OFFSHORE TECHNOLOGY CONFERENCE PAPER

6200 North Central Expressway NUMBER aTe 2671

Dallas, Texas 75206

Behavior of Soft Clays Under Earthquake

Loading Conditions
By

I. M. Idriss, R. Dobry, E. H. Doyle, and R. D. Singh, Shell Development Co.

THIS PAPER IS SUBJECT TO CORRECTION

©Copyright 1976
Offshore Technology Conference on behalf of the American Institute of Mining, Metallurgical, and Petroleum
Engineers, Inc. (Society of Mining Engineers, The Metallurgical Society and Society of Petroleum Engineers),
American Association of Petroleum Geologists, American Institute of Chemical Engineers, American Society
of Civil Engineers, American Society of Mechanical Engineers, Institute of Electrical and Electronics En-
gineers, Marine Technology Society, Society of Exploration Geophysicists, and Society of Naval Architects
and Marine Engineers.
This paper was prepared for presentation at the Eighth Annual Offshore Technology Conference, Houston,
Tex., May 3-6, 1976. Permission to copy is restricted to an abstract of not more than 300 words. Illustrations
may not be copied. Such use of an abstract should contain conspicuous acknowledgment of where and by
whom the paper is presented.

ABSTRACT be used as a basis for assessing the performance

of pile foundations; and (c) evaluation of
A new method is present~dto compute the e~~k~~ ground failure potential of the soil, which may
quake response of soft clay deposits. In this be manifested in phenomena such as liquefaction
method, the change of properties of tha soil of sands and increase in pore pressures or re-
during cyclic loading is directly incorporated molding of clays during seismic loading.
into a.nonlinear response calculation in ,the
time domain. The parameters of the stress- A previous publication (1) included a review of
strain model proposed for the clay are obtained geotechnical aspects related to the seismic de-
from cyclic shear tests. -A sample problem of_ sign of offshore structures; that reference also
earthquake response of a soft clay offshore discussed available observations and analytical
site is discussed. procedures used to estimate ground response to
an assumed baserock motion. For a site where
INTRODUCTION subsurface conditions consist of horizontal,
or almost horizontal, parallel soil layers, the
Development of seismic design parameters for response of the soil column to horizontal base-
offshore structures, such as fixed platforms, rock excitation can be reasonably considered as
is a complex problem that requires incorporation that of a shear beam. Therefore, calculation
of the effects of local site conditions, struc- of this response reduces to the solution of one-
tural characteristics, and water. An important dimensional vertically propagating shear waves.
consideration in developing these parameters Computer programs like SHAKE can solve the wave
is the response in thef:ree field (ie, assuming propagation equation and compute the desired
no structure present). Studies _related to ground response parameters; an equivalent linear
free field resporise aurlng-a postulatea seismic model of the profile is used, with the damping
event provide results relevallt to: (a) esti- and stiffness properties of the soil adjusted
mates of ground motions at sea floor or at to the seismic strain level, to approximate the
some depth below sea floor--these motions may actual nonlinear hysteretic behavior of the
then be used as a basis for the seismic design soil (2). Most soil response analyses to date,
of the structure; (b) estimates of stresses onshore or offshore, have been done using the
and displacements induced in the soil by these equivalent linear method.
motions--the estimated displacements may then

References and illustrations at end of naner.

- - - - - - - - - - - - - - - - - - - - - - - - ----- - -
 606 BEHAVIOR OF SOFT CLAYS UNDER EARTHOUAKE LOADING CONDITIONS OTC 2671

The results of cyclic loading tests on soft initial curve described by Eq. 1. The proposed
~la~s, such as found in many marine deposits, model assumes that this degradation process
lndlcate a degradation in soil stiffness due occurs only at the times when the stress changes
to cyclic loading, which may·be important for sign (ie~ 7v7ry half cycle). T~erefore, Eq. 1
the seismic design of an offshore structure and the lnltlal backbone curve ln Fig. 2 are
located at a soft clay ;ite. This effect is associated with the first half cycle of loading.
associated with an increase in pore pressure For subsequent half cycles, new backbone curves
during cyclic loading and also with deteriora- are defined, each having stresses that are a
tion of the structure of the clay due to re- fraction, 0, of the initial stresses; ie,
molding (3, 4, 5, 6). Modulus degradation
of the soil may affect the response character- (2)
istics of.the site and therefore should be con-
sidered in the selection of seismic design
parameters for the offshore structure and its o = 1 for the first half cycle, and its value
foundations. decreases continuously during cyclic loading;
o is defined as a Degradation Index. This
Degrad~tion Index may be considered as a mea-
To account for this modulus degradation in re-
sponse analyses, a nonlinear stress-strain model sure of the current value of the stiffness of
was developed for soft clays (7). Th~ develop- the clay, expressed as a fraction of the initial
men~ of this model was based on the results of
stiffness.
a series of cyclic tests on soft San Francisco·
Bay Mud. A brief description of the basic For any controlled-strain triaxial cyclic test,
characteristics of this model and its use in the stress, ad, or the secant modulus, ad/E, as
nonlinear response calculations together with a function of number of cycles plots as a -
results for a sample problem form the subject straight line on log-log paper, with the modulus
of this paper. . decreasing with increasing number.of cycles.
The slope of this line, t, is a measure of the
BEHAVIOR OF SOFT CLAYS DURING CYCLIC LOADING rate of modulus degradation. The line is
steeper and the val~e of t is larger when a
The stress-strain behavior of a soil during one higher value of strain, E, is applied in the
cycle of shear loading can be approximately test. Figure 3 shows the relation between t
described by a closed hysteretic loop' such as and cyclic strain obtained for San Francisco
shown in Fig. 1.Th~ curve defined by the tips Bay Mud using the same series of tests mentioned
of hysteretic loops associated with different previously. Six of the cyclic tests shown in
strain levels is called a "backbone curve", - _. Fig, 3 were perf2rme~ after consolidating the
Figure 1 presents one possible mathematical re- sample under a confining pressure, a c = 1.5 ksf,
presentatinn of..the backbone curve and hyster- and two tests under ac = 3.0 ksf. The limited
etic curve of a soil, based on the use of the amount of data in Fig. 3 suggests that the
Ramberg-Osgood equation. Figure 2 shows a value of t may be independent of confining
backbone curve ·obtained experimentally from a pressure.
series of controlled-strain, cyclic triaxial
tests on samples of normally consolidated San The stress-strain behavior ofa soft clay, such
Francisco Bay Mud. In each test, the sample as San Francisco Bay Mud, during controlled-
was first consolidated under a confining pres- strain tests is fully defined by the initial
sure of 1.5 ksf and then subjected to 100 cycles backbone curve, associated with the first half
of a uniform cyclic axial strain, with a maxi- cycle (Fig. 2), together with the relation be-
mum amplitude.:. E; for each cycle, the maximum tween t and strain (Fig. 3). For transient
deviator. stress, ad,· was measured. Each data loading, the proposed model assumes that the
point in Fig. 2 corresponds to a different test , value of Degradation Index, corresponding to
-" -

and the values of ad and E: presented in the each new half cycle, depends only on two para-
figure were obtained from the first cycle of meters: (a) the value of Degradation Index, 0,
loading. Therefore, the backbone curve in F~g. for the previous half cycle; and (b) the maximum
2 can be considered representative of the ini- strain, E, developed during the previous half ,.
tial condition of the soil, at the outset of cycle. With this assumption, and after some
cyclic loading. algebra, the (negative) increment, ~o, from one
half cycle to the next, is given by:
The initial backbone curve for a cyclic triaxial
l+t
test, such as shown in Fig. 2, can be considered ,,_ 1 ""t
a stress-strain function of the form: ~u - -- zt (0) (3)

where t is determined from a curve such as Fig ..

3, for the maximum strain develope~ during the _
During cyclic loading, the backbone curve de- previous half cycle.
grades and adopts a position lower than the

-- - - - - ------- - - - - - -
 I. M. IDRISS, R. DOBRY, E. H. DOYLE. AND R. D. SINGH 607

During the new half cycle, the current backbone the backbone equation in conjunction with the
curve, O"d(E), and also the:f:llil stress.":"strain use of the method of characteristics to solve
behavior of.the soil, is completely determined the wave propagation equations.for vertically
by Eq. 2, except that the current value, 0 + ~o, propagating shear waves. This program was modi-
should be used inEq. 2. fied by the introduction of the proposed soil
degradation mode~ and the new program, thus
To verify the proposed model, the cyclic tests specialized to study the seismic response of
on San Francisco Bay Mud were extended beyond soft clay deposits, was designated DCHARM
100 cycles using varying values of cyclic strain (Method of Characteristics Including Soil
Figure 4 shows the results of one of these tests Degradation). In this new program, Eq. 4 is
together with the stresses predicted by the assumed to be valid for the initial backbone
proposed model. In this. test, _~ cyclic axial curve of the soil which is used for the first
strq.in of ,:,2%. was applied during 100 cycles, half cycle of seismic loading. For subsequent
then .:.1% was used for the following 100 cycles, half cycles, the equation of the backbone curve
etc., for a total of 800 cycles. The cyclic is modified as follows:
stress, O"d; was measuredat different cycles,

~ ~)J
and the results are plotted in Fig. 4 together _ Y T T
with the predicted values of stresses. The Y- 8 Gmax Y + a 0 Gmax yy (5)
effect of degradation is very pronounced in Y y
this test, with 0 dropping to a value of 0.4
at the end of the test; and the agreement be- where o is the current value of the Degradation
tween measured and predicted stresses is very Index at each level within the soil profile.
good. These results, and other tests on San
Francisco Bay Mud not presented her~appear to The use of program DCHARM to analyze the re-
confirm the validity of the proposed model for _ sponse of soft clay deposits requires defining
transient cyclic loading. the following input soil parameters :

All results for the proposed model have been 1. Total unit weight of soil at different
presented in terms of deviator stress, O"d, and depths.
axial strain, E, measured during triaxial tests. 2. Gm~x at differentdepth~ for the initial
Similar conclusions are assumed to be valid for condition of the soil.
cyclic loading in the field and can be expressed 3. Ramberg-Osgood parameters Yy ' a, and R.
in terms of shear stress, T, and shear strain, y. These parameters, together with Gmax '
completely define the .initial bacKDone
ANALYTICAL MODEL curve to be used in the first half cycle
of loading.
Available nonlinear ground response computer 4. Relationship between degradation parameter,
programs (8, g, 10) characterize the shear stress- t, and shear strain, y.
strain behavior of the soil by its backbone 5. Values of shear viscosity at different
curve; for generalized transient loading, the depths, if a viscosity is to be added to
soil is assumed to behave as a Massing Solid the profile (9).
(1.1, 12). One of the expressions used to repre- 6. A minimum value of d that the clay can adopt.
sent actual soil backbone curves is the Ramberg-
Osgood equation (9): EXAMPLE PROBLEM
The example selected to illustrate the use of
the method is an uffshore site, where the soil
(4) conditions consist of approximately 200 ft of
silty clay underlain by rock. This site was
described with some detail in a previous publi-.
where T shear stress cation, which also included results of a ground
y = shear strain response analysis of the site using the equi-
valent linear method (1).
G
max = shear modulus at very small
strains Selection of Soil Parameters
Yy' a, R soil parameters
Figures 5 and 6 reproduce the experimental
The hysteretic loop plotted in Fig. 1 was de- curves showing the influence of cyclic shear
rived from the application of Massing rules to strain on the secant modulus, G/Gmax , and on
the Ramberg-Osgood backbone curve, shown in the the damping ratio, A. These curves were ob-
same figure. tained from controlled-strain cyclic triaxial
and torsion laboratory tests conducted on re-
Nonlinear ground response program CHARSOIL (9) presentative samples of the clay from the site.
utilizes the Ramberg-Osgood equation to describe The experimental curves in Figs. 5 and 6 were

- - - - - - - - - - - - - - - - - - - - - - - - - - - _ . -
 608 S13FT CT.AYS TTNllFREARTHQUAKE LOADING CONDITIONS
--. -., ------ OF ---------
RT3T-IAVTOR OTC 2671

used to define the Ramberg-Osgoodparameters of as described above. Some results of the cal-
the initial backbone curve, used for DCW culations are presented in Figs. 7 and 8, where
calculations. The Rambexg-Osgoodexpressions they are compared with the results of the anal-
for G/Gmax and A obtaitiedfrom Eq. 4 are as yses for the same rock excitation and soil pro-
follows: ““ file, computed by the equivalent linear method
(1).
—.G .1..
(6al There are two basic differencesbetween these
G R-1
max two procedures. One differencepertains to
l+aG~
()max ‘y the use of the equivalent linear (SHAKE) or
nonlinear (DCHARM)procedure, and the other
difference is the effect of degradation,allowed
R by DCHARM but not by SHAKE. To separate the
T influence of these two effects on the results,
G an additional analysis was performed with
A(%) _ 100C4 ()max ‘y
‘r -Y_ (6b) DCW but without degradation (ie, 6 = 1 for
() all strain leveis); this is equivalent to using
‘Y the original program CHARSOIL. The results
computed with this additionalnonlinear cal-
The followingparameters were selected for the. culation (CHARSOIL)are also included in Figs.
clay at this site: Yy = 0.02%, R= 3, a= 1. 7 and 8.
The theoreticalcurves for G/Gmax and A, com-
puted from Eqs. 4 and 6 for this set of para- Figure 7 presents the input rock accelerogram
meters, are compared in Figs. 5 and 6 with the used and the accelerationtime histories com-
experimentalcurves. The agreement for the puted at depths of 25 ft (SHAKE) and 30 ft
modulus in Fig. 5 is good. The agreement for (CHARSOILand DCHARM). As can be noted, the
the damping in Fig. 6.is good for large strains; three computed time histories are very similar
at small strains, the theoreticalA tends to and significantlydifferent from the applied
zero, while the experimentalcurve has a minimum baserock motions both in amplitude and frequency
value of A = 4%.. To “accountfor this difference content. The three computed maximum ground
at small-strains,and to avoid unrealistichigh accelerationsat the depths of 25 or 30 ft
frequency oscillationsin the computed ground range between 0.11 and 0.13 g, which represents
motions, a shear viscosity was added to the a significantdeamplificationof the 0.33 g
soil, with a value equivalent to a damping ratio maximum rock accelerationused as input. High
of 4% at small strains and at the fundamental frequency motions predominate in the input rock
period of the soil deposit. excitation,while low frequency (long period)
motions predominate in the ground response com-
The initial backbone curve was completely de- puted by all methods.
fined by the selection of the Ramberg-Osgood
parameters, described aboye; andbythe val_ues Figure 8 presents the accelerationresponse
of Gmax at different.depthsof the soil profile. spectra, for a 5% spectral damping ratio, cor-
For the calculationspresented in this paper, responding to the accelerationtime histories
the same values of Gmax,used for the equivalent shown in Fig. 7. The three response spectra
linear model calculationsin ref. (l), were correspondingto ground motions at depths of
utilized. 25 to 30 ft are very similar and significantly
different from the input spectrum. While the
In-additionto the initial backbone curve, valid input rock spectrum has a predominant period
for the first half cycle of loading, a relation- of approximately0.55 see, all”computed ground
ship between the degradationparameter, t, and spectra have a predominant period of about 2.5
shear strain, such as shown in Fig. 3, was de- sec, indicatingthat the effect of the soft
fined for the clays in the 200-ft deposit and clay profile was to deamplify the motions at
used in the calculations.‘ This relationship low periods while amplifying them at a period
was based on the results of controlled-strain of about 2.5 sec. This amplificationat a soft
and controlled-stresscyclic triaxial tests site may be significant for a long period struc-
performed on clay s~ples_from the site. A ture, such as an offshore platform. The ordi-
minimum value of the Degradation Index, 6 = 0.4, nate of the maximum spectral accelerationsin
was selected for the clay. Fig. 8 at a period of about 2.5 sec is 0.38 g
using program DCHARM and 0.63 g using program
Computed Ground Response - SHAKE, which can be a significantdifference if
the spectrum is to be used for the seismic de-
The response of t_he200-ft clay profile to a sign of a structure. The correspondingvalue
baserock horizontal excitationhaving a maximum using program CHARSOIL is 0.54 g, indicating
accelerationof 0.33 g was computed using pro- that most of this difference between SHAKE and
gram DCW, with the soil parameters selected DCHARM is due to the effect of degradation

.
 I. M. IDIUSS. ‘R. DO13RY1
,—.l?. .---
H. nfwT,ti.
=.-+ ,.-Arvn”
-,-. R.. n
=.”
STNTCU
-,, u .L
6

rather than to the use of a nonlinear”instead ACKNOWLEDGMENTS

of equivalent linear method.
The theoretical studies and the laboratory.
Additional results and comparisonsincluded: testing leading to the development of the basic
maximum acceleration,maximum shear stress, approach outlined in this paper were supported
maximum shear strain, and maximum soil displace- by a research grant from the ProfessionalDe-
meritwith respect to rock at various depths in velopment Progran.of lfoodward-ClydeConsultants
the soil profile. The three sets of values (WCC]. The new program has been used in as-
were generally similar, with the effect of de- sessing ground response characteristicsfox a
gradation being to decrease calculatedaccel- number of cases for a group of oil companies.
erations and stresses at shallow depths below The authors gratefully acknowledgethis support
sea”floor and to increase calculated strains The authors are also indebted to Professor-
an-ddisplacementsat all depths. J. Lysmer, University of California at Berkeley
for his helpful suggestions on the use and li-
Figure 9 shows the evolution of the Degradation mitations of nonlinear numerical procedures.
Index, 8, with time during the earthquakeat a Other WCC personnel who contributedto this
depth of 30 ft. Figure -lO”pre&entsthe value work included J. Wilson, director of WCC’S
of 6 at the end of the earthquake at different laboratoriesin Oakland, California;Dr. R. Akk]
depths. The value of 6 at most depths in Fig. E. Ng; R. Youngs; and R. Murarka.
10 is between 0.4 and 0.6, suggesting a sub-
stantial increase in pore pressures or partial REFERENCES
remolding of the clay in the profile caused by
the prescribed input motions. 1. Idriss, I. M., Dobry, R., and Power, M. S.:
!IsoilResponse Considerationsin Seismic
SUMMARY AND CONCLUSIONS Design of Offshore Platforms,ttJournal-of
Petroleum Engineering,March 1976.
1. A new stress-strainmodel, which incorpo-
rates the effect of.modulus degradation for 2. Schnabel, P. B., Lysmer, J., and Seed, H. B,
soft clays subjected to cyclic shear loading, !ls~KE--A Computer program for Earthquake
was developed and verified experimentallyfor Response Analysis of Horizontally Layered
normally consolidatedSan Francisco Bay Mud. Sites,” Report No. EERC 72-12, Earthquake
EngineeringResearch Center, University of
2. A nonlinear ground response computer pro- California, Berkeley, 1972.
gram, DCHARM, was developed by incorporating
the degrading soil model into an existing pro- 3. Thiers, G. R., and Seed, H. B.: ‘tCyclic
gram. The program uses the method of charac- Stress-StrainCharacteristicsof Clay,”
teristics to solve the one-dimensionalwave’ Journal of the Soil Mechanics and Foundation
equation, and it allows modification of material Division, ASCE, vol. 94, No. SM2, March
properties with tirne,.due
to increase in pore 1968, p. 555-569.
pressure or remolding. It is hoped that this
program would provide an improvementin the 4. Esrig, M. I., Ladd, R. S., and Bea; R. G.:
engineeringtools available for assessing-the !!Materialproperties of SubmarineMissis-
seismic response at sites underlain.bysoft sippi Delta Sediments Under Simulated Wave
clay sediments. Loadings,??paPer “presentedin Offshore
Technology Conference,OTC 2188, Dallas,
3. A sample problem was selected of a soft Texas, 1975.
clay offshore site. Results of response cal-
culations using the same baserock excitation 5. Sangrey, D. A., Hankel, D. J., and Esrig,
but different methods of analysis were presented M. 1.: !t~e Effective Stress Response of a
and compared. The’analyticalmethods used were: Saturated Clay Soil to Repeated Loading,”
(a) Equivalent Linear (programSHAKE), (b) Non- Canadian GeotechnicalJournal,6,241,”1969.
linear (progiarnCHARSOIL), and _(c)Nonlinear
with Modulus Degradation (programDC-). For 6. Wilson, N. E., and Greenwood, J. R.: ltPore
the sample problem, the main effect of degra- Pressures and Strains After Repeated Load-
dation is to significantlydecrease spectral ing of Saturated Clay,” Canadian Geotechnical
accelerationsin the long period range at shal- Journal,ll,269, 1974.
low depths below sea floor. Other effects,
such as decrease in induced stresses and in- 7. Idriss, I. M., Dobry, R., and Singh, R. D.:
crease in relative displacements,were not I!EarthquakeResponse of Soft ClaY Deposits
fully addressed in this paper. These latter IncludingModulus DegradationEffects,” in
effects would be taken into account in assessing preparation,to be submitted for possible
performanceat a specific site _andin evaluating publication in the Journal of the Geotech-
the consequencesof the occurrence of a seismic nical EngineeringDivision, ASCE, 1976.
event.
10 BEHAVIOR OF SOFT CLAYS UNDER LRTHQUAKELOADING CONDITIONS OTC 2671

8. Idriss, I. M., and Seed, H. B.: ‘tSeismic Earthquakes,” Bulletin “of the Seismological
Response_ofHorizontal Soil Layers,” Journal SOCietY of America, Vol. 65, No. 5, p. 1315-
of.Soil Mechanics and Foundation Division, 1336, 1975.
ASCE, Vol. 94, No. SM4, July-1968,”p. 1003-
1031. 11. Massing, G.: !!Eigenspannungen.
and Verfesti-
gung beim Messing,11proceedings Of the
9. Streeter, V. L., Wylie, E. B., and Richart, Second InternationalCongress of Applied
F. E.: IIsoilMotion Computationsby Char- Mechanics, p. 332-335, 1926.
acteristicMethod,” Journal of the Geotech-
nical Engineering Division, ASCE~ Vol. 100, 12. Iwan, W. D.: !t~ a Class of Models for the
No. GT3, March 1974. Yielding Behavior of Continuous and Compo-
site Systems,” Journal of Applied Mechanics,
10. Joyner, W. B., and Chen, A. T. F.: !!Cal- Paper No. 67-APM-S, Transactionsof ASME,”
culation of Nonlinear Ground Response in 1967.

Backbone Curve
-r A’ ‘max

Y=yy &&+a (Gm:xyyR)]

[

Y
●

FIG, 1 - RAMBERG-OSGOOD BACKBONE CURVE AND STEADY-STATE HYSTERESIS LOOP,

 FC = 1.5 ksf

Cycle No. = 1

-“ ●
_/-
/-. -
/ /“
-0
(D - /-
4 x’ ‘ =
!s(%) 0.02[0.5 Ud +0.2 (0.5”@l
/ “
/
/

/
/
/
P ● Experimental
Data

o 0.5 1.0 .1.5 2.0 2,5

AXIALSTiAIN(kc) - PERCENT
FIG, 2 -}ACKBONE CURVE FROM STRAIN-CONTROLLED CYCLIC TRIAXIAL TEST, SAN
FRANCISCO BAY MUD,

0.25 t
● UC = 1.5 ksf

A rC = 3.0 ksf

0.20

0.15
A

0.10 /

●
0.05 db

0
o 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0
-CYCLIC AXIAL STRAIN, 6 -- PERCENT
FIG, 3 - DEGRADATION PARAMETER T vs STRAIN-STRAIN-CONTROLLED CYCLIC TRIAXIAL TESTS,
SAN FRANCISCO BAY MUD,
 E=?2%. *1% *0.03% *2% ?1% f3.03%
- 1.0
m
Y
u“’0.8

0.6

0.4

0.2

0 ,

~
1600

----- 1 ~------ ~ = cdl (Degradationnot al1owed)

1 ~

‘f
I
1 I 1
1200
‘1 I I
1 ~ I
L -----. , L-
: ----: m.
! 1
.800 I
I I
● I I
t) I ; Ud . Small(Degradationallowed)
/ 4I

400 b--
I .
() ● ()
~---- 4) ,/---
/0 ● -,
●
● t)
It

—-_
, ‘MEASUREO
0
o 100 200- 300 400 500 600 700 800
NUMBEROF CYCLES,N
FIG, 4”- TRANSIENT CYCLIC TRIAXIAL TEST, SAN FRANCISCO BAY MUD,

1.0
—~’-
W.

E
-%b
‘ .>
12.8 \ \
‘1,\ FROMTESTRESULTS
1, Y
0.6
x Used for calculations by Ii
& DCHARM and CHARSOIL \
(Ramberg-Osgoodwith:~\ \
2 Yy = ~.02%, R = 3, ci= 1) \
\J
0.4
\\

\
k>%
0.2’

\\
-L \ .
----- _
----
0.
10-4 10-3 10-1 1 10
SHEAR STRAIN
, y - PERCENT
FIG, 5 - EXAMPLE RELATIONSHIP - REDUCTION IN MODULUS vs SHEAR STRAIN, SILTY CLAYS,
,
3!

//--
30

7’”
25
<7”
20
~7’
/
//
15
,//
FROM TEST RESULTS /’

10 /‘
~Used for calculation by
DCHARMand CHARSOIL
/ (Ramberg -OsgDod with:
,/ ,ij Yy = 0.02%, R= 3, a= 1)
5
/’ /
———
/

/0’
———- -—-
0 .A 3 .7 .
10”7 10-J 10-G 1o-1
1 10
SHEAR STRAIN , Y - PERCENT
FIG, 6 - EXAMPLE RELATIONSHIP - DAMPING RATIO vs SHEAR STRA N, SILTY CLAYS,
C@h,
Feat ,Sam moor

-0.2 EO 70.80’
0 10 20 30 40= 93 G

0.2 .. .. .
Non-linear
(CHARSOIL)

: 0.1

5-

+-!3 0.0
s
3:
-0.1

-0,2
9 10 20 30 T3 5a . ~. 70 w’%

-!3,2
4 I
o 10 20 30 40 50 60 70 . 80 92
.

1 1
0.’3

0.2

0.1

-25 0.0.
!=
s
ij -0.1

-0.2

-0.3

BEDROCK
-0,4
q 10 20 30 40 xl m 70 _ 80 90
TIME Iil SECO!IDS

FIGi 7 - EXAMPLE GROUND RESPONSE CALCULATIONS FOR AN

OFFSHORE PLATFORM SITE,
“T—57777
S$AC.1Dmn$d
=5%
0.7
— CW1v.k. L..r (SHAKE!

‘— NC.I-lkiw G* CWraMim @CHARMl

. P

Sas

1
u

k%

6! — - —

—
L32

,.*

—i

! 0.64

0,s3

0,,0, 0,s Q.m 1,(4 ml tam

IMM.oaxcm

FIG, 8 - EXAMPLE GROUND RESPONSE CALCULATIONS FOR AN OFFSHORE PLATFORM SITE:

RESPONSE SPECTRA,

1.1.
1.0
Depth = 30 Feet
0.9

0.8 \

0.7

0.6
\

0.5

0.4

0.3

0.2

0.1

o
0 10 20 30 40 50 60 70 80 90
TIME, SECONDS
FIG, 9 --EXAMPLE RESULTS: DEGRADATION INDEX VS TIME DURING EARTHQUAKE EXCITATION,
 DEGRADATION INDEX , ~ , AT THE END OF EARTHQUAKE (t = 85 sec. )
0 “0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1
0

20

40 — — — —

60

80

l--
U
?
100
.
r
+
L
m
D
120

d
140

160

180

200 9

220
FIG, 10 - EXAMPLE RESULTS: DEGRADATION INDEX AT
END OF EARTHQUAKE EXCITATION,