10 20 30 40 50

Corrected Blow Count, (N 1)60

Figure 8 Empirical Liquefaction Data from Magnitude 8 Earthquakes

with Scaled Simplified Base Curves Using Various MSF from Table 1

164
 References

Ambraseys, N.N., 1988, Engineering Seismology: Earthquake Engineering and Structural

Dynamics, Vol. 17, p. 1-105.

Arango, I., 1996, Magnitude Scaling Factors for Soil Liquefaction Evaluations: Journal of
Geotechnical Engineering, ASCE, Vol. 122, No. 11, p. 929-36.

Loertscher, T.W. and Youd, T.L., 1994, Magnitude Scaling Factors for Analysis of Liquefaction:
Proceedings from the Fifth U.S.-Japan Workshop on Earthquake Resistant Design of Lifeline
Facilities and Countermeasures Against Soil Liquefaction, National Center of Earthquake
Engineering Research Technical Report NCEER-94-0026, p. 703-15.

Seed, H.B., and Idriss, I.M., 1982, Ground Motions and Soil Liquefaction During Earthquakes:
Earthquake Engineering Research Institute Monograph.

Seed, H.B., Tokimatsu, K, Harder, L.F., and Chung, R.F., 1984, "The Influence of SPT Procedures
in Soil Liquefaction Resistance Evaluations, Report No. UBC/EERC-84/15, Earthquake
Engineering Research Center, University of California, Berkeley, California.

165
166
 Application of K 0 and Ka Correction Factors

Leslie F. Harder, Jr.

Chief, Division of Engineering

California Department of Water Resources
Sacramento, California 94236-0001

Ross Boulanger

Assistant Professor
Department of Civil and Environmental Engineering
University of California, Davis
Davis, California 95616

Abstract

The K0 and K" correction factors were originally developed by Seed (1983) to adjust cyclic resistance
ratios obtained from empirical SPT and CPT correlations to different static stress conditions. The
Ko correction factor extends cyclic resistance ratios to high confining stresses, while the Kc,
correction factor adjusts cyclic resistance ratios to sloping ground conditions. Because there are
virtually no case histories available to help define these correction factors, the results from laboratory
test programs have been used to develop these factors. This section describes the development of
these correction factors, describes the results of recent laboratory programs, and recommends
modified correction factors for use in liquefaction evaluations.

167
 INTRODUCTION

Liquefaction evaluations generally employ correlations between liquefaction resistance and corrected
SPT blowcount or CPT tip resistance. One of the most commonly used SPT correlations is the one
developed by Seed et al. (1985). In this correlation, the performance of over one hundred sites
following earthquake loading was evaluated. For the critical layer at each site, the estimated cyclic
stress ratio induced by the earthquake (r:a/ov0 ') was determined and plotted as a function of the
corrected SPT blowcount, (N 1) 60 (see Figure 1). Sites which had liquefied (solid symbols) plotted
in areas with low SPT resistance and/or high cyclic stress ratio. Sites which showed no evidence of
liquefaction (hollow points) plotted in areas with high SPT resistance and/or low cyclic stress ratio.
Three dividing lines between areas ofliquefaction and no liquefaction were drawn for fines contents
of 5, 15, and 35 percent fines. These dividing lines are curved and have been used with corrected
SPT blowcounts to predict the liquefaction resistance of sandy and silty soils.

This liquefaction resistance is expressed in a normalized form as a cyclic stress ratio causing
liquefaction. As pointed out by Peter Byrne, use of the Seed et al. (1985) SPT correlation gives a
cyclic resistance ratio (CRR) and should properly be noted as such. Although this and similar
correlations are invaluable for the evaluations of liquefaction resistance, the database for the
correlations are limited to flat ground under relatively low levels of confining stress. To extend the
correlations for use with soils under sloping ground or high confining stresses, correction factors
must be used. Because SPT and CPT correlations are used in conjunction with the Seed-Lee-Idriss
approach, the correction factors are based on the pre-earthquake static stress conditions on the
horizontal plane. The use of correction factors to extend the existing correlations to different stress
conditions is performed as follows:

(1)

where:

CRR = the cyclic resistance ratio (r:a/ov0 ') at the current stress state (av 0 ', r: 5).

(CRR) 1 = the cyclic resistance ratio at the reference state:

(SPT correlation; av0 ' = 1 tsf, r: 5 = 0).

K0 = a correction factor for the level of vertical effective confining stress, ov0 '.

Ka = a correction for the level of static horizontal shear stress, (a=r:/av 0 ').

ov 0 ' initial vertical effective confining stress.

r:5 = static shear stress on the horizontal plane.

168
The Ka and Ka correction factors have generally been based on the normalized results of cyclic
triaxial and cyclic simple shear tests. Several studies (e.g. Seed, 1983, and Seed and Harder, 1990)
have offered suggested ranges for the factors based on available laboratory test results. The
magnitudes of these corrections can be extremely large for deep soil layers under either flat or
sloping ground. However, there is a wide range and scatter in the available data and, because the
correction factors often have a controlling influence on the final evaluation ofliquefaction resistance,
their use has been questioned for particular projects.

Ka CORRECTION FACTOR FOR CONFINING STRESS

General

In early liquefaction evaluations, liquefaction resistance was determined by performing cyclic

laboratory tests (e.g. cyclic triaxial) on "undisturbed" soil specimens. The tests were performed for
a range of static stress conditions. To represent level ground conditions, isotropically-consolidated
cyclic triaxial tests were commonly performed. To correct for the incorrect boundary conditions
inherent in the cyclic triaxial test, a Cr correction factor was used.

The results of tests performed on "undisturbed" isotropically-consolidated specimens of several

different soils are shown in Figure 2. The results show that, while the cyclic resistance of a soil
increases with increasing confinement, the cyclic resistance ratio decreases. This is because the
cyclic resistance curves for the different soils are not straight lines passing through the origin, but
rather are curves that flatten out at increasing consolidation stresses. As illustrated in Figure 3, the
cyclic resistance ratio is simply the secant line intersecting the cyclic resistance curve at the
consolidation stress of interest. At higher and higher levels of confinement, Figures 2 and 3 show
that the slope of the secant line will flatten and that the cyclic resistance ratio will decrease.

Studies by Seed (1983)

In recent years, the use of cyclic laboratory tests for direct evaluations of liquefaction resistance has
dramatically decreased because of concerns regarding sample disturbance and other factors. Such
evaluations now generally employ empirical correlations such as the Seed et al. (1985) SPT-
liquefaction correlation. The Seed SPT correlation and others generally define the cyclic resistance
of soils to liquefaction in the form of a cyclic resistance ratio (r:a/ov0 '). This normalized cyclic
resistance is based on the performance of soils which were under relatively low effective confining
stresses, typically less than 1 tsf (-100 kPa). However, many liquefaction evaluations require the
determination of cyclic resistance for significantly higher confining stresses. To extend SPT
liquefaction correlations to higher confinement, Seed (1983) developed the Ka correction factor.
The approach used by Seed is illustrated in Figure 3 and the range of Ka values recommended by
Seed (1983) is shown in Figure 4.

169
The Ka values developed by Seed (1983) were the result of normalizing cyclic resistance ratios of
isotropically-consolidated cyclic triaxial tests to the CRR values associated with an effective
confining stress of 1 tsf. For confining stresses greater than 1 tsf, the Ka correction factor is less
than one and decreases with increasing confining stress. Using this suggested range for Ka, the CRR
value at 8 tsfbecomes only about 40 to 60 percent of the CRR value at 1 tsf.

Studies by Seed and Harder (1990)

Later studies by Harder (1988) and Seed and Harder (1990) employed the same approach and used
the results of available cyclic triaxial testing programs. The Seed and Harder (1990) Ka values are
shown in Figure 5 and show a suggested curve for the Ka correction factor. This later curve gives
a Ka value of 0.44 at a confining stress of 8 tsf and is generally somewhat lower than the range
suggested by Seed (1983). For confining stresses above 8 tsf, the Seed (1983) curve suggests a
continuing linear decrease in Ka whereas the later Seed and Harder (1990) curve is concave up,
suggesting a limiting K 0 value of about 0.4 for very high confining stresses.

The Seed and Harder (1990) Ka curve is based on laboratory tests performed on both "undisturbed"
and reconstituted sand specimens. The "undisturbed" specimens generally consisted of silty sands
typically recovered in thin-walled tubes with Piston samplers as part of dam safety investigations.
The reconstituted specimens generally consisted of clean sands prepared by pluviation or moist
tamping as part of university research studies. Failure criteria generally consisted of 5 percent axial
strain within either 10 or 15 cycles ofloading.

There is considerable scatter in the Seed and Harder Ka database and this is partly due to the
different sources of the data. For the "undisturbed" specimens, there was often a wide range in
densities and materials due to the heterogeneity of both natural deposits and fills. In addition, there
was generally no attempt to compensate for the increased densities of "undisturbed" specimens that
would be expected to occur with higher confining stresses. Compensating for the higher densities
would be expected to result in even lower Ka values. On the other hand, consolidation of
"undisturbed" specimens in the laboratory to confining stresses greater than their in situ values may
have adversely affected their particle fabrics (including the beneficial effects of aging and prior
seismic history), and possibly resulted in an unrepresentative decrease in both cyclic resistance ratio
and Ka. Reconstituted specimens prepared in the laboratory, however, can be constructed to
maintain similar relative densities and particle fabric over the full range of confining stresses.

Studies by Vaid et al. (1985) and Vaid and Thomas (1994)

Research performed in recent years by Y. Vaid and his colleagues on clean sands have indicated
relatively little decrease in CRR values for even very high confining pressures. This corresponds
to relatively high K 0 values. Data developed by Dr. Vaid at the University of British Columbia has

170
been obtained using a constant volume cyclic simple shear device to test dry sand specimens
prepared by pluviation. The results have been reported in the studies by Vaid et al. (1985) and Vaid
and Thomas (1994). Shown in Figure 6 are the UBC Ka results for Tailings, Ottawa, and Fraser
sands for confining stresses ofup to 26 tsf (2500 kPa). This data has been developed for specimens
with the same relative densities after consolidation (i.e. the increase in density produced by
increasing the confining stress has been compensated for). For confining stresses of up to 26 tsf,
the UBC Ka values are above 0.67. For confining stresses of only about 10 tsf, the UBC Ka values
are generally between 0.75 and 0.9. The results also showed that lower relative densities produced
higher Ka values.

The UBC Ka values are significantly higher than those suggested by either Seed (1983) or Seed and
Harder (1990). This suggests that use of the Seed and Harder (1990) Ka curve could be very
conservative for some soils, particularly clean sands.

Discussion

The available results from cyclic laboratory tests show a wide potential range for Ka values. At
confining stresses of about 8 tsf, the Seed (1983) and the Seed and Harder (1990) curves (see
Figures 4 and 5) give K 0 values that are only half of those indicated by recent UBC testing (see
Figure 6). In 1991, portions of the Seed and Harder (1990) database were reexamined and the results
for clean sands were found to be significantly higher than those for "undisturbed" silty sand
specimens. Shown in Figure 7 are the clean sand data from Seed and Harder (1990) compared with
some of the UBC data. The clean sand data points from the Seed and Harder database show far
better agreement with the UBC results than did the silty sand data.

As a result of this comparison, Byrne and Harder (1991) recommended a clean sand Ka curve for
use in evaluating liquefaction resistance of clean sands and gravels at Terzaghi Dam. This clean
sand curve gave a Ka value about 45 percent higher than the Seed and Harder (1990) curve at a
confining stress of8 ts£ However, due to the lack of new data for silty soils, it was recommended
at that time that liquefaction evaluations employ the original Seed and Harder Ka curve for soils
with significant fines contents.

The large range in potential Ka values may require that critical projects such as large dams employ
site specific corrections developed using laboratory tests performed on project soils. Pillai and
Byrne (1994) presented the results of a site specific investigation of confining stress effects for the
fine sand beneath Duncan Dam. In this investigation, frozen samples were obtained near the
downstream toe of the dam, taken to the laboratory, and tested in cyclic triaxial and cyclic simple
shear devices after thawing. The test results showed that the CRR values determined in both sets
of tests were independent of confining stress. This was because of compensating factors that
essentially balanced each other:

171
 • Increasing confining stress decreased the CRR (i.e. the Ka effect).

• Increasing confining stress also increased the consolidated density of the specimens
before cyclic loading and this increased the CRR.

Pillai and Byrne used SPT test results to estimate the effect of the increased density on CRR and
were able to back calculate a set of site-specific Ka values for the Duncan Dam fine sand (see
Figure 8). By comparing Figures 7 and 8, it may be observed that the Duncan Dam K 0 values were
just slightly less than the "clean sand" Ka values previously recommended by Bryne and Harder
(1991 ). In addition to this, Arango (1996) describes recent Ka data obtained from cyclic tests of silty
and clayey sands that plot between the Duncan Dam and UBC curves.

Recommendations

Recent laboratory test programs on a range of potentially liquefiable soil types have provided a basis
for refining the relationship between Ka and effective confining stress. The available information
indicates that the curve shown in Figure 9 will provide a reasonably conservative estimate of Ka
for use in evaluations of liquefaction resistance. This curve is considered appropriate for both clean
and silty sands and gravels.

For any given value of confining stress, K 0 generally decreases with increasing relative density.
However, the decrease in K 0 with increasing relative density is likely to be offset by the decrease
in post-liquefaction deformability that accompanies increases in relative density. For high risk and
critical projects where there are potentially liquefiable layers under high effective confining stresses,
site-specific laboratory testing may be necessary to develop reliable corrections. Samples of the
highest possible quality should be obtained for this purpose.

Ka; CORRECTION FACTOR FOR SLOPING GROUND

General

The Ka correction factor was suggested by Seed (1983) to extend SPT and CPT correlations to
sloping ground conditions. Sloping ground induces static shear stresses on horizontal planes within
a soil mass prior to the onset of earthquake shaking (see Figure I 0). The relative magnitude of the
static shear stress ('tJ on the horizontal plane can be assessed by normalizing it with respect to the
vertical effective confining stress. The resulting parameter is called the alpha value (rx=r:/ov0 '). For
level ground conditions, the alpha value is zero. Early research suggested that the presence of a
static shear stress always improved the cyclic resistance of a soil because higher cyclic shear stresses
were required to cause shear stress reversal. This conclusion still appears to hold for moderately

172
dense and dense cohesionless soils under relatively low confining pressures (say< 3 tsf). However,
the presence of a static shear stress has since been shown to decrease the cyclic resistance of loose
sandy soils and/or some moderately dense soils under very high confining stresses.

Studies by Seed (1983)

Seed and his colleagues at the University of California, Berkeley performed both cyclic simple shear
and anisotropically-consolidated cyclic triaxial tests to investigate the effect of static shear stresses
on cyclic resistance. Shown in Figure 11 are typical results from cyclic simple shear tests showing
increases in cyclic resistance for increasing levels of static shear. This result was considered typical
for soils with relative densities of 50 percent or higher.

The results of early research led Seed (1983) to develop the Ka correction for the presence of static
shear. The Ka correction factor is used to extend the SPT and CPT correlations developed for level
ground to conditions where sloping ground would induce static shear stresses on horizontal planes.
Presented in Figure 12 is the Ka range suggested by Seed (1983) for relative densities equal to 50
percent or higher. As may be noted, the Ka correction factors lead to very large increases in cyclic
resistance. For an alpha value of about 0.2, comparable to a soil layer beneath a moderate slope,
the Ka correction factor would increase the cyclic resistance by approximately 70 percent.

Studies by Vaid et al., Yoshimi et al., Szerdy, Jong and Seed

Continued research performed at various universities showed that the effect of the static shear stress
on cyclic resistance was very complicated. Several of these studies were completed in the 1970's and
1980's, including those performed by Yoshimi and Oh-Oka (1975), Vaid and Finn (1979), Tatsuoka
et al. (1982), Vaid and Chem (1983, 1985), Szerdy (1986), and Jong and Seed (1989). These
research programs involved the use of cyclic simple shear, cyclic ring torsional shear, and cyclic
triaxial equipment. All of these research programs resulted in very consistent results regardless of
the type oftest equipment employed:

• The presence of a static shear stress increased the cyclic resistance (i.e. Ka> 1) of
moderately dense or dense sandy soils for confining stresses less than about 3 tsf.

• The presence of a static shear stress decreased the cyclic resistance of loose, sandy
soils (i.e. Ka <1 ).

• The effect of ~tatic shear stresses on cyclic resistance was significantly dependent on
the failure criterion used to define failure during testing.

• The effect of static shear stresses on cyclic resistance was significantly dependent on
the confining stresses used during testing.

173
Figure 13 presents results obtained by Vaid and Finn (1979) and Vaid and Chern (1983), which
illustrate the trends observed at confining stresses less than about 3 tsf. Figure 14 presents results
by Vaid and Chern (1985) that illustrate how very high confining stresses can result in Ka values less
than 1 for a dense sand. This effect is attributed to the sandy soils becoming more contractive with
increasing confining stress.

Studies by Seed and Harder (1990)

The studies by Seed and Harder (1990) compiled much of the available laboratory research and
developed a set of Ka correction factors for a range of relative densities (see Figure 15). The
derivation of these Ka factors used a failure criterion of7.5 percent shear strain (5 percent axial
strain) in 10 or 15 cycles. This criterion was chosen because it was in common use in the late 1980's.
The Seed and Harder (1990) Ka values shown in Figure 15 reflect the large range in laboratory
results available at the time and show the effect of relative density. The Seed and Harder Ka values
were based on data obtained for effective confining stresses of 3 tsf or less. Higher confinement
makes soils more contractive and could lead to lower Ka factors, as illustrated by the Vaid and
Chern (1985) data shown in Figure 14.

Studies by Boulanger et al. (1991) and Boulanger and Seed (1995)

Boulanger and Seed (1995) presented the results of bi-directional cyclic simple shear tests performed
to investigate the effects of static shear stresses both parallel and perpendicular to the direction of
cyclic loading. Their work verified many of the results previously developed by other researchers
(e.g. Vaid and Finn, 1979) regarding the effects of density and failure criteria on the determined
cyclic resistance. Their studies also showed the following results regarding the direction of shaking
relative to the direction of the static shear (i.e. the direction of sloping ground):

• For cyclic loading perpendicular to the direction of the static shear stress, the effect
of static shear was to reduce the cyclic shear resistance of sand specimens with
relative densities of 35 and 45 percent, regardless of failure criterion. Reduced cyclic
shear resistance was also found in the perpendicular direction for a relative density
of 55 percent for failure criteria of less than 3 percent shear strain.

• Cyclic shear resistance in the direction perpendicular to the static shear stress was up
to 30 percent less than the cyclic resistance parallel to the static shear stress. This
difference increased with higher alpha values, and did not appear to be dependent on
either relative density or number of loading cycles (see Figure 16).

The studies by Boulanger and Seed (1995) noted that, as for similar studies, the residual pore
pressure ratios were limited to values far less than unity with increasing levels of static shear stress.
Residual pore pressures for alpha values of about 0.1 never exceeded about 80 percent of the initial

174
vertical effective stresses, and residual pore pressures for alpha values of about 0.3 never exceeded
about 40 percent of the initial vertical effective stresses (It should be noted that residual pore
pressure refers to the pore pressure value after cyclic loading has ended, and that higher peak pore
pressures often temporarily occur during cyclic loading).

Boulanger et al. (1991) also found that specimens tested with static shear stresses developed most
of their limiting excess pore pressure ratios relatively early within individual tests. For this reason,
Boulanger et al. (1991) tentatively recommended the use of revised Ka values. These revised Ka ·
values are somewhat lower than those developed by Seed and Harder (1990), and are generally based
on a failure criterion of3 percent shear strain rather than 7.5 percent. Use of the 3 percent shear
strain failure criterion is believed to better capture the onset of large pore pressure generation. The
revised Ka values of Boulanger et al. (1991) were also weighted more heavily towards the data from
simple shear or torsional simple shear tests rather than those from triaxial tests. This was because
the former allow continuous rotation of principal stress directions, and thus a better representation
of how static shear stresses affect shear stress reversal on all possible planes.

Figure 17 presents a slightly modified version of the Ka values previously recommended by

Boulanger et al. (1991). In general, the new Ka values shown in Figure 17 for relative densities of
about 35 percent are about the same as those suggested by Seed and Harder (1990). However, the
Ka values for higher relative densities are up to 40 percent lower than comparable values suggested
by Seed and Harder (1990). However, there is some overlap in the ranges suggested by both sets of
studies. Also shown in Figure 17 are SPT (N 1) 60 values that are considered to approximately
correspond to the different ranges in relative density.

Discussion

The wide ranges in potential Ka values show the need for continued research and field verification
of the effects of static shear stresses on liquefiable soils. The different rates of pore pressure
generation and different limiting values of potential excess pore pressure for different locations
within a slope show that the evaluation of liquefaction resistance of sloping ground is an extremely
complicated endeavor. Considerable research is needed into what constitutes liquefaction under
sloping ground conditions and how to determine the potential for triggering liquefaction in a
practical manner. It is essential that case histories illustrating the liquefaction performance of
sloping ground and soil beneath and adjacent to foundations be studied to enable assessment of Ka
factors under field conditions.

Recommendations

Recent laboratory studies have yielded a wide range of Ka correction factors. After evaluating the
results from recent test programs, it is recommended that the Ka vs. alpha relationships shown in

175
Figure 17 be used for liquefaction evaluations of soils. The values sho~ in Figure 17 are
appropriate for effective confining stresses ofless than 3 tsf Significantly higher confining stresses
will cause sandy soils to behave more contractive, and thus lower values of Ko: may be appropriate
for such conditions (see Figure 14).

For high risk and critical projects on sloping ground or where there are high initial static shear
stresses, site-specific laboratory testing should be considered using high quality samples in order to
develop reliable values of Ko:.

CONCLUSIONS/RECOMMENDATIONS

1. The empirical data available for use in predicting liquefaction resistance is generally limited
to relatively flat ground under relatively small levels of confining stress. There remains
considerable uncertainty regarding the appropriateness of various correction factors to extend
the available SPT and CPT empirical correlations to conditions with sloping ground or high
confining stresses.

2. It is essential that case histories regarding the seismic performance of earth structures with
liquefiable soils under high confining stresses and sloping ground be investigated and
documented.

3. Laboratory test programs have shown a large range of potential K0 and Ko: correction factors.
Some of the available data is relatively old and oflesser quality than more recent studies.
Additional research examining the effect of fines content and higher confinement on both
corrections would be of great benefit.

4. The K 0 curve sho~ in Figure 9 is considered to provide a reasonably conservative estimate

of K 0 values for use in liquefaction analyses of clean or silty sands and gravels.

5. The Ko: relationships shown in Figure 17 are recommended for effective confining stresses
less than 3 tsf. For significantly higher confining stresses, lower Ko: values may be
appropriate.

6. For critical projects, site specific laboratory testing may be necessary to develop corrections
for both sloping ground and high confining stresses. Samples of the highest possible quality
should be obtained for this purpose. For soils with low fines contents, in situ freezing of
samples for laboratory testing may be the best approach. Alternatively, considerable insight
may still be gained from tests performed on reconstituted specimens.

176
 REFERENCES

1. Arango, Ignacio (1996) "Correction Factor K 0 - Derivation of Site-Specific K 0 for an Old

Clayey Sand Deposit," working paper prepared for Workshop on Evaluation of Liquefaction
Resistance, Salt Lake City, Utah, January 4-5, 1996.

2. Boulanger, Ross W. and Seed, Raymond B. (1995) "Liquefaction of Sand under Bidirectional
Monotonic and Cyclic Loading," ASCE Journal ofGeotechnical Engineering, Vol. 121, No.
12, December.

3. Boulanger, Ross W., Seed, R. B., Chan, C. K., Seed, H. B., and Sousa, J. (1991)
"Liquefaction Behavior of Saturated Sands under Uni-directional and Bi-Directional
Monotonic and Cyclic Simple Shear Loading," Geotechnical Engineering Report No.
UCB/GT/91-08, University of California, Berkeley.

4. Byrne, Peter M. and Harder, Leslie F., (1991) "Terzaghi Dam, Review of Deficiency
Investigation, Report No. 3," prepared for BC Hydro, Vancouver, British Columbia.

5. Harder, L. F., Jr. (1988) "Use of Penetration Tests to Determine the Cyclic Load Resistance
of Gravelly Soils," Dissertation submitted as partial satisfaction for the degree of Doctor of
Philosophy, University of California, Berkeley.

6. Harder, L. F., Jr. and Seed, H. Bolton (1986), "Determination of Penetration Resistance for
Coarse-Grained Soils using the Becker Hammer Drill," Earthquake Engineering Research
Center, Report No. UCB/EERC-86/06, University of California, Berkeley.

7. Hynes, M. E. (1988) "Pore Pressure Generation Characteristics of Gravel under Undrained

Cyclic Loading," Dissertation submitted as partial satisfaction for the degree of Doctor of
Philosophy, University of California, Berkeley.

8. Pillai, V. S. and Byrne, P. M. (1994) "Effect of Overburden Pressure on Liquefaction

Resistance of Sand," Canadian Geotechnical Journal, Vol. 31, pp. 53-60.

9. Seed, H. Bolton (1983) "Earthquake-Resistant Design of Earth Dams," Proceedings of a

Symposium on Seismic Design of Embankments and Caverns, ASCE, Philadelphia,
Pennsylvania, May 6-10, 1983.

10. Seed, H. B., Tokimatsu, K., Harder, L. F., and Chung, R. (1985) "Influence of SPT
Procedures in Soil Liquefaction Resistance Evaluations," ASCE Journal of Geotechnical
Engineering, Vol. 111, No. 12, December.

177
11. Seed, R. B., and Harder, L. F., Jr. (1990) "SPT-based analysis of Cyclic Pore Pressure
Generation and Undrained Residual Strength," Proceedings of the H. Bolton Seed Memorial
Symposium, May, 1990.

12. Vaid, Y. P. and Chern, J.C. (1983) "Effect of Static Shear on Resistance to Liquefaction,"
Soils and Foundations, JSSMFE, Vol. 23, No. 1, March.

13. Vaid, Y. P. and Chern, J.C. (1985) "Cyclic and Monotonic Undrained Response of Saturated
Sands," Advances in the Art of Testing Soils under Cyclic Conditions, ASCE Convention,
Detroit.

14. Vaid, Yoginder P., Chern, Jing C., and Tumi, Hadi (1985) "Confining Pressure, Grain
Angularity, and Liquefaction," ASCE Journal ofGeotechnical Engineering, Vol. 111, No.
10, October.

15. Vaid, Y. P. and Finn, W. D. L. (1979) "Static Shear and Liquefaction Potential," ASCE
Journal of the Geotechnical Division, Vol. 105, No. GTlO, October.

16. Vaid, Y. P. and Thomas, J. (1994) "Post Liquefaction Behaviour of Sand," Proceedings of
the 13th International Conference on Soil Mechanics and Foundation Engineering, New
Delhi, India.

178
 0.6
.~ 8 37

25 8

Percent Fines = 35 15 :a5

I I
0.5
I
.I
I
I
I
l
I I I
I I I
I I I
I I I
1 I
I I
.:.,o
0.4
I I I
I

I I

T0 v
I
I
rro• I
I
I

I
I
I
I

.31 J I I
o;'
0
I I I
I I I
20 I II
03 el2 I
■ So+ I
I

~27
■ao
so•
0.

FINES CONTENT 2 5 %
Modified Chinese Code Proposal (cioy conteni=5%.) ®

M:Jrginal No
Liquefaction Liquefaction Liquefoctico
Pon-American data 1:11 0
Japanese data

0L------..J--------'-------'------ I
Chinese data .
'9 Q 0

"'
0 10 20 30 40 50
( Nr )60

Figure 1: Relationship Behveen Cyclic Stress Ratios Causing Liquefaction During

M==7.5 Earthquakes and Corrected SPT Blowcounts in Silty Sands
(after Seed et al., 1985)

179
 ·"g 1.75--------------,------.----,-----,
Ill

Chobot Dom
(K2lmox~65
Son Fernondo Dem -
foundation
Lower Franklin Dom
(K2lmox"'S0
Fairmont Dom
(K2lma," 50
. Lower Son FernonOO Dom
(Kzlmo, .,45
Upper Son Fernando Dam
(K2lmax" 30

Ambient Consolidation Pressure - kg per sq. cm.

Cyclic Loading Resistance of Shell Materials for

Selected Dams (see Table 2)

3.5
Cyclic Triaxial Tests
K,=10
Oroville Dam.H=70.
(K 2lm.,=170.

, Los Angeles Dam .

.,..,,,.-...... H=120:, (K )m,a,c:=-70.
2
Upper San Leandro Dam,
H=17:i, (K 2lm,,,70.
.,,,. ... ---- Lake Arrowhead Dam,
,, H,200'.,(K )m•x•60.
2

All Dams Evaluated for Magnitude 81'.

Earthquakes With Om.ax.=0.4 to 0.59
and Found to be Satisfactory

1. 2. 3. 4. 5. 6. 7. 8. 9.
Ambient Consolidation Pressure - Kg/cm 2

Cyclic Leading Characteristics of Shell Materials for

Several Dams Shown by Analyses to be Capable of with-
standing Strong Earthquake Shaking

Figure 2: Cyclic Loading Resistance of Selected Soils as Determined by Isotropically-

Consolidated Cyclic Triaxial Tests (from Seed, 1983)

180
 0.6r--------------------------,

0.4
CYCLIC
SHEAR
STRENGTH
'c
(Is f)
0.2

FAILURE IN 10 CYCLES
o~------'-----,_J_____ ..J_,_ _ _ _,____ _ ___.1

0 2 3 4 5
EFFECTIVE CONSOLIDATION STRESS ON FAILURE PLANE, <Ttc' (tsf)

NOTE: 0"3 .° for ISOTROPICALLY- CONSOLIDATED CYCLIC TR/AXIAL TESTS

C CJ"dp
r 2

C, CJ"dp
CYCLIC STRESS RATIO
20"3e

[C~~3d:] 0"3; = I tsf

>
r~3
C,CJ"dp]
< a3e'> I tsf

Figure 3: Typical Cyclic Resistance Determined from Isotropically-

Consolidated Cyclic Triaxial Tests

181
 "o 1.0
:,:: 1/J
L Q)
05
~:llQ) .8
• 0
\:5 r..
--;; a.
>

--~
).,~ Cll

-c
0;;::
.6

1/J C
Q) 0
:::,U .4
ro...,
>cQ)
Q) r..
·--
>-2!
....,._
ro o
.2

&1ii
.o
Q 1. 2. 3. 7 4. 5. 6. 8.
Effective Overburd~n Pressure(6;,')- Kg/cm 2

Figure 4: Relationship between Effective Confining Pressure

and Ko Suggested by Seed (1983)

tu so
1.2
0
-
•o
1.0 2.0 3.0 4. 0 60 7.0 80

•o

' ,.,.
1.0
a •
..... -
• e

0.8
0 'l,,.. • C

,, ;-...,__
Ka-
, ~ ...... -...;
0 '

0.6
••
VJ
0
V
6
r.a1.itWOHT OAW
l ~ &RAOWHEAO 0&111
Srl(fn(t.0 DAM S"£L.L
Uf"P(II SAN l.CANORO 0,..1 SHCLL
LO•E• SAN FtRN.&N00 0,111 SKELL
Ul"P(R $AN t£1Ull.&.N00 GAW SHELL
"
I·
r--- ......
--. ----
I:> a

,__

...•0
0.4 LOS ANCitL.£$ O&M SM£L.L T
P'[IIAIS DAU SHELL., IIC• ts,100-4

0
S&IIDIS DAM SM[LL
S&ROIS 0,11111 f'0U"OAT10N .,
• TM[,.....LITO U'T[RUT DAIi fOUNO&.TION

0.2 •
0
THElh,..LlfO FOAU.&l' 0.ul tCUHDATION
&l,IJ[lOP( QUI IVP!JltVIDUS MAT[ALl,L
e FORT P(CK DAU SH[LL
0 SACRAMENTO RIVER $.6.NO, O,. •)I,'°, Tl, !00%
•
C
WOHT(AEY O $.lNO, 01'. 1 ~o/•
A[IC 1£0,0RO $ANO, Or. • 40, 10%
0 NE'tll' J[RS[l' RACICFILL. f'PI tter'l!•t..

1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0

EFFECTIVE CONFINING PRESSURE (tsfl or (ksc)

Figure 5: Relationshiop between Effective Confining Pressure

and Ka Suggested by Seed and Harder (1990)

182
 1.1

[
g,
,,, Dr=60%

.
1 l--,Q.-<J---<>-~<J------------------!--+i:X-60%-
f
! ~
'O
-
·--o--••
g@ C 0r=65%
0 & -·•o---•
Dr=70%
. . --,,0..--..
0.9

0.8 . ··-c_
......... '"'.. ..................
...... ....
............. .. ..... __
.............. .. ......
.... ........

0.6
0 500 1,000 1,500 2,000 2,500 3,000
Effective Confining Stress (kPa)

Figure 6: Relationship between Effective Overburden Pressure and Ko

(UBC data from Vaid et al., 1985 and Vaid and Thomas, 1994)

183
 1.2 llfV

lo
□
1 -□,
..: i-,-,-,;, ... - .. . . . .. . . ... . . . . . - . - - .. - - . -·Z
♦ K--~9 : _: _: _= _: _:_:
#"'••
I ••• • : : ! : , •
• •••a: .. - ... - -
0.8 "'

Ka :iR:_ x--·-- c:r ••

....... -~ ~ □
-
- ••• ..
-• --- -... - - ,.. - - - "Cl
. . . . . ... _._._._._. -~

0.6
c SARDIS DAM SHELL
r----.... ◊ "" ean Sand"
1----i--

-
a SARDIS DAM FOUNDATION
Seed1~
0 SAC. RIVER SAND ii ""d a.rdel:' I
El MONTEREY O SAND ll9 90 J;
0.4 ◊ REID BEDFORD SAND

♦ NEW JERSEY BACKFILL

*TAILINGS SAND DR=60

0.2 X TAILINGS SAND DR=65

Z OTTAWA SANO DR=60
-AonAWA SAND DR=70

0
0 1 2 3 4 5 6 7 8 9 10
EFFECTIVE CONFINING PRESSURE (tsf)

Figure 7: Comparison of UBC Ko Values with Clean Sand Data

from Seed and Harder (1990)

,.2

0.8

,;1 0.6

0.4

0.2

0
0 500 1,000 1,500

Effective Confining Stress (!<Pa)

Figure 8: Comparison of Duncan Dam Ko Values with other Ko Data

(after Pillai and Byrne, 1994)

184
 1.2

1
~ ...........
0.8
"-..r---...
Ka 0.6
I
I
0.4

0.2

0
0 1 2 3 4 5 6 7 8 9 10
EFFECTIVE CONFINING PRESSURE (tsf)

Figure 9: Recommended Ko Values for Clean and Silty Sands and Gravels

185
 (a)
Perpendicular to Dip Direction
Parallel to Dip Direction

Ov

t re
(b)
/ ...
I ~
I

(c)

Figure 10: Static and Cyclic Stress Conditions on Horizontal Planes

Beneath Sloping Ground (from Boulanger and Seed, 1995)

186
 b0 .6 r--,---,---,-----r----,-----r--,---,
•.,
1...0

~ .5
l'l CC:= "t"re
0( =0.3
s o;,
~ .4 ""=02
I...
·5
CT oe =0.1
&! 3
"'
1.'"' Q(;::O.
vi .2
I...
.,
ro
.c
Vl
.!,!
.1 D,~ 50 %
~
un~--'----'---~---'--~--=---::-~
0. 1. 2. 3. 4. 5. 6. 7. 8.
Initial Effective Vertical Stress 'tc -Kg/cm 2

Figure 11: Typical Effects of Initial Static Shear Stresses on

the Cyclic Resistance of Sands (after Seed, 1983)

~
::s::
t
..., 1.8
u
~
o 1.6
...,
ro
a::
Ill
a'.; 1.4
...,I..
V)
D,~ 50 %
u
-~ 1.2
>,
u
.05 :10 .15 20 .25 .30
Static Stress Ratio , 0<
Static Stress Ratio.

Figure 12: Ko: Correction Factors Suggested by Seed (1983)

187
 o, o,-so•,. 2
O'"v~ • 2 kg/cm
Y::a 10 o/.
0.6
N • 10
N .. ro
02 2%
5 g,.

j
0.1

z .,.
1o) Io I

.,.{
i
O.~:
o
' -.J8 o,
!
Dr - 50 o/.,
2 )',:; 10 'l'.1 .: o• Oro■ 68 o/.
a-V0-2kq/cm
C"vd-2 k,;ilcm
2 i
o,f Ne 30
2 %,
N =a 30
J
5'¾
o,I

(bl '¼
lbl
o
o 01 02 o., 04 05
Ts IC"vo'
00 0.1 0.2 o., 0.4 0.5
rs ICT"vo'

a. Effect of Static Shear on Cyclic Resistance Determined in Constant Volume Cyclic Shear Tests
(After Vaid and Finn, 1979)

0.,3 0.3
,. ,Lco1•
C, • ◄~
I
0,•55•t.,M1dium 'drnse
,.,.
V p,2 V 0.2 •;.
.b.,. •be u

..v-
.....

0.1
..u-
....
0.1
"be
.....
..u 0.1 1---+---+---l
lo 1 (bl IC I
0 0
o'---__.___....__ __.
0 0.1 0.2 0.3 0 0.1 0.2 0.3 0 0.1 0.2 0.3
T 1 /u;c T',lc-,.'c

b. Effect of Static Shear on Cyclic Resistance Determined in Cyclic Triaxial Tests

(after Vaid and Chern, 1983)

Figure 13: Typical Effects of Static Shear on Cyclic Resistance

188
 2.5~-----.----~---.----,--------------.

CYCLIC TRIAXIAL TESTS

2.0
~CTdp)f /2 Ke7CTmc'J
K~m~-.=-=------, «
CTCTdp)i /2KcrCTmc].,,,
0

- { ) - - TAILING SAND !Dr =70¾)

1.5 after Vaid and Chern (19851

Dr = 70%
1
O"mc = 16 -21 ksc

oo---~---~--~=----=----:-------~--~
0.1 0.2 0.3 0.4 0.5 0.6 0. 7
ex

Figure 14: Effect of Static Shear on Cyclic Resistance for Moderately

Dense Sand (Dr = 70 % ) at High Confinement
(adapted from Vaid and Chern, 1985)

K c,c:
1.0

0.5

a;, I::; 3 tsf

o~---'------'-----'-----L------'
0 0.1 0.2 0.3 0.4 0.5
o(

Figure 15: Ka. Correction Factors Suggested by Seed and Harder (1990)

189
 1.0 1.0

a:.a.
rr.· Ul

-.
u
Ul
~ 0.5 0.5
. 7=J,O o
4
D,=35,.
D,=45it rr.·
7=7.5:,; o
•
Dr=3~
D,=45,.
0,=55%
rr.· N=10, JO [] D,=55~ Ul
N=10, 30 0

Ul
~
~
0.0 0.0
0.0 0.1 0.2 O.J 0.0 0.1 0.2 O.J

Figure 16: Ratio of "Perpendicular" to "Parallel" Cyclic Loading Resistance for Failure
Criteria of3% and 7.5% Shear Strain (from Boulanger and Seed, 1995)

2.0

CTv6:::,; 3 tsf
D,"" 55-70%
(N1)so"' 14-22
1.5

Ka 1.0
D, :e 45-50%
(N1)so"' B-12

0.5 o,., 35%

(N1)so"" 4-6

0
0 0.1 0.2 0.3 0.4

Figure 17: Recommended Ka Values for Effective Confining Pressures Less than 3 tsf

190
 Seismic Factors for Use in Evaluating
Liquefaction Resistance

T. Leslie Youd

Professor of Civil Engineering

Brigham Young University
Provo, Utah, 84602-4081

Abstract

Application of the simplified procedure for evaluation of liquefaction resistance requires estimates
of earthquake magnitude and peak acceleration as input seismic parameters. These factors
characterize duration and intensity, respectively, of earthquake shaking at a site. The following
comments summarize consensus statements developed at the workshop for determining these
seismic factors.

• The simple correlation between magnitude and duration incorporated within the simplified
procedure is adequate and conservative for typical liquefaction hazard analyses. This
relationship is adequate for application in the eastern U.S. as well as the western U.S.

• Magnitudes based on the moment magnitude scale, Mw, are the preferred for calculation of
liquefaction resistance. Where estimates ofMw are not available, magnitudes from other scales
may be substituted within the following limits: ML< 6, m 8 <7.5, and 6 <M, < 8.

• The preferred procedure for estimating peak accelerations for use in liquefaction resistance
analyses is through application of empirical attenuation relationships specifically developed for
soil conditions representative of the site. For sites with soft soils or other soils not characterized
in attenuation relationships, local site response calculations using programs such as SHAKE or
DESRA may be used. The least desirable procedure is application of amplification factors or
transfer functions.

• Use of attenuation relationships based on the geometric mean of two orthogonal peak horizontal
components is consistent with development of the simplified procedure and should be used in
liquefaction resistance evaluations. Peak accelerations estimated from attenuation relationships
incorporating the larger of the two peaks, however, are conservative and allowable.
• High-frequency acceleration spikes that occur in some strong motion records may be safely
ignored for calculation of liquefaction resistance.

191
 Introduction

Application of the simplified procedure for evaluating liquefaction resistance of soil layers requires
estimates of earthquake magnitude, peak acceleration, and a magnitude scaling factor as seismic
input factors for the analysis. These factors characterize duration and intensity of earthquake
shaking at a site. Magnitude and peak acceleration are discussed in this section. Magnitude scaling
factors are discussed in a separate section entitled "Magnitude Scaling Factors" (Youd and Noble,
this report).

Earthquake Magnitude

In the development of the simplified procedure, Seed and Idriss (1982) used magnitude as a measure
of duration of strong seismic shaking. To incorporate duration, they introduced the concept of an
equivalent number of significant stress cycles contained in a strong motion accelerogram (Seed et
al., 197 5; Seed and Idriss, 1982). From an evaluation of the number of significant stress cycles
contained in accelerograms recorded at free field sites, Seed et al. (1975) developed the generalized
relationship between earthquake magnitude and number of equivalent stress cycles listed in Table 1.
Thus magnitude is effectively a measure of shaking duration as used in the simplified procedure.

Records from past earthquakes indicate that the relationship between duration and magnitude is
rather uncertain and that several factors other than magnitude influence duration. For example,
recorded durations have varied by a factor of two or more for earthquakes of a given magnitude.
This variation is partially dependent on the mechanism of faulting. Unilateral faulting with rupture
beginning at one end of the fault and propagating to the other end usually produces much longer
durations than bilateral faulting in which slip begins near the midpoint on the fault and propagates
in both directions. Duration also generally increases with distance from the seismic energy source
and may vary with topography of subsurface bedrock (basin effect) and with site conditions. To
compensate for uncertainty in duration and other factors, Seed and Idriss (1982) drew a rather
conservative bound on the standard CRR plot for magnitude 7.5 earthquakes to separate data
indicative of liquefaction from data indicative of nonliquefaction (Figure 1). The conservative
placement of this bound assures that liquefaction will seldom occur at sites where nonliquefaction
is predicted, even where longer than normal durations of shaking occur. Conversely, in some ·
instances liquefaction may not occur because of shorter than normal duration even though
liquefaction is predicted by the conservative bound plotted on Figure I.

Because of uncertainty between magnitude and duration, the workshop addressed the following
questions with respect to magnitude as an index of shaking duration and developed the following
consensus answers:

Question A: Is magnitude a sufficient parameter or should additional correction factors or

adjustments be developed to account for unusually long or short durations?

192
 0.6.------.---------------------~
37
.29 25 8

Percent Fines = 35 15 .::;;5

I
I
0.51-------,1-------+l-----.......,l.------+--------I
I I
I I
I I
I I
I I
I I
I
I

CRR curves for 5,15, and

35 percent fines, respectively

10 20 30 40 50
Corrected Blow Count, (N1) 60
Figure 1. Base Curve and Liquefaction Data from the
Simplified Procedure (modified from Seed et al., 1984).

Answer A: Variations of shaking duration for a given magnitude are commonly a function of the
mechanics and style of faulting. For example, bilateral faulting associated with the 1989 Loma
Prieta (Mw = 6.9) and 1995 Hyogoken-Nanbu (Kobe) (Mw = 6.8) earthquakes generated durations
(7 to IO seconds) that were less than half of durations normally expected for those magnitudes of
earthquakes. Conversely, the 1988 Armenian earthquake (Mw = 6.8) was characterized by a very
long duration of about 50 seconds, which is more than twice the duration normally expected.

193
 Table 1. Number of Equivalent Stress Cycles as a Function of Earthquake Magnitude
(after Seed and Idriss, 1982)

Earthquake Magnitude Number of Equivalent

Stress Cycles

5 1/2 2-3

6 5
6 3/4 10
7 1/2 15

8 1/2 26

Faulting characteristics and variations in shaking duration for a given magnitude of earthquake,
however, are difficult to predict in advance of the event. Thus use of a conservative simplified
relationship between magnitude and duration is an acceptable approach for routine evaluations of
liquefaction resistance.

Question B: A primary difference between eastern U.S. earthquakes and western U.S. earthquakes
is that strong ground motions generated by eastern earthquakes are generally richer in high-
frequency components. Consequently, eastern earthquakes of the same magnitude and duration may
generate more significant stress cycles than western events. Because of the likely larger number of
loading cycles in the east, should an additional correction factor be introduced for eastern U.S.
earthquakes?

Answer B: Although higher frequencies of motion and greater numbers of loading cycles may occur
at bedrock outcrops in the eastern U.S., their influence on the development of liquefaction may not
be significant. High-frequency motions generally attenuate or damp out rather quickly as they
propagate through soil layers. This filtering action would reduce differences in numbers of
significant loading cycles between eastern and western earthquakes at soil sites. Because
liquefaction occurs only within soil strata, differences in numbers ofloading cycles between eastern
and western earthquakes are not likely to be great. Without more recorded strong motion data from
soil sites in the eastern U.S. from which numbers of significant loading cycles can be quantified,
there is little basis for development of additional correction factors for eastern localities.

Another difference between eastern and western U.S. earthquakes is that strong ground motions
generally propagate to greater distances in the east. By applying state-of-the-art procedures for the
eastern U.S. for estimating peak ground accelerations, differences in ground motion propagation
between western and eastern earthquakes are properly accounted for. Thus use of standard criteria
developed largely for western U.S. earthquakes is acceptable for routine application in the eastern
U.S., provided peak accelerations are estimated using attenuation functions specifically developed
for eastern U.S. earthquakes.

194
 1/----- ----
9

Ms
..
8

~.,- ------
_1':'.!IMA-
,,..- .. ---
L ,:-_:::.-- mB
i------
7 ~ .,- ML
_., .,-'

----- ---·- ----

/~
,1/,&,-''
hV .,,,,,- mb

v·.,,-

//
/~#
'/~/
;' I I
I/
I
•
Scale Magnitude
I
4 I ML Local or Richter

3 ~
:7 ~II
I
M,

mb
Surface wave

Short-period body wave

I/
mB Long-period body wave

MJMA Japanese Meteorological Agency

3 4 5 6 7 8 9 10
Moment Magnitude, Mw

Figure 2. Relationships Between Moment Magnitude and Various

Magnitude Scales (after Heaton et al., 1986)

Question C: Which magnitude scale should be used by engineers in selecting a magnitude

parameter for liquefaction resistance analyses?

Answer C: Seismologists commonly calculate earthquake magnitudes using five different scales:
(1) local or Richter magnitude, ML; (2) surface-wave magnitude, M,; (3) short-period body-wave
magnitude, mb; (4) long-period body-wave magnitude, m 8 ; and (5) moment magnitude, Mw.
Moment magnitude, Mw, is the scale most commonly used for engineering applications and is the
scale preferred for calculation of liquefaction resistance. For example, engineering seismologists
have found that Mw generally correlates better than other scales with attenuation of peak ground
motions and spectral ordinates, and, consequently, with earthquake damage. Moment magnitude is
calculated from the amount of seismic energy released in earthquakes rather than peak displacement
of an accelerogram trace. Thus Mw is a more theoretically correct measure of earthquake strength
and duration than other magnitude estimates for engineering purposes.

Where estimates of moment magnitude are not available, magnitudes from other scales may be
substituted directly for Mw within the following limits: ML < 6, m 8 <7.5, and 6 <Ms < 8 (Figure 2).

195
The short-period body-wave magnitude, mb, a scale commonly used in the eastern U.S., may be used
for magnitudes between 5 and 6 provided such magnitudes are corrected to Mw using the curves
plotted in Figure 2. Because all magnitude scales, except Mw, saturate at some level (do not increase
in magnitude with increased seismic energy), substitutions, as listed above, should not be applied
beyond the given limits (Idriss, 1985).

Peak Acceleration

For calculation ofliquefaction resistance using the simplified procedure, peak horizontal acceleration
at ground surface is used to characterize the intensity of ground shaking at a site. Specific guidance
was not given in the original procedure for defining this parameter. To provide some guidance on
estimating peak acceleration for evaluation of liquefaction resistance, the workshop participants
addressed the following questions and developed the accompanying consensus answers.

Question D: What procedures are preferred for estimating peak horizontal ground acceleration at
potentially liquefiable sites?

Answer D: In practice, peak accelerations for use in analysis of liquefaction resistance are generally
evaluated from ground motion attenuation relationships. Estimates of peak acceleration may be
either deterministic or probabilistic. Deterministic estimates are based on magnitude and the
distance between the site and the seismic source zone for the design earthquake. Probabilistic
estimates are based on the largest acceleration likely to excite a site in a given exposure time (usually
50 years with two to ten percent probability of exceedance). Because the associated earthquake
magnitude is not generally known, a separate estimate of magnitude is generally required. Future
probabilistic hazard maps will likely include information on earthquake magnitude as well as peak
ground motion parameters.

The following three methods, in order of preference, may be used for estimating peak acceleration.

(1) The preferred method for estimating peak acceleration at a site is through direct application of
correlations between peak horizontal acceleration, earthquake magnitude, and distance from the
seismic energy source. Several investigators have developed and routinely apply correlations for
bedrock and stiff to moderately stiff soil sites. Which of these relationships should be used for
liquefaction hazard assessment requires consideration of such factors as region of the country, type
of faulting, local preference, etc. Some preliminary attenuation relationships have also been
developed for soft soil sites. For example, Idriss (1991) developed the attenuation relationships
plotted on Figure 3b from response of soft soil sites during the 1989 Loma Prieta earthquake. Also
shown in the figure (3a) are attenuation relationships developed by Idriss for rock sites.

In summary, the preferred procedure for estimating peak acceleration is through statistically reliable
attenuation relationships that were developed for site conditions similar to those at the site in
question.

196
 1
=(a)
on
~

·p
C
0
r- --- r-- ~ ......
--.... r-- :-- ,.._r·
;--.. ..._ ~
ro ~

,..::, 'r-
-
;..

' 'r- ~ C'.......

........
V
V

N~
u
u r-~~ ['.._
-....
~
ro 0.1
C
'
'
~
''
' '
I\.
'
'
Earthquake
, Magnitude = r7"""
0 ' " " "
·-::c:
N
; ..
0
' " '\
"\. \. \ "I\
'
I\
'
/ .'.1
J£

'\ '- \ r--

..::.:: \
ro \ \
V \. [\ ' fi ')
p..
\ ~
[\ ,
0.01 \
1 10 100
Distance - km

C
- ·- ~

·-~
r- . : "'I-- :
~ ::: - Earthquake _
0 I..,
Magnitude - =-
"'",.i,..
- .... r""'-,-. ' ..._ - ' ..._ •

-
.... "'t- 1--
V r--- ,,
V
u
u
' ' ' ...... ,.... /.':)
~
.......
' '
' '
·......
.....
.....
"'r-,.
,...

ca
.... 0.1
r-, ' ,
..
' ' ,:.,:
·-
C . ' "
0 ' '
N
·c::0 '
::c: '
..::.::
ro
'
~

0.01 - - - - - ~ ~ ~ ~ - - - - ~ ~ ~ - - - ~
1 10 100
Distance - km

Figure 3. Mean Peak Horizontal Accelerations at (a) Rock Sites and

(b) Soft Soil Sites (After Idriss, 1991)

197
(2) For site conditions that are not compatible with those specified for available attenuation
relationships, peak acceleration should be estimated from local site response analyses. Computer
programs such as SHAKE, DESRA, etc., may be used for these calculations. Input information
required for these analyses include well-documented soil profiles, measured or estimated shear-wave
velocity profiles, and time histories of acceleration (accelerograms). The time histories should
approximate anticipated future ground shaking either on a bedrock outcrop near the site or in stiff
soil layers or bedrock beneath the site. Recorded accelerograms are preferable to synthetic records.
Accelerograms derived from white noise should be avoided because of unnatural characteristics
commonly assimilated into these records. A suite of plausible earthquake records should be used
in the analysis, including as many records as feasible from earthquakes with similar magnitudes
recorded at similar distances from the seismic source as the site in question.

(3) A third and least desirable method for estimating peak ground acceleration is through
amplification ratios, such as those developed by Idriss (1990; 1991) and Seed et al.(1994 ). These
factors, commonly incorporated in building code provisions, use a multiplier or ratio by which
bedrock outcrop or stiff-site motions are amplified to estimate ground motions on soft sites. Because
amplification ratios are magnitude and perhaps frequency dependent, caution and considerable
engineering judgment are required in their application. Amplification ratios by Idriss (1991) for soft
soil sites are plotted on Figure 4. This plot shows ratios for magnitude 5.5 and 7 earthquakes that
are appropriate for sites underlain by several meters or more of soft soil similar in consistency to San
Francisco Bay mud.

·-.....
0
ro
~ 4
I Soft soil sites
c:: I rrna Tnitude"" 7
--
0
~ 3

~
u
·--E
t;::
0.
2
<r:
1
~~
Soft soil sites I
magnitude"" 51/2
~

---
0
0.1 0.2 0.3 0.4 0.5 0.6
Acceleration at Rock Sites - g

Figure 4. Magnitude-Dependence variations of Amplification of Peak

Horizontal Accelerations at Soft Soil Sites (After Idriss, 1991)

198
Question E: Which peak acceleration should be used? (a) the largest horizontal acceleration
recorded on a three-component accelerogram; (b) the geometric mean of the peaks from the
horizontal records; or (c) a vectorial combination of horizontal accelerations.

Answer E: According to I.M. Idriss (oral report at workshop), the larger of the two horizontal peak
components of acceleration were used in the original development of the simplified procedure, where
recorded values were available. Where recorded values were not available, which was the
circumstance for most sites in the data base, peak acceleration values were estimated from
attenuation relationships incorporating the geometric mean of the two orthogonal peak horizontal
accelerations. The geometric mean is defined as the square root of the product of the two orthogonal
peaks. In nearly all instances where recorded motions were used, the peaks from the two horizontal
records were approximately equal. Thus where a single peak was used, that peak and the geometric
mean of the two peaks were near the same value. Based on this information, the workshop
concurred that the geometric mean is more consistent with the derivation of the procedure, and
recommended use of the peak accelerations derived from attenuation relationships incorporating the
geometric mean. The larger of the two peaks, however, is always equal to or larger than the
geometric mean; thus use of the larger peak, a common procedure in engineering practice, is
conservative and allowable. Vectorial accelerations are seldom calculated and should not be used.

Question F: Liquefaction usually develops at soil sites where ground motion amplification may
occur, but also at localities where sediments may soften as excess pore pressures develop, reducing
peak acceleration. How should investigators account for both amplification and softening in
estimations of peak acceleration?

Answer F: With respect to site amplification and increased pore-water pressure, the recommended
procedure is to calculate or estimate a peak acceleration that incorporates the influence of site
amplification, but neglects the influence of excess pore pressure. Simply stated, the peak
acceleration to be used in liquefaction resistance evaluations is the peak horizontal acceleration that
would have occurred at ground surface in the absence of increased pore water pressure or
liquefaction.

Question G: Should high-frequency spikes in acceleration records be considered or ignored?

Answer G: In general, short duration, high-frequency acceleration spikes should be ignored for
liquefaction resistance evaluations. By using attenuation relationships for estimation of peak
acceleration, as noted above, high-frequency spikes are essentially neglected because few high-
frequency peaks occur in data bases from which attenuation relationships were derived. Similarly,
ground response analyses using programs such as SHAKE or DESRA generally attenuate or filter
out high-frequency spikes, reducing the influence of these spikes on site response. Where
amplification ratios are used to estimate peak accelerations, engineering judgment should be used
as to which bedrock accelerations should be amplified. Generally, however, high-frequency spikes
in bedrock records can be safely ignored for evaluation of liquefaction resistance.

199
 References

Idriss, I.M. (1985). "Evaluating Seismic Risk in Engineering Practice." Proceedings, 11th
International Conference on Soil Mechanics and Foundation Engineering, San Francisco, Vol.
1, p. 255-320.

Idriss, I.M. (1990). "Response of Soft Soil Sites During Earthquakes." Proceedings, H. Bolton Seed
Memorial Symposium, Vol. 2, BiTech Publishers, LTD, Vancouver, B.C. Canada, p. 273-90.

Heaton, T.H., Tajima, F., and Mori, A.W., 1986, "Estimating Ground Motions Using Recorded
Accelerograms." Surveys in Geophysics, Vol. 8, p. 25-83.

Idriss, I.M. (1991). "Earthquake Ground Motions at Soft Soil Sites." Proceedings, 2nd International
Conference on Recent Advances in Geotechnical Earthquake Engineering and Soil Dynamics,
Vol 3, p. 2265-71.

Seed, H.B., Idriss, I.M., Makdisi, F. and Banerjee, N., 1975, "Representation of Irregular Stress
Time Histories by Equivalent Uniform Stress Series in Liquefaction Analyses." Report No.
EERC 75-29, Earthquake Engineering Research Center, University of California, Berkeley,
Calif.

Seed, H.B., Tokimatsu, K, Harder, L.F., and Chung, R.F., 1984, "The Influence ofSPT Procedures
in Soil Liquefaction Resistance Evaluations, Report No. UBC/EERC-84/15, Earthquake
Engineering Research Center, University of California, Berkeley, California.

Seed, H.B., and Idriss, I.M. (1982). Ground Motions and Soil Liquefaction During Earthquakes.
Earthquake Engineering Research Institute Monograph.

Seed, R.B., Dickenson, S.E., Rau, G.A., White, R.K., and Mok, C.M., (1994) "Site Effects on Strong
Shaking and Seismic Risk: Recent Developments and Their Impact on Seismic Design Codes
and Practice." Proceedings, Structural Congress II, ASCE, (I ):573-78.

200
 LIQUEFACTION CRITERIA BASED ON STATISTICAL
AND PROBABILISTIC ANALYSES
T. Leslie Youd

Professor of Civil Engineering

Brigham Young University
Provo, Utah, 84602-4081

Steven K. Noble

Project Engineer
DOWL Engineers, Inc.
Anchorage, Alaska, 99503-5999

Abstract

Probabilistic procedures for evaluating liquefaction resistance have the advantage of allowing an
acceptable level of risk to be specified by the user. Liao and his colleagues used a logistic procedure
to develop probabilistic CRR curves. The original Seed and Idriss magnitude scaling factors,
however, were used to correct for magnitude. Youd and Noble (herein) use the logistic procedure
to analyze liquefaction resistance with a magnitude added as an independent variable. New case
history data and (N 1) 6ocs (corrected for fines content) were added to enlarge the case history data set.
Primary conclusions from the study are:

• The probabilistic procedure allows direct incorporation of an appropriate probability, or risk

factor in liquefaction hazard analyses. The procedure also provides a more scientifically rigorous
method of analysis of the data than the hand-shaped curves used in the simplified procedure.

• The analyses by Liao and his colleagues indicate, for clean sands, that the standard criteria from
the simplified procedure provide a probability of occurrence of about 20% for corrected blow
counts (N,) 60 between 11 and 28. Below an (N,) 60 of 11, the original simplified base curve is
characterized by a probability ofliquefaction smaller than 20%. Above an (N 1) 60 of 28, the curves
of Liao et al. indicate a probability of liquefaction greater than 20%. The curves in the upper par
of the range, however, are near the limit of liquefaction occurrences and are not well constrained
by empirical data.

• The analyses by Youd and Noble include magnitude as an independent variable eliminating the
need for magnitude scaling factors in the analysis. The Youd and Nobel results are more
conservative than those of Liao et al. for (N 1) 60cs less than 20 and characterize the simplified base
curve by probabilities ranging from 20% to 50%.

201
 Introduction

Liao et al. (1988) reviewed several studies involving statistical or probabilistic analyses of cyclic
stress ratio, standard penetration resistance, and field performance data that have been reported in
the geotechnical literature. Although useful for estimating probable error in evaluating liquefaction
resistance at field sites, these methods do not provide adequate quantification of conditional
probability for use in risk-type analyses. To provide a more quantitative and direct model for risk-
type analyses, Liao et al. used a logistic function along with statistical regression to quantify the
probability ofliquefaction as a function of cyclic resistance ratio (CRR) and (N 1) 60 •

As part of their study, Liao et al. reevaluated all known liquefaction case-history data as of 1988,
using data contained in several catalogues, including the 125 case histories compiled by Seed et al.
(1985). Because many case study sites are characterized by incomplete or uncertain information,
Liao et al. disregarded some site data. They finally selected 278 sites that they classified as
"reliable" case studies that could be used for statistical regression analyses. Even so, the data from
many of these sites were imperfect, requiring engineering judgment in selecting some property
values. In their reanalysis of borehole data, Liao et al. generally used the minimal standard-
penetration blow-count measured in granular layers as the critical blow count for their analyses.

Loertscher and Youd (1994) and Youd and Noble (this paper) extended the analyses of Liao et al.
to include magnitude as an additional independent variable and added several new case histories to
the data set. The work of Youd and Noble was specifically conducted for this workshop to evaluate
the use of probabilistic and statistical procedures both directly for evaluation of liquefaction
resistance (this paper), and as a method for evaluating magnitude scaling factors for use with the
simplified procedure (Magnitude Scaling Factors, this report). The analyses and results reported
herein were developed after the formal workshop event, and hence were not discussed or approved
during the workshop discussions.

Logistic Analysis by Liao and His Colleagues

Clean Sands

Liao et al. (1988) conducted statistical regression analyses using data from sites underlain by clean
sands (fines content equal to or less than 12%) and sites underlain by silty sands (fines content
greater than 12%). Probabilistic regression curves from the analysis on clean sand sites are
reproduced in Figure l. The regression equation for these curves, which could be applied directly
in practice, follows:

QL = 16.477 + 6.4603ln(CSRN) - 0.39760(N 1) 60 (1)

where QL = logit (PL)= ln[(PL/(1-Pd]. PL is defined as the probability that liquefaction will occur,
1-PL is the probability that liquefaction will not occur, and CSRN is the cyclic stress ratio generated
at the site normalized to a magnitude of 7.5. The magnitude scaling factors published by Seed and

202
 0.6 r----------------------,

~
t:tJ 0.5
.99 .95 .9 .5 .1 .05
u +
·-~
0

~ 0.4 +

~
t:IJ +

-·-
U
(.)

~
0.3
+ ++

't::I ++
+
~
·-
~ + 0

1 0.2
0

] 0

1·-
0
i::
0 0 0
0.1 00 0 0
0
0
0
0

0
0
0.0
0 10 20 30 40 50

Corrected / Normalized SPT Value (N 1)60

Figure 1 Contours of Equal Probability of Liquefaction, Pu for Clean Sand (Fines
Content Less than 12%); Plus Symbols Indicate Data from Sites where
Liquefaction Occurred; Circles Indicate Data where Surface Evidence of
Liquefaction Did Not Occur (modified from Liao, 1996)

Idriss (1982), which have been conventionally used in liquefaction hazard analyses, were applied
by Liao et al. to normalize CSR to CSRN. PL can be explicitly calculated as

(2)

The primary advantage of Equation 1 for engineering applications is that the user can select an
appropriate probability of exceedance or risk of occurrence for analyzing liquefaction hazard. For
example, for noncritical sites, a probability ofliquefaction of20% to 30% might be appropriate.

203
 0.6
I I
/
Liao et al. (1988), PL =20% I ,'
l I I I
Liao et al. (1988), PL =50% I I
0.5 ' I ,' I
/
Youd and Noble, PL=20% I I
I I
Youd and Noble, PL =50% I
I I
I /

0.4 Seed et al. (1985) ,I

0 Simplified Base Curve

!oo
r/l
0.3
II)
.!=
r:r:,

·-u
u
;;,-.
0.2 , ,
u ,,

0.1

0 _ _ __.__ _ _....._______.....__ __.__ _ __.__ ___._ _ _ ~

0 5 10 15 20 25 30 35
Corrected Blowcount, (N 1)60
Figure 2 Comparison of Probabilistic Results from Equation 1 (Liao et al., 1988),
Equation 6, and the Simplified Base Curve

If more conservative criteria are being applied in other aspects of the seismic analysis, such as
earthquake selection, a probability on the order of 5% to 10% might be appropriate.

Because the cyclic resistance ratio versus (N 1\ 0 curves developed by Seed et al. (1985) for the
simplified procedure were conservatively drawn by hand, the probability of liquefaction was not
determined, nor was the probability of occurrence constant across the trajectory of the curves.
Figure 2 shows clean sand curves developed by Liao et al. (1988) for magnitude 7.5 earthquakes
along with the base curve for clean sands from the simplified procedure, hereafter called the
simplified base curve. (Probabilistic curves from Youd and Noble are also included on this
diagram.) Between corrected blow counts, (N 1) 60 , of 11 and 28, the simplified base curve lies near
the 20% probability curve of Liao et al. Below an (N,) 60 of 11, the original simplified base curve
becomes very conservative compared to the curves of Liao et al. The primary reason for this
divergence is that Seed et al. (1985) projected their curve through the origin of the plot, whereas Liao
et al. let the regression analysis shape the trajectories of the probabilistic curves. As a consequence,
the trajectories of probabilistic curves flatten to meet the (N 1) 60 = 0 ordinate at cyclic stress ratios of

204
 0.6 .-------r------,---"'T"'""-----.------,
Silty Sand Contours
~
.9 .5 .1
0.5
,''
I
6 .9 .' .5 .l ,:
-~ ' ,
~ ,'' ,I
, ,
,' ,
,, ,'
,,

0.3 ,
,,
, I

0.2

,,
,,
0.1
--
--
0.0
0 10 20 30 40 50
Corrected / Normalized SPT Value (N 1)60
Figure 3 Comparison of Probabilistic CRR Curves from Clean and Silty Sands
(modified from Liao, 1996)

0.055 and 0.078, respectively. Above an (N 1) 60 of 28, the curves diverge with the simplified base
curve rising steeply to become asymptotic with the (N 1) 60=30 ordinate, while the 20% probability
curve continues to curve gently upward. There is very little case history data to constrain the curves
in either the lower or upper regions where they diverge.

Silty Sands

The analysis of silty sands (fines contents greater than 12%) by Liao et al. are plotted on Figure 3
along with the previously reported curves for clean sands. In this instance, the 50% curve lies
significantly to the left of the equivalent clean sand curve, indicating that silty sands are generally
more resistant to liquefaction than clean sands based on (N 1) 60 criteria. The curves for silty sand are
reproduced on Figure 4 along with CRR curves for various fines contents developed by Seed et al.
(1985). The 50% probability curve for silty sands from Liao et al. lies between the 15% and 35%
fines content curves of Seed et al. Thus the mean or 50% curve from Liao et al. is in reasonable
agreement with the higher fines content curves of Seed et al.

205
The width of the band of probabilistic curves between 10% and 90% probability, however, is very
wide for silty sands. These curves extend well beyond the equivalent clean sand curves on either
side of the plot (Figure 3). The latter relationship indicates that at the I 0% probability level, greater
(N 1) 60 is required to prevent silty sands from liquefying than is required for clean sands. This
relation is opposite to the relationship suggested by Seed et al. and the relationship at the 50%
probability level.

The wide range of the probabilistic curves for silty sands is largely a consequence of the large
amount of scatter in the observational data. Possible contributors to this scatter include imprecisions
in reported fines contents and the possible influence of soil plasticity. Fines contents at many
investigated sites may be too variable within soil layers to be easily or adequately characterized by
a single number. The plasticity of the fines may also have a significant influence on liquefaction
resistance that is not accounted for in the present criterion. The scatter may also indicate large
variances in the liquefaction behavior of silty sands and sandy silts. In any event, the uncertainty,
as demonstrated by the curves for silty sands in Figure 3, is so large that the use of the probabilistic

0.6 ---~-----,.---...--------.------,

35% 15%
6 0.5
.9 : : .5 .1
·.g
~
<Ll
<Ll

t
f/J
0.4

-.....
u
u
U>-> 0.3
"O
11)
.....N
~
§ 0.2
0
::::
11)
"O
Seed et. al. (1984, 1985)
.S
..... 0.1 curves for FC = 15% and
::::
OJ)
FC =35%
Cll
::;E
0.0 ..___ _...___ ___,___ ____._ _ _..___ ____.
0 10 20 30 40 50
Corrected / Normalized SPT Value (N 1)60
Figure 4 Comparison of Probabilistic CRR Curves for Silty Sands (Fines Contents
Greater Than 12%) with Deterministic CRR Curves from Seed et al., (1985)
for Sands with Fines Contents of 15% and 35% (modified from Liao, 1996)

206
curves is questionable for silty materials, at least at low- or high-probability levels. This scatter also
raises doubts concerning the reliability of simple correction factors for fines content as presently
applied in evaluations of liquefaction resistance of silty soils.

Applications

Following the workshop, S.S.C. Liao submitted the following information: There have been many
applications where a probabilistic evaluation of liquefaction has been incorporated as part of a risk
analysis framework. Projects include tailings and earth dams (Vick, 1994) and regional hazard
studies (Budhu et al., 1987; Hashash, 1987). In these studies, liquefaction probability was based on
the logistic regression equations derived by Liao et al. (1988). Ostadan et al. (1991) incorporated
the equations of Liao et al. (1988) into a computer program combined with earthquake hazard
models, which were used in risk analyses for a nuclear material storage facility (Arango et al., 1996).
Similarly, the Liao et al. (1988) model formed the basis of the liquefaction probability calculations
(RMS, 1997) in a project to develop a standardized earthquake loss estimation methodology
embodied in computer program HAZUS. This program was funded by the Federal Emergency
Management Agency (FEMA) and managed by the National Institute of Building Sciences (NIBS).
This program has been used in earthquake economic loss studies for Portland, Oregon (Dames and
Moore, 1996) and Boston, Massachusetts (EQE/PB, 1997). In general, however, application of
probabilistic analysis is beyond the normal practice of most geotechnical engineers. Hence the
workshop participants did not approve recommendations for engineering practice, but did encourage
continued development of these concepts.

Probabilistic Analyses by Loertscher and Youd and by Youd and Noble

Loertscher and Youd ( 1994) extended the analyses of Liao et al. to include earthquake magnitude
as an independent variable and added new case histories to the data base from earthquakes with
magnitudes less than 7. Most of the data analyzed by Loertscher and Youd were taken from
previous compilations of case history data, including Seed et al. (1985), Liao (1986), Ambraseys
(1988), and Bartlett and Youd (1992). Loertscher and Youd added new case histories from the
Marina District of San Francisco, where abundant liquefaction effects developed in response to the
1989 Loma Prieta earthquake (Mw = 6.8), but where liquefaction effects were not reported following
the 1957 Daly City earthquake (Mw = 5.3). The Loma Prieta event generated peak accelerations
estimated at about 0.15 g to 0.25 g in the Marina District (Bartlet et al., 1992). Peak accelerations
during the Daly City event were estimated at about 0.20 g (Loertscher and Youd, 1994). Data were
also added from sites where liquefaction effects were observed or were not observed after several
Imperial Valley, California earthquakes: 1979 Imperial Valley (Mw = 6.6), 1981 Westmorland (Mw
= 6.0), 1987 Elmore Ranch (Mw = 6.2), and 1987 Superstition Hills (Mw = 6.6). New data were also
added from several sites underlain by saturated granular materials that were strongly shaken by the
1987 Whittier Narrows event (Mw = 5. 9), but liquefaction effects were not reported.

Loertscher and Youd ( 1994) analyzed the compiled data set using a logistic regression analysis
similar to that used by Liao et al. (1988). The analyses, however, yielded unexpected differences

207
with results published by Liao et al. (1988) and also conflicted with results embodied in the
simplified procedure. Youd and Noble (herein) corrected some inconsistencies in the Loertscher and
Youd data set, and reanalyzed the data to verify the results of Loertscher and Youd. They also
expanded the analysis to more fully consider issues such as the form of the probabilistic equation.

Following the procedures of Liao et al., Youd and Noble (herein) analyzed data from sands with
fines content less than 12%. Unfortunately, there were insufficient data in that set to constrain the
analysis. To increase the amount of data, sands with fines contents up to 35% were used by
correcting (N 1) 60 for fines content. Corrections were made by increasing (N 1) 60 to an equivalent
clean-sand corrected blow count, (N 1) 60c,, using the fines content correction factor recommended by
Idriss and Seed as noted in the Summary Report (this report).

(3)

where a and p are coefficients determined from the following equations:

a= 0 forFC :$ 5% (4a)
a= exp[l.76 - (190/FC 2)] for 5% < FC < 35% (4b)
a= 5.0 forFC ~ 35% (4c)

P= 1.0 for FC :s, 5% (5a)

p = [0.99 + (FCu/1000)] for 5% < FC < 35% (5b)
p = 1.2 forFC ~ 35% (5c)

where FC is fines content measured from laboratory gradation tests on retrieved soil samples. The
final data set is available at http://www.et.byu.edu/-cewww/faculty/youd/papers/liqdata.xls.

The addition of the fines-content corrected penetration resistances, (N 1) 60c., provided sufficient data
to adequately constrain the analyses for magnitudes between 5.75 and 7.75 to yield meaningful
results. At lower magnitudes, between 5.25 and 5.75, constraint is provided by approximately 25
sites where effects of liquefaction were not observed, but only one site where effects were observed.
Case history data are not available for magnitudes smaller than 5.25, which is approximately the
threshold magnitude for generation of liquefaction.

The results of the Youd and Noble reanalysis are incorporated in the following equation, which is
similar in form to the equation developed by Liao et al. (Equation 1), except that magnitude, Mw, is
added as an independent variable.

(6)

(N 1) 6ocs is the corrected blow count, including a correction for fines content. Equation 6 can be
rewritten for calculation of CRR as follows:

208
 Table 1 Case History Data from Liquefied Sites Predicted as Nonliquefiable
by the Simplified Base Curve for Magnitude 7.5 Earthquakes

Year Locality/Site or Hole No. Data MW Depth Fines (N,)60 (N,)60cs CSR
Set (m) Content

1964 Niigata/G 10-22 B-Y' 7.5 4 12 13.5 15.l 0.165

1964 N iigata/H 10-45 B-Y 1 7.5 3 8 12.6 13.3 0.140

1976 Tangshan/Luan Nan Seed2 7.6 5.4 3 22.2 22.2 0.169

1976 Tangshan/Coastal Region Seed 2 7.6 3 10 11.4 12.5 0.124

1978 Miyagiken-Oki/Yurlag Br. 2 Seed2 7.7 3.3 7 22.5 22.9 0.229

'Data from Bartlett and Youd (1992) comprlat10n of case h1stones
2
Data from Seed et al. ( 1984) compilation of case histories

In CRR = 2.466 - O.7289Mw + 0.O834(N 1)6ocs + O.3231 ln[PL/(1-PL)] (7)

This form is useful for routine engineering calculations ofliquefaction resistance where Mw, (N 1) 60 c,,
and PL are determined from seismic, site, and risk evaluations, respectively. Curves derived from
Equation 6 for magnitude 7.5 earthquakes are plotted on Figure 5. In this instance, the abscissa for
the plot is the fines-content corrected blow count, (N 1) 6oc,, rather than (N 1) 60 . Case history data, in
terms of CSR and (N,\ow for earthquakes with magnitudes between 7.25 and 7.75 are also shown
on the plot. The simplified base curve is plotted on Figure 5 for reference.

Unexpected results from this analysis are that the probabilistic curves of Youd and Noble generally
plot much lower on the diagram than those of Liao et al. for clean sands and (N 1) 600, less than 25
(Figure 2), and that the simplified base curve is enveloped between the 20% and 50% probability
curves for (N 1) 60 between 4 and 25 (Figure 5). The latter result indicates that the simplified base
curve may be characterized by higher probabilities ofliquefaction than previously thought for (N 1) 60c,
between 5 and 25. The argument that the simplified base curve is characterized by probabilities of
liquefaction between 20% and 50% in this (N 1) 6ocs range is supported by several data points
indicative of liquefaction that are misclassified by the simplified base curve. Information on the
misclassified data are listed in Table l. Three of these data are from case histories compiled by Seed
et al. (1984) but are characterized by fines contents greater than 5%, and hence were not plotted on
diagrams such as Figure 1. These data most likely had strong influence on the results of the logistic
analyses as well as similar data from other magnitude ranges in the data base. As noted, the
implication of this result is that the simplified curve may not be as conservative as previously
thought and perhaps not as conservative as generally desired for engineering practice.

Probabilistic CRR curves derived from Equation 6 are also plotted on Figure 2 for comparison with
the clean-sand curves of Liao et al. The curves of Youd and Noble have the same general shape, but

209
 •
••
•
0

0
Magnitude, Mw, = 7.5 ± 0.25
Legend
Liquefaction datum o
Non-liquefaction datum •
Seed, et. al. base curve - - •

0.00
0 10 20 30 40 50

Corrected Blow Count, (N 1)60cs

Figure 5 Probabilistic CRR Curves Derived from Equation 6 with Simplified

Base Curve and History Data Plotted for Reference

are generally steeper in aspect and intersect the origin [(N 1) 60=0 ordinate] at lower CSR than those
of Liao et al. For (N 1) 60 less than 20 and 27, respectively, the 50% and 20% curves of Youd and
Noble lie below and are more conservative than the equivalent curves of Liao et al. Conversely,
above blow counts of 20 and 27, respectively, the curves of Youd and Noble lie above, are less
conservative, and are much steeper in aspect than the equivalent curves of Liao et al. Also, the
spread between the 20% and 50% curves of Youd and Noble is greater across the plot than those of
Liao et al. The increased spread may be due in part to the wider range of fines contents incorporated
in the Youd and Noble data set. Other factors that could have caused differences between results of
the two probabilistic investigations include the additional case history data used by Youd and Noble

210
 Magnitude, Mw, = 9.0 ± 0.25
Legend
Liquefaction datum o
I
Non-liquefaction datum • I
PL= 50% - - - · I
I I
PL= 32% - ·- · I
I
PL= 20% · ··-· I

Simplified base curve --

• I
I

I
I
• I
I
I
0.30 •
-+-----+-,,.-----t-----1'------t----,t---t---t-------i
I

• I
I
I

I
I

I
I I

• , , I

, , 0

, ,, /
/

0.10
,,;

- -
-
., , . .... ..
.,,. .. ..,, ......
.. - . . .
0.00 ~--+----1---+---+----+----+---+----t---+---I
0 10 20 30 40 50

Corrected Blow Count, (N 1)6ocs

Figure 6 Probabilistic CRR Curves Developed from Equation 6 for Mw = 9.0
with Empirical Data from the 1964 Great Alaska Earthquake (Mw = 9.2)

and the incorporation of magnitude as an independent variable. Youd and Noble incorporated 369
data points into their analysis compared to 278 points by Liao et al. Most of the additional points
are from earthquakes with magnitudes less than 7.

The lack of agreement with the results of Liao et al. and the indication that the simplified base curve
is characterized by probabilities of liquefaction as great as 50% was the primary reason for the
reanalysis of the Loertscher and Youd data by Youd and Noble (herein). In the reanalysis, the form

211
of the logit equation was simplified (cross terms removed) and a few errors in the data set were
corrected. The results of the reanalysis, however, are in rather close agreement with results of
Loertscher and Youd (1994). That is, for a given PL, Mw, and (N1\ 0 , CRR calculated from Equations
5 or 6 are consistently larger than those estimated by Loertscher and Youd, but by no more than a
few percent for PL between 20% and 80%, (N 1)6ocs less than 30, and Mw between 5.5 and 7.5. Based
on this reevaluation, Youd and Noble conclude that Equation 6 is a correct probabilistic assessment
of the case history data set, assuming the forms of Equations I and 6 are properly formulated.

Equation 6 is sufficiently constrained by data to be valid for assessing liquefaction resistance (CRR)
for magnitudes between 5.75 and 7.75 and CRR less than 0.4. Extrapolation to magnitudes less than
5. 75 is justified by case history data from sites that did not liquefy and from the fact that the
threshold magnitude for generation of liquefaction is about magnitude 5.

Insufficient case history data are available to adequately constrain the probabilistic regression
analysis for magnitudes greater than 7.75. For example, Figure 6 shows probabilistic curves for
magnitude 9 earthquakes along with empirical data for earthquakes with magnitudes between 8.75
and 9.25. The empirical data for this magnitude range is sparse, with data from only a few sites
where liquefaction effects were observed following the 1964 great Alaska earthquake (Mw = 9.2).
The data are even more sparse for magnitudes between 8.25 to 8.75 where no case histories are
included in present data bases. The observational data for magnitude 7.75 to 8.25 earthquakes are
also rather weak with data coming primarily from three pre-1925 earthquakes: the 1991 Mino-Owari,
Japan (M = 7.8), 1906 San Francisco, California (Mw = 7.9), and 1923 Kwanto, Japan (Mw = 7.9).

The probabilistic curves shown on Figure 5 appear overly conservative with respect to the limited
data plotted on the figure. Similarly conservative results are estimated for all magnitudes greater
than 8. Because the probabilistic estimates are not adequately constrained by empirical data at these
higher magnitudes, Equation 6 should not be used for magnitudes larger than 7.75. For magnitudes
greater than 7.75, the workshop participants recommend use of the simplified procedure with the
magnitude scaling factors (MSF) proposed by Idriss (Summary Report, this report).

Based on the above arguments, Youd and Noble (herein) recommend that Equations 6 and 7 with
PL of 20% to 30% should be used for evaluation of liquefaction resistance for earthquakes with
magnitudes less than 7.75. This conservatism yields PL with less than mean minus one standard
deviation (32%) probabilities that liquefaction will occur. That level of conservatism is generally
acceptable in engineering practice for noncritical structures.

Conclusions

The probabilistic analyses reviewed and conducted herein lead to the following conclusions:

l. The probabilistic procedure allows direct incorporation of probability of liquefaction

occurrence or acceptable risk into liquefaction hazard analyses. Thus the engineer is given
the option of specifying a level of risk as part of the analysis of liquefaction hazard.

212
 References

Ambraseys, N.N., 1988, "Engineering Seismology," Earthquake Engineering and Structural

Dynamics, Vol. 17, p. 1-105.

Arango I., Ostadan, F ., Lewis, M.R., and Gutierrez, 1996, "Quantification of Seismic Liquefaction
Risk", Proceedings, ASME Pressure Vessel and Piping Conference, Montreal, Canada, 1996.

Bardet, J.P., Kapuskar, M., Martin, G.R., and Proubet, J., 1992, "Site Response Analysis," The Loma
Prieta, California, Earthquake of October 17, 1989--Marina District (T.D. O'Rourke, ed.): U.S.
Geological Survey Professional Paper 1551-F, p. 84-140.

Bartlett, S.F., and Youd, T.L., 1992, Empirical Analysis of Horizontal Ground Displacement
Generated by Liquefaction-Induced Lateral Spread, National Center for Earthquake Engineering
Research, Technical Report NCEER-92-0021.

Budhu, M. Vijayakumar, V. Giese, R. F., Baumgras, L., 1987, Liquefaction Potential For New York
State: A Preliminary Report on Sites in Manhattan and Buffalo, Technical Report NCEER-87-
0009, August 1987), National Center for Earthquake Engineering Research (NCEER), State
University of New York, Buffalo, NY.

Dames and Moore, 1996, Summary Results, Earthquake Loss Estimation Pilot Study for the
Portland Metropolitan Region, Report prepared by Dames & Moore, Inc. (D&M), for the
National Institute of Building Sciences, under the sponsorship of the Federal Emergency
Management Agency, December 1996.

EQE/PB, 1997, Summary Results, Earthquake Loss Estimation Pilot Study for the City of Boston,
Report prepared by EQE International, Inc. and Parsons Brinckerhoff (PB), for the National
Institute of Building Sciences,for Federal Emergency Management Agency, in press.

Hashash, Y.M.A., 1987, "Liquefaction Probability Mapping in Greater Boston" unpublished M.S.
Thesis, Department of Civil Engineering, Massachusetts Institute of Technology Cambridge, ·
Massachusetts, 1987.
Liao, S.S.C., 1986, Statistical Modeling in Earthquake-Induced Liquefaction, unpublished PhD
Dissertation, Massachusetts Institute of Technology, Cambridge, Massachusetts.

Liao, S.S.C., 1996, Discussion of "Reconsideration oflnitiation of Liquefaction in Sandy Soils" by

C.E. Fear and E.C. McRoberts, Journal ofGeotechnical Engineering, Vol. 122, No. 11, p. 957-
959.

Liao, S.S.C., Veneziano, D., and Whitman, R.V., 1988, "Regression Models for Evaluating
Liquefaction Probability," Journal ofGeotechnical Engineering, Vol. 114, No. 4, p. 389-411.

214
Loertscher, T. W., and Youd, T.L., 1994, Magnitude Scaling Factors for Analysis of Liquefaction
Hazard, unpublished Research Report No. CEG.94-02, Department of Civil and Environmental
Engineering, Brigham Young University, Provo, Utah.

Ostadan, F., Marrone, J., Liteheiser, J.J., Arango, I., 1991, "Liquefaction Hazard Evaluation:
Methodology and Application," Proceedings, Third US. Conference on Lifeline Earthquake
Engineering, Los Angeles, CA, ASCE, pp. 591-600.

RMS, 1996, Development of a Standardized Earthquake Loss Estimation Methodology, Draft

Technical Manual, Prepared for National Institute of Building Sciences, under the sponsorship
of Federal Emergency Management Agency, 1996.

Seed, H.B., and Idriss, I.M., 1982, Ground Motions and Soil Liquefaction During Earthquakes,
Earthquake Engineering Research Institute Monograph.

Seed, H.B., Tokimatsu, K, Harder, L.F., and Chung, R.F., 1984, "The Influence ofSPT Procedures
in Soil Liquefaction Resistance Evaluations," Report No. UBC/EERC-84/15, Earthquake
Engineering Research Center, University of California, Berkeley, California.

Seed, H.B., Tokimatsu, K., Harder, L.F., and Chung, R.M., 1985, "The Influence ofSPT Procedures
in Soil Liquefaction Resistance Evaluations," Journal ofGeotechnical Engineering, ASCE, Vol.
111, No. 12, p. 1425-45.

Vick, S.G., 1995, "Geotechnical Risk and Reliability--From Theory To Practice in Dam Safety,"
Proceedings, The Earth, Engineers, and Education--A Symposium in Honor of Robert V
Whitman, October 7 and 8, 1994, Dept. Of Civil Engineering, Massachusetts Institute of
Technology, Cambridge, Massachusetts, pp. 45-58.

215
216
 Liquefaction Criteria Based on Energy
Content of Seismograms

T. Leslie Youd
Professor of Civil Engineering
Brigham Young University
Provo, Utah 84602-4081

Robert E. Kayen
Research Civil Engineer
U.S. Geological Survey
Menlo Park, California 94025

James K. Mitchell
University Distinguished Professor
Virginia Tech
Blacksburg, Virginia 24061-0105

Abstract

Because energy is a more fundamental property or measure of earthquake excitation than peak
acceleration and magnitude, several investigators have suggested that the amount of energy generated
at specific points in a soil layer in response to earthquake shaking may correlate better with the
development of liquefaction than with cyclic stress ratio. Kayen (1993) reviewed the literature on
this topic and investigated the use of accelerogram energy, expressed in terms of Arias intensity, as
a parameter for use in evaluating liquefaction resistance of soils. The use of Arias intensity as a
measure of seismic excitation has the advantage of eliminating other seismic parameters from the
analysis, including peak acceleration, magnitude, and a magnitude scaling factor. Although the
workshop agreed that use of accelerogram energy presents an important new direction in analysis
ofliquefaction resistance, the procedure is not yet sufficiently verified to be recommended for use
in engineering practice. The following questions need to be addressed: What procedures should be
used, and how accurately can Arias intensity be estimated for application at field sites where strong
motion records are not available? Has sufficient case history data and experience been compiled and
analyzed to provide adequate verification of the procedure? The workshop encourages continued
research to answer these questions and further develop energy procedures for use in engineering
practice.

217
 Introduction

Several investigators have proposed the use of energy content of accelerograms as a measure of the
excitation or demand placed on a liquefiable soil by earthquake shaking (Davis and Berrill, 1978;
Figueroa and Dahisaria, 1991). Use of energy input at specific points in a soil layer obviates the
need for other seismic parameters, such as peak acceleration, earthquake magnitude, and magnitude
scaling factors as presently used in the simplified procedure. These investigators claim that energy
is a more fundamental measure of earthquake excitation and should correlate better with insitu
generation of pore pressures than the presently used cyclic stress ratio. In a Ph.D. dissertation, Kay en
(1993) reviews the literature on this topic and investigated the use of accelerogram energy, expressed
in terms of Arias intensity, as a tool for evaluating liquefaction resistance of soils. The workshop
participants briefly considered the use of energy content of accelerogram records in evaluating
liquefaction resistance, but concurred that the present state of development and verification is
insufficient to recommend the procedure for general use in engineering practice. Nevertheless, the
workshop encourages further research and development of the procedure for possible future
implementation.

The following text, largely summarized from Kayen's Ph.D. dissertation (Kayen, 1993), provided
the basis for the workshop discussion. Kay en and Mitchell ( 1997) further summarize these analyses
and propose procedures for engineering application.

Energy Approach

Rather than using cyclic stress ratio, Kayen (1993) correlated liquefaction resistance with Arias
intensity as a measure of the energy content of accelerograms recorded at several localities where
surface effects of liquefaction were or were not observed following major earthquakes. Arias
intensity is defined as the total energy per unit weight absorbed by single-degree-of-freedom-
undamped oscillators evenly spaced in frequency from O to oo when excited by a time history
(accelerogram) of earthquake motions. The total horizontal Arias intensity is calculated from the
following equation:

Ih =!xx+ IYY = 2:_f

2g O
1
.a;(t) dt + 2:_J .a2ct) dt
2g O
1

y
(1)

where Ih is total two-component horizontal Arias intensity, Ixx and IYY are horizontal components of
Arias intensity in the x and y directions, respectively, a,,(t) and a,,(t) are acceleration time histories
from strong motion accelerograms in the x and y directions, and g is the acceleration of gravity. The
parameter Ih represents the sum of intensities contributed by the two horizontal components of
motion. Because the time integral of the squares of the accelerations are normalized by the
acceleration of gravity, the dimensional units of Arias intensity are length divided by time, or
velocity (Kayen, 1993, p. 29). The general practice is to use the metric system of units to define
Arias intensity, yielding intensities in units of meters-per-second. Intensity values should be
calculated from corrected accelerograms by the trapezoidal integration method.

218
 2 seismogram statistics and
profiles field data 10
4 I
/
/
/
6 I
I
/
I
8 I
I
I
10 I
I
I
12 Afean
1
40
-lo
(P= -1) (P-0) ,' (P= !11)
14
I ( :
0 10

Depth Reduction Factor for Arias Intensity, rb

Figure 1 Normalized Arias Intensity versus Depth-Reduction Profiles Modeled Using
SHAKE: (a) Profiles from Several Synthetic Seismograms; (b) Synthesis of
Seismogram Profiles (modified from Kayen and Mitchell, 1997)

Kayen calculated Arias intensity values for a number of localities where strong motions were
recorded and where surface effects of liquefaction were or were not reported. From these
calculations, a data set was compiled that included Arias intensities, site soil profiles with measured
penetration resistances, soil types, grain size properties, etc. To estimate the severity of earthquake
motions within the soil column, a depth-reduction factor for Arias intensity, rb, which is analogous
to rd, was developed with the ground response program SHAKE (Kayen and Mitchell, 1997). This
analysis showed that rb on average diminishes from 1.0 at ground surface to 0.58 at a depth of 6 m,
and further diminishes to a value of0.46 at depths of 10 m or below (Figure 1). The Arias intensity
at depth in the soil column, Ihb• can be calculated as follows:

From analyses of the compiled data set, Kayen and Mitchell (1997, p. 1169) reported the following:

Based on these case studies of known field-behavior during earthquakes, a relation

was developed between the measured liquefaction resistance of the soil and the Arias
intensity characteristics of nearby strong motion recordings. The data included in
this study are limited to sites where a direct measure or reasonably well-constrained
estimate of the Arias intensity could be made. ... The resultant association finds
excellent segregation of liquefaction and non-liquefaction points on a plot of the

219
 (b) Liquefaction 0
23
60 61
62 6764
♦ ♦ ♦ ♦♦
66+ ♦2422*
25 +z7 59
•
1

3 9
• 10
* 4 ••
7 •

17
•
16
~ 3~i ~
38
f ♦4
31 i
◊i
36 37 29 ◊ 3J
9
34 ·
◊ •J2 o H ~
50 72 53
♦ 51 il~J ◊ 58
52 ◊ ◊ 55 ◊
"' 56 49 ◊
Clean-sand boundary
No liquefaction

0 .1 L.J...L~,-J....UL.J...L..1,().1...L()J...u..J~L.J..L~.J....U .....................1-QL..L..LJU-J,.,Q<;.........U-LLU...L..LJ~J..I 0.1

0 5 10 15 20 25 30 0 5 10 15 20 25 30
(N 1)60 (N 1)6ocs

Figure 2 Ihb versus (N 1} 60 from Compiled Case Histories: (a) Plot Without Fines
Content Correction; (b) Plot with Fines Content Correction
(modified from Kayen and Mitchell, 1997)

estimated Arias intensity at depth (Ihb) versus values of (N 1) 60 [Figure 2a]. A

boundary separating the liquefaction and non-liquefaction fields has an ordinate
intercept (Ni) 60 =0, of approximately Ih=0.22 meters/second. The lower-bound
threshold indicates that a minimum Arias intensity is needed to trigger liquefaction
even in extremely loose soil.

Kayen (1993, p.195) notes the following regarding the boundary curves presented in Figure 2:

From the plot presented in Figure 2, it can be seen that the field penetration resistance
used as a measure of soil resistance to liquefaction correlates with accelerogram
energy in a similar way as it does with cyclic stress ratio. That is, the liquefaction
boundary in Ihb-(N 1) 60 space has an upwardly concave curvature similar to that of the
familiar CSR-(N 1) 60 and CSR-qc 1 correlation. This curvature indicates that there is
a limiting upper-bound penetration resistance above which liquefaction cannot occur,
regardless of the strong motion input.

220
In Figure 2b, fines content correction factors for silty-sand were applied to the SPT data set
following the guidelines given in the Summary Report (this report). Kayen and Mitchell (1997)
report the following conclusions from this study:

The correction factors for silty sands were used to convert the SPT data set ... to
equivalent 'clean sand' SPT values. The clean-sand boundary curve presented in
Figure 2b is a singular earthquake magnitude-independent boundary curve that
delineates the threshold condition of initial liquefaction in Ihb-(N 1) 60 space. We found
that the application of the (NCEER Workshop) fines-content correction factor above,
segregates our data set almost entirely into distinct zones of liquefaction and non-
liquefaction occurrence, such that the boundary clearly envelopes the liquefaction
frontier.

The qc, relation of Robertson and Campanella (1985) and CPT data collected during post-earthquake
investigations of Loma Prieta liquefaction test sites allowed for the development of normalized-cone
penetration resistance-Arias Intensity-space boundary curves as shown in Figure 3 (Kayen and
Mitchell, 1997). To address the need for predictive models for Arias intensity at sites where strong
motion records are not available, or for design-basis earthquakes, Kayen and Mitchell (1997) present
equations describing surface-measured Arias intensity in terms of moment magnitude, Mw, and the
Pythagorean source distance, r*, from the investigation site to the closest point to the fault rupture
plane at the focal depth:

... earthquake motion and site characteristics data were tabulated for 66 earthquake
records in the western United States, primarily from California (Kayen, 1993) and
segregated the sites into three representative profiles--rock, alluvium, and soft soil--to
regress the following relations between (average) two-component Arias intensity,
moment magnitude, and source distance.

Log Ih=Mw - 4.0 - 2 Log r* [rock sites] (3)

Log Ih=Mw - 3.8 - 2 Log r* [alluvial sites] (4)

Log Ih=Mw - 3.4 - 2 Log r* [soft sites] (5)

For alluvium and soft soil sites, the regressed value of Arias intensity is higher than
for rock sites owing to the effect oflocal soil amplification .... Figure 4 presents the
predicted mean intensity for rock, alluvium, and soft soil sites versus the horizontal
(surface) distance to the fault rupture plane in kilometers, based on an earthquake
focal depth of 10 kilometers.

Procedures used to develop these equations for attenuation of Arias intensity as a function of
earthquake magnitude, distance from a seismic energy source, and local site condition (rock,
alluvium, and soft soil sites) are similar to those of other investigators (e.g., Joyner and Boore, 1988)
to develop relations between peak acceleration, peak velocity, etc., as a function of these same
parameters.

221
 5
' '

Lirefactitn
-----
(.)
I!)

a
(f.J
D 5o>0.l5 mm
'-' : \,.
-
,s 0.
.D
,.!:;

1 6
l 10
♦ 3
uJ
♦ 8
2 :
i
················1 .........~~,i........ .. ( ................
~
I!)
Cl /1 7j :
73' i ' 16
~ 7: : 1\ 17,19. '
_.>-,
..... 18[ 77 '.5

_.
(f.J
i:: i7t ;
-
I!)

i::
( f.J

.....Cs::I
Nr liquef~ction
I-<
<r::

0.1
0 5 10 15 20
qcl (MPa)

Figure 3 Ihb versus 4c1 for Case History Sites (modified from Kayen and Mitchell, 1997)

Conclusions

Kayen (1993, p. 200) gives the following conclusions to this work:

In conclusion, ... we have collected and evaluated a data set of points from which a
boundary in Ihb-(N 1) 60 space was established, separating liquefaction occurrence from
non-liquefaction occurrence. This boundary is based on the field performance of
sites during moderate to large earthquakes. The most notable aspects of this
boundary are that it is magnitude independent, and is constructed from an earthquake
intensity parameter that is calculated directly from strong motion data measurable at
a site, has a clear physical meaning, and contains no arbitrary elements.

The use of Arias intensity as a measure of seismic excitation for evaluating liquefaction resistance
has the advantage of eliminating the need for other seismic parameters including peak acceleration,
magnitude, and a magnitude scaling factor. Thus, use of Arias intensity could simplify the
incorporation of seismic factors into the simplified procedure. Use of Arias intensity also has the
advantage of using the more physically fundamental parameter of energy in the analysis.

222
 Predicted mean surface-arias intensity (mis)

100 - -
· · · · · · · · · · · · • . . . for focal depth of IO km.
- .
- - - - - .. - - -

· · · · · · Soft soil
0.001
- Alluvium
--Rock
0.00011

Surface Distance to Fault-Rupture Plane, (km)

Figure 4 Ih for 50 th percentile (P = 0) Earthquake response Plotted as a Function of

Surface Distance to Fault Rupture (modified from Kayen and Mitchell, 1997)

Although the workshop agreed that the work by Kayen introduces an important new direction in
analysis ofliquefaction resistance, the procedure is not yet sufficiently verified to be recommended
for immediate use in engineering practice. The following questions introduce primary issues that
need to be addressed: What procedures should be used, and how accurately can Arias intensity be
calculated for application at field sites where strong motion records are not available? Has sufficient
case history data and experience been compiled and analyzed to provide verification of the
procedure? Kayen and Mitchell (1997) have partially addressed the above questions by developing
the attenuation relations above and compiling and comparing results from many case histories to
further verify the applicability of the procedure. The workshop encouraged further research to
answer these questions.

223
 References

Davis, R.O., and Berrill, J.B., 1978, "Energy Dissipation and Seismic Liquefaction in Sands,"
Earthquake Engineering and Structural Dynamics, Vol. 10, p. 69-68.

Figueroa, J.L., and Dahisaria, N, 1991, "An Energy Approach in Defining Soil Liquefaction,"
Proceedings, 2nd International Conference on Recent Advances in Geotechnical Earthquake
Engineering and Soil Dynamics, St. Louis, Missouri, Vol. l. p. 407-410.

Joyner, W.B., and Boore, D.M., 1988, "Measurement, Characterization, and Prediction of Strong
Ground Motion," Earthquake Engineering and Soil Dynamics !!--Recent Advances in
Ground Motion Evaluation, Geotechnical Special Publication 20, ASCE, New York, p. 43-
102.

Kayen, R.E., 1993, "Accelerogram-Energy Approach for Prediction of Earthquake-Induced Ground

Liquefaction," unpublished Ph.D. dissertation, University of California at Berkeley,
Berkeley, California.

Kayen, R.E., and Mitchell, J.K., 1997, Assessment of Liquefaction Potential During Earthquakes by
Arias Intensity: Journal of Geotechnical Engineering, ASCE, Vol 123, No. 12, p. 1162-
1174.

Robertson, P.K, and Campanella, R.G., 1985, "Liquefaction Potential of Sands Using the CPT,"
Journal of Geotechnical Engineering, ASCE, Vol. 111, No. 3, p. 384-403.

224
 Cyclic Liquefaction based on the Cone Penetrometer Test

Richard S. Olsen

Research Civil Engineer

Earthquake Engineering and Geophysics Branch
Geotechnical Laboratory
USAE Waterways Experiment Station
Vicksburg, MS

Abstract

This paper describes suggested recommendations for CPT-based estimation ofliquefaction

potential. Basically there are two approaches, those techniques based singularly on cone
resistance and the CPI soil characterization chart technique. Selecting the optimum technique is
dependent on the encountered soil type and if soil index tests are measured. Background
information and limitations are described for each technique. Stress normalization is also fully
described because its importance is not well known.

225
 Introduction

The Cone Penetrometer Test (CPI) can now be considered the primary field test for assessing
liquefaction resistance. The CPI became more accepted in the l 990's because it is a more
accurate and repeatable test when compared with the Standard Penetration Test (SPT) and is less
expensive and provides a continuous record. CPT-based techniques for estimating liquefaction
resistance, which were developed and refined in the l 980's, have been verified in the l 990's. In
concept, the SPT was an excellent tool for estimating liquefaction potential. However,
performing repeatable and accurate SPT measurements is very difficult. The CPI has more
potential for accurate estimation ofliquefaction potential because it provides two accurate
independent measurements; the cone resistance and sleeve resistance. However, unlike the SPT,
soil samples are not retrieved during CPI soundings. Consequently, for CPI-based evaluations,
some effort should be expended toward soil sampling, preferably using the SPT, for confirmation
of soil type and for soil index testing.

This paper will fully describe the cone resistance based and the CPT soil characterization chart
based techniques for estimating liquefaction resistance. Various techniques for stress
normalization of the CPT cone resistance will also be fully described. Examples and a procedural
cookbook are described at the end of the paper.

CPT Measurements and Relevance to Estimating Liquefaction Potential

The CPT independently measures tip stress (cone resistance) and side friction (sleeve friction
resistance) which in combination can be used to estimate SPT blow count or liquefaction
potential. CPI cone resistance is a bearing stress influenced by many factors, of which the
drained friction angle is the most dominant. The CPT sleeve friction resistance is an index of
remolded strength after breakage of the soil structure and after the soil has undergone large strain.
Historically, CPT cone resistance singularly has been used to estimate liquefaction potential, but
this is a limiting approach. Many factors influence liquefaction resistance such as confining stress,
residual strength, density, soil type, fabric, etc. Using both CPI measurements to estimate
liquefaction potential has more promise than using only cone resistance.

Stress Normalization for the CPT Cone Resistance

Stress normalization is required for all CPI -based techniques for estimating normalized
liquefaction Cyclic Resistance Ratio (CRR1). Stress normalization is very important for low
confining stresses (depths less than 2 meters) and very high confining stresses (depths greater than
25 meters). For vertical effective stresses greater than one atmosphere (atm), an approximating
linear stress normalization technique produces resultants which are increasingly overconservative.
Always use the best available stress normalization technique when estimating liquefaction
potential, regardless of the stress normalization technique used to develop the predictive
technique.

226
CPT-based Techniques for Estimating Liquefaction CRR1

Techniques for CPT-based estimation ofliquefaction potential can be divided into two different
approaches: 1) techniques based on the cone resistance and 2) the CPT soil characterization
chart-based technique. If soil indices, such as fines content, are measured then they are used
either I) as an ingredient for estimating CRR1 using the cone resistance-based technique, or 2) to
verify the CPT soil characterization chart-based technique. Table I shows the decision process
for selecting the CPT-based technique. Technique selection is based on the encountered soil type
and the purpose of measured soil indices (if soil indices are measured). For clean sand, either
technique can be used. For non clean sands, the cone resistance-based technique requires nearby
soil samples for soil index tests. The CPT soil characterization technique can be used with any
soil type and if soil indices are measured, then the indices are used to confirm the technique. Each
of these three approaches will be fully described in the proceeding sections.

Table 1 Criterion for selecting the CPT-based technique for estimation of the normalized
liquefaction cyclic resistance ratio (CRR 1)

Are CPT-based technique for estimating liquefaction potential

General soil index
soil type tests Cone resistance CPT soil characterization chart
measured? based techniques based technique

Clean sands No
.., ..,
Easy to use procedure
.., ..,
Soil index tests are
Non clean sands Yes Soil index tests are
part of the estimating
(fines< 40%) used to confirm the technique
process

No ..,
..,
Other Yes Soil index tests are
soil types used to confirm the technique
( sandy silts
to clay) No
..,

227
Definitions of Cyclic Shear Stress and Calculating the Liquefaction Factor of Safety

Resistance to cyclic loading is generally represented in terms of the liquefaction cyclic resistance
ratio (CRR). The earthquake load is now defined as the average induced cyclic shear stress ratio
(CSR). CSR can be calculated using the rd factor or computed with computer software such as
SHAKE. The factor of safety against liquefaction is defined as

FS. == CRR (1)

hq CSR

CPT and SPT-based techniques can estimate a normalized cyclic resistance ratio, CRR1 (also
referenced as CRR7_5 in other publications). The CRR at the in situ vertical effective stress and
design earthquake magnitude can be determined from CRR1 with the following equation:

(2)

where
MSF= Earthquake magnitude scaling factor
K,, = Confining stress scaling factor
Ka= Initial shear stress scaling factor

The earthquake magnitude scaling factor (MSF) can be estimated with the following equation
(Idriss 1998):
102.24
MSF == - - (3)
M2.s6

where
M = Earthquake magnitude

This equation is an update to the Seed and Idriss ( 1982) formulation. An earthquake with a
magnitude 7.5 represents approximately 15 cycles of shear (Seed and Idriss 1982). CPT and
SPT-based techniques for estimating CRR1 assume that the soil will experience 15 cycles of shear.
For earthquakes with lower magnitudes, the soil experiences a lessor number of applied cycles of
shear. Consequently, for the same earthquake applied shear stress level, the soil will have a higher
apparent strength for lower magnitude earthquakes. The MSF in Equation 3 is equal to 1 for a
earthquake magnitude 7.5 and is greater than 1 for smaller magnitude earthquakes.

For the laboratory cyclic triaxial test, CRR is given as the a ratio of maximum cyclic shear stress

228
to cause liquefaction to the initial effective confining stress, adj2a' 3 , times an correction factor of
approximately 0.57. The 0.57 factor converts triaxial result to equivalent horizontal field results.
The state-of-practice is to define liquefaction in laboratory cyclic tests as 5% double-amplitude
axial strain. The laboratory CRR1 is defined as the CRR for liquefaction at 15 cycles of uniform
loading (to represent an earthquake magnitude (M) of7.5) at a vertical effective stress of I atm
(I atm = I 00 kPa) and using a 5% double-amplitude failure criterion.

The K0 scaling factor corrects for failure envelope curvature of the liquefaction resistance to
vertical effective stress envelope (Harder and Boulanger 1998; Olsen 1996). The resultant is that
CRR decreases with increased vertical effective stress. For a vertical effective stress range of0.5
to 2 atm, the K 0 scaling factor ranges from 1. 05 to 0. 9. Therefore for most non critical structures
K0 can be assumed equal to 1 for vertical effective stresses near I atm. The K" scaling factor is
for conditions near earth slopes. For level ground, K" is equal to one.

Stress Normalization for CPT and SPT Measurements

The best means of estimating a geotechnical property with CPI data is to start by normalizing the
CPT cone resistance to a standard vertical effective stress of I atmosphere (atm). The next step is
to estimate the normalized geotechnical property (such as liquefaction resistance) using the
normalized CPI data. All techniques for estimating liquefaction resistance do so for a condition
representing a vertical effective stress of one atm. Thus the task of determining liquefaction
resistance is greatly simplified because we only estimate liquefaction resistance for a single vertical
stress level. The liquefaction resistance for other overburden stress levels is calculated using the
inverse of the stress normalization technique. In all cases, using the best available stress
normalization technique produces the most accurate predicted value.

There are numerous techniques for stress normalization of CPI and SPT measurements and
liquefaction resistance. Approximating linear techniques are easy to use, while the stress focus-
based technique uses a nonlinear stress exponent and requires computer-based processing. In
most situations, a technique between these two extremes will provide results of sufficient
accuracy.

To illustrate the difference between approximating and sophisticated stress normalization

techniques, let's assume a uniform thick sand deposit having an equivalent friction angle of 35°.
A sophisticated stress normalization technique would predict, using CPI data, an equivalent
friction angle of 3 5° for all depths in this soil deposit. However, an approximating linear stress
normalization technique would predict an equivalent value less than 35° (say 32°) for the bottom
of the deposit and values greater than 35° (say 40°) for the top of the deposit. To estimate the
most realistic equivalent value requires using the best stress normalization technique.

Techniques for estimating soil properties are not married to a particular stress normalization
technique. Stated another way, the better the stress normalization technique the more accurate

229
the estimated equivalent value. The best possible stress normalization technique will enhance the
accuracy of any predictive technique (which was developed using an approximating stress
normalization technique).

Two general factors dictate the selection of the stress normalization technique: 1) the criticalness
of the structure, and 2) overburden stress range. Any stress normalization technique gives the
correct equivalent value for a soil element having a vertical effective stress of one atm. Similarly,
linear and exponent techniques provide approximately the same result (within a few percent) for
soil elements having a vertical effective stress between 0.8 and 1.2 atm. The need for the most
accurate stress normalization technique is more critical for estimation of liquefaction resistance
compared to estimating soil classification. The stress focus-based technique, using an iterative
nonlinear stress exponent for stress normalization, should be required for estimating liquefaction
resistance at overburden stress extremes (i.e., shallow offshore sediments or foundations for large
earth dams) and for critical structures. Generally, start with the approximating linear-based stress
normalization technique and progress to more sophisticated techniques, if required. While the
stress focus based technique is technically the most accurate, the other techniques can be justified
because a higher level of sophistication is only needed for shallow and deep conditions, and
critical structures.

Stress normalization can be divided into the following categories (from the least sophisticated
(and less accurate) to the most sophisticated (and more accurate):
1) linear relationship
(Douglas and Olsen, 1981; Robertson 1981, 1986, 1988, 1996),
2) Constant exponent (i.e., 0.5 or 0.61) (Robertson 1994, 1995, 1996),
3) Variable stress exponent (i.e., 0.1 to 1) (Olsen 1984, 1986, 1988, 1990),
4) Stress focus technique (variable stress exponent based on soil type and relative strength)
(Olsen 1994a, 1994b, 1995a, 1995b, 1996)

Each of these stress normalization techniques will be described using the CPT cone resistance.
The same stress normalization factor determined for the CPT can be used for the SPT.

Approximating Linear Stress Normalization Technique

The approximating linear stress normalization technique shown in Equation 4 (Douglas and Olsen
1981; Robertson 1982, 1984, 1996) should only be used for general estimating purposes or when
the overburden vertical effective stress is near one atm. The "linear relationship" technique is
rarely justified except for quick field "back of the envelope" estimations. A linear stress
normalization technique can be justified only for clay because clays exhibit little if any failure
envelope curvature (Olsen 1994).

230
 (4)

where
q01L = normalized cone resistance (qc1) using the approximating linear technique
(equivalent value at a vertical effective stress of 1 atm)
q 0
= measured cone resistance (in atm units)
(1 atm" 100 KPa" O.l l\1Pa " 1 ton/ft2 " 14.7 psi)
av = vertical effective stress (in atm units)

Constant Stress Exponent Technique for Stress Normalization

The constant stress exponent equation for stress normalization is shown in Equation 5, using a
stress exponent of 0. 7. This technique dates back to the early 1980's when Professor
Schmertmann (1978) observed that the CPT cone resistance for a given relative density in large
scale chambers, at differing confinement stresses, could be related using a constant stress
exponent. This constant stress exponent is now sometimes published as 0.5. A stress exponent of
0. 5 is incorrect because the trend of CPT chamber data with vertical effective stress for normally
consolidated medium dense sand shows a stress exponent of approximately 0.7. The SPT
chamber data trends show a general stress exponent of O. 5 which is inaccurate because of the
short column of drilling mud (Olsen 1994). Many researchers would rather report the SPT
chamber-based stress exponent of0.5 rather than the CPT chamber-based stress exponent of0.7.
Recent research (Olsen 1994) indicates that a stress exponent of0.7 represents a typical sand at a
medium dense to loose consistency.

(5)

where
q010 = Normalized cone resistance (q01 ) based on a contact stress exponent
(equivalent value at a vertical effective stress of 1 atm)

A well-published alternative (Skempton 1986; Kayen et al. 1992) to Equation 5 that also matches
the combined scatter of field data is shown below;

qclc = ( qc
0.7 + 0~
l (6)
1.7

231
The only potential problem with this equation is at very low and very high stress conditions. The
equation was designed to match data from 0.8 to 2 atm (80 KPa to 200 KPa) and will deviate
from the constant stress exponent at extreme overburden stresses. However, the constant stress
exponent technique in Equation 6 is easy to perform on a scientific calculator. This category of
stress normalization should only be reserved for the most basic calculator or quick check
calculations.

Stress Normalization using a Stress Exponent Dependent on Soil Type

The stress normalization technique using a stress exponent dependent on soil type is shown in
Equation 7.

(7)

where
q 015 = normalized cone resistance (q 01 ) using a stress exponent based on soil type
c = stress exponent based on soil type

This technique requires an approximate knowledge of the soil type (Olsen 1984, 1988a, 1988b).
For sands, the stress exponent is given a value of approximately 0.6. For clays, field data trends
suggest a stress exponent of one. More specifically, the CPI estimated soil type can be used to
directly estimate the stress exponent. A several step iterative solution is required when using
software but one or two steps are sufficient if solved by hand. This technique has been
superseded by the stress focus-based stress normalization technique.

CPT Stress Normalization Using the Stress Focus Based Variable Stress Exponent
Technique

Recent research on the influence of confining stress on CPI and SPT measurements has resulted
in a new theory- The stress focus theory (Olsen 1994; Olsen and Mitchell 1995). The stress
focus theory uses a variable stress exponent for stress normalization as shown in Equation 8.

(8)
(a~Y

232
where
qcle = normalized cone resistance (q01 ) using a nonlinear variable stress exponent
(equivalent value at a vertical effective stress of 1 atm)
(Jtota! = total vertical stress in atm units
C = stress exponent dependent on soil type sand relative strength level
( see contours of stress exponent in Figure 1)

The stress exponent, c, shown in Figure I, is dependent on soil type and relative strength level.
For sands, it defines the log-log slope for a constant relative density trend as shown in Figure 2.
The cone resistance stress exponent, c, decreases as sand relative density increases and can be
approximated (for sand) as shown below using relative density, D, (Olsen and Mitchell 1995):

c = 1 - (D, - 10%)0.007 (9)

The stress focus theory explains why the stress exponent for sands is dependent on initial relative
density. For all overburden stress conditions, this "variable stress exponent" provides the most
accurate CPT cone resistance normalization.

SPT Stress Normalization

The SPT blow count (N) is normalized to the equivalent value (N1) at a vertical effective stress of
I atm using Equation 10.

(10)

Then stress exponent in Equation 10 was suggested by Seed and Idriss (1984) to equal 0.45 for
sands having relative densities of60 to 80% and equal to 0.55 for sands having relative densities
of 40 to 60%. These recommendations were established based on data from SPT chamber tests
reported by Marcuson and Bieganousky (1977). Recent research by Olsen (1994) suggests that
the constant low mud pressure in the SPT chamber boreholes, at all confining stress levels,
reduces the confining stress at the SPT sampler. This reduced stress causes the inferred stress
exponents for sands to be too low by a value of0.15. Stress focus theory explains why the SPT
chamber-based stress exponents are lower than the CPT chamber values. The consequence is that
the CPT determined stress exponent from Figure 1 or Equation 9 should also be used for SPT
normalization. The 1997 draft ASTM SPT standard for SPT stress normalization references
Olsen (1994) as an alternative technique where a higher level of sophistication is required. The
CPT determined CPT stress exponent should be used with the SPT stress exponent and in general
ranges from 0.6 for medium dense sand, 0.7 for loose sand, to 1 for clay.

233
 Cone
resistance
stress

-
£3
·2:
:::I 100
E Loose

--~
+-'
ro

C:
ro
+-'
(/J
.(/J

~
Q)
C:
0
(.)
"O
Q)
.N
Soil
ro 10
E
I...
0 ~
z §
-~~

Olsen & Mitchell (1995)

Olsen (1988)
1 '-----.L..---'----.L..-_._.....L..-'-J.......L_.__ _ _- ' -_ __,__...,____,___,__,_...,__L..J
0.1 1.0 10.0

Friction Ratio(%)
Figure 1 Stress exponents for cone resistance on the CPT soil characterization chart
(Olsen and .Mitchell 1995).

234
 Net Cone Resistance (atm units)
10 JOO 1000 10,000

Increasing

r
Relative
Density
(Increasing
dilative
behavior)
Normalized Cone
Resistance
is at this vertical
effective stress
level Horizontal
Expansion
(Cavity
Expansion)
Bearing
Failure
o;c = 140 atm
(Approximately 1 km depth) _/

Stress Focus
for a given sand

Sedimentary Rock Behavior

(with minimal dilative behavior)

Figure 2 Description of how cone resistance trends toward the stress focus for any relative
density (Olsen 1994).

235
 CPT-based Estimation of Soil Classification

Knowing the soil classification of soils penetrated by the CPT is as important as estimating the
liquefaction resistance. The soil type and relative strength infers how the soil reacts ifliquefaction
is triggered. In the past, soil type could be estimated from a CPT soil classification chart either
manually or with non-linear chart lookup software. Robertson (1998) recently converted soil
classification contours into an equation form (soil index, IJ using a modification of the Olsen soil
characterization chart (Olsen 1988). However, Robertson's le is an arbitrary value requiring
memorization of the soil type for each l range. The most recent CPI-based chart for estimating
0

soil type is shown in Figure 3 (Olsen and Mitchell 1995). The CPT Soil Characterization Number
(SCN), shown on the soil classification contours, represents unique and definable soil types. The
SCN has been part of all the CPT soil characterization charts (Olsen 1984, 1986, 1988, 1990,
1994; Olsen and Mitchell 1995). The CPT SCN can be approximated (for normally consolidted
soils) by the following equation:

(11)

where
SCN = Soil Classification Number (SCN) (normally consolidated soil) (see Figure 3)
+1 = boundary between a silty sand and fine sand
(where the silt is effecting the strength behavior of sand,
i.e., 10 to 15% fines content)
0 = behavior ofa pure silt (ML behavior with fines content= 50 to 60%)
-1 = boundary between a silty clay (CL or CH) and clayey silt (ML)
(fines content= 80% to 100%)
-2 = boundary to organic peats and unstable soils
qc 1 = Normalized cone resistance (in atm units) (using the best possible technique)
Rr = Calculated friction ratio (percentage)
The calculated friction ratio (Rr) is defined as:

R = f, 100 (12)
f qc

where
f. = Measured CPT sleeve friction resistance in atm units (1 atm = lOOKPa)

The SCN, from Equation 11, is only for normally consolidated soils. It produces SCN values too
high for overconsolidated conditions. The following equation estimates the SCN for
overconsolidated conditions which occur at friction ratios (Rr) generally greater than 1.9 %.

236
 .,
I "'
;;;
Olsen & Mitchell (1995) ~
Olsen (1988) 0

z
_,,,,
w

0
t
0,

:i
"'
0
r-.
I i~

--·-c
1
Fines < 5%\
, 0 ense 1
en
Medium ~ I
:J dense I
E 100

-
rn
~ Very
Loose <2,'o'(\O

-
C: loose
m
en
en
~
Q)
C:
0
(..)

-0
Q)
Soil
.!::l 10 Classification
m Number (SCN)
E
s...
(Olsen 1995
version)
0
z
.... ....
~- -
.sg
u

1 L------'-----'--'--._.&......L_._.L.L._ _ _.,____.,_____.____.___.__---'--L..J...J
0.1 1.0 10.0
Friction Ratio(%)

Figure 3 CPT soil characterization chart (Olsen and Mitchell 1995).

237
 SCN* = ((log 10 qJ1. 5 - 1.1 (log 10 Rf - 0.27)) 1.34 - 1.87 (For Rf> 1.9%) (13)

where

ScN• = Soil Classification Number for over consolidation conditions (for~> 1.9%)

The CPT Soil SCN was designed to represent unique properties at integer values (i.e. -1, 0 and 1)
(Olsen 1984, 1986, 1988, 1994, 1995). At SCN equal to -1, the soil should have a classification
and strength behavior between that of a silty clay and clayey silt. More specifically, SCN equal
to -1 represents a Unified Soil Classification System (USCS) classification between CL (or CH or
MH) and a ML. At SCN equal to 0, the soil behavior should be that of pure silt and the USCS
equal to ML. At SCN equal to 1, the soil should have a classification and behavior between that
of a fine sand (or slightly silty sand) and silty sand where the silt is starting to influence the
strength level of the sand. The USCS for SCN equal to 1 is between a classification of SP/SM
and SM. By USCS definitions, the boundary between an SP and SM is 5% fines, the boundary
between SP/SM and SM is 12% fines, and the boundary between ML and SM is 50% fines.
Therefore, an SCN equal to 1 (between SM/SP and SM) represents a fines content of
approximately 12% and a SCN equal to O (pure silt) represents a fines content of approximately
70%. Over the last 15 years the location of these SCN boundaries have been adjusted as new
field CPT data were included into the database.

Estimating Liquefaction Resistance of Clean Sand

Using only the Cone Resistance

The CPT cone resistance-based technique for estimating liquefaction potential originated in the
early 1980s as simple conversions from SPT-based techniques. The original correlation by
Robertson (1982) and Robertson and Campanella (1985) for clean sands, shown in Figure 4, has
been shown to be historically correct based on accumulated cyclic laboratory data and historical
field performance data. A critical point for any cone resistance-based technique is the intersection
of the liquefaction boundary line for a normalized cone resistance at CRR1=0.2. Robertson
(1982) established a normalized cone resistance of 118 atm for this intersection point and this
value has been proven over the last 15 years. Figure 5 illustrates various cone resistance-based
techniques for determining liquefaction potential of clean sand together with recent field
performance data and the NCEER recommended relationship. Figure 6 is a simplification of
Figure 5 showing only the NCEER recommended correlation and field performance data.

Field performance data points from Suzuki, et al. (1985) together with the NCEER-recommended
correlations are shown in Figure 7. Three of Suzuki's field performance data points plot beyond
the proposed liquefaction resistance relationship for clean sand. These data points represent 9%
of the field performance data where liquefaction was observed but where the CPT-based
technique estimates non liquefaction. The Suzuki database has not been scrutinized and the
source of these four outlying data points are unknown. The NCEER recommended correlation in

238
 -
0:::
0:::
T"""

0.6
()
_.
0
+-'
~ 0.5
Cl)
(.)
C
ro
+-'
en 0.4
en
Cl)
0:::
(.) Liquefaction
(.)
0.3 Zone
>-
()
C
0
t5
~ 0.2
Cl)
:::, Non
C" Liquefaction
...J
"'O Zone
Q) 0.1
N
1

m \_(Robertson, 1982)
E
s...
0
z 0.0
0 50 100 150 200

Normalized Cone Resistance (atm units)

Figure 4 Relationship between normalized liquefaction cyclic resistance ratio (CRR1) and
normalized cone resistance by Robertson (1982).

239
 o. 7 .........
T""T""T""T""T""T""T""T""T""T""T""T""T""T""-rr-T""T""T""T""T""T""T""T""T""T""T""T""T""T""T""T"",-,-,-,-.....,...,.....,...,.....,...,,-,-,-,-.....,....., I
8
♦ A ■ Liquefied Deposit Field Performance Data
◊ t:. □ Non Liquefied Deposit
■ Shibata et al (1988) data
0.6 Clean sands (Fines< 5%) ♦ Stark et al (1995) data

-
0:::
0:::
A Suzuki et al (1995)

-
(,)
o
:.:;
«l
0.5
Stark et al 1995
0::: Liquefaction
Q) NCEER recommendation
0
C
■ 111 for clean sand
«l 0.4 ■
+' A
.!!? ♦
en A ■
Q)
0:::
·00 o.3
• A
A

>, ◊A
(,)
C A ◊
0 A !::,.
u«s 0.2
..,_ ◊
◊
Dw. ◊
i
Q) 8 □
::J ◊
0-
::J I:,.

0.1 .
~~~i_et al J!~~
!::,.l:,,.
Do t8 No
Liquefaction
"'--NCEER recommendation
for clean sand
0.0 ~~~~~~~~~~~~~~~~~~~~~~~~~~

0 50 100 150 200 250

Normalized Cone Resistance (atm units)

Figure 5 Comparison between CPI cone resistance-based methods to determine the

normalized liquefaction cyclic resistance ratio (CRR1) and recent field performance
data.

240
 0.7

♦ A ■ Liquefied Deposit
r
Field Performance Data ' i
z
~
◊ Ll □ Non Liquefied Deposi
■
0,
Shibata el al ( 1988) data ;i
0.6 •
- Clean sands ♦ Stark et al (1995) data 8
(Fines content< 5%) A Suzuki et al (1995) data
0::: M=7.5 ~
,
0:::
u
.__.
0 0.5
:;:::;
co
0::: Liquefaction
Q)
0
• 11 • ■

-
C:
co
-~
(/)
Q)
0.4

.A
•• A A
+■ 1/ NCEER 1996
Salt Lake
workshop
0::: .A
.Q

u
0
>.
0.3

.A
~
◊A
...
A

..:
C: A ◊
0
:;:::;
A
A • r■ ◊

~
0
.E 0.2
◊◊ ◊ □ ¢◊.i
Q)
~
0-
:.:J
1,.• .A ••
A_. -\.e_... ~ ~
□ ◊
8

0.1 ◊
Do □ No
Liquefaction

0.0
0 50 100 150 200 250

Normalized Cone Resistance (atm units)

Figure 6 Comparison between the NCEER recommended cone resistance-based technique

for determining the normalized liquefaction cyclic resistance ratio (CRR1) and
recent field performance data.

241
 -
0::
0::
.....
0.6
r '\
(.) ♦ A Liquefied deposit Field Performance Data
.........
0 ◊ A Non liquefied deposit ♦ Stark et al (1995) data
:.:;
ro 0.5 Clean sands A Suzuki et al (1995) data
0:: (Fines content< 5%) ,,,
M=7.5
~ yNCEER1996
C ♦
19 Salt Lake workship
U) 0.4 Liquefaction ♦
·w
Q) ♦
A
0:: ♦ A

...
.2 No Liquefaction
u
>,
0.3
◊A L:,.
(.) A
C
0
\ ◊
t;;,.
A ♦ ~ i
t 0.2
A. A
◊
◊
~ ◊
~ ♦A. ♦·A
Q)
A ♦• ♦ 1:+
::J
C" A A A Jilt ~♦ AA These 3 data points are
outside'ihe NCEER
:.:J recommended
"O 0.1 relationship and represent
Q)
-~ro 9% of the liquefied
field performance
E data points
s....
0 0.0
z 0 50 100 150 200 250
Normalized Cone Resistance (atm units)

Figure 7 Comparison for clean sand between the NCEER-based CPT cone resistance
technique to determine the normalized liquefaction cyclic resistance ratio (CRR1)
and recent field performance data.

242
Figure 5 represents the best available correlation for CPT cone resistance-based estimation of
liquefaction resistance of clean sands.

Estimating Liquefaction Resistance of Non Clean Sands

using the Cone Resistance

Inclusion of silt within a sand matrix decreases cone resistance and may decrease liquefaction
resistance. Estimating the liquefaction resistance of non clean sands using cone resistance-based
techniques can be accomplished by two means, either 1) the calculated equivalent clean sand (silt
corrected) cone resistance, or 2) the chart-based solution in terms of a measured soil index.
Limitations of the silt correction concept will be discussed in the next section.

The chart-based solution in Figure 8 is the most viable approach for cone resistance-based
estimation ofliquefaction resistance for non clean sands. This figure is based on fines content
and/or mean grain size. There are only two fines content trends in Figure 8, namely 5% and 35%.
The influence of fines content on SPT-based liquefaction resistance was observed by
Seed, et al. (I 983) to be non linear and it is likely that the CPT relationship is also non linear.
At least three fines content trends are required for non linear interpolation purposes. Therefore,
at present, there are insufficient fines content trends to establish a non linear relationship.

The relationship of mean grain size to liquefaction resistance represents a good correlation
because of the large number of independently developed mean grain size trends in Figure 8.
However, mean grain size is a less desirable index because a full gradation test is more expensive
than a fines content test. Also, for a given soil composition, it is likely that the CPT estimated
liquefaction CRR1 value will be different when using a fines content or mean grain size criteria in
Figure 8.

Equivalent Clean Sand Cone Resistance, (qc 1)c,

The equivalent clean sand cone resistance, (q01 ) 0,, is the same concept as the SPT equivalent clean
sand cone resistance, ((N1) 60) 0,. One means of defining(q 01 ) 0, is shown below (Robertson and
Wride 1998):
(14)

where
K0 = correction factor that is a function of grain characteristics

CRR1 then estimated using clean sand relationships ofCRR 1 versus(q01 ) 0, such as Figure 6. There
is no physical or theoretical meaning for (q01 ),, or ((N1) 60) 0,. They do not represent the equivalent
value a sand if the silt content was removed. They are only a convenience for calculating CRR1.

243
 0.6

-
0:::
....
Fines= 35% NCEER recommendations for

-
0::: modified from D50 = 0.25 mm
() Stark & Olson (1986) and/or fines= 5%
0
0.5
~
co
0::: D50 = 0.05 mm
a.> from D50 = 0.10 mm
C Shibata et al (1988)
(.) from
co
..... 0.4 Shibata et al (1988)
•5Q
u,
a.> D50 = 0.15 mm
0::: modified from
.5::! Robertson (1982)
~ 0.3
()
C
0
t, Non
Liquefaction
'ffl 0.2
Side
::J
C"
~
-0
a.>
.N 0.1
co
E
I..
0
z
0.0
0 50 100 150 200
Normalized Cone Resistance (atm)

Figure 8 Comparison between various CPT-based techniques for estimating the normalized
liquefaction cyclic resistance ratio (CRR1).

244
The equivalent clean sand concept was originally developed for the SPT as a correction for grain
size differences by the late Prof H. Bolton Seed (1984). He then introduced a chart solution, in
terms of fines content, which removed the need to determine the equivalent clean sand value.
Olsen (1988) then reintroduced equivalent clean sand technique with the SPT silt correction chart
(Figure 9) based on Seed's 1986 SPT liquefaction determination chart. The purpose for this silt
correction chart was only to simplify development of the CPT soil characterization chart-based
technique for estimating liquefaction potential (to be introduced in the next section). Ironically,
this SPT silt correction chart has been widely referenced and copied, and has now been extended
to the CPT equivalent clean sand cone resistance. It is important to repeat that there is no
physical meaning for the equivalent clean sand cone resistance or SPT blow count. It is therefore
recommended that cone resistance-based liquefaction potential should be determined using chart-
based solutions rather than by means of equivalent clean sand cone resistance.

CPT Estimation of Fines Content

The fines content of non clean sands is an important ingredient for cone resistance-based
techniques for estimating liquefaction resistance. Fines content can be either 1) measured from
samples taken from nearby boreholes or 2) estimated using CPI-based techniques. Fines content
can be estimated based on the CPT estimated soil type as shown in Figure 3 (Olsen and l'vfitchell
1995; Olsen and Koester 1995; Olsen 1988). However, techniques for estimating liquefaction
resistance based on the equivalent clean sand normalized cone resistance are an over
generalization of the CPT soil characterization-based chart technique (to be discussed in the next
section). Estimating fines content using CPT measurements must be considered a crude
prediction having a high standard deviation. Liquefaction resistance is also more complex than
can be generalized with a simplistic correlation of CPT estimated soil classification to CPT ·
estimated fines content. The CPT is a strength measurement test which is influenced rather than
dominated by fines content.

Problems with Geologically Complex Sites

Unlike the SPT where the sampler recovers a soil sample of the tested soil, the CPT does not
provide a soil sample. The CPT cone resistance-based technique for non clean sands however
requires a soil sample, typically from a nearby boring several meters away. The disadvantages of
measuring soil index tests are the economics of requiring nearby boreholes and the additional
costs oflaboratory testing. Also, soil conditions at geologically complex sites can quickly change
over short lateral distances. Consequently, measured soil indices from a nearby borehole may not
match the soils penetrated with the CPT a short distance away. Measured soil indices should only
be used when the potentially critical soil layers are thick and uniform. The calculated liquefaction
potential will be conservative if the sampled soil is finer than the soil probed; similarly, the
calculated value will be unconservative if the sampled soil is coarser than the soil probed.
CPI-based techniques for estimating liquefaction potential that require soil samples should not be
used at geologically complex sites.

245
 10

....
C
8
....
Q)
C
0 7
(.)
1/)
Q)
C 6
I.:
.....
.g 5
1/)
C
0
~ 4
~
.....
0
(.) 3
I-
a..
Cf) 2
Data from Seed, Tokimatsu,
1 Harder, and Chung (1986)
Correlation from Olsen (1988)

0
0 5 10 15 20 25 30 35 40 45 50
Fines content
(percent passing #200 sieve)

Figure 9 Estimating the SPT silt correction based on fines content (Olsen 1988)
(data from Seed, Tokimatsu, Harder, and Chung 1986).

246
 Estimating Liquefaction Resistance for All Soil Types
using the CPT Soil Characterization Chart Technique

The CPT soil characterization chart technique for estimating liquefaction resistance is shown in
Figure IO (Olsen and Koester 1995). The normalized liquefaction cyclic resistance ratio (CRR1)
is determined for any depth and any soil type based on the combination of normalized cone
resistance and friction ratio. This technique does not require laboratory measured soil index tests
to estimate liquefaction resistance, unlike other SPT and CPT-based techniques.

Historical Development of the CPT Soil Characterization Chart Technique

The CPT soil characterization chart technique for estimating liquefaction potential originated in
the early 1980s (Olsen 1984) and has been constantly refined and improved since then
(Olsen 1988; Olsen and Farr 1986; Olsen and Koester 1995; Olsen, Koester, and Hynes 1996).
This technique indirectly includes the effects of soil type, fines content influence, peak strength,
high strain strength, and lateral stress influence. Specifically, it was developed based on:
1) correlations to cyclic laboratory tests, 2) trends of CPT estimated normalized SPT values, 3)
trends ofSPT silt corrections using CPT estimated silt content, 4) the Seed, et al. (1984, 1986)
SPT to CRR1 correlations, and finally 5) field performance data (Tokimatsu, et al. 1990; Kayen,
et al. 1992; Suzuki, et al. 1995, 1995b; file data; project data; etc.).

Estimating SPT Blow Count Using the CPT Soil Characterization Chart Technique

The critical starting point for this chart technique was to using an accurate technique for
CPT-based estimation of SPT blow count Contours of CPT-estimated SPT-normalized blow
counts, N 1, can be established on the CPT soil characterization chart as shown in Figure 11.
These SPT contours were developed using both CPT measurements (Olsen 1984, 1988, 1994)
and are more accurate than if based on the q/N ratio (Robertson 1983; Seed, et al. 1986). Seed,
et al. (1986) used the q/N technique to establish a cone resistance-based technique for estimating
CRR1; we now know that this approach results in unconservative estimates ofCRR1. The next
step, after establishing SPT contours (Figure 11 ), was establishing contours of CPT-estimated
equivalent clean sand normalized blow counts, (N 1)c,, using the procedure shown in Figure 12
(Olsen 1988). Equivalent clean sand SPT contours were calculated based on the combination of
SPT N 1 contours and CPT estimated fines content together with the SPT-based silt correction
relationship shown in Figure 9. CRR 1 contours were then approximated by converting the (N1) 0,
contours to CRR1 contours based on the Seed (Ni)0, to CRR1 relationship (Figure 13). These
SPT-estimated CRR1 contours are the framework for further refinements based on cyclic
laboratory and field performance data.

CPT-based Cyclic Laboratory Data Trends of Liquefaction Resistance

The cyclic laboratory data (also shown Figure 12) were used to refine the position of
SPT-developed CRR1 contours on the CPT soil characterization chart. The cyclic laboratory

247
 Olsen, Koester & Hynes (1996)
Olsen (1988, 1984), Olsen & Mitchell (1995)

Dense

§'
·c
:::, 100
E
-
1a

10

IO IO
.......... ~C\I~
0 0 0 0 0
CPT predicted Normalized
Liquefaction Cyclic Resistance Ratio (CRR1)
1 ..__ _ _.,___ _.__ _.___._-'---'--'--'--'-----'-----'---'----'----'---'--'-~
0.1 1.0 10.0
Friction Ratio (%)

Figure 10 Estimation of the normalized liquefaction cyclic resistance ratio (CRRi) using the
CPT soil characterization techniques (Olsen, Koester, & Hynes 1996).

248
 JOOO
a Contours of SPT Nl I
6 I
= 50 b,;, I
4
N1
'l7,r;;, I
C;,
I I
~ 35
---_,s 2
N1
I
<C
..__.,
Nl = 25 I
Q.) N1 = 15
C,) 100
i:: =7
_,
·(tJ

.....C/l
U)
0)
~
Q.) /
C
0 e,S
._,◊-.
u ~\i-
. '"O i
(l)
<;,'<>-~

-a
.N /
...... 10
/
ro / -:'.es
6 .,..'-◊
h ~'\.
0 /
z <f>v" /
4
/
/
/ c"-'lr'1
2
,,,,,,...
CPT soil characterization Jines
/ from Olsen (1988)
1.
2 4 6 8 2 4 6 8
0.1 ·1.0 10.0
Normalized Friction Ratio (%)

Figure 11 CPT estimation ofSPT N 1 using both CPT measurements (Olsen 1994, 1988,
1986, 1984).

249
 1000,---------'------'--'-............._ ___._--J._,__'-'--'-"-'-t
800 Developing the CPT procedure
600
for estimating (( 1avfa,)5 ,J1
400

....en
- 0
II)
200

...E
2
SPT N 1 contour$

_i;;
. N, .. -..=N,-t-li.NLlq

ctl~ II
18=13+5

18=9-+9
40
0
0-

Q) 20
u
C
co
iii
"iii
Q)
a:
O>
C
0
()
"O
~
u
...0...O>
.
() 2
((.].v). ) =0 2
l'y5,Y.1.

1
0.1 0.4 2 4 6 8 10

Corrected Friction Ratio (%) in terms of tsf

Figure 12 Graphical procedural steps for CPT-based determination of the SPT blow count,
fines content, equivalent clean sand SPT blow count, and normalized liquefaction
cyclic resistance ratio (CRR1) (Olsen 1988).

250
 ( l J $ , , - - - - - - , - - - - . - - -.......- - - , - - - - ,

0,5
(
[
II
OJ!
Jt
II
a..:: ... r!

• I fl
Trtv
<:;'
O,!,
•

.1
•
.
•
• I
,
I
J
I
I
,~
I

• 0
J'

• • ,;9 ~1;
Q ''I
'I
0.2
• .,,Oe

FiNES CON1ENT:s:; S%
Chinese l:lriild;n9 Cclcle {~C:I CMlant•Ol @ -
~tu,,:i/lCI :.:0
Llq.re!i,cii:::o:i ue.ie,~ i.,'l;,g~tiQol
Poa~M1;:rlco:t1 dar:i • a n
Jc panese oota
• 9 0

0
Cl'line $e data ,... I
:.
'
0 10 20 30 40 50
{N11150

Figure 13 Relationship between normalized liquefaction cyclic resistance ratio (CRR 1) and N 1
for clean sand (Seed et al. 1984).

251
database contains soil types from clean sand to silty clay and CRR1 levels from 0.12 to 0.32.
Each laboratory-based data point represents a normalized cyclic resistance ratio value (CRR1) and
was typically determined from a series oflaboratory cyclic tests. The cyclic laboratory data
confirm that the CRR1 contours curve down into the clay region of the CPT soil characterization
chart.

Strain Potential of Clays

This technique uses a strain based criterion for liquefaction designation rather than a pore pressure
generation criterion. Clays can strain during earthquakes and expressing cyclic mobility.
Normally consolidated soft clays can accumulate 5% cyclic strain during an earthquake if the
average earthquake induced shear stresses approach the static strength. Consequently, if
earthquake induced stresses are greater than the clay undrained strength, the clay will deform and
meet the liquefaction criterion. The static clay undrained strength is best expressed as the
undrained strength divided by vertical effective stress, designated as the strength ratio, and
signified as the c/p ratio; for normally consolidated clay, the c/p is approximately 0.31. An
earthquake-induced CSR of0.3 represents a large nearby earthquake and if applied to a normally
consolidated clay deposit will induce dynamic deformation. Normally consolidated soft clays
therefore can have an equivalent liquefaction resistance CRR1 of at least 0.28 as is reflected in
Figure 10.

This CPT soil characterization technique also includes estimates CRR1 contours for sensitive clay
and sensitive soil mixture in Figure 10. For increasing soil sensitivity, the CPI estimated CRR1
contour change from CRR1=0.28 (for normally consolidated clays and silts) down to CRR1= 0.1
(highly sensitive clays and silts). The CRR1 contours for sensitive soils were developed primarily
based on laboratory cyclic tests.

If a clay or silt mixture and is flagged as liquefiable (based on this technique) than the soil layer
should be sampled for further evaluation. An excellent index is the liquidity index (LI), defined
as:
w - PL
LI= _n_ __
(15)
LL - PL

where
LL = Liquid limit (in percentage)
PL = Plastic limit (in percentage)
w 0 = In situ water content (in percentage)

The LI is a relative strength index because the LL and PL represent the water contents for two
strength levels. The "China Criterion" specifies that soils are potentially liquefiable if LI is greater
than 0.9. Alternatively, for clays and clayey silts, ifin situ or laboratory-based vane shear tests

252
indicate a sensitivity greater than 4 than the soil should be labeled suspect. The purpose of these
index tests are to flag non sands that may experience severe softening and strength loss during an
earthquake. The MSF should not be applied to clays and silt mixtures for earthquake magnitude
less than 7.5. The MSF factor may not apply to clays and silt mixtures because MSF applies to
cyclic induced pore pressure generation in sands; the sensitive behavior of weak clays and soil
mixtures developed primarily by over straining.

Limitations of Field Performance Data

Field performance data is defined as are soil deposits having in situ measurements together with
observations on whether liquefaction effects occurred during an earthquake. For the last 20
years, the preferred means of correlating SPT to liquefaction potential was with field performance
data. This procedure has been extended to cone resistance-based technique (Stark and Olson
1995) for sands. It is also possible to use CPT field performance data to establish the CPT soil
characterization chart. However, this task is more difficult than collecting a statistically
significant number of data points. The data that actually defines the correlations are based only
on field performance data that have a liquefaction factor of safety near one. Soil deposits with
obvious liquefaction (i.e., factor of safety <O. 7) do not contribute toward establishing the
liquefaction resistance contour lines. Sand boil expression at ground surface during an
earthquake signifies that a sandy soil layer is densifying due to liquefaction; the problem is
identifying which soil layer(s) liquefied. Also, not all liquefied deposits express sand boils at the
ground surface. Establishing a useable field performance database that contains marginal
liquefaction data is difficult. A good field performance database must represent all soil types and
all relative strength levels and have a sufficient number of data points representing marginal
liquefaction. However, only part of the worldwide field performance database represents
marginal liquefaction. The required field performance database to establish a CPT based
correlation ofliquefaction potential is not likely to exist for the foreseeable future.

Using Field Performance Data to Prove the CPT Soil Characterization Chart Technique

The best use for field performance data is to confirm the CPT soil characterization chart technique
because the quantity of field performance data to establish a CPT soil chart technique is not likely
to exist for the foreseeable future. On the other hand, cyclic laboratory data represents a factor of
safety of one and can be used directly to establish contours ofliquefaction CRR1 . Cyclic
laboratory data should only be considered an indicator ofliquefaction potential while field
performance data should be considered direct data. Figures 14 and 15 show the field performance
data from Suzuki, et al. (1995) together with the CPT soil characterization chart-based contours
for CRR1 equal to 0.25 and 0.15, respectively. The field performance data for CRR1=0.25
matches the CPT soil characterization chart CRR1=0.25 contour in Figure 14. A few of the field
performance data are beyond the CPT soil characterization chart CRR1=0.15 contour in
Figure 15.

253
 1000 .,
I!?
"'le
l3
zw
_,
<I)

0
No Liquefaction 't:
Cl
.,;
~
IP- 0::
u
u.
b,. b,. 0
IS.

-t i)
:::
C:
:J
&
A.
b,.b,.
Field Performance Data
Suzuki et al (1995)
Liquefaction data
100
,,_.
--
E
ro
Q)
AA
/ Non Liqufaction data

.......
-
C: A A
ro A A
t i)
'iii Liquefaction .A
AA ~i
~ 6
A
Q)
C:
0
0
A A /
"C
Q)
.!:::! Proposed relationship
ro for Liquefaction
10
E
L.
Cyclic Resistance Ratio
0 (CRR1) = 0.25
z (Olsen 1988)

1
0.1 1.0 10.0
Friction Ratio(%)

Figure 14 Comparison of CPI soil characterization chart technique (Olsen 1988) to field
performance data (Suzuki, et al.. (1995) for a normalized liquefaction cyclic
resistance ratio (CRR1) equal to 0.25.

254
 1000
Three da1a points are outside the CRR=0.15 relationship and represent
15% of the displayed liquefied field performance data. All of the displayed
CRR=0.15 data points were averaged from field data performance data
ranging from CRR=0.1 to 0.2, thus the data having CRR between 0.15 to 0.2
are also plotting beyond the CRR=0.15 contour. Some of this averaged data •
therefore has build-in bias. ~
i

-( /)
:!=
C:
:J

-E
ca
'--'
(I}
(.)
100

C:

~ •"'
Liq uefoctio_£
1::,. Field
Performance
·en
(I}
1::,.~
Data
0::: Suzuki et al (1995)
(I} Non Liquefaction data
C: Liquefaction data
0
0
"O
(I}
.!::!
ca 10
E
,._
0 Proposed relationship/
z for Liquefaction
Cyclic Resistance Ratio
(CRR1) = 0.15
(Olsen, 1988)

1
0.1 1.0 10.0
Friction Ratio(%)

Figure 15 Comparison ofCPT soil characterization chart technique (Olsen 1988) to field
performance data (Suzuki, et al .. 1995) for a normalized liquefaction cyclic
resistance ratio (CRR1) equal to 0.15. Three of the data points fall outside of the
CRR1=0. l 5 bounds, most likely because the field performance data is averaged.

255
Suzuki Field Performance Data to Prove the CPT Soil Characterization Chart Technique

Three of the Suzuki et al (1995) field performance data points plot beyond the CPT soil
characterization chart-based CRR1=0.15 contour in Figure 15. These three data points represent
15% of the 19 liquefied field performance data for CRR1=0.15. These three points may not
represent the critical data point such as from Niigata or San Fernando Dam on the H. Bolton Seed
SPT liquefaction potential chart. The CRR1=0.15 field performance designation represents an
average corresponding to data within the range of 0.1 to 0.2. Figure 16 graphically illustrates the
probable width of the CRR1=0.15 designation. The upper end of this CRR1=0.15 contour has
CRR1 levels from 0.15 to 0.2. It is therefore logical that some of the CRR=0.15 field
performance data should plot between the 0.15 and 0.20 contours in Figure 15. The same number
of data points from Suzuki also fall outside of the normalized cone resistance versus CRR1
relationship for clean sand in Figure 7. It therefore appears that these outlying data from Suzuki
in Figure 7, Figure 15, and Figure 14 may have built-in conservatism.

New Graphic Format for the CPT Soil Characterization Chart Technique

The CPT soil characterization chart technique for determining liquefaction potential is an
unfamiliar format to many geotechnical engineers. A more familiar format might enhance the
acceptance of this technique. The CPT soil characterization chart-based technique can be
translated to the same graphic format as the standard plot of normalized cone resistance versus
CRR1. Contours ofCRR1 from the CPI soil characterization chart (Figure 10) were transformed
into Rr contours on a chart of qc1 versus CRR1 as shown in Figure 17 (Olsen, Koester, and Hynes
1995). This is an exact transformation, either chart (Figure 10 or Figure 17) will estimate the
same CRR1. Liquefaction CRR1 is estimated (from the vertical axis in Figure 17) by scaling the
CPI qc 1 on the horizontal axis and intersecting the corresponding CPT Rr contour.

Soil classification contours (Olsen and Mitchell 1995) and corresponding CPI estimated fines
content (from Figure 3) are shown in Figure 18. The CPT characterization chart technique can be
confirmed on a project basis by comparing measured soil classification to the contours of soil
classification in Figure 18. Remember, soil samples retrieved several meters from a CPT sounding
may not reflect the probed soil.

Figure 19 displays the various cone resistance-based techniques for estimating liquefaction
resistance of clean and non clean sands together with the soil characterization chart technique.
Measured soil indices are used with the cone resistance-based techniques for estimating
liquefaction resistance. Measured soil indices now have a double purpose: l) to estimate
liquefaction resistance using the cone resistance technique (using Figure 19), and/or 2) to confirm
the CPT soil characterization chart technique (in Figure 18). These charts allow the results of
both techniques to be compared when soil indices are measured.

Figure 20 is a complex compilation including the cone resistance-based technique for estimating
liquefaction resistance, the CPT estimated soil types, inferred fines contents trends, and finally

256
 Olsen, Koester & Hynes ( 1996)
Olsen (1988, 1984), Olsen & Mitchell (1995)

Dense
£. '3,°/0.

..
'f\ne~ ..
The CRR 1=0.15
contour was
developed using
field performance
data covering a
(I} range of CRR data
0 between 0.1 and 0.2
C
Jg
-~
(/)

~
(I}
C
0
0
"C
(I}
.!::!
~
,_ 10
0
z

LO
.,.... .,.... N N (")
ci ci cicio
CPT predicted Normalized
Liquefaction Cyclic Resistance Ratio (CRR 1)
1 L------'------l..-.l...--'--...l....-L...l......L.J._ _ __,__ _.___,JL_.l._1-...1.-'-.1.-J

0.1 1.0 10.0

Friction Ratio(%)

Figure 16 The data range for the CRR1=0. l 5 contour which Suzuki, et al (I 995) used to
establish the CRR1=0.15 designation.

257
-c::c::
,..... 0.6

--
(.)
0
:;:;
a:s 0.5
c::
a,
(.J
C
.....a:s
en 0.4
en
a,
c::
(.J

0 0.3
>,
(.)
C
0
13 0.2
~
a,
:::I
C"
:.:::i
"C
a, 0.1
.!:::!
a:s
E
I..
0
z 0.0
0 50 100 150 200 250
Normalized Cone Resistance (atm units)

Figure 17 Relationship between the normalized liquefaction cyclic resistance ratio (CRR1)
and normalized cone resistance in terms of the CPT friction ratio contours
proposed by Olsen, Koester, and Hynes (1996) ..

258
 0.6
CPT predicted soil classification (Olsen, 1988; Olsen & Mitchell, 1995)
>-
(ll based on the CPT predicted liquefaction Cyclic Resistance Ratio (CRR)
0

-
0::
0::
0.5
"C

~
'.Q
0<I)
CPT predicted fines
content (Olsen, 1984, 1988)
-~ ~o"'
~ C-:,0 '!!...i,(/.,'<f. c,\c,\

--
C)

~
0
m
C:
0
(.)
c? ~'o-c

~'<>
Ci lf'.(\ c,\~
~ \ec e\'ll-
,;/> ~~e ~
0""'<> {'-'O,z. e \O 0~
-~o.~ .o"'
0:: '\, -.:,.,o.~ ~~
0.4 ~e 01
Q)
(..) o'"'e r>.;\99
C 1...."-g'o
.....m ro
C),
·o(\'
\C\~ 10!01
.!:!2 ~<t ~o'
Cl) C)~ ~'lJ.trJ'.l).e'=>
Q) 0.3 ~~
0:: o:\
.2
c3
>,
C)
C 0.2
.Q
.....
(..)
~
Q)
::I
.Q'" 0.1
....J

0.0
0 50 100 150 200 250
Normalized Cone Resistance (atm units)

Figure 18 Inclusion of CPT estimated soil classification on the cone resistance versus CRR1
chart using the CPT soil characterization technique for estimating liquefaction
CRR1

259
 .......
er: 0.5
er:
-
C)

0
:.:;
ro
.
.'
er: 0.4 •
. :.
o I
I 0

23
C:
'

-
ro
en
·u;
Q) 0.3
er:
0
·c:;
(']
C: 0.2
0
u
~
Q)
::I
er 0.1
::::i

0.0 .__..__..__.,___..__..__..__..__..__..__..__..__..__..__..__.__.__.__.__.__.__.__.__.__,.___,
0 50 100 150 200 250
Normalized Cone Resistance (atm units)

Figure 19 Comparison of the cone resistance based technique and the CPT soil
characterization chart based technique on the cone resistance versus CRR1 chart

260
 0.6
CPT predicted soil classification (Olsen, 1988; Olsen & Mitchell, 1995)
~based on the CPT predicted liquefaction Cyclic Resistance Ratio (CRR)
(.)

-
0::
0::
0.5
zw
~
0

--
(..)

0
:;:::; &
s
~

co ... C,
0:: 0.4
Q) "
"'
(.) "'
C:
co
1n
·u5
Q) 0.3
0::
(.)
(.)
>,
(..)
C:
0
0.2
NCEER Recommended
I '
And/or

u
.1!!
Cone Resistance based
technique for
Q) prediction of CRR for sands
:::I D 50 "' 0.25 and/or Fines< 5%
0- 0.1
.....J .____::.._:=:::::=:__----:------= D 50 = 0.15 mm
L--==------------- D 50 = 0.10 mm
~ - - - - - - - - - - - - - - - - F i n e s = 35%
D50"' 0.05 mm
0.0
0 50 100 150 200 250

Normalized Cone Resistance (atm units)

Figure 20 Inclusion of I) CPI estimated soil type, 2) CPT soil characterization based
technique for predicting CRR1, and 3) the cone resistance technique for estimating
CRR1 on the cone resistance versus CRR1 chart

261
friction ratio contours for estimating liquefaction potential using the CPT soil characterization
chart technique. Note how the CPT estimated fines contents from the soil classification contours
match the contours at different fines contents for the cone resistance-based technique. For
example, the soil characterization technique estimated liquefaction resistance contour for "fines <
5%" matches the cone resistance-based contour labeled "D50=0.25 mm or Fines<5%". Also, the
cone resistance-based technique contour labeled "Fines=35%" is between the CPT soil
characterization contours labeled "Fines= 40 to 60%" and "Fines= 10 to 15%" at high CRR1
levels. Figure 20 illustrates that the fines content trends with the soil characterization chart
technique match the trends from the cone resistance-based technique.

CPT-Based Estimation of Liquefaction CRR1 Using a Single Equation

Specialized software is required to effectively estimate liquefaction CRR1 for all data points in a
CPT sounding using Figure 10. A single equation was therefore developed which matches all the
non-linear contours in Figure 10 and is applicable to all soil types. CPT based estimation of CRR1
in Figure 10 (or Figure 17) can be approximated and simplified with the following equation:

CRR 1 = [0.00128 :c ] - 0.025

(ov)°.7
+ (0.17 R1) - (0.028Rj) + (0.0016Rj) (16)

= Generalized nonnalized cone resistance ( qc1 )

( a'v )o.7 (see Equations 7 and 8)

CRR1 = Normalized liquefaction cyclic resistance ratio

Rr= Calculated friction ratio (percentage)
qc,= CPT measured cone resistance (in atm units) (1 atm = 100 kPa)
av= Vertical effective stress (in atm units)

Equation 16 is a simplification of Figure 10 because nonnalization of the CPT cone resistance can
be simplified and the CRR1 curves have been generalized. These assumptions are adequate for
non critical conditions and for depths between 5 and 13 meters. The estimated CRR1 from
Equation 16 for clays and clayey silts is more conservative than from Figure 10. This
conservatism may flag soil mixture and clay layers as liquefiable and therefore necessitate soil
sampling (to determine the liquidity indices or for vane shear testing to determine sensitivity). For
deep deposits of clay, the stress nonnalization portion of this equation generates CRR1 values too
high. The equation is also conservative for overconsolidated conditions. The uniqueness of this
technique is that it can estimate CRR1 for all soil types and now is in an equation fonn.

262
 Consequence of Liquefaction Resistance as Estimated by the CPT
The consequence ofliquefaction can be divided into three broad categories: 1) flow slides, 2)
deformation/cracking, and 3) sand boil expression. These categories can be differentiated if they
all represent a soil near the bottom toe of a slope. A liquefaction-induced flow slide typically
represents I 00% pore pressure generation, large strain softening, a dramatic drop in shear
strength, and large slope movements. Deformation and cracking of a slope without massive
movement suggests that I) slope movement causes the soil to dilate which increased the soil
resistance to movement, or 2) slope movement stopped when the earthquake motions stopped
(i.e., Newmark sliding). Sand boil expression on level ground surface reflects grain matrix
densification, resulting in water expulsion. Sand boil expression at the bottom of a slope without
slope cracking/movement indicates that an isolated liquefied sand zone generated excess
pore fluid.

Consequences of liquefaction can be depicted on the CPT soil characterization chart, as shown in
Figure 21, based on relative density (for sand) and soil type. Specifically, liquefaction flow
potential with large slope movement can occur with very loose sands. Cyclic liquefaction
potential (having low slope movement potential) can occur with medium dense sand. Dense sand
can experience cyclic mobility where a slight slope movement causes the sand to dilate, resulting
in a dramatic strength gain and decrease of slope movement. Finally, cyclic mobility potential
with little potential for post earthquake sliding can occur for normally consolidated clays. These
trends from cyclic mobility to liquefaction flow potential depicted in Figure 21 are inversely
proportional to the CPI friction ratio. The potential for slope movement is low to moderate for
Rr=l % and increases to major for Rr=0.2%. This Rr trend to liquefaction flow potential is not
dependent on soil type. If a soil layer has a calculated liquefaction factor of safety near or less
than 1 then the next step is to check for liquefaction flow potential using the CPT friction ratio. If
a soil layer is estimated to liquefy and the friction ratio is low (such as lower than 0.6%), and
ground geometry is not level, then there may be potential for slope movement.

Examples of CPT Estimated Liquefaction CRR1

Examples of the CPI soil characterization technique for estimating liquefaction resistance are
shown in the next several figures for sites which have experienced earthquakes. Each chart
presents raw CPT data ( cone resistance, sleeve friction resistance, and calculated friction ratio)
and CPT estimated geotechnical properties (soil classification, liquefaction CRR, and fines
content). Also shown are estimates of the average induced-earthquake cyclic stress ratio (CSR)
from a recent earthquake for comparison to estimated liquefaction resistance CRR.

The first two examples are for the Moss Land site that experienced liquefaction during the 1989
Loma Prieta earthquake. The first example, in Figure 22, is from Sandholdt Road next to Moss
Landing Harbor at CPI sounding UC-4 (data from Boulanger, et al. 1997, Boulanger 1998).
Slope inclinometer measurements show that 26 cm oflateral deformations occurred between a

263
 1000 .-----,--.--~~-r-,--.-,r,------r----r-----r-----r-----r-....,.......-,-,
Olsen (1988, 1984)
Olsen and Koester (1995) If the average earthquake induced cyclic stress ratio ~
Olsen & Mltchell (1995) (ISR or CSR) is close or greater than CRR ~
then check for liquefaction consequence 0

(i.e. flow) by examing the CPT friction ratio(%) z

w
rn
..J
0
Friction ratio is inversely proportional
to flow liquefaction potential

-
~
UJ
C
::J

.....E 100

-ca
~
C
.....ca
.SQ
UJ
~
Q)
C
8
"O
Q)
.N 10
ca
E
5...
0
z

IO IO
,.- ,.- NNM
c:i c:i c:i ci c:i
CPT predicted Normalized
Liquefaction Cyclic Resistance Ratio (CRR1)
1
0.1 1.0 10.0

Friction Ratio (%)

Figure 21 Fully annotated chart for estimating normalized liquefaction cyclic resistance ratio
(CRR1) using the CPI soil characterization chart technique.

264
depth of 2 and 4. 5 meters. Between a depth of 3. 5 and 6 meters the earthquake induced CSR is
at the low extreme of the CPT estimated CRR range, i.e., marginal liquefaction. This zone is
composed of soil mixtures and sand. A relatively clean sand layer at a depth of 4 to 4. 5 meters
has a friction ratio of0.2% but a CRR>0.3. It appears that slope movement occurred within this
soil mixture soil zone between a depth of3.5 and 4 meters.

The second example for Moss Landing are three CPT soundings near the Marine laboratory
building (adjacent to a volleyball court) as shown in Figure 23 (data from Boulanger, et al. 1997;
Boulanger 1998; Youd 1997). Silt was observed in the sand boil expressions next to the
volleyball court (Boulanger, et al. 1997). The CPT-estimated data infers extensive liquefaction at
numerous depth zones composed of sands and soil mixtures. The excess pore fluid for the sand
boil expressions probably originated from the clean sands at a depth of I. 5 to 1O+ meters. The
soil mixture layer composed of silts to dirty sand located at a depth between 1.5 and 4 meter,
were probably liquefied during the earthquake and "dragged" to the ground surface with flow of
excess pore fluid from all the liquefied sand layers.

The next example, in Figure 24, is from the USGS Wildlife site, CPI sounding c3g, which
experienced sand boil expression on the ground surface and minor lateral spreading (data from
Youd 1997). Based on the CPI-estimated data, three soil zones appear to have experienced
liquefaction: a thin soil mixture zone layer at 2 meters, a dirty sand zone from 2.5 to 3.6 meters,
and clean sand to dirty sand from 5.8 to 6.2 meters. The soil zone at 2.5 to 3.6 meters has a
friction ratio less than 0.5% and therefore a high potential for slope movement. Sand boil
expression on the ground surface probably originated from a depth of5 to 6.5 meters because of
the low fines content.

Cross Sections of CPT-Estimated Liquefaction resistance

Cross sections of CPI-estimated liquefaction resistance are the ultimate means of evaluating
liquefaction potential. Individual depth plots of CPT-based estimation ofliquefaction resistance
have been the state-of-the-art for some time (Olsen 1988), but individually provide no spatial
information. However, cross sections of CPT-estimated liquefaction are the best means of
geotechnical site characterization for the purpose of evaluating sand boil expression potential,
building foundation instability, and sliding potential. Isolated liquefiable soil lenses at different
elevations can be identified with cross sections and are generally not a problem. Continuous
liquefiable lenses or layers can be a major stability problem. If a continuous soil layer or zone is
estimated to experience liquefaction, then CPI-estimated soil classification ( or friction ratio) can
be used to estimate the potential for slope movement.

A cross section of CPT-estimated liquefaction resistance for Moss Landing is shown in Figure 25.
This cross section was developed using the soil layer tracing technique. Note how continuous soil
layers oflow liquefaction resistance can be tracked across the site.

265
 CPT sounding: UC4
Boring: UC-B10 CPT CPT
Site: Moss Landing, CA , USA
estimated* estimated* soil
Data Source: Prof Boulanger (UC Davis)
liquefaction classification
Cone resistance
(atm units) Sleeve friction Calculated Cyclic Resistance
(1 aim~ 100 kPa)
resistance friction ratio Ratio, CRR
(percentage) (M= 6) CPT estimated*
(atm units)
(MSF=1.3) fines content(%)
0 0 0
0 0 .- ~ ci 0 <;)"-'1,'!>~
.- 0 0 0
ci .- ..- <;)· <;)· <;)· <;)· <;)· O 7-0 AO roO io '\oO
0
. . .
i
1
i
I
I

i
I
I
,
I
2
3
-¥- ,;
I
I
I

i
I
I
!
1;,
4 ' I ~
~ ~,.
~

5 '
....... 6
~ 7 ~
.
*
N
8
°'°' E 9
..__., I I

£ 10
C.
•.
0
a, 11
12
13 - ll
I
I
14 I
~ I
15 I
I
;;,1
16 #1
N1
ol

17 iii.,1 "''Kl:
Ill

cl
18 "''. . .F,
.
- - - - - - · Normalized Cone Resistance, q 01 • = Olsen Soil Characterization Technique • measured

Figure 22 Use of the CPT soil characterization technique for estimating normalized liquefaction cyclic resistance ratio (CRR1}
for CPT soundings UC-4 at Moss Landing which experienced liquefaction and lateral spreading during the 1989
Loma Prieta Earthquake (Data from Boulanger, 1998).
 CPT soundings: C-3, UC-8, & UC-7
Boring:none CPT CPT
Site:Moss Landing (Volleyball Court) CA, USA estimated* estimated* soil
Data Source: Youd (1997) & Boulanger (1998) liquefaction classification -
Cyclic Resistance ~
Sleeve friction
Cone resistance
(atm units) resistance
(atm units)
Calculated
friction ratio
Ratio, CRR
(M=7)
e
.a .a
"'
I!!
11
E
~
'ti
(1 aim= 100 kPa) (percentage) (MSF=1.19) -~ -~ : re
0
0 0 0 0 8 0 .,E'tl~c~
0 8 0 d C! d 0 d ~ .... cc:Bc
d d SCN= -2 U -1 ~ 0 8J 1 8J 2 U 3 8J

3
,;;-
._
4
....
<I)
<I)
g 5 edium
.c
..... Clean
a.
6
""
en
-..J
0
<I)

10
,. = Olsen Soil Characterization Technique

Figure 23 Use of the CPT soil characterization technique for estimating normalized liquefaction cyclic resistance ratio (CRR1)
for three CPT soundings at volleyball court next to the Moss Landing Marine Laboratory where "silt" boil
expressions were observed (next to the volleyball court) during the 1989 Loma Prieta Earthquake (Data from Youd
1997 and Boulanger 1998).
 CPT sounding: C3g
Boring: C3g CPT CPT
Site: Wildlife CA, USA
estimated* estimated* soil
Data Source: Prof Youd (B.Y.Univ)
liquefaction classification
Cone resistance
(atm units) Sleeve friction Calculated Cyclic Resistance
(1 atm = 100 kPa) resistance friction ratio Ratio, CRR
(atm units) (percentage) (M=6. 6 ) C PT estimated*
....0 8 C! (MSF=1.38) fines content(%)
0
....0 0
....
.... ....
0
0
....
0
0
....
0
c:i c:i ...q ....c:i c:i ....0 ~ r:::,~ r:::,':- r:::,'),, r:::,'> r:::,~ O 7-0 AO e,O sO '\QO
0
1
2

rv
3
a,
00

-...
en
4

5
-
Q)
Q)
E
6
---
.c 7
0..
Q)
8
0
9

10
11
12

13
------·Normalized Cone Resistance, q01 •=Olsen Soil Characterization Chart Technique • measured

Figure 24 Use of the CPT soil characterization technique for estimating normalized liquefaction cyclic resistance ratio (CRR1)
for CPT soundings c3g at the USGS Wildlife research site which experienced liquefaction (Data from Youd 1997).
·•·.·.b;§~~~~t:I,gt4ii~ft1#r1\;1:i•··••
.• Harbor Master bulidlng ..
• Moss Landing, California

Figure 25 Cross section example (using the soil trace method) ofCPT estimated liquefaction resistance for the Harbor Master
building area at Moss Landing, California (CPT data from Boulanger 1998).
 A Procedure for CPT-Based Estimation of Liquefaction Resistance

This section describes the step by step procedure for using the CPT soil characterization
technique to estimate liquefaction potential. Details ofthis technique are described in the body of
this paper.

The two critical parts of this procedure are 1) calculating the normalized CPT cone resistance and
2) estimating the liquefaction resistance using the CPT soil characterization technique. The
procedure starts by normalizing the CPT cone resistance to a standard vertical effective stress of
1 atmosphere (atm) (1 atm = 100 kPa "" bar "" tsf). The next step is CPT-based estimation of
normalized liquefaction cyclic resistance ratio (CRR1) using normalized CPT data for all data in
each CPT sounding. The CPT soil characterization technique does not use a silt correction factor
to determine an equivalent clean sand cone resistance for the purpose of estimating liquefaction
resistance because there is no physical meaning for an equivalent clean value. CRR1 is then
converted to an equivalent liquefaction resistance ratio (CRR) representing the in situ vertical
effective stress and for the design earthquake. CRR is then compared to the earthquake induced
cyclic shear stress ratio (CSR) to calculate the liquefaction factor of safety and to evaluate soil
behavior ifliquefaction is predicted to occur.

The seven step procedure is:

1) Determine the earthquake induced cyclic stress ratio (CSR) with the following steps:
a) Establish the earthquake magnitude (M)
b) Determine the surface ground acceleration (ag)
c) Establish the CSR versus depth profile, by one of two methods:
i) with software evaluation ( such as with SHAKE software), or
ii) with the rd technique

2) Estimate CRR1 using the CPT soil characterization technique.

This step requires normalizing the CPT cone resistance which is used to estimate the
CRR1 for all data points in a CPT sounding. Using the best available stress normalization
technique produces the most accurate predicted value. For many geotechnical situations,
an approximating stress normalization techniques will achieve good results. There are
numerous techniques for stress normalization of the CPT cone resistance, ranging from
constant exponent techniques to the soil characterization chart technique requiring
specialized software. The CPT estimated CRR1 can be accomplished by one of the three
following procedures (from a single equation procedure to a technique requiring computer
software):

270
 a) CPI-based procedure for estimation ofCRR1 using a generalized equation.
This procedure shown in Equation 16 is a generalization and conservative
representation of the CPI soil characterization chart technique shown in Figure 10
(or Figure 17) (and described in item c below). This simplified procedure uses a
constant stress exponent of O. 7 (for all soil types) to normalize the cone resistance
in Equation 16. This approach can be justified for most non critical liquefaction
evaluations and for a vertical effective stress range of0.8 to 1.3 atm (80 to 130
kPa). This formulation was designed to produce conservative CRR1 values for
clays and silt mixtures when compared to the chart solution in Figure 10. For deep
deposits of clay this formulation generates CRR1 values close to the chart solution.

b) Modification of the simplified equation.

The above procedure can be improved by using more accurate estimates of the
stress exponent. For example, the 0.7 stress exponent in Equation 16 can be
replaced with soil type dependent exponents, such as 0.6 for sand and 1 for clay.
Alternatively, the stress exponent for sands can be estimated using Equation 9
based on the estimated relative density.

c) The CPI soil characterization chart technique for estimating liquefaction resistance.
The CPI soil characterization chart based technique for stress normalization is
described by Olsen (1994, 1998, 1995) and Olsen & Mitchell (1995) and requires a
software program because it's an iterative technique. This technique is usable for
all soil types, strength levels, and confining stress levels.

3) Calculate the CPI soil classification number (SCN) using Equation 11 (see Figure 3).
Estimating soil type is as important as estimating liquefaction resistance because soil type
controls soil behavior ifliquefaction is triggered. The CPT estimated soil classification
number (SCN) shown in Figure 3 is an index of soil classification (Olsen and Mitchell
1995) and can be estimated by Equation 11. The SCN was developed to represent unique
soil behavior at integer values (i.e., -1, 0, and 1) and is important for liquefaction
evaluation. At SCN equal to -1, the soil should have a behavior between that of a silty
clay and clayey silt. At SCN equal to I, the soil should have a classification and behavior
between that fine sand and silty sand.

4) Calculate CRR (in situ value) from CRR1 (normalized value) based on Equation 2 using
the Earthquake magnitude scale factor (MSF) (see Equation 3) and Ka

5) Calculate liquefaction factor of safety (p,liq) using Equation 1.

6) If the FSliq is less than 1.2 then sand boil and slope movement potential should be
evaluated using the procedure in Figure 21. If the friction ratio is less than 0.8% and FSliq
< 1.2 than there is potential for slope movement. If the CPI friction ratio is less than
0.5% and FSliq < 1.2 then there is major potential for movement. The potential for sand

271
 boil expression is high when the SCN is high, such as SCN>l. If the SCN is low, such as
SCN<O (i.e. silt to clay), the soil layer should be sampled and evaluated using Liquidity
Index criterion (Equation 15) or the strength sensitive determined, such as with the vane
shear device. If the earthquake induced stress is greater than the static strength of a
normally consolidated soft clay (i.e. large nearby earthquake with CSR=0.28) then the
soil can dynamically deform to meet the liquefaction criterion. Any continuous low FSliq
soil layer should be independently sampled, using SPT procedures, to at least verify CPT
predicted SPT values and gradation characteristics.

7) The CPT estimated data should then be displayed in a depth or elevation format such as
shown in Figure 24. CPT soundings should be positioned along strings (Olsen and Farr,
1986) with the CPT estimated CRR1 shown in cross sections such as shown in Figure 25.

Conclusions

The CPT soil characterization-based chart technique for estimating liquefaction resistance of any
soil type is shown in Figure 10 (or Figure 17) and can be expressed by Equation 16. This
technique is comprehensive, technically mature, and well published and was developed based on
many contributing factors. Numerous stress normalization techniques were described in this
paper for calculating the normalized CPT cone resistance. The criticalness of the structure should
dictate which stress normalization technique is required. Estimated liquefaction resistance based
on the CPT characterization chart technique can be confirmed on a project basis by comparing
measured fines content to the CPT-estimated fines contents from the CPT soil characterization
chart in Figure 3 . The best means of verifying the CPT soil characterization chart approach is if
CPT-predicted SPT (using Figure 11) matches field measured values. The uniqueness of this
technique is that both CPT measurements are used to estimate liquefaction resistance for all soil
types, all strength levels, and for all overburden stresses. All other techniques require estimating
the soil fines content, calculating a silt correction factor in order to determine the equivalent clean
sand value which is then used to estimate liquefaction resistance using equivalent clean
relationships. There is no physical or theoretical meaning for the equivalent clean sand value; it's
just a convenience for determining liquefaction resistance of non clean sands. The CPT soil
characterization technique directly estimates liquefaction resistance using both CPT strength
measurements because these measurements are unique.

References

Boulanger, R.W., Mejia, L.H., and Idriss, I.M. (1997) "Liquefaction at Moss Landing during
Loma Prieta Earthquake," Journal ofGeotechnical and Geoenvironmental Engineering, ASCE,
May 1997, 123 (5), pg 454-467

Boulanger, R.W. (1998). Personal communication to R.S.Olsen (CPT and SPT data for Moss
Landing).

272
Casagrande, Arthur. (1932) "The structure of clay and its importance in foundation engineering,"
Contributions to Soil Mechanics 1925-1940, Boston Society of Civil Engineers, pg 257-276.

Campanella, R. G. and Robertson, P. K. (1984). "A seismic cone penetrometer to measure

engineering properties of soil," 54th Annual International Meeting and Exposition of the Society
ofExploration Geophysicists, Atlanta, GA, pg 138-141.

Douglas, B.J., Olsen R.S., and Martin, G.R. (1981). "Evaluation of the cone penetrometer test
for SPT liquefaction assessment." Proceedings of in situ testing to evaluate liquefaction
susceptibility, ASCE: New York.

Douglas, B. J., and Olsen, R. S. (1981). "Soil classification using the electric cone
penetrometer," Proceedings of a Session Entitled Cone Penetration Testing and Experience -
ASCE Fall Convention, ASCE, New York, 209-227.

Douglas, B. J., Olsen R. S., and Martin, G. R. (1981). "Evaluation of the Cone Penetrometer
Test for SPT Liquefaction Assessment," Proceedings of In Situ Testing to Evaluate Liquefaction
Susceptibility, ASCE, New York.

Marcuson, S.F. III and Bieganousky, W.A. (1977), "SPT and Relative Density in Coarse Sands,"
Journal ofGeotechnical Engineering Division, ASCE, Vol. 103, No. GTl 1, pp. 1295-1309.

Martin, G. R., and Douglas, B. J. (1980). "Evaluation of the Cone Penetrometer for
Liquefaction Hazard Assessment," for U.S. Geological Survey, Contract No. 14-08-0001-17790,
October, Project No. 79-153, Long Beach, CA.

Olsen, R. S. ( 1996) "The influence of confining stress on Liquefaction Resistance",

US-Japan workshop on Advanced Research on Earthquake Engineering for Dams, at Waterways
Experiment Station, Vicksburg, MS, Nov 12-14, 1996.

Olsen, R.S., Koester, J.P., and Hynes, M.E. (1996) "Evaluation of Liquefaction Potential using
the CPT", Proceedings of the 28th Joint meeting of the US-Japan Cooperative Program in
Natural Resources- Panel on Wind and Seismic Effects, U.S. National Institute of Standards and
Technology, Gaithersburg, Maryland, May 1996

Olsen, R.S. and Mitchell, J.K. (1995) "CPT Stress Normalization and Prediction of Soil
Classification", Proceedings of the International Symposium on Cone Penetrometer Testing -
CPT'95, Linkoping, Sweden, October 1995

Olsen, R.S. and Koester, J.P. (1995) "Prediction of Liquefaction Resistance using the CPT",
Proceedings of the International Symposium on Cone Penetrometer Testing - CPT'95, Linkoping,
Sweden, October 1995

Olsen, R. S. ( 1994) "Normalization and Prediction of Geotechnical Properties using the Cone

273
Penetrometer Test", Technical Report GL-94-29, USAE Waterways Experiment Station,
Vicksburg, MS 39180, August 1994

Olsen, RS. (1994) "Normalization and Prediction ofGeotechnical Properties using the Cone
Penetrometer Test (CPT)", Ph.D. Dissertation submitted to the University of California, Berkeley,
May 1994

Olsen, R.S. (1988) "Using the CPT for dynamic site response characterization". Proceedings of
the Earthquake Engineering and Soil Dynamic II Conference, Geotechnical Special Publication
Number 2, ed. J. Lawrence Von Thun, 374-388. ASCE: New York.

Olsen, RS. (1988) "Soil classification and site characterization using the cone penetrometer test"
Proceedings of the First International Symposium on Penetration Testing (JSOPT-1), ed. J. de
Ruiter, 887-893. A. A. Balkema: Rotterdam, Netherlands.

Olsen, R. S. and Farr, J. V. (1986) "Site Characterization using the Cone Penetrometer Test,"
Proceedings ofln Situ 86 - Use ofln Situ Tests in Geotechnical Engineering, Geotechnical
Special Publication Number 6, S. P. Clemence, ed., ASCE, New York, pg 854-868.

Olsen, R.S. (1984) "Liquefaction analysis using the cone penetrometer test" Proceedings of the
Eighth World Conference on Earthquake Engineering, Volume III, 247-254. Prentice-Hall Inc.:
Englewood Cliffs, NJ.

Robertson, P. K. (1982) "In-Situ Testing of Soil with Emphasis on its application to Liquefaction
Assessment," Ph.D. Dissertation submitted to the University of British Columbia, December,
1982

Robertson, P. K., Campanella, R G., Gillespie D., and Greig, J. (1986). "Use of Piezometer
Cone Data," Proceedings ofln Situ 86 - Use ofln Situ Tests in Geotechnical Engineering,
Geotechnical Special Publication Number 6, S. P. Clemence, ed., ASCE, New York, 1263-1280.

Robertson, P. K. and Campanella, R G. (1983). "Interpretation of Cone Penetration Tests -

Part I (Sand)," Canadian Geotechnical Journal, Vol. 20, No. 4, pp. 734-745.

Robertson, P. K. and Campanella, R G. (1983). "Interpretation of Cone Penetration Tests -

Part U (Clay)," Canadian Geotechnical Journal, Vol. 20, No. 4.

Robertson, P. K., Campanella, R G., and Wightman, A. (1983). "SPT-CPT Correlations,"

Journal of the Geotechnical Division, ASCE, Vol. 109, November.

Schmertmann, J. H. (1976). "Predicting the q/N Ratio," Final Report D-636, Engineering and
Industrial Experiment Station, Department of Civil Engineering, University ofFlorida,
Gainesville.

274
Schmertmann, J. H. (1978a). "Guidelines for Cone Penetration Test, Performance and Design,"
U.S. Department of Transportation, Federal Highway Administration, Report FHWA-TS-78-209,
Washington, July 1978, 145 pgs.

Schmertmann, J. H. (1978b). "Study of Feasibility of Using Wissa-Type Piezometer Probe to

Identify Liquefaction Potential of Saturated Sands," U.S. Army Engineer Waterways Experiment
Station, Report S-78-2.

Schmertmann, J. H. (1979a). "Statics ofSPT," Journal of the Geotechnical Engineering

Division, ASCE, Vol. 105, No. GT5, May.

Schmertmann, J. H. (1979b). "Energy Dynamics ofSPT," Journal of the Geotechnical

Engineering Division, ASCE, Vol. 105, No. GT&, August.

Seed, H.B. (1976). "Evaluation of Soil Liquefaction Effects on Level Ground During
Earthquakes," Liquefaction Problems in Geotechnical Engineering, ASCE Preprint 2752,
Philadelphia, PA

Seed, H.B. and De Alba, P. (1986). "Use ofSPT and CPT tests for evaluating the liquefaction
resistance of soils, 11 Proceedings of the Specialty Conference on the Use ofln Situ Tests in
Geotechnical Engineering, Blacksburg, VA, ASCE Geotechnical Special Publication No. 6, pp.
120-134.

Seed, H.B. and Idriss, I. M. (1970). "Soil Moduli and Damping Factors for Dynamics Response
Analysis," Report No. EERC 70-10, Univ. of California, Berkeley, December.

Seed, H.B. and Idriss, I. M. (1981). "Evaluation ofLiquefaction Potential of Sand Deposits
Based on Observations of Performance in Previous Earthquakes," Geotechnical Engineering
Division, ASCE National Convention, St. Louis, Session No. 24.

Seed, H. B. and Idriss, I. M. (1982). Ground Motions and Soil Liquefaction during Earth-
quakes, Earthquake Engineering Research Institute, Berkeley, CA

Seed, H.B., Idriss, I. M. and Arango, T. (1983). "Evaluation ofLiquefaction Potential Using
Field Performance Data," Journal ofGeotechnical Engineering Division, ASCE, Vol. 109, No. 3,
March 1983, pp. 458-492.

Seed, H.B., Seed, R. B., Harder, L. F., and Jong, H. L. (1988). "Re-Evaluation of the Slide in
the Lower San Fernando Dam in the Earthquake of February 9, 1971," Report No. UCB/
EERC-88/04, April, Earthquake Engineering Research Center, University of California at
Berkeley, Berkeley, CA.

Seed, H.B., Seed, R. B., Harder, L. F., and Jong, H. L. (1989). "Re-Evaluation of the Lower
San Fernando Dam - Report 2 - Examination of the Post-Earthquake Slide ofFebruary 9, 1971,"

275
Contract Report GL-89-2, U.S. Army Engineer Waterways Experiment Station, Vicksburg, MS.

Skempton, A.W. (1986) "Standard Penetration Test Procedures and the Effects in Sands of
Overburden Pressure, Relative Density, Particle Size, Aging, and Over consolidation,"
Geotechnique, Vol. 36, No. 3, pp 425-447.

Suzuki, Y., Tokimatsu, K., Koyamada K., Taya, Y., Kurbota, Y. (1995) "Field Correlation of Soil
Liquefaction based on CPT Data," Proceedings of the International Symposium on Cone
Penetrometer Testing - CPT'95", Linkoping, Sweden, October 1995.

Suzuki, Y., Koyamada K., Tokimatsu, K., Taya, Y., Kurbota, Y. (1995) "Empirical Correlations
of Soil Liquefaction based on Cone Penetrometer Test," First International Conference on
Earthquake Geotechnical Engineering", K. Ishihara Editor, 1985.

Youd, T.L. and Idriss, I. M. editors. (1998) "Workshop on Evaluation of Liquefaction

Resistance," Proceedings of a 4-5 January 1996 workshop, sponsored by FHWA, NSF and WES,
to be published by the National Center for Earthquake Engineering Research, Salt Lake City.

Youd, T.L. (1997). Personal communications to RS. Olsen (CPT data for several sites)

276
 Appendix A
List of Participants

Chair of Workshop W.D. Liam Finn

T. Leslie Youd University of British Columbia
368 CB Civil Engineering Department
Brigham Young University 2324 Main Mall
Provo, UT 84602 Vancouver, BC
CANADA V6T 124
Co-Chair of Workshop
Izzat M. Idriss Leslie F. Harder
P.O. Box 330 California Dept. of Water Res.
Davis, CA 95617-0330 P.O. Box 942836
Sacramento, CA 94236-0001
Participants
Ronald D. Andrus Mary Ellen Hynes
Research Civil Engineer Earthquake Engr. and Geo. Div.
NIST Waterways Experiment Station
Bldg and Fire Rsrch Lab 3909 Halls Ferry Road
Bldg 226, Room Bl58 Vicksburg, MS 39180-6199
Gaithersburg, MD 20899
Kenji Ishihara
Ignacio Arango 2415-123 Nara-cho
22 Bowling Drive Aoba-ku, Yokohama 123
Oakland, CA 94618 JAPAN

Gonzalo Castro Sam Liao

GEI Consultants, Inc. Parsons Brinkerhoff
1021 Main Street One South Station
Winchester, MA 01890 Boston, MA 02110

John T. Christian William F. Marcuson, ill

Consulting Engineer GeotechLab
23 Fredana Road USAE Waterways Experimental Station
Waban, MA 02168-1103 3 909 Halls Ferry Road
Vicksburg, MS 39180-6199
Ricardo Dobry
Rensselaer Polytechnic Institute Geoffrey R. Martin
Civil & Environmental Engineering University of Southern California
JEC 4040 Civil Engineering Department
Troy, NY 12180-3590 Los Angeles, CA 90089-2531

A-1
James K. Mitchell P.K. Robertson
Virginia Tech University of Alberta
Civil Engineering Department Department of Civil Engineering
Patton Hall Edmonton
Blacksburg, VA 24061 CANADA T6G 2G7

Yoshiharu Moriwaki Raymond Seed

Woodward-Clyde Consultants University of California
2020 E. First Street, Suite 400 Department of Civil Engineering
Santa Ana, CA 92705 434C Davis Hall
Berkeley, CA 94720
Maurice S. Power
Geomatrix Consultants Kenneth H. Stakoe
100 Pine St., Suite 1000 University of Texas
San Francisco, CA 94111 Civil Engineering Department
Austin, TX 78712

A-2
 Appendix B
Workshop Agenda

WORKSHOP ON EVALUATION OF LIQUEFACTION RESISTANCE

The Inn at Temple Square
Salt Lake City, Utah, January 4 and 5, 1996
Chair: T.L. Youd, Co-Chair: I.M. Idriss

FINAL AGENDA

January 4, 1996

8:00 am Opening Ceremonies East Brunswick Room

Conducting: Les Youd, Brigham Young University
Opening Remarks: GeoffMartin, USC
Keynote Remarks: I.M. Idriss, University of Calif at Davis
Keynote Remarks: Les Youd

8:20 am Special Lecture Liquefaction Lessons Learned from Recent

Earthquakes In Japan - Kenji Ishihara, Science
University of Tokyo

8:50 am Topic 1: Liquefaction Resistance Criteria Based on Cone Penetration

Measurements

Session Moderator: John Christian

Presenter: Peter Robertson, University of Alberta
Discussion: James Mitchell, Virginia Tech
Discussion: Mary Ellen Hynes, Corps of Engineers
Discussion: Yoshi Moriwaki, Woodward-Clyde Consultants
Open Discussion:
Consensus Building: Maury Power, Geomatrix

10:40 am Break

10:50 am Topic 2: Liquefaction Resistance Criteria Based on Standard Penetration (SPT)

Measurements, Including the Influence of Fines Content

Session Moderator: I.M. Idriss

Presenter: Peter Robertson, University of Alberta
Discussion: Ricardo Dobry, Rensselaer Polytechnic Institute
Discussion: John Christian, Private Consultant
Discussion: Gonzalo Castro, GEI, Inc.
Open Discussion:
Consensus Building: Bill Marcuson

8-1
12:30 pm Lunch

1:15 pm Topic 3: Criteria Based on Probabilistic Evaluation of SPT Measurements

Session Moderator: Ricardo Dobry

Presenter: Les Youd, BYU
Discussion: Sam Liao, Parsons Brinkerhoff
Discussion: I.M. Idriss, University of Calif at Davis
Discussion: Bill Marcuson, Corps of Engineers
Open Discussion:

3:15 Break

3:40pm Topic 4: Magnitude Scaling Factors

Session Moderator Ricardo Dobry

Presenter: Les Youd, BYU
Discussion: Ignacio Arango, Bechtel
Discussion: I.M. Idriss, UCD
Discussion: Liam Finn, UBC
Open Discussion:
Consensus Building: John Christian
(Topics 3 and 4)

5:30 pm End Day-1 Discussions

6:30 pm Dinner at a local restaurant

Informal after-dinner group discussions at various localities in hotel

January 5

8:00 am Opening Remarks Day-2: Les Youd

8:05 am Topic 5: Criteria Based on Measured Shear Wave Velocity

Session Moderator -Liam Finn

Presenter: Ron Andrus, NIST
Discussion: Ken Stakoe, Univ. of Texas at Austin
Discussion: Peter Robertson
Discussion: Yoshi Moriwaki
Open Discussion:
Consensus Building: Ricardo Dobry

10:00 am Break
B-2
10:20 am Topic 6: Criteria Based on Becker Hammer Measurements

Session Moderator: Jim Mitchell

Presenter: Les Harder, Calif Division of Water Resources
Discussion: Mary Ellen Hynes
Discussion: Ron Andrus
Discussion: J.M. Idriss
Open Discussion:
Consensus Building: Liam Finn

12:00 pm Lunch

12:45 pm Topic 7: Correction Factors, Ka, K,,, age of deposit, etc.

Session Moderator Yoshi Moriwaki

Presenter: Les Harder
Discussion: Ignacio Arango
Discussion: Ray Seed
Discussion: Maury Power
Open Discussion:
Consensus Building: Jim Mitchell

2:10 pm Break

2:30 pm Topic 8: Estimates of Peak Acceleration and Magnitude

Session Moderator: Maury Power

Presenter: Les Youd
Discussion: Bill Marcuson
Discussion: Liam Finn
Discussion: Ignacio Arango
Open Discussion:
Consensus Building: Yoshi Moriwaki

3:45 pm Closing Remarks:

Summary and Research Needs Identified: Geoff Martin

Summary and Guidance for Final Report: Sam Liao
Summary and Guidance for Final Report: J.M. Idriss
Final Comments and Appreciation: Les Youd

4:30pm Workshop Adjourns

B-3
 Appendix C
Definitions of Terms
Definitions

Although definitions were not a major topic at the workshop, clarification of a few terms is
necessary for communication and for incorporation of new information and methods into the
simplified procedure. Three definitions widely used in the report are listed below. Other
definitions are listed by authors of various sections to clarify their use of terms.

Liquefaction: The term liquefaction as used in this report refers to a change of state from a solid
granular material to a dense viscous-like liquid without consideration of possible deformation or
instability of the liquefied material. Thus, evaluation ofliquefaction resistance as used herein
refers to the determination of the capacity of a soil to resist this change of state, or in other words
triggering of the liquefied condition.

Cyclic Stress Ratio (CSR): As used in the original development of the simplified procedure, the
term cyclic stress ratio (CSR) refers to both the cyclic stress ratio generated by the earthquake
and the cyclic stress ratio required to generate a change of state in the soil to a liquefied condition.
To avoid confusion between these two uses, cyclic stress ratio in this report refers only to the
cyclic stress ratios generated by the ean:hquake.

Cyclic Resistance Ratio (CRR): The stress ratio required to cause a change of state of the soil to
a liquefied condition is referred to throughout this report as the cyclic resistance ratio (CRR).
This change of terminology is recommended for standard use in engineering practice.

C-1
 NATIONAL CENTER FOR EARTHQUAKE ENGINEERING RESEARCH
LIST OF TECHNICAL REPORTS

The National Center for Earthquake Engineering Research (NCEER) publishes technical reports on a variety of subjects related
to earthquake engineering written by authors funded through NCEER. These reports are available from both NCEER
Publications and the National Technical Information Service (NTIS). Requests for reports should be wrected to NCEER
Publications, National Center for Earthquake Engineering Research, State University of New York at Buffalo, Red Jacket
Quadrangle, Buffalo, New York 14261. Reports can also be requested through NTIS, 5285 Port Royal Road, Springfield,
Virginia 22161. NTIS accession numbers are shown in parenthesis, if available.

NCEER-87-0001 "First-Year Program in Research, Education and Technology Transfer," 3/5/87, (PB88-134275, A04, MF-
AOJ).

NCEER-87-0002 "Experimental Evaluation of Instantaneous Optimal Algorithms for Structural Control," by R.C. Lin, T.T.
Soong and A.M. Reinhorn, 4/20/87, (PB88-134341, A04, MF-AO!).

NCEER-87-0003 "Experimentation Using the Earthquake Simulation Facilities at University at Buffalo," by A.M. Reinhorn
and R.L. Ketter, to be published.

NCEER-87-0004 "The System Characteristics and Performance of a Shaking Table," by J.S. Hwang, K.C. Chang and G.C. Lee,
6/1/87, (PB88-134259, A03, MF-AO!). This report is available only through NTIS (see address given
above).

NCEER-87-0005 "A Finite Element Formulation for Nonlinear Viscoplastic Material Using a Q Model," by 0. Gyebi and G.
Dasgupta, 11/2/87, (PB88-213764, A08, MF-AOJ).

NCEER-87-0006 "Symbolic Manipulation Program (SMP) - Algebraic Codes for Two and Three Dimensional Finite Element
Formulations," by X. Lee and G. Dasgupta, 11/9/87, (PB88-218522, A05, MF-AO!).

NCEER-87-0007 "Instantaneous Optimal Control Laws for Tall Buildings Under Seismic Excitations," by J.N. Yang, A.
Akbarpour and P. Ghaemmaghami. 6/10/87, (PB88-134333, A06, MF-AO!). This report is only available
through NTIS (see address given above).

NCEER-87-0008 "IDARC: Inelastic Damage Analysis of Reinforced Concrete Frame - Shear-Wall Structures," by Y.J. Park,
A.M. Reinhorn and S.K. Kunnath, 7/20/87, (PB88-134325, A09, MF-AOl). This report is only available
through NTIS (see address given above).

NCEER-87-0009 "Liquefaction Potential for New York State: A Preliminary Report on Sites in Manhattan and Buffalo," by M.
Budhu, V. Vijayakumar, R.F. Giese and L. Baumgras, 8/31/87, (PB88-!63704, A03, MF-AO!). This report
is available only through NTIS (see address given above).

NCEER-87-0010 "Vertical and Torsional Vibration of Foundations in Inhomogeneous Media," by A.S. Veletsos and K.W.
Dotson, 6/1/87, (PB88-134291, A03, MF-AOl). This report is only available through NTIS (see address
given above).

NCEER-87-0011 "Seismic Probabilistic Risk Assessment and Seismic Margins Studies for Nuclear Power Plants," by Howard
H.M. Hwang, 6/15/87, (PB88-134267, A03, MF-AO!). This report is only available through NTIS (see
address given above).

NCEER-87-0012 "Parametric Studies of Frequency Response of Secondary Systems Under Ground-Acceleration Excitations,"
by Y. Yong and Y.K. Lin, 6/10/87, (PBSS-134309, A03, MF-AOl). This report is only available through
NTIS (see address given above).

NCEER-87-0013 "Frequency Response of Secondary Systems Under Seismic Excitation," by J.A. HoLung, J. Cai and Y.K.
Lin, 7/31/87, (PB88-134317, AOS, MF-AOl). This report is only available through NTIS (see address given
above).

D-1
NCEER-87-0014 "Modelling Earthquake Ground Motions in Seismically Active Regions Using Parametric Time Series
Methods," by G.W. Ellis and A.S. Cakmak, 8/25/87, (PB88-134283, A08, MF-AOl). This report is only
available through NTIS (see address given above).

NCEER-87-0015 "Detection and Assessment of Seismic Structural Damage," by E. DiPasquale and A.S. Cakmak, 8/25/87,
(PB88-163712, A05, MF-AOl). This report is only available through NTIS (see address given above).

NCEER-87-0016 "Pipeline Experiment at Parkfield, California," by J. Isenberg and E. Richardson, 9/15/87, (PB88-163720,
A03, MF-AO!). This report is available only through NTIS (see address given above).

NCEER-87-0017 "Digital Simulation of Seismic Ground Motion," by M. Shinozuka, G. Deodatis and T. Harada, 8/31/87,
(PB88-155197, A04, MF-AOl). This report is available only through NTIS (see address given above).

NCEER-87-0018 "Practical Considerations for Structural Control: System Uncertainty, System Time Delay and Truncation of
Small Control Forces," J.N. Yang and A. Akbarpour, 8/10/87, (PB88-163738, A08, MF-AO!). This report is
only available through NTIS (see address given above).

NCEER-87-0019 "Modal Analysis of Nonclassically Damped Structural Systems Using Canonical Transformation," by J.N.
Yang, S. Sarkani and F.X. Long, 9/27/87, (PB88-187851, A04, MF-AO!).

NCEER-87-0020 "A Nonstationary Solution in Random Vibration Theory," by J.R. Red-Horse and P.D. Spanos, 11/3/87,
(PB88-163746, A03, MF-AO!).

NCEER-87-0021 "Horizontal Impedances for Radially Inhomogeneous Viscoelastic Soil Layers," by A.S. Veletsos and K.W.
Dotson, 10/15/87, (PB88-150859, A04, MF-AO!).

NCEER-87-0022 "Seismic Damage Assessment of Reinforced Concrete Members," by Y.S. Chung, C. Meyer and M.
Shinozuka, 10/9/87, (PB88-150867, A05, MF-AO!). This report is available only through NTIS (see address
given above).

NCEER-8.7-0023 "Active Structural Control in Civil Engineering," by T.T. Soong, 11/11/87, (PB88-187778, A03, MF-AO!).

NCEER-87-0024 "Vertical and Torsional Impedances for Radially Inhomogeneous Viscoelastic Soil Layers," by K.W. Dotson
and A.S. Veletsos, 12/87, (PB88-187786, A03, MF-AO!).

NCEER-87-0025 "Proceedings from the Symposium on Seismic Hazards, Ground Motions, Soil-Liquefaction and Engineering
Practice in Eastern North America," October 20-22, 1987, edited by K.H. Jacob, 12/87, (PB88-188115, A23,
MF-AO!).

NCEER-87-0026 "Report on the Whittier-Narrows, California, Earthquake of October 1, 1987," by J. Pantelic and A.
Reinhorn, 11/87, (PB88-187752, A03, MF-AO!). This report is available only through NTIS (see address
given above).

NCEER-87-0027 "Design of a Modular Program for Transient Nonlinear Analysis of Large 3-D Building Structures," by S.
Srivastav and J.F. Abel, 12/30/87, (PB88-187950, AOS, MF-AO!). This report is only available through
NTIS (see address given above).

NCEER-87-0028 "Second-Year Program in Research, Education and Technology Transfer," 3/8/88, (PB88-219480, A04, MF-
AOl).

NCEER-88-0001 "Workshop on Seismic Computer Analysis and Design of Buildings With Interactive Graphics," by W.
McGuire, J.F. Abel and C.H. Conley, 1/18/88, (PB88-187760, A03, MF-AO!). This report is only available
through NTIS (see address given above).

NCEER-88-0002 "Optimal Control of Nonlinear Flexible Structures," by J.N. Yang, F.X. Long and D. Wong, 1/22/88, (PB88-
213772, A06, MF-AOI).

D-2
NCEER-88-0003 "Substructuring Techniques in the Time Domain for Primary-Secondary Structural Systems," by G.D.
Manolis and G. Juhn, 2/10/88, (PB88-213780, A04, MF-AO!).

NCEER-88-0004 "Iterative Seismic Analysis of Primary-Secondary Systems," by A. Singha!, L.D. Lutes and P.D. Spanos,
2/23/88, (PB88-213798, A04, MF-A0l).

NCEER-88-0005 "Stochastic Finite Element Expansion for Random Media," by P.O. Spanos and R. Ghanem, 3/14/88, (PB88-
213806, A03, MF-A0l).

NCEER-88-0006 "Combining Structural Optimization and Structural Control," by F.Y. Cheng and C.P. Pantelides, 1/10/88,
(PB88-213814, A05, MF-A0l).

NCEER-88-0007 "Seismic Performance Assessment of Code-Designed Structures," by H.H-M. Hwang, J-W. Jaw and H-J.
Shau, 3/20/88, (PB88-219423, A04, MF-AO!). This report is only available through NTIS (see address given
above).

NCEER-88-0008 "Reliability Analysis of Code-Designed Structures Under Natural Hazards," by H.H-M. Hwang, H. Ushiba
and M. Shinozuka, 2/29/88, (PBSS-229471, A07, MF-A0l). This report is only available through NTIS (see
address given above).

NCEER-88-0009 "Seismic Fragility Analysis of Shear Wall Structures," by J-W Jaw and H.H-M. Hwang, 4/30/88, (PB89-
102867, A04, MF-AO!).

NCEER-88-0010 "Base Isolation of a Multi-Story Building Under a Harmonic Ground Motion • A Comparison of
Performances of Various Systems," by F-G Fan, G. Ahmadi and LG. Tadjbakhsh, 5/18/88, (PB89-122238,
A06, MF-AO!). This report is only available through NTIS (see address given above).

NCEER-88-0011 "Seismic Floor Response Spectra for a Combined System by Green's Functions," by F.M. Lavelle, L.A.
Bergman and P.O. Spanos, 5/1/88, (PB89-102875, A03, MF-AOI).

NCEER-88-0012 "A New Solution Technique for Randomly Excited Hysteretic Structures," by G.Q. Cai and Y.K. Lin,
5/16/88, (PB89-102883, A03, MF-AO!).

NCEER-88-0013 "A Study of Radiation Damping and Soil-Structure Interaction Effects in the Centrifuge," by K. Weissman,
supervised by J.H. Prevost, 5/24/88, (PB89-144703, A06, MF-AO!).

NCEER-88-0014 "Parameter Identification and Implementation of a Kinematic Plasticity Model for Frictional Soils," by J.H.
Prevost and D.V. Griffiths, to be published.

NCEER-88-0015 "Two- and Three- Dimensional Dynamic Finite Element Analyses of the Long Valley Dam," by D.V.
Griffiths and J.H. Prevost, 6/17/88, (PB89-144711, A04, MF-AO!).

NCEER-88-0016 "Damage Assessment of Reinforced Concrete Structures in Eastern United States," by A.M. Reinhom, M.J.
Seidel, S.K. Kunnath and Y.J. Park, 6/15/88, (PB89-122220, A04, MF-AO!). This report is only available
through NTIS (see address given above).

NCEER-88-0017 "Dynamic Compliance of Vertically Loaded Strip Foundations in Multilayered Viscoelastic Soils," by S.
Ahmad and A.S.M. Israil, 6/17/88, (PB89-102891, A04, MF-A0I).

NCEER-88-0018 "An Experimental Study of Seismic Structural Response With Added Viscoelastic Dampers," by R.C. Lin, Z.
Liang, T.T. Soong and R.H. Zhang, 6/30/88, (PB89-122212, A05, MF-AO!). This report is available only
through NTIS (see address given above).

NCEER-88-0019 "Experimental Investigation of Primary - Secondary System Interaction," by G.D. Manolis, G. Juhn and A.M.
Reinhom, 5/27/88, (PB89-122204, A04, MF-AO!).

NCEER-88-0020 "A Response Spectrum Approach For Analysis of Nonclassically Damped Structures," by J.N. Yang, S.
Sarkani and F.X. Long, 4/22/88, (PB89-i02909, A04, MF-AO!).

D-3
NCEER-88-0021 "Seismic Interaction of Structures and Soils: Stochastic Approach," by AS. Veletsos and AM. Prasad,
7/21/88, (PB89-122196, A04, MF-A0I). This report is only available through NTIS (see address given
above).

NCEER-88-0022 "Identification of the Serviceability Limit State and Detection of Seismic Structural Damage," by E.
DiPasquale and A.S. Cakmak, 6/15/88, (PB89-122188, A05, MF-AO!). This report is available only through
NTIS (see address given above).

NCEER-88-0023 "Multi-Hazard Risk Analysis: Case of a Simple Offshore Structure," by B.K. Bhartia and E.H. Vanmarcke,
7/21/88, (PB89-145213, A05, MF-AO!).

NCEER-88-0024 "Automated Seismic Design of Reinforced Concrete Buildings," by Y.S. Chung, C. Meyer and M.
Shinozuka, 7/5/88, (PB89-122170, A06, MF-AO!). This report is available only through NTIS (see address
given above).

NCEER-88-0025 "Experimental Study of Active Control of MDOF Structures Under Seismic Excitations," by L.L. Chung,
R.C. Lin, T.T. Soong and A.M. Reinhom, 7/10/88, (PB89-122600, A04, MF-AO!).

NCEER-88-0026 "Earthquake Simulation Tests of a Low-Rise Metal Structure," by J.S. Hwang, K.C. Chang, G.C. Lee and
R.L. Ketter, 8/1/88, (PB89-102917, A04, MF-AO!).

NCEER-88-0027 "Systems Study of Urban Response and Reconstruction Due to Catastrophic Earthquakes," by F. Kozin and
H.K. Zhou, 9/22/88, (PB90-162348, A04, MF-AO!).

NCEER-88-0028 "Seismic Fragility Analysis of Plane Frame Structures," by H.H-M. Hwang and Y.K. Low, 7/31/88, (PB89-
131445, A06, MF-AO!).

NCEER-88-0029 "Response Analysis of Stochastic Structures," by A Kardara, C. Bucher and M. Shinozuka, 9/22/88, (PB89-
174429, A04, MF-AO!).

NCEER-88-0030 "Nonnormal Accelerations Due to Yielding in a Primary Structure," by D.C.K. Chen and L.D. Lutes,
9/19/88, (PB89-131437, A04, MF-AO!).

NCEER-88-0031 "Design Approaches for Soil-Structure Interaction," by A.S. Veletsos, AM. Prasad and Y. Tang, 12/30/88,
(PB89-174437, A03, MF-AO!). This report is available only through NTIS (see address given above).

NCEER-88-0032 "A Re-evaluation of Design Spectra for Seismic Damage Control," by CJ. Turkstra and A.G. Tallin, 1 tn/88,
(PB89-145221, A05, MF-AO!).

NCEER-88-0033 "The Behavior and Design of Noncontact Lap Splices Subjected to Repeated Inelastic Tensile Loading," by
V.E. Sagan, P. Gergely and R.N. White, 12/8/88, (PB89-163737, A08, MF-AO!).

NCEER-88-0034 "Seismic Response of Pile Foundations," by S.M. Mamoon, P.K. Banerjee and S. Ahmad, 11/1/88, (PB89-
145239, A04, MF-AO!).

NCEER-88-0035 "Modeling ofR/C Building Structures With Flexible Floor Diaphragms (IDARC2)," by A.M. Reinhom, S.K.
Kunnath and N. Panahshahi, 9n/88, (PB89-207153, AO?, MF-AO!).

NCEER-88-0036 "Solution of the Dam-Reservoir Interaction Problem Using a Combination of FEM, BEM with Particular
Integrals, Modal Analysis, and Substructuring," by C-S. Tsai, G.C. Lee and R.L. Ketter, 12/31/88, (PB89-
207146, A04, MF-AO!).

NCEER-88-0037 "Optimal Placement of Actuators for Structural Control," by F.Y. Cheng and C.P. Pantelides, 8/15/88,
(PB89-162846, A05, MF-AO!).

D-4
NCEER-88-0038 "Teflon Bearings in Aseismic Base Isolation: Experimental Studies and Mathematical Modeling," by A.
Mokha, M.C. Constantinou and AM. Reinhom, 12/5/88, (PB89-2!8457, AIO, MF-AO!). This report is
available only through NTIS (see address given above).

NCEER-88-0039 "Seismic Behavior of Rat Slab High-Rise Buildings in the New York City Area," by P. Weidlinger and M.
Ettouney, 10/15/88, (PB90-145681, A04, MF-AO!).

NCEER-88-0040 "Evaluation of the Earthquake Resistance of Existing Buildings in New York City," by P. Weidlinger and M.
Ettouney, 10/15/88, to be published.

NCEER-88-0041 "Small-Scale Modeling Techniques for Reinforced Concrete Structures Subjected to Seismic Loads," by W.
Kim, A. El-Attar and R.N. White, I 1/22/88, (PB89-!89625, AOS, MF-AO!).

NCEER-88-0042 "Modeling Strong Ground Motion from Multiple Event Earthquakes," by G.W. Ellis and A.S. Cakmak,
10/15/88, (PB89-174445, A03, MF-AO!).

NCEER-88-0043 "Nonstationary Models of Seismic Ground Acceleration," by M. Grigoriu, S.E. Ruiz and E. Rosenblueth,
7/15/88, (PB89-!89617, A04, MF-AO!).

NCEER-88-0044 "SARCF User's Guide: Seismic Analysis of Reinforced Concrete Frames," by Y.S. Chung, C. Meyer and M.
Shinozuka, 11/9/88, (PB89-174452, A08, MF-AO!).

NCEER-88-0045 "First Expert Panel Meeting on Disaster Research and Planning," edited by J. Pantelic and J. Stoyle, 9/15/88,
(PB89-174460, A05, MF-AO!). This report is only available through NTIS (see address given above).

NCEER-88-0046 "Preliminary Studies of the Effect of Degrading Infill Walls on the Nonlinear Seismic Response of Steel
Frames," by C.Z. Chrysostomou, P. Gergely and J.F. Abel, 12/19/88, (PB89-208383, AOS, MF-AO!).

NCEER-88-0047 "Reinforced Concrete Frame Component Testing Facility - Design, Construction, Instrumentation and
Operation," by S.P. Pessiki, C. Conley, T. Bond, P. Gergely and R.N. White, 12/16/88, (PB89-l 74478, A04,
MF-AO!).

NCEER-89-0001 "Effects of Protective Cushion and Soil Compliancy on the Response of Equipment Within a Seismically
Excited Building," by J.A. HoLung, 2/16/89, (PB89-207179, A04, MF-AO!).

NCEER-89-0002 "Statistical Evaluation of Response Modification Factors for Reinforced Concrete Structures," by H.H-M.
Hwang and J-W. Jaw, 2/17/89, (PB89-207187, ADS, MF-AO!).

NCEER-89-0003 "Hysteretic Columns Under Random Excitation," by G-Q. Cai and Y.K. Lin, 1/9/89, (PB89-196513, A03,
MF-AO!).

NCEER-89-0004 "Experimental Study of'Elephant Foot Bulge' Instability of Thin-Walled Metal Tanks," by Z-H. Jia and R.L.
Ketter, 2/22/89, (PB89-207195, A03, MF-AO!).

NCEER-89-0005 "Experiment on Performance of Buried Pipelines Across San Andreas Fault," by J. Isenberg, E. Richardson
and T.D. O'Rourke, 3/10/89, (PB89-218440, A04, MF-AO!). This report is available only through NTIS (see
address given above).

NCEER-89-0006 "A Knowledge-Based Approach to Structural Design of Earthquake-Resistant Buildings," by M. Subramani,

P. Gergely, C.H. Conley, J.F. Abel and A.H. Zaghw, 1/15/89, (PB89-218465, A06, MF-AO!).

NCEER-89-0007 "Liquefaction Hazards and Their Effects on Buried Pipelines," by T.D. O'Rourke and P.A. Lane, 2/1/89,
(PB89-218481, A09, MF-AO!).

NCEER-89-0008 "Fundamentals of System Identification in Structural Dynamics," by H. Imai, C-B. Yun, 0. Maruyama and
M. Shinozuka, 1/26/89, (PB89-207211, A04, MF-AO!).

D-5
NCEER-89-0009 "Effects of the 1985 Michoacan Earthquake on Water Systems and Other Buried Lifelines in Mexico," by
A.G. Ayala and M.J. O'Rourke, 3/8/89, (PB89-207229, A06, MF-A0l).

NCEER-89-R0lO "NCEER Bibliography of Earthquake Education Materials," by K.E.K. Ross, Second Revision, 9/1/89,
(PB90-125352, A05, MF-A0l). This report is replaced by NCEER-92-0018.

NCEER-89-0011 "Inelastic Three-Dimensional Response Analysis of Reinforced Concrete Building Structures (IDARC-3D),
Part I- Modeling," by S.K. Kunnath and A.M. Reinhom, 4/17/89, (PB90-114612, A07, MF-A0I).

NCEER-89-0012 "Recommended Modifications to ATC-14," by C.D. Poland and J.O. Malley, 4/12/89, (PB90-108648, Al5,
MF-A0l).

NCEER-89-0013 "Repair and Strengthening of Beam-to-Column Connections Subjected to Earthquake Loading," by M.

Corazao and A.J. Durrani, 2/28/89, (PB90-109885, A06, MF-A0l).

NCEER-89-0014 "Program EXKAL2 for Identification of Structural Dynamic Systems," by 0. Maruyama, C-B. Yun, M.
Hoshiya and M. Shinozuka, 5/19/89, (PB90-109877, A09, MF-A0I).

NCEER-89-0015 "Response of Frames With Bolted Semi-Rigid Connections, Part I - Experimental Study and Analytical
Predictions," by P.J. DiCorso, A.M. Reinhorn, J.R. Dickerson, J.B. Radziminski and W.L. Harper, 6/1/89, to
be published.

NCEER-89-0016 "ARMA Monte Carlo Simulation in Probabilistic Structural Analysis," by P.O. Spanos and M.P. Mignolet,
7/10/89, (PB90-109893, A03, MF-A0l).

NCEER-89-P0l 7 "Preliminary Proceedings from the Conference on Disaster Preparedness - The Place of Earthquake
Education in Our Schools," Edited by K.E.K. Ross, 6/23/89, (PB90-108606, A03, MF-AOI).

NCEER-89-0017 "Proceedings from the Conference on Disaster Preparedness - The Place of Earthquake Education in Our
Schools," Edited by K.E.K. Ross, 12/31/89, (PB90-207895, A012, MF-A02). This report is available only
through NTIS (see address given above).

NCEER-89-0018 "Multidimensional Models of Hysteretic Material Behavior for Vibration Analysis of Shape Memory Energy
Absorbing Devices, by E.J. Graesser and F.A. Cozzarelli, 6/7/89, (PB90-164146, A04, MF-A0l).

NCEER-89-0019 "Nonlinear Dynamic Analysis of Three-Dimensional Base Isolated Structures (3D-BASIS)," by S.

Nagarajaiah, A.M. Reinhom and M.C. Constantinou, 8/3/89, (PB90-161936, A06, MF-AOl). This report has
been replaced by NCEER-93-0011.

NCEER-89-0020 "Structural Control Considering Time-Rate of Control Forces and Control Rate Constraints," by F.Y. Cheng
and C.P. Pantelides, 8/3/89, (PB90-120445, A04, MF-A0I).

NCEER-89-0021 "Subsurface Conditions of Memphis and Shelby County," by K.W. Ng, T-S. Chang and H-H.M. Hwang,
7/26/89, (PB90-120437, A03, MF-A0I).

NCEER-89-0022 "Seismic Wave Propagation Effects on Straight Jointed Buried Pipelines," by K. Elhmadi and M.J. O'Rourke,
8/24/89, (PB90-162322, AIO, MF-A02).
NCEER-89-0023 "Workshop on Serviceability Analysis of Water Delivery Systems," edited by M. Grigoriu, 3/6/89, (PB90-
127424, A03, MF-AOl).

NCEER-89-0024 "Shaking Table Study of a 1/5 Scale Steel Frame Composed of Tapered Members," by K.C. Chang, J.S.
Hwang and G.C. Lee, 9/18/89, (PB90-160169, A04, MF-AOl).

NCEER-89-0025 "DYNAID: A Computer Program for Nonlinear Seismic Site Response Analysis - Technical
Documentation," by Jean H. Prevost, 9/14/89, (PB90-161944, A07, MF-A0I). This report is available only
through NTIS (see address given above).

D-6
NCEER-89-0026 "1:4 Scale Model Studies of Active Tendon Systems and Active Mass Dampers for Aseismic Protection," by
AM. Reinhorn, T.T. Soong, R.C. Lin, Y.P. Yang, Y. Fukao, H. Abe and M. Nakai, 9/15/89, (PB90-173246,
AIO, MF-A02).

NCEER-89-0027 "Scattering of Waves by Inclusions in a Nonhomogeneous Elastic Half Space Solved by Boundary Element
Methods," by P.K. Hadley, A Askar and AS. Cakmak, 6/15/89, (PB90-145699, A07, MF-A0l).

NCEER-89-0028 "Statistical Evaluation of Deflection Amplification Factors for Reinforced Concrete Structures," by H.H.M.
Hwang, J-W. Jaw and AL. Ch'ng, 8/31/89, (PB90-J64633, A05, MF-A0I).

NCEER-89-0029 "Bedrock Accelerations in Memphis Area Due to Large New Madrid Earthquakes," by H.H.M. Hwang,
C.H.S. Chen and G. Yu, 11/7/89, (PB90-162330, A04, MF-AO!).

NCEER-89-0030 "Seismic Behavior and Response Sensitivity of Secondary Structural Systems," by Y.Q. Chen and T.T.
Soong, 10/23/89, (PB90-164658, A08, MF-AO!).

NCEER-89-0031 "Random Vibration and Reliability Analysis of Primary-Secondary Structural Systems," by Y. Ibrahim, M.
Grigoriu and T.T. Soong, I 1/10/89, (PB90-161951, A04, MF-A0l).

NCEER-89-0032 "Proceedings from the Second U.S. - Japan Workshop on Liquefaction, Large Ground Deformation and Their
Effects on Lifelines, September 26-29, 1989," Edited by T.D. O'Rourke and M. Hamada, 12/1/89, (PB90-
209388, A22, MF-A03).

NCEER-89-0033 "Deterministic Model for Seismic Damage Evaluation of Reinforced Concrete Structures," by J.M. Bracci,
A.M. Reinhom, J.B. Mander and S.K. Kunnath, 9/27/89, (PB91-108803, A06, MF-AO!).

NCEER-89-0034 "On the Relation Between Local and Global Damage Indices," by E. DiPasquale and AS. Cakmak, 8/15/89,
(PB90-173865, A05, MF-AO!).

NCEER-89-0035 "Cyclic Undrained Behavior of Nonplastic and Low Plasticity Silts," by A.J. Walker and H.E. Stewart,
7/26/89, (PB90-183518, AlO, MF-AO!).

NCEER-89-0036 "Liquefaction Potential of Surficial Deposits in the City of Buffalo, New York," by M. Budhu, R. Giese and
L. Baumgrass, 1/17/89, (PB90-208455, A04, MF-A0I).

NCEER-89-0037 "A Deterministic Assessment of Effects of Ground Motion Incoherence," by AS. Veletsos and Y. Tang,
7/15/89, (PB90-164294, A03, MF-AO!).

NCEER-89-0038 "Workshop on Ground Motion Parameters for Seismic Hazard Mapping," July 17-18, 1989, edited by R.V.
Whitman, 12/1/89, (PB90-173923, A04, MF-AOl).

NCEER-89-0039 "Seismic Effects on Elevated Transit Lines of the New York City Transit Authority," by C.J. Costantino,
C.A. Miller and E. Heymsfield, 12/26/89, (PB90-207887, A06, MF-A0l).

NCEER-89-0040 "Centrifugal Modeling of Dynamic Soil-Structure Interaction," by K. Weissman, Supervised by J.H. Prevost,
5/10/89, (PB90-207879, A07, MF-A0l).
NCEER-89-0041 "Linearized Identification of Buildings With Cores for Seismic Vulnerability Assessment," by I-K. Ho and
A.E. Aktan, 11/1/89, (PB90-251943, A07, MF-AO!).

NCEER-90-0001 "Geotechnical and Lifeline Aspects of the October 17, 1989 Loma Prieta Earthquake in San Francisco," by
T.D. O'Rourke, H.E. Stewart, F.T. Blackburn and T.S. Dickerman, 1/90, (PB90-208596, A05, MF-AO!).

NCEER-90-0002 "Nonnormal Secondary Response Due to Yielding in a Primary Structure," by D.C.K. Chen and L.D. Lutes,
2/28/90, (PB90-251976, A07, MF-AO!).

D-7
NCEER-90-0003 "Earthquake Education Materials for Grades K-12," by K.E.K. Ross, 4/16/90, (PB91-251984, A0S, MF-
A05). This report has been replaced by NCEER-92-0018.

NCEER-90-0004 "Catalog of Strong Motion Stations in Eastern North America," by R.W. Busby, 4/3/90, (PB90-251984, A0S,
MF-AO!).

NCEER-90-0005 "NCEER Strong-Motion Data Base: A User Manual for the GeoBase Release (Version 1.0 for the Sun3)," by
P. Friberg and K. Jacob, 3/31/90 (PB90-258062, A04, MF-A0l).

NCEER-90-0006 "Seismic Hazard Along a Crude Oil Pipeline in the Event of an 1811-1812 Type New Madrid Earthquake,"
by H.H.M. Hwang and C-H.S. Chen, 4/16/90, (PB90-258054, A04, MF-A0l).

NCEER-90-0007 "Site-Specific Response Spectra for Memphis Sheahan Pumping Station," by H.H.M. Hwang and C.S. Lee,
5/15/90, (PB91-108811, A0S, MF-AO!).

NCEER-90-0008 "Pilot Study on Seismic Vulnerability of Crude Oil Transmission Systems," by T. Ariman, R. Dobry, M.
Grigoriu, F. Kozin, M. O'Rourke, T. O'Rourke and M. Shinozuka, 5/25/90, (PB91-108837, A06, MF-AO!).

NCEER-90-0009 "A Program to Generate Site Dependent Time Histories: EQGEN," by G.W. Ellis, M. Srinivasan and A.S.
Cakmak, 1/30/90, (PB91-108829, A04, MF-AO!).

NCEER-90-0010 "Active Isolation for Seismic Protection of Operating Rooms," by M.E. Talbott, Supervised by M.
Shinozuka, 6/8/9, (PB91-l 10205, A0S, MF-AO!).

NCEER-90-0011 "Program LINEARID for Identification of Linear Structural Dynamic Systems," by C-B. Yun and M.
Shinozuka, 6/25/90, (PB91-110312, A08, MF-AO!).

NCEER-90-0012 "Two-Dimensional Two-Phase Elasto-Plastic Seismic Response of Earth Dams," by A.N. Yiagos, Supervised
by J.H. Prevost, 6/20/90, (PB91-110197, Al3, MF-A02).

NCEER-90-0013 "Secondary Systems in Base-Isolated Structures: Experimental Investigation, Stochastic Response and
Stochastic Sensitivity," by G.D. Manolis, G. Juhn, M.C. Constantinou and A.M. Reinhorn, 7/1/90, (PB91-
110320, A08, MF-AO!).

NCEER-90-0014 "Seismic Behavior of Lightly-Reinforced Concrete Column and Beam-Column Joint Details," by S.P.
Pessiki, C.H. Conley, P. Gergely and R.N. White, 8/22/90, (PB91-108795, Al 1, MF-A02).

NCEER-90-0015 "Two Hybrid Control Systems for Building Structures Under Strong Earthquakes," by J.N. Yang and A.
Danielians, 6/29/90, (PB91-125393, A04, MF-A0l).

NCEER-90-0016 "Instantaneous Optimal Control with Acceleration and Velocity Feedback," by J.N. Yang and Z. Li, 6/29/90,
(PB91-125401, A03, MF-A0l).

NCEER-90-0017 "Reconnaissance Report on the Northern Iran Earthquake of June 21, 1990," by M. Mehrain, 10/4/90,
(PB91-125377, A03, MF-A0l).

NCEER-90-0018 "Evaluation of Liquefaction Potential in Memphis and Shelby County," by T.S. Chang, P.S. Tang, C.S. Lee
and H. Hwang, 8/10/90, (PB91-125427, A09, MF-AO!).

NCEER-90-0019 "Experimental and Analytical Study of a Combined Sliding Disc Bearing and Helical Steel Spring Isolation
System," by M.C. Constantinou, A.S. Mokha and A.M. Reinhorn, 10/4/90, (PB91-125385, A06, MF-A0l).
This report is available only through NTIS (see address given above).

NCEER-90-0020 "Experimental Study and Analytical Prediction of Earthquake Response of a Sliding Isolation System with a
Spherical Surface," by A.S. Mokha, M.C. Constantinou and A.M. Reinhorn, 10/11/90, (PB91-125419, A05,
MF-A0l).

D-8
NCEER-90-0021 "Dynamic Interaction Factors for Floating Pile Groups," by G. Gazetas, K. Fan, A. Kaynia and E. Kausel,
9/10/90, (PB9!-170381, A05, MF-AO!).

NCEER-90-0022 "Evaluation of Seismic Damage Indices for Reinforced Concrete Structures," by S. Rodriguez-Gomez and
A.S. Cakmak, 9/30/90, PB91-171322, A06, MF-AO!).

NCEER-90-0023 "Study of Site Response at a Selected Memphis Site," by H. Desai, S. Ahmad, E.S. Gazetas and M.R. Oh,
10/11/90, (PB91-196857, A03, MF-AO!).

NCEER-90-0024 "A User's Guide to Strongmo: Version 1.0 of NCEER's Strong-Motion Data Access Tool for PCs and
Terminals," by P.A. Friberg and C.A.T. Susch, 11/15/90, (PB91-171272, A03, MF-AO!).

NCEER-90-0025 "A Three-Dimensional Analytical Study of Spatial Variability of Seismic Ground Motions," by L-L. Hong
and A.H.-S. Ang, 10/30/90, (PB91-! 70399, A09, MF-AO!).

NCEER-90-0026 "MUMOID User's Guide - A Program for the Identification of Modal Parameters," by S. Rodriguez-Gomez
and E. DiPasquale, 9/30/90, (PB91-!71298, A04, MF-AO!).

NCEER-90-0027 "SARCF-11 User's Guide - Seismic Analysis of Reinforced Concrete Frames," by S. Rodriguez-Gomez, Y.S.
Chung and C. Meyer, 9/30/90, (PB91-171280, A05, MF-AO!). .

NCEER-90-0028 "Viscous Dampers: Testing, Modeling and Application in Vibration and Seismic Isolation," by N. Makris
and M.C. Constantinou, 12/20/90 (PB91-!90561, A06, MF-AO!).

NCEER-90-0029 "Soil Effects on Earthquake Ground Motions in the Memphis Area," by H. Hwang, C.S. Lee, K.W. Ng and
T.S. Chang, 8/2/90, (PB9!-190751, A05, MF-AO!).

NCEER-91-0001 "Proceedings from the Third Japan-U.S. Workshop on Earthquake Resistant Design of Lifeline Facilities and
Countermeasures for Soil Liquefaction, December 17-19, 1990," edited by T.D. O'Rourke and M. Hamada,
2/1/91, (PB91-l 79259, A99, MF-A04).

NCEER-91-0002 "Physical Space Solutions of Non-Proportionally Damped Systems," by M. Tong, Z. Liang and G.C. Lee,
1/15/91, (PB9!-179242, A04, MF-AO!).

NCEER-91-0003 "Seismic Response of Single Piles and Pile Groups," by K. Fan and G. Gazetas, 1/10/91, (PB92-174994,
A04, MF-AO!).

NCEER-91-0004 "Damping of Structures: Part I - Theory of Complex Damping," by Z. Liang and G. Lee, 10/10/91, (PB92-
197235, A12, MF-A03).

NCEER-91-0005 "3D-BASIS - Nonlinear Dynamic Analysis of Three Dimensional Base Isolated Structures: Part II," by S.
Nagarajaiah, A.M. Reinhom and M.C. Constantinou, 2/28/91, (PB91-!90553, A07, MF-AO!). This report
has been replaced by NCEER-93-001 I.

NCEER-91-0006 "A Multidimensional Hysteretic Model for Plasticity Deforming Metals in Energy Absorbing Devices," by
E.J. Graesser and F.A. Cozzarelli, 4/9/91, (PB92-!08364, A04, MF-AO!).

NCEER-91-0007 "A Framework for Customizable Knowledge-Based Expert Systems with an Application to a KBES for
Evaluating the Seismic Resistance of Existing Buildings," by E.G. Ibarra-Anaya and S.J. Fenves, 4/9/91,
(PB91-210930, A08, MF-AO!).

NCEER-91-0008 "Nonlinear Analysis of Steel Frames with Semi-Rigid Connections Using the Capacity Spectrum Method,"
by G.G. Deierlein, S-H. Hsieh, Y-J. Shen and J.F. Abel, 7/2/91, (PB92-!13828, A05, MF-AO!).

NCEER-91-0009 "Earthquake Education Materials for Grades K-12," by K.E.K. Ross, 4/30/91, (PB91-212l42, A06, MF-
AOI). This report has been replaced by NCEER-92-0018.

D-9
NCEER-91-0010 "Phase Wave Velocities and Displacement Phase Differences in a Harmonically Oscillating Pile," by N.
Makris and G. Gazetas, 7/8/91, (PB92-108356, A04, MF-AOl).

NCEER-91-0011 "Dynamic Characteristics of a Full-Size Five-Story Steel Structure and a 2/5 Scale Model," by K.C. Chang,
G.C. Yao, G.C. Lee, D.S. Hao and Y.C. Yeh," 7/2/91, (PB93-116648, A06, MF-A02).

NCEER-91-0012 "Seismic Response of a 2/5 Scale Steel Structure with Added Viscoelastic Dampers," by K.C. Chang, T.T.
Soong, S-T. Oh and M.L. Lai, 5/17/91, (PB92-l 10816, A05, MF-A0l).

NCEER-91-0013 "Earthquake Response of Retaining Walls; Full-Scale Testing and Computational ·Modeling," by S.
Alarnpalli and A-W.M. Elgamal, 6/20/91, to be published.

NCEER-91-0014 "3D-BASIS-M: Nonlinear Dynamic Analysis of Multiple Building Base Isolated Structures," by P.C.
Tsopelas, S. Nagarajaiah, M.C. Constantinou and A.M. Reinhorn, 5/28/91, (PB92-113885, A09, MF-A02).

NCEER-91-0015 "Evaluation of SEAOC Design Requirements for Sliding Isolated Structures," by D. Theodossiou and M.C.
Constantinou, 6/10/91, (PB92-114602, All, MF-A03).

NCEER-91-0016 "Closed-Loop Modal Testing of a 27-Story Reinforced Concrete Flat Plate-Core Building,., by H.R.
Somaprasad, T. Toksoy, H. Yoshiyuki and A.E. Aktan, 7/15/91, (PB92-129980, A07, MF-A02).

NCEER-91-0017 "Shake Table Test of a 1/6 Scale Two-Story Lightly Reinforced Concrete Building," by A.G. EI-Attar, R.N.
White and P. Gergely, 2/28/91, (PB92-222447, A06, MF-A02).

NCEER-91-0018 "Shake Table Test of a 1/8 Scale Three-Story Lightly Reinforced Concrete Building," by A.G. El-Attar, R.N.
White and P. Gergely, 2/28/91, (PB93-l 16630, A08, MF-A02).

NCEER-91-0019 "Transfer Functions for Rigid Rectangular Foundations," by A.S. Veletsos, A.M. Prasad and W.H. Wu,
7/31/91, to be published.

NCEER-91-0020 "Hybrid Control of Seismic-Excited Nonlinear and Inelastic Structural Systems," by J.N. Yang, Z. Li and A.
Danielians, 8/1/91, (PB92-143171, A06, MF-A02).

NCEER-91-0021 "The NCEER-91 Earthquake Catalog: Improved Intensity-Based Magnitudes and Recurrence Relations for
U.S. Earthquakes East of New Madrid," by L. Seeber and J.G. Armbruster, 8/28/91, (PB92-176742, A06,
MF-A02).

NCEER-91-0022 "Proceedings from the Implementation of Earthquake Planning and Education in Schools: The Need for
Change - The Roles of the Changemakers," by K.E.K. Ross and F. Winslow, 7/23/91, (PB92-129998, A12,
MF-A03).

NCEER-91-0023 "A Study .of Reliability-Based Criteria for Seismic Design of Reinforced Concrete Frame Buildings," by
H.H.M. Hwang and H-M. Hsu, 8/10/91, (PB92-140235, A09, MF-A02).
NCEER-91-0024 "Experimental Verification of a Number of Structural System Identification Algorithms," by R.G. Ghanem,
H. Gavin and M. Shinozuka, 9/18/91, (PB92-176577, A18, MF-A04).

NCEER-91-0025 "Probabilistic Evaluation of Liquefaction Potential," by H.H.M. Hwang and C.S. Lee," 11/25/91, (PB92-
143429, A05, MF-AOl).

NCEER-91-0026 "Instantaneous Optimal Control for Linear, Nonlinear and Hysteretic Structures - Stable Controllers," by J.N.
Yang and Z. Li, 11/15/91, (PB92-163807, A04, MF-A0l).

NCEER-91-0027 "Experimental and Theoretical Study of a Sliding Isolation System for Bridges," by M.C. Constantinou, A.
Kartoum, A.M. Reinhom and P. Bradford, 11/15/91, (PB92-176973, AlO, MF-A03).

NCEER-92-0001 "Case Studies of Liquefaction and Lifeline Performance During Past Earthquakes, Volume l: Japanese Case
Studies," Edited by M. Hamada and T. O'Rourke, 2/17/92, (PB92-197243, Al 8, MF-A04).

D-10
NCEER-92-0002 "Case Studies of Liquefaction and Lifeline Performance During Past Earthquakes, Volume 2: United States
Case Studies," Edited by T. O'Rourke and M. Hamada, 2/17/92, (PB92-197250, A20, MF-A04).

NCEER-92-0003 "Issues in Earthquake Education," Edited by K. Ross, 2/3/92, (PB92-222389, A07, MF-A02).

NCEER-92-0004 "Proceedings from the First U.S. - Japan Workshop on Earthquake Protective Systems for Bridges," Edited
by LG. Buckle, 2/4/92, (PB94-142239, A99, MF-A06).

NCEER-92-0005 "Seismic Ground Motion from a Haskell-Type Source in a Multiple-Layered Half-Space," A.P. Theoharis, G.
Deodatis and M. Shinozuka, 1/2/92, to be published.

NCEER-92-0006 "Proceedings from the Site Effects Workshop," Edited by R. Whitman, 2/29/92, (PB92-197201, A04, MF-
A0l).

NCEER-92-0007 "Engineering Evaluation of Permanent Ground Deformations Due to Seismically-Induced Liquefaction," by

M.H. Baziar, R. Dobry and A-W.M. Elgamal, 3/24/92, (PB92-222421, A13, MF-A03).

NCEER-92-0008 "A Procedure for the Seismic Evaluation of Buildings in the Central and Eastern United States," by C.D.
Poland and J.O. Malley, 4/2/92, (PB92-222439, A20, MF-A04).

NCEER-92-0009 "Experimental and Analytical Study of a Hybrid Isolation System Using Friction Controllable Sliding
Bearings," by M.Q. Feng, S. Fujii and M. Shinozuka, 5/15/92, (PB93-150282, A06, MF-A02).

NCEER-92-0010 "Seismic Resistance of Slab-Column Connections in Existing Non-Ductile Flat-Plate Buildings," by A.J.
Durrani and Y. Du, 5/18/92, (PB93-I 16812, A06, MF-A02).

NCEER-92-001 I "The Hysteretic and Dynamic Behavior of Brick Masonry Walls Upgraded by Ferrocement Coatings Under
Cyclic Loading and Strong Simulated Ground Motion," by H. Lee and S.P. Prawel, 5/11/92, to be published.

NCEER-92-0012 "Study of Wire Rope Systems for Seismic Protection of Equipment in Buildings," by G.F. Demetriades, M.C.
Constantinou and A.M. Reinhom, 5/20/92, (PB93-! ! 6655, A08, MF-A02).

NCEER-92-0013 "Shape Memory Structural Dampers: Material Properties, Design and Seismic Testing," by P.R. Witting and
F.A. Cozzarelli, 5/26/92, (PB93-1 I 6663, A05, MF-AO!).

NCEER-92-0014 "Longitudinal Permanent Ground Deformation Effects on Buried Continuous Pipelines," by M.J. O'Rourke,
and C. Nordberg, 6/15/92, (PB93-116671, A08, MF-A02).

NCEER-92-0015 "A Simulation Method for Stationary Gaussian Random Functions Based on the Sampling Theorem," by M.
Grigoriu and S. Balopoulou, 6/11/92, (PB93-127496, A05, MF-A0I).
NCEER-92-0016 "Gravity-Load-Designed Reinforced Concrete Buildings: Seismic Evaluation of Existing Construction and
Detailing Strategies for Improved Seismic Resistance," by G.W. Hoffmann, S.K. Kunnath, A.M. Reinhom
and J.B. Mander, 7/15/92, (PB94-142007, A08, MF-A02).

NCEER-92-0017 "Observations on Water System and Pipeline Performance in the Lim6n Area of Costa Rica Due to the April
22, 1991 Earthquake," by M. O'Rourke and D. Ballantyne, 6/30/92, (PB93-12681 l, A06, MF-A02).

NCEER-92-0018 "Fourth Edition of Earthquake Education Materials for Grades K-12," Edited by K.E.K. Ross, 8/10/92,
(PB93-114023, A07, MF-A02).

NCEER-92-0019 "Proceedings from the Fourth Japan-U.S. Workshop on Earthquake Resistant Design of Lifeline Facilities
and Countermeasures for Soil Liquefaction," Edited by M. Hamada and T.D. O'Rourke, 8/12/92, (PB93-
163939, A99, MF-El I).

NCEER-92-0020 "Active Bracing System: A Full Scale Implementation of Active Control," by A.M. Reinhom, T.T. Soong,
R.C. Lin, M.A. Riley, Y.P. Wang, S. Aizawa and M. Higashino, 8/14/92, (PB93-127512, A06, MF-A02).

D-11
NCEER-92-0021 "Empirical Analysis of Horizontal Ground Displacement Generated by Liquefaction-Induced Lateral
Spreads," by S.F. Bartlett and T.L. Youd, 8/17/92, (PB93-188241, A06, MF-A02).

NCEER-92-0022 "IDARC Version 3.0: Inelastic Damage Analysis of Reinforced Concrete Structures," by S.K. Kunnath, A.M.
Reinhom and R.F. Lobo, 8/31/92, (PB93-227502, A07, MF-A02).

NCEER-92-0023 "A Semi-Empirical Analysis of Strong-Motion Peaks in Terms of Seismic Source, Propagation Path and
Local Site Conditions, by M. Kamiyama, M.J. O'Rourke and R. Flores-Berrones, 9/9/92, (PB93-150266,
AOS, MF-A02).

NCEER-92-0024 "Seismic Behavior of Reinforced Concrete Frame Structures with Nonductile Details, Part I: Summary of
Experimental Findings of Full Scale Beam-Column Joint Tests," by A. Beres, R.N. White and P. Gergely,
9/30/92, (PB93-227783, A05, MF-A0l).

NCEER-92-0025 "Experimental Results of Repaired and Retrofitted Beam-Column Joint Tests in Lightly Reinforced Concrete
Frame Buildings," by A. Beres, S. El-Borgi, R.N. White and P. Gergely, 10/29/92, (PB93-227791, A05, MF-
A0I).

NCEER-92-0026 "A Generalization of Optimal Control Theory: Linear and Nonlinear Structures," by J.N. Yang, Z. Li and S.
Vongchavalitkul, 11/2/92, (PB93-188621, A0S, MF-AO!).

NCEER-92-0027 "Seismic Resistance of Reinforced Concrete Frame Structures Designed Only for Gravity Loads: Part I -
Design and Properties of a One-Third Scale Model Structure," by J.M. Bracci, A.M. Reinhom and J.B.
Mander, 12/1/92, (PB94-104502, A08, MF-A02).

NCEER-92-0028 "Seismic Resistance of Reinforced Concrete Frame Structures Designed Only for Gravity Loads: Part II -
Experimental Performance of Subassemblages," by LE. Aycardi, J.B. Mander and A.M. Reinhom, 12/1/92,
(PB94-104510, A08, MF-A02).

NCEER-92-0029 "Seismic Resistance of Reinforced Concrete Frame Structures Designed Only for Gravity Loads: Part ID -
Experimental Performance and Analytical Study of a Structural Model," by J.M. Bracci, A.M. Reinhom and
J.B. Mander, 12/1/92, (PB93-227528, A09, MF-AO!).

NCEER-92-0030 "Evaluation of Seismic Retrofit of Reinforced Concrete Frame Structures: Part I - Experimental Performance
of Retrofitted Subassemblages," by D. Choudhuri, J.B. Mander and A.M. Reinhom, 12/8/92, (PB93-198307,
A07, MF-A02).

NCEER-92-0031 "Evaluation of Seismic Retrofit of Reinforced Concrete Frame Structures: Part II - Experimental Performance
and Analytical Study of a Retrofitted Structural Model," by J.M. Bracci, A.M. Reinhom and J.B. Mander,
12/8/92, (PB93-198315, A09, MF-A03).

NCEER-92-0032 "Experimental and Analytical Investigation of Seismic Response of Structures with Supplemental Fluid
Viscous Dampers," by M.C. Constantinou and M.D. Symans, 12/21/92, (PB93-191435, Al 0, MF-A03).

NCEER-92-0033 "Reconnaissance Report on the Cairo, Egypt Earthquake of October 12, 1992," by M. Khater, 12/23/92,
(PB93-188621, A03, MF-AO!).

NCEER-92-0034 "Low-Level Dynamic Characteristics of Four Tall Flat-Plate Buildings in New York City," by H. Gavin, S.
Yuan, J. Grossman, E. Pekelis and K. Jacob, 12/28/92, (PB93-188217, A07, MF-A02).

NCEER-93-0001 "An Experimental Study on the Seismic Performance of Brick-Infilled Steel Frames With and Without
Retrofit," by J.B. Mander, B. Nair, K. Wojtkowski and J. Ma, 1/29/93, (PB93-227510, A07, MF-A02).

NCEER-93-0002 "Social Accounting for Disaster Preparedness and Recovery Planning," by S. Cole, E. Pantoja and V. Razak,
2/22/93, (PB94-142114, A12, MF-A03).

D-12
NCEER-93-0003 "Assessment of 1991 NEHRP Provisions for Nonstructural Components and Recommended Revisions," by
T.T. Soong, G. Chen, Z. Wu, R-H. Zhang and M. Grigoriu, 3/1/93, (PB93-188639, A06, MF-A02).

NCEER-93-0004 "Evaluation of Static and Response Spectrum Analysis Procedures of SEAOC/UBC for Seismic Isolated
Structures," byC.W. Winters and M.C. Constantinou, 3/23/93, (PB93-198299, AIO, MF-A03).

NCEER-93-0005 "Earthquakes in the Northeast - Are We Ignoring the Hazard? A Workshop on Earthquake Science and
Safety for Educators," edited by K.E.K. Ross, 4/2/93, (PB94-I03066, A09, MF-A02).

NCEER-93-0006 "Inelastic Response of Reinforced Concrete Structures with Viscoelastic Braces," by R.F. Lobo, J.M. Bracci,
K.L. Shen, A.M. Reinhorn and T.T. Soong, 4/5/93, (PB93-227486, A05, MF-A02).

NCEER-93-0007 "Seismic Testing of Installation Methods for Computers and Data Processing Equipment," by K. Kosar, T.T.
Soong, K.L. Shen, J.A. HoLung and Y.K. Lin, 4/12/93, (PB93-198299, AO?, MF-A02).

NCEER-93-0008 "Retrofit of Reinforced Concrete Frames Using Added Dampers," by A. Reinhom, M. Constantinou and C.
Li, to be published.

NCEER-93-0009 "Seismic Behavior and Design Guidelines for Steel Frame Structures with Added Viscoelastic Dampers," by
K.C. Chang, M.L. Lai, T.T. Soong, D.S. Hao and Y.C. Yeh, 5/1/93, (PB94-141959, A07, MF-A02).

NCEER-93-0010 "Seismic Performance of Shear-Critical Reinforced Concrete Bridge Piers," by J.B. Mander, S.M. Waheed,
M.T.A. Chaudhary and S.S. Chen, 5/12/93, (PB93-227494, A08, MF-A02).

NCEER-93-0011 "3D-BASIS-TABS: Computer Program for Nonlinear Dynamic Analysis of Three Dimensional Base Isolated
Structures," by S. Nagarajaiah, C. Li, A.M. Reinhorn and M.C. Constantinou, 8/2/93, (PB94-141819, A09,
MF-A02).

NCEER-93-0012 "Effects of Hydrocarbon Spills from an Oil Pipeline Break on Ground Water," by O.J. Helweg and H.H.M.
Hwang, 8/3/93, (PB94-141942, A06, MF-A02).

NCEER-93-0013 "Simplified Procedures for Seismic Design of Nonstructural Components and Assessment of Current Code
Provisions," by M.P. Singh, L.E. Suarez, E.E. Matheu and G.O. Maldonado, 8/4/93, (PB94-141827, A09,
MF-A02).

NCEER-93-0014 "An Energy Approach to Seismic Analysis and Design of Secondary Systems," by G. Chen and T.T. Soong,
8/6/93, (PB94-142767, All, MF-A03).

NCEER-93-0015 "Proceedings from School Sites: Becoming Prepared for Earthquakes - Commemorating the Third
Anniversary of the Loma Prieta Earthquake," Edited by F.E. Winslow and K.E.K. Ross, 8/16/93, (PB94-
154275, Al 6, MF-A02).

NCEER-93-0016 "Reconnaissance Report of Damage to Historic Monuments in Cairo, Egypt Following the October 12, 1992
Dahshur Earthquake," by D. Sykora, D. Look, G. Croci, E. Karaesmen and E. Karaesmen, 8/19/93, (PB94-
142221, A08, MF-A02).

NCEER-93-0017 "The Island of Guam Earthquake of August 8, 1993," by S.W. Swan and S.K. Harris, 9/30/93, (PB94-
14!843, A04, MF-AO!).

NCEER-93-0018 "Engineering Aspects of the October 12, 1992 Egyptian Earthquake," by A.W. Elgamal, M. Amer, K. Adalier
and A. Abul-Fadl, 10/7193, (PB94-141983, AOS, MF-AO!).

NCEER-93-0019 "Development ofan Earthquake Motion Simulator and its Application in Dynamic Centrifuge Testing," by I.
Krstelj, Supervised by J.H. Prevost, 10/23/93, (PB94-181773, A-IO, MF-A03).

NCEER-93-0020 "NCEER-Taisei Corporation Research Program on Sliding Seismic Isolation Systems for Bridges:
Experimental and Analytical Study of a Friction Pendulum System (FPS)," by M.C. Constantinou, P.
Tsopelas, Y-S. Kim and S. Okamoto, 11/1/93, (PB94-142775, A08, MF-A02).

D-13
NCEER-93-0021 "Finite Element Modeling of Elastomeric Seismic Isolation Bearings," by L.J. Billings, Supervised by R.
Shepherd, 11/8/93, to be published.

NCEER-93-0022 "Seismic Vulnerability of Equipment in Critical Facilities: Life-Safety and Operational Consequences," by K.
Porter, G.S. Johnson, M.M. Zadeh, C. Scawthom and S. Eder, 11/24/93, (PB94-181765, A16, MF-A03).

NCEER-93-0023 "Hokkaido Nansei-oki, Japan Earthquake of July 12, 1993, by P.I. Yanev and C.R. Scawthorn, 12/23/93,
(PB94-181500, A07, MF-AOI).

NCEER-94-0001 "An Evaluation of Seismic Serviceability of Water Supply Networks with Application to the San Francisco
Auxiliary Water Supply System," by I. Markov, Supervised by M. Grigoriu and T. O'Rourke, 1/21/94,
(PB94-204013, A07, MF-A02).

NCEER-94-0002 "NCEER-Taisei Corporation Research Program on Sliding Seismic Isolation Systems for Bridges:
Experimental and Analytical Study of Systems Consisting of Sliding Bearings, Rubber Restoring Force
Devices and Fluid Dampers," Volumes I and II, by P. Tsopelas, S. Okamoto, M.C. Constantinou, D. Ozaki
and S. Fujii, 2/4/94, (PB94-181740, A09, MF-A02 and PB94-18! 757, Al 2, MF-A03).

NCEER-94-0003 "A Markov Model for Local and Global Damage Indices in Seismic Analysis," by S. Rahman and M.
Grigoriu, 2/18/94, (PB94-206000, Al 2, MF-A03).

NCEER-94-0004 "Proceedings from the NCEER Workshop on Seismic Response of Masonry lnfills," edited by D.P. Abrams,
3/1/94, (PB94-180783, AO?, MF-A02).

NCEER-94-0005 "The Northridge, California Earthquake of January 17, 1994: General Reconnaissance Report," edited by J.D.
Goltz, 3/11/94, (PB193943, AlO, MF-A03).

NCEER-94-0006 "Seismic Energy Based Fatigue Damage Analysis of Bridge Columns: Part I - Evaluation of Seismic
Capacity," by G.A. Chang and J.B. Mander, 3/14/94, (PB94-219185, Al 1, MF-A03).

NCEER-94-0007 "Seismic Isolation of Multi-Story Frame Structures Using Spherical Sliding Isolation Systems," by T.M. Al-
Hussaini, V.A. Zayas and M.C. Constantinou, 3/17/94, (PB193745, A09, MF-A02).

NCEER-94-0008 "The Northridge, California Earthquake of January 17, 1994: Performance of Highway Bridges," edited by
I.G. Buckle, 3/24/94, (PB94-193851, A06, MF-A02).

NCEER-94-0009 "Proceedings of the Third U.S.-Japan Workshop on Earthquake Protective Systems for Bridges," edited by
LG. Buckle and I. Friedland, 3/31/94, (PB94-195815, A99, MF-A06).

NCEER-94-0010 "3D-BASIS-ME: Computer Program for Nonlinear Dynamic Analysis of Seismically Isolated Single and
Multiple Structures and Liquid Storage Tanks," by P.C. Tsopelas, M.C. Constantinou and A.M. Reinhorn,
4/12/94, (PB94-204922, A09, MF-A02).

NCEER-94-0011 "The Northridge, California Earthquake of January 17, 1994: Performance of Gas Transmission Pipelines,"
by T.D. O'Rourke and M.C. Palmer, 5/16/94, (PB94-204989, AOS, MF-AO!).

NCEER-94-0012 "Feasibility Study of Replacement Procedures and Earthquake Performance Related to Gas Transmission
Pipelines," by T.D. O'Rourke and M.C. Palmer, 5/25/94, (PB94-206638, A09, MF-A02).

NCEER-94-0013 "Seismic Energy Based Fatigue Damage Analysis of Bridge Columns: Part II - Evaluation of Seismic
Demand," by G.A. Chang and J.B. Mander, 6/1/94, (PB95-18106, A08, MF-A02).

NCEER-94-0014 "NCEER-Taisei Corporation Research Program on Sliding Seismic Isolation Systems for Bridges:
Experimental and Analytical Study of a System Consisting of Sliding Bearings and Fluid Restoring
Force/Damping Devices," by P. Tsopelas and M.C. Constantinou, 6/13/94, (PB94-219144, AlO, MF-A03).

D-14
NCEER-94-0015 "Generation of Hazard-Consistent Fragility Curves for Seisnuc Loss Estimation Studies," by H. Hwang and
J-R. Huo, 6/14/94, (PB95-18!996, A09, MF-A02).

NCEER-94-0016 "Seisnuc Study of Building Frames with Added Energy-Absorbing Devices," by W.S. Pong, C.S. Tsai and
G.C. Lee, 6/20/94, (PB94-219136, A!O, A03).

NCEER-94-0017 "Sliding Mode Control for Seisnuc-Excited Linear and Nonlinear Civil Engineering Structures," by J. Yang,
J. Wu, A. Agrawal and Z. Li, 6/21/94, (PB95-138483, A06, MF-A02).

NCEER-94-0018 "3D-BASIS-TABS Version 2.0: Computer Program for Nonlinear Dynanuc Analysis of Three Dimensional
Base Isolated Structures," by AM. Reinhorn, S. Nagarajaiah, M.C. Constantinou, P. Tsopelas and R. Li,
6/22/94, (PB95-182176, A08, MF-A02).

NCEER-94-0019 "Proceedings of the International Workshop on Civil Infrastructure Systems: Application of Intelligent
Systems and Advanced Materials on Bridge Systems," Edited by G.C. Lee and K.C. Chang, 7/18/94, (PB95-
252474, A20, MF-A04).

NCEER-94-0020 "Study of Seisnuc Isolation Systems for Computer Floors," by V. Lambrou and M.C. Constantinou, 7/19/94,
(PB95-138533, AIO, MF-A03).

NCEER-94-0021 "Proceedings of the U.S.-Italian Workshop on Guidelines for Seisnuc Evaluation and Rehabilitation of
Unreinforced Masonry Buildings," Edited by D.P. Abrams and G.M. Calvi, 7/20/94, (PB95-138749, Al3,
MF-A03).

NCEER-94-0022 "NCEER-Taisei Corporation Research Program on Sliding Seisnuc Isolation Systems for Bridges:
Experimental and Analytical Study of a System Consisting of Lubricated PTFE Sliding Bearings and Mild
Steel Dampers," by P. Tsopelas and M.C. Constantinou, 7/22/94, (PB95-182!84, A08, MF-A02).

NCEER-94-0023 "Development of Reliability-Based Design Criteria for Buildings Under Seisnuc Load," by Y.K. Wen, H.
Hwang and M. Shinozuka, 8/1/94, (PB95-211934, A08, MF-A02).

NCEER-94-0024 "Experimental Verification of Acceleration Feedback Control Strategies for an Active Tendon System," by
S.J. Dyke, B.F. Spencer, Jr., P. Quast, M.K. Sain, D.C. Kaspari, Jr. and T.T. Soong, 8/29/94, (PB95-212320,
AOS, MF-AO!).

NCEER-94-0025 "Seisnuc Retrofitting Manual for Highway Bridges," Edited by LG. Buckle and I.F. Friedland, published by
the Federal Highway Adnunistration (PB95-212676, A15, MF-A03).

NCEER-94-0026 "Proceedings from the Fifth U.S.-Japan Workshop on Earthquake Resistant Design of Lifeline Facilities and
Countermeasures Against Soil Liquefaction," Edited by T.D. O'Rourke and M. Hamada, 1 ln/94, (PB95-
220802, A99, MF-E08).

NCEER-95-0001 "Experimental and Analytical Investigation of Seisnuc Retrofit of Structures with Supplemental Damping:
Part I - Fluid Viscous Damping Devices," by A.M. Reinhom, C. Li and M.C. Constantinou, 1/3/95, (PB95-
266599, A09, MF-A02).

NCEER-95-0002 "Experimental and Analytical Study of Low-Cycle Fatigue Behavior of Senu-Rigid Top-And-Seat Angle
Connections," by G. Pekcan, J.B. Mander and S.S. Chen, 1/5/95, (PB95-220042, A07, MF-A02).

NCEER-95-0003 "NCEER-ATC Joint Study on Fragility of Buildings," by T. Anagnos, C. Rojahn and A.S. Kiremidjian,
1/20/95, (PB95-220026, A06, MF-A02).

NCEER-95-0004 "Nonlinear Control Algorithms for Peak Response Reduction," by Z. Wu, T.T. Soong, V. Gattulli and R.C.
Lin, 2/16/95, (PB95-220349, AOS, MF-AO!).

D-15
NCEER-95-0005 "Pipeline Replacement Feasibility Study: A Methodology for Minimizing Seismic and Corrosion Risks to
Underground Natural Gas Pipelines," by R.T. Eguchi, H.A. Seligson and D.G. Honegger, 3/2/95, (PB95-
252326, A06, MF-A02).

NCEER-95-0006 "Evaluation of Seismic Performance of an II-Story Frame Building During the 1994 Northridge
Earthquake," by F. Naeim, R. DiSulio, K. Benuska, A. Reinhom and C. Li, to be published.

NCEER-95-0007 "Prioritization of Bridges for Seismic Retrofitting," by N. Basoz and A.S. Kiremidjian, 4/24/95, (PB95-
252300, A08, MF-A02).

NCEER-95-0008 "Method for Developing Motion Damage Relationships for Reinforced Concrete Frames," by A. Singha) and
A.S. Kiremidjian, 5/11/95, (PB95-266607, A06, MF-A02).

NCEER-95-0009 "Experimental and Analytical Investigation of Seismic Retrofit of Structures with Supplemental Damping:
Part II- Friction Devices," by C. Li and A.M. Reinhom, 7/6/95, (PB96-128087, All, MF-A03).

NCEER-95-0010 "Experimental Performance and Analytical Study of a Non-Ductile Reinforced Concrete Frame Structure
Retrofitted with Elastomeric Spring Dampers," by G. Pekcan, J.B. Mander and S.S, Chen, 7/14/95, (PB96-
137161, A08, MF-A02).

NCEER-95-0011 "Development and Experimental Study of Semi-Active Fluid Damping Devices for Seismic Protection of
Structures," by M.D. Syrnans and M.C. Constantinou, 8/3/95, (PB96-136940, A23, MF-A04).

NCEER-95-0012 "Real-Time Structural Parameter Modification (RSPM): Development of Innervated Structures," by Z.

Liang, M. Tong and G.C. Lee, 4/11/95, (PB96-137153, A06, MF-AO!).

NCEER-95-0013 "Experimental and Analytical Investigation of Seismic Retrofit of Structures with Supplemental Damping:
Part III - Viscous Damping Walls," by A.M. Reinhom and C. Li, 10/1/95, (PB96-176409, Al 1, MF-A03).

NCEER-95-0014 "Seismic Fragility Analysis of Equipment and Structures in a Memphis Electric Substation," by J-R. Huo and
H.H.M. Hwang, (PB96-128087, A09, MF-A02), 8/10/95.

NCEER-95-0015 "The Hanshin-Awaji Earthquake of January 17, 1995: Performance of Lifelines," Edited by M. Shinozuka,
11/3/95, (PB96-176383, Al5, MF-A03).

NCEER-95-0016 "Highway Culvert Performance During Earthquakes," by T.L. Youd and C.J. Beckman, available as NCEER-
96-0015.

NCEER-95-0017 "The Hanshin-Awaji Earthquake of January 17, 1995: Performance of Highway Bridges," Edited by I.G.
Buckle, 12/1/95, to be published.

NCEER-95-0018 "Modeling of Masonry Infill Panels for Structural Analysis," by A.M. Reinhom, A. Madan, R.E. Valles, Y.
Reichmann and J.B. Mander, 12/8/95.

NCEER-95-0019 "Optimal Polynomial Control for Linear and Nonlinear Structures," by A.K. Agrawal and J.N. Yang,
12/11/95, (PB96-168737, A07, MF-A02).

NCEER-95-0020 "Retrofit of Non-Ductile Reinforced Concrete Frames Using Friction Dampers," by R.S. Rao, P. Gergely and
R.N. White, 12/22/95, (PB97-133508, AlO, MF-A02).

NCEER-95-0021 "Parametric Results for Seismic Response of Pile-Supported Bridge Bents," by G. Mylonakis, A. Nikolaou
and G. Gazetas, 12/22/95, (PB97-100242, Al2, MF-A03).

NCEER-95-0022 "Kinematic Bending Moments in Seismically Stressed Piles," by A. Nikolaou, G. Mylonakis and G. Gazetas,
12/23/95.

D-16
NCEER-96-0001 "Dynamic Response of Unreinforced Masonry Buildings with Flexible Diaphragms," by A.C. Costley and
D.P. Abrams," 10/10/96.

NCEER-96-0002 "State of the Art Review: Foundations and Retaining Structures," by I. Po Lam, to be published.

NCEER-96-0003 "Ductility of Rectangular Reinforced Concrete Bridge Columns with Moderate Confinement," by N. Wehbe,
M. Saiidi, D. Sanders and B. Douglas, 11/7/96, (PB97-133557, A06, MF-A02).

NCEER-96-0004 "Proceedings of the Long-Span Bridge Seismic Research Workshop," edited by I.G. Buckle and J.M.
Friedland, to be published.

NCEER-96-0005 "Establish Representative Pier Types for Comprehensive Study: Eastern United States," by J. Kulicki and Z.
Prucz, 5/28/96.

NCEER-96-0006 "Establish Representative Pier Types for Comprehensive Study: Western Urrited States," by R. Imbsen, R.A.
Schamber and T.A. Osterkamp, 5/28/96.

NCEER-96-0007 "Nonlinear Control Techniques for Dynamical Systems with Uncertain Parameters," by R.G. Ghanem and
M.I. Bujakov, 5/27/96, (PB97-!00259, A17, MF-A03).

NCEER-96-0008 "Seismic Evaluation of a 30-Year Old Non-Ductile Highway Bridge Pier and Its Retrofit," by J.B. Mander,
B. Mahmoodzadegan, S. Bhadra and S.S. Chen, 5/31/96.

NCEER-96-0009 "Seismic Performance of a Model Reinforced Concrete Bridge Pier Before and After Retrafit," by J.B.
Mander, J.H. Kirn and C.A. Ligozio, 5/31/96.

NCEER-96-0010 "IDARC2D Version 4.0: A Computer Program for the Inelastic Damage Analysis of Buildings," by R.E.
Valles, A.M. Reinhorn, S.K. Kunnath, C. Li and A. Madan, 6/3/96, (PB97-I00234, Al 7, MF-A03).

NCEER-96-0011 "Estimation of the Economic Impact of Multiple Lifeline Disruption: Memphis Light, Gas and Water
Division Case Study," by S.E. Chang, H.A. Seligson and R.T. Eguchi, 8/16/96, (PB97-133490, All, MF-
A03).

NCEER-96-0012 "Proceedings from the Sixth Japan-U.S. Workshop on Earthquake Resistant Design of Lifeline Facilities and
Countermeasures Against Soil Liquefaction, Edited by M. Hamada and T. O'Rourke, 9/11/96, (PB97-
13358 l, A99, MF-A06).

NCEER-96-0013 "Chemical Hazards, Mitigation and Preparedness in Areas of High Seismic Risk: A Methodology for
Estimating the Risk of Post-Earthquake Hazardous Materials Release," by H.A. Seligson, R.T. Eguchi, K.J.
Tierney and K. Richmond, 11/7/96.

NCEER-96-0014 "Response of Steel Bridge Bearings to Reversed Cyclic Loading," by J.B. Mander, D-K. Kirn, S.S. Chen and
G.J. Prernus, 11/13/96, (PB97-140735, Al2, MF-A03).

NCEER-96-0015 "Highway Culvert Performance During Past Earthquakes," by T.L. Youd and C.J. Beckman, 11/25/96,
(PB97-133532, A06, MF-AO!).

NCEER-97-0001 "Evaluation, Prevention and Mitigation of Pounding Effects in Building Structures," by R.E. Valles and
A.M. Reinhorn, 2/20/97, (PB97-!59552, AJ4, MF-A03).

NCEER-97-0002 "Seismic Design Criteria for Bridges and Other Highway Structures," by C. Rojahn, R. Mayes, D.G.
Anderson, J. Clark, J.H. Horn, R.V. Nutt and M.J. O'Rourke, 4/30/97, (PB97-194658, A06, MF-A03).

NCEER-97-0003 "Proceedings of the U.S.-ltalian Workshop on Seismic Evaluation and Retrofit," Edited by D.P. Abrams and
G.M. Calvi, 3/19/97, (PB97-194666, A13, MF-A03).

D-17
NCEER-97-0004 "Investigation of Seismic Response of Buildings with Linear and Nonlinear Fluid Viscous Dampers," by
A.A. Seleernah and M.C. Constantinou, 5/21/97, (PB98-109002, Al 5, MF-A03).

NCEER-97-0005 "Proceedings of the Workshop on Earthquake Engineering Frontiers in Transportation Facilities," edited by
G.C. Lee and I.M Friedland, 8/29/97, (PB98-128911, A25, MR-A04).

NCEER-97-0006 "Cumulative Seismic Damage of Reinforced Concrete Bridge Piers," by S.K. Kunnath, A. El-Bahy, A
Taylor and W. Stone, 9/2/97, (PB98-108814, Al 1, MF-A03).

NCEER-97-0007 "Structural Details to Accommodate Seismic Movements of Highway Bridges and Retaining Walls," by
R.A. hnbsen, RA Schamber, E. Thorkildsen, A Kartoum, B.T. Martin, T.N. Rosser and J.M. Kulicki,
9/3/97.

NCEER-97-0008 "A Method for Earthquake Motion-Damage Relationships with Application to Reinforced Concrete
Frames," by A. Singha! and AS. Kirernidjian, 9/10/97, (PB98-108988, A13, MF-A03).

NCEER-97-0009 "Seismic Analysis and Design ofBridge Abutments Considering Sliding and Rotation," by K. Fishman and
R. Richards, Jr., 9/15/97, (PB98-108897, A06, MF-A02).

NCEER-97-0010 "Proceedings of the FHWA/NCEER Workshop on the National Representation of Seismic Ground Motion
for New and Existing Highway Facilities," edited by I.M. Friedland, M.S. Power and RL. Mayes, 9/22/97.

NCEER-97-0011 "Seismic Analysis for Design or Retrofit of Gravity Bridge Abutments," by K.L. Fishman, R. Richards, Jr.
and R.C. Divito, 10/2/97, (PB98-128937, A08, MF-A02).

NCEER-97-0012 "Evaluation of Simplified Methods of Analysis for Yielding Structures," by P. Tsopelas, MC.
Constantinou, C.A. Kircher and AS. Whittaker, 10/31/97, (PB98-128929, Al 0, MF-A03).

NCEER-97-0013 "Seismic Design of Bridge Columns Based on Control and Repairability of Damage," by C-T. Cheng and
J.B. Mander, 12/8/97.

NCEER-97-0014 "Seismic Resistance of Bridge Piers Based on Damage Avoidance Design," by J.B. Mander and C-T.
Cheng, 12/10/97.

NCEER-97-0015 "Seismic Response of Nominally Symmetric Systems with Strength Uncertainty," by S. Balopoulou and M.
Grigoriu, 12/23/97.

NCEER-97-0016 "Evaluation of Seismic Retrofit Methods for Reinforced Concrete Bridge Columns," by T.J. Wipf, F.W.
Klaiber and F.M. Russo, 12/28/97.

NCEER-97-0017 "Seismic Fragility of Existing Conventional Reinforced Concrete Highway Bridges," by C.L. Mullen and
A.S. Cakmak, 12/30/97.

NCEER-97-0018 "Loss Asssessment of Memphis Buildings," edited by D.P. Abrams and M. Shinozuka, 12/31/97.

NCEER-97-0019 "Seismic Evaluation ofFrarnes with Infill Walls Using Quasi-static Experiments," by K.M. Mosalarn, RN.
White and P. Gergely, 12/31/97.

NCEER-97-0020 "Seismic Evaluation ofFrarnes with Infill Walls Using Pseudo-dynamic Experiments," by K.M. Mosalarn,
RN. White and P. Gergely, 12/31/97.

NCEER-97-0021 "Computational Strategies for Frames with Infill Walls: Discrete and Smeared Crack Analyses and Seismic
Fragility," by K.M. Mosalarn, RN. White and P. Gergely, 12/31/97.

NCEER-97-0022 "Proceedings of the NCEER Workshop on Evaluation of Liquefaction Resistance of Soils," edited by T.L.
Youd and I.M. ldriss, 12/31/97.

D-18