.

JNI'SNOS

ry AA'IL{\ NHOI

AAlMN

//?-\

rx\)
Y-/

1t731"r6'guaqdalE

uq)unclpowlw 'f

,{4gqu1sedols
puB rll8uerls IIos

CHAPTER3

Soil Mechanics Principles

For slope stability analysesto be useful, they must represent the correct problem, correctly' .fttrmuLoted.This
requires(l) mastery of the principles of soil mechanLcs.(2) knowledge of geology and site conditions. and
-31knowledge of the propertiesof the soils at the site.
In thrs chapter we deal with the principles of soil meihanics that are needed to understand and to fbrmulate
.inalysesof slope stability problems correctly.

DRAINED AND UNDRAINED CONDITIONS

The conceptsof drained and undrained conditions are
,rf fundamentalimportance in the mechanicalbehavior
,rf soils, and it is worthwhile to review these concepts
.it the beginning of this examinationof soil mechanics
principles. The lay definitions of drained antl un.lruined (drained : dry or emptied, undrained - not
,lry or not ernptied) do not describe the way these
t ords are used in soil mechanics.The definitions used
in soil mechanicsare relatedtcl the easeand speedwith
n hich water rroves in or out of soil in comparisonwith
rhe length of tirne that the soil is subjected to some
;hange in load. The crux of the issue is whether or not
ihanges in changes in load cause changes in pore
J)ressure:
. Druined is the condition under which water is able
to flow into or out of a mass of soil in the length
o1 time that the soil is subjectedto some change
in load. Under drained conditions, changesin the
loads on the soil do not causechangesin the water
pressurein the voids in the soil, becausethe water
can move in or out of the soil frecly when the
volume of voids increasesor decreasesin responsc
to the changing loads.
. IJndruined is the condrtion under which there is
no flow of water into or out of a mass of soil in

the length of time that the soil is subjectedto some

change in load. Changes in the loads on the soil
cause changesin the water pressurein the voids,
b e c a u s et h e w a t e r c a n n o t m o v e i n o r o u t i n r e sponse to the tendency lor the volume of voids to
change.
An example that illustrates these conditions is
shown in Figure 3.1, which shows a clay test specimen
in a direct shear test apparatus.The permeability of the
clay is low, and its compressibility is high. When the
normal load P and the shear load Z are increased,there
is a tendency for the volume of the clay to decrease.
This decreasein volume of the clay would take place
entirely by reduction of the volume of the voids becausethe clay particlesthemselvesare virtually incompressible.However, fbr the volume of the voids in the
clay to decrease,water would have to run out of the
clay becausewater is also virtually incornpressible.
If the loads P and T were increasedquickly, say in
1 second,the clay specimenwould be in an undrained
state lbr some period of time. Within the period of I
secondinvolved in increasingP andT, there would not
be enough time for any si-qnificantamount of water to
flow out of the clay. It is true that even in a period of
1 second there would be some small amount of flow,
but this would be insignificant. For all practical purposes the clay would be undrained immediately afler
the loads were changed.
If the loads P and I were held constantfor a longer
period. say one day, the state of the clay specimen
would change fiom undrained to drained. This is because within a period of one day. there would be sufficient time tbr water to flow out of the clay. Within
this time the volume of the voids would decreaseand
come essentiallytct equilibriLrm.It is true that equilibasyrnptotically.and strictly
rium would be approacl-rcd

l9

20

sotl

MECHANICs PRTNCTPLES

Metalplate
Porousstone
Normal load P

Shearbox

Shear loadT

F'igure3.1

Dircct sheal test apparatus

woulcl be approachedclosely bLrt

spcaking.ecluilibrir-rm
However.
firr all practical purposcs
never bc rcached.
would
be
clraincd
alfcr
the loacls were helcl
thc clay
constant lirr one day.
It is clear from this exarnple that the dil'l'erencebetween undriiined and drained. as these words are r.rsed
in soif mechanics,is time. Every mass of soil has cherracteristics that determinc how long is required fbr transition liom an undrained to a drained condition. A
the tinte reqr"rired
practical rneasurcclf this tirne is 1,),r,
to achieve 99% o1' thc equilibrium volume change,
which 1or practical plrrposes,we consider to be equi, e
l i b r i u m . U s i n g T e r z a g h i ' st h e o r y o f c o n s o l i d a t i o n w
can estimatethe value of /.,.,:

D)
t,,,,: 4 -

(1,,

unclrainedshear tests on soils. Direct shear tests are

sr"ritahlcfirr drained shcar tests on soils, but not firr
tests. Drair.redclirecttcsts are perlbrrncd Llsr-rndrained
i n g t h i n s p e c i r n e n s o t h a t 1 ) i s s r n a l l ,a n d u s i n g a s l t l w
ratc o['shearing so that the specimen is drainecl
throughout thc tcst. Direct shcar tests are not good fbr
undrainedtests.becausethe only way to preventdrainagc is to apply the loads very quickly, which can rcsult
in higher r.ncasuredstrength due to strain rate efl'ects.
Triaxial tests are better sr-ritcdto undrained testing in
thc laboratory,bccausedrainagccan be preventcdctlntp l e t e l y b y s c a l i n g t h e t e s t s p c c i m e n si n i m p e r m e a b l e
lletrbranes. [Jndrained triaxial tests can therefirre be
perfbrrnedslowly enough to eliminate undesirablerate
ell'ects.and still be undraincd.

( - 1I.)

where /,,.,is the time required for 994/r,of the equilibrium volume change,D the greatestdistancethat water
must travel to flow out of the soil mass (length units),
and r:,.the coefhcient of consolidation(length squared
p e r u n i t o f t i m e ) . F o r t h e t e s t s p e c i m e ni n F i g u r e 3 . 1 ,
D w o u l d b e h a l f t h e s p e c i m e nt h i c k n e s sa. b o u t 1 . 0c m ,
and c,,would be about 2 cm)/h ( l9 ft':lyr). Using these
numbers. we would estimate that /,,.,would be 2.0 h.
One second after the new loads were applied, the test
specimenwould be undrained.Afier 2 hours or longer,
the test specimenwould be drained.
Parenthetically,it should be noted that the use of
the direct shear test as an example of drained and undrained conditions is not meant to indicate that the
direct shear apparatus is suitable fbr both drained and

Recapitulation
. The difl'erencebetweenundrainedand drained
conditionsis time.
. Undrainedsignifiesa conditionwherechangesin
loadsoccur more rapidly than watercan flow in
increaseor
or out of the soil. The pore pressures
to the changesin loads.
decrease
in response
. Drained signifiesa conditionwhere changesin
load are slow enough,or remain in place long
enough,so that wateris ableto flow in or out of
the soil, permittingthe soil to reach a stateof
equilibriurnwith regard to water flow. The pore
pressures
in the drainedconditionare controlled
by the hydraulicboundaryconditions,andareunaffectedby the changesin loads.

21

ToTAL AND EFFECTIVE STRESSES

TO'TAL AND EFFECTIVE STRESSES

.Ii'e.rs is delined as force per unit area. Total s/ress is
:hc sum of all fbrces, including those transmitted
.hrough interparticle contacts and those transmitted
:hrough water pressures,divided by the total area.Total
ri'r,c includes both the area of voids and the area of
.rrlid.
Effective s/ress includes only the forces that are
:ransmitted through particle contacts. It is equal to the
:,rtal stressminus the water pressure.The total normaL
'ri?.r.ron the potential shear plane in the test specimen
n Fi-eure3.1 is equal to

W+P

(.3.2)

.,.hereo is the total stress(force per unit of area);W

:hc weight of the upper half of the specimen,porous
.tone,metalplate,and the steelball throughwhich the
.rradis applied;P the appliednormalload (F); and A
:he total area(L2).For a typical direct shearapparatus,
.rith a 102-mmr(4-inr) shearbox, W would be about
] ] . 4 N ( 2 . 8r b ) .
Before any load is applied to the specimen(when
P : 0), the normalstresson the horizontalplaneis

*': #.,,*l\:

1 . 2k P a

(3.3)

The eff-ectivestressis equalto the total stressminus

lhe water pressure.Considerthe condition before any
load is appliedto the specimen(when P : 0): If the
\pecimenhas had enoughtime to come to a drained
;ondition, the waterpressurewould be hydrostaticand
rts value would be governedby the depth of water in
rhereservoiraroundthe shearbox. For a typical direct
.hear apparatusthe depth of water (h in Figure 3.1)
rrouldbe about2 in. (about0.051m). The correspondrng hydrostaticwater pressureat the level of the horjzontalplanewould be

where oi. is the initial eff'ectivestress.If a load P :

200 N is appliedto the specimen,the changein normal
stresswould be
Ao:

200N :
19.4kPa
0 . 0 1 0m
3,

and the total stressafter the load is appliedwould be

(r:

(r,.+ Arr - 1.2kPa + 19.4kPa: 20.6kPa

(3.7)

The valuesof total stressare definedwithout ref'erence to how much of the fbrce might be carried by
contactsbetweenparticlesor to how much is transmitted throughwaterpressure.Total stressis the same
fbr the undrainedand drainedconditions.The valueof
total stressdependsonly on equilibrium;it is equal to
the total of all normal forcesdivided by the total area.
When the load P is appliedrapidly and the specimen
The specimen
is undrained,
the porepressure
changes.
is confined within the shearbox and cannot defbrm.
The clay is saturated(the voids are lilled rvith water),
so the volume of the specimencannotchangeuntil
water flows out. In this condition. the addedload is
The soil
carriedentirelyby increasedwater pressure.
skeleton(the assemblageof particlesin contact with
one another)does not changeshape,does not change
volume,and carriesnone of the new appliedload.
Under theseconditionsthe increasein water pressureis equal to the changein total stress:
L , u: L , o : 1 9 . 4k P a

- ou - u, - 1.2kPa - 0.5 kPa : 0.7 kPa

rr,i,,

(3.5)

(3.8)

whereAa is the increasein waterpressuredue to the

changein load in the undrainedcondition.The water
pressureafter the load is appliedis equalto the initial
water pressureplus this changein pressure:
L t : u o * A u : 0 . 5k P a + 1 9 . 4k P a : 1 9 . 9k P a

(3.e)

it,,: y,,h: (9.81kN/mr.110.051

m,;: 0.-5kPa (.34)
n here a,,is the initial water pressurein the specimen.
y,, the unit weight of water : 9.81 kN/m3, and li the
heightof waterabovethe horizontalplans : 0.051m.
With o : 1.2 kPa and nn - 0.5 kPa, the effective
\rfessis equalto 0.7 kPa:

(3.6)

The effective stress is equal to the total stress [Eq.

(3.7)l minus the water pressure[Eq. (3.9)]:
o' : 20.6kPa -

1 9 . 9k P a : 0 . 7 k P a

(3.10)

Becausethe increasein water pressurecaused by the

200-N load is equal to the increasein total stress,the
efl'ective stress does not change.

22

sort- MECHANICS PRINCIPLITS

The effective stress after the load is applied [Eq.

(3.10)l is the same as the efl'ective stress before the
l o a d i s a p p l i e d t E q . ( 3 . 5 ) 1 .T h r s i s b e c a u s et h e s p e c i men is undrained. Water does not have time to drain
as the load is applied. so there is no volume change in
the saturatedspecimen. As a result, the soil skeleton
does not strain. The load carried by the soil skeleton,
which is rneasuredby the value of efl'ectivestress,does
not change.
If the load is maintained over a period of time,
clrainagewill occur. and eventually the specimen will
be drained. The drained condition is achieved when
there is no difl-erencebetween the water pressuresinside the specimen (the pore pressure) and the water
pressure outside. governed by the water level in the
reservoir aror-rndthe direct shear apparatus.This condition will be achieved(tbr practical purposes)in about
2 hours and will persistuntil the loird is changedagain.
Afier 2 hours the specimen will have achieved 997o
equilibrium, the volume change will be essentially
complete,and the pore pressureon the horizontal plane
will be equal to the hydrostiitic head at that level. tt :
0.5 kPa.
In this drained condition the ell'ectivestressis

c' : 20.6kPa - 0.5 kPa : 20.1kPa (3.11)

andall of the 200-Nloadis carriedby the soil skeleton.
Recapitulation
. TotaL s/res.r is the sum of all forces. including
those transmitted through particle contacts and
those transmittedthrough water pressures.divided
by total area.
. E/|et'tive .s/rz.ssis equal ttt the totirl stress tninus
the water pressure. It is the torce transmitted
through particle contacts,divided by total area.

DRAINED AND UNDRAINED SHEAR

STRENGTHS
Sl'terrrstrenglft is defincd as the maximum value of
shear stress that the soil ciin withstand. Thc slteur
.rr,?.r.ron the hclrizontalplane in the direct shear test
specirnenin Fi-gure3.1 is equal to the shear firrce clivided bv the area:

r-i

( 3 .I 2 )

The shear strength of soils is controlled by effective stress.whether tailure occurs under drained or unilrained conditions. The relationship between shear
strength and eflective stress can be represented by a
Mohr-Coulomb strengthenvelope,as shown in Figure
3.2. The relationshipbetween r and o' shown in Figure
3.2 can be expressedas
r' : c' + ,rLtan(b'

(3.13)

where c'' is the efl'ective stress cohesion. oi, the efl'ective stress on the failure plane at failure, and rf' the
efTectivestressangle of internal friction.
Sources of Shear Strength
I1 a shear load 7 is applied to the test specimenshown
i n F i g u r e 3 . 1 , t h e t o p o f t h e s h e a rb o x w i l l m o v e t o
the lefi relative to the bottom of the box. If the shear
load is large enough, the clay will fail by shearingon
the horizontal plane, and the displacementwould be
very large. Failure would be accompaniedby development of a ruptr-rrezone, or break through the soil,
along the horizontal plane.
As the upper half of the specimenmoved to the lefi
with respect to the lower half and the stren-ethof the
soil was mobilized, the particles within the rupture
zone would be displaced fiom their original positions
relative to adjacentpalticles. lnterparticlebonds would
be broken. some individual particles would be broken'
particles would rotate and be reoriented into new positions" and particles would slide across their contacts
with nei-ehboringparticles. These movements of the
particles would be resistedby the strengthof interparticle bonds, by fiictional resistanceto sliding, and by
fbrces fiom adjacent particles resistin-edisplacement
ancl reorientation. These types of resistanceare the
sourccsof shear strength in soils.
The two most inlportant factors -eoverning the
strengthsof soils are the magnitudeof the interparticle
contact firrces and the density of the soil. Larger interparticle contact fbrces (larger values of efl'ective
stress)and higher densitiesresult in higher stren-eths.
As r increases.the shear displacement(A"r) between
the top and the bottom of the shearbox would increase.
as shown in Ftgure 3.3. This sheardisplacementresults
from shear strains in the rupture zone. The shear displacementsin direct sheartestscan be measuredeasily.
but shear strains cannot be determincd. because the
thickness of the shear zttne is not known. While the
clirect shear tcst can be used to measllre the shear
slrengths of soils. it providcs only qr-ralitativcinlbrrnation about strcss strain bchavior. It is ptlssible tcr
c l e t e r r n i n ew h c t h e r s o i l s a r e d u c t i l e ( s h e a r r e s i s t a n c e

DRAINED AND UNDRAINED SHEAR S'fRENGTHS

23

F o rp r e s s u r e>sp r e c o n s o l i d a t i o n
p r e s s u r e ,n v e l o peex t e n d b
sack
t h r o u g ho r i g i n( c ' = 0 )

(D

H i g hp r e c o n s o l i d a t i o n

o
(g
q)
c

Effectivestress- o'

la)
E n v e l o p e sa r e s l i g h t l y
c u r v e o ,m o r e s o a t
h i g h e rd e n s i t y

Q loo""

c)
@
$
C)
g

Effectivestress- o'

(b)
Figure3.2 Ellbctive stressshearstrengthenvelopes:(a) for clay; (b) fbr sands,gravels.and
rockfill.

femains high afier failure) or brittle (shear resistance

decreasesafter tailure).
Drained Strength
Drained strength is the strength of the soil when it is
loaded slowly enough so that no excesspore pressures
are induced by applied loads. In the tield. drained con.litions result when loads are applied slowly to a mass
trf soil, or where they persist fbr a long enough time
.o that the soil can drain. In the laboratory, drained
conditions are achieved by loading test specimens
.lowly so that excesspore pressuresdo not develop as
the soil is loaded.
Imagine that the direct sheartest specimenshown in
Figure 3.1 reacheda drained condition under the load
of 200 N and was then loaded to failure by increasing

Z slowly so thatexcessporepressures
did not develop.
As shown by Eq. (3.11).the eff'ectivestresson the
horizontalplaneat equilibriumunderthe 200-N load
would be 20.1 kPa. and it would remainconstantas
the clay was shearedslowly.
The strengthof the specimen
canbe calculated
using
Eq. (3.13). If the clay is normally consolidated,
c'
would be zero. The value of rf' would probablybe
between25 and 35" tor normallyconsolidated
sandy
or silty clay. As an example,supposethat @' is equal
to 30". The drainedstrengthof the clay would be
s:

c ' t c i , t a n6 ' : 0

+ ( 2 0 . 1 ) ( 0 . 5 8 )1: 1 . 6k P a
( 3 .1 4 )

wherec' : 0 and tan 6' : tan 30" : 0.-58.

24

sott- MECHANTCS PRTNCIPLES

I

+l

tll

t<-

Duclile
(."gg'"9
_/ _
/-

a
a

Brittle
(sensitive)

a
(g
q)
-c

normally consolidatedclays tend to compresswhen

sheared.
clay is one that was subjected
An overconsolidated
previously to higher eff'ectivestressand thus has a
higherdensitythanthat of a normallyconsolidatedsoil
at the sameeffective stress.As a result, overconsolidated soils compressless when shearedthan do norsoils,or if the previousmaximum
mally consolidated
effective stresswas much higher than the effective
stressduring shearing,the clay will dilate.
Pore PressureChangesDuring UndrainedShear

- Ax
Horizontaldisplacement
Figure 3.3 Shear stress-sheardisplacementcurvesfor direct
shear test.

Volume Changes During Drained Shear

Whether a soil tends to compress or dilate when
sheared depends on its density and the effective stress
that confines it. In dense soils the particles are packed
tightly together, and tight packing results in a great
deal of interference between particles when they move
relative to one another. In very dense soils, particles
cannot move relative to each other unless they ride up
over each other. which causes dilation.
Higher effective stressestend to prevent dilation, because work is required to cause the soil to expand
against the effective con{ining pressure.If the eft-ective
confining pressure is high enough, the soil may not
dilate. Instead,as shearingtakes place, individual particles will be broken.
ln soils with low densities the soil particles are farther apart on average, in a loose assemblage.As a
loose soil is sheared,particlestend to tall into the gaps
between adjacentparticles.and the volume of the soil
decreases.
The lower the density and the higher the efl'ective
stress,the more likely the soil is to compress when
sheared. Conversely, the higher the density and the
lower the confining pressure,the more likely the soil
is to dilate. In clays, density is governed primarily by
the hi,ehestef-fbctivestressto which the clay has been
sub.jected.
A tnrntully consoliduted soll is one that has not
been subjected to an efl'ective stress hi-{her than the
present efl'ective stress, and its density is the lowest
possible for any given efl'ective stress. As ii result,

The tendencyof normally consolidatedand lightly ovclays to compresswhen shearedresults

erconsolidated
in increasedpore pressureswhen shearstressesincreaseunder undrainedconditions.The tendencyof
soils to dilate when sheared
heavily overconsolidated
resultsin negativechangesin pore pressureswhen
shear stressesincrease under undrained conditions.
Thus, when clays are shearedunder undrainedconditions, the effectivestresson the potentialfailure plane
changes,becoming lower in normally consolidated
soils.
soils and higherin heavilyoverconsolidated
UndrainedStrength
Undrained strength is the strengthof the soil when
loaded to failure under undrainedconditions.In the
field, conditions closely approximatingundrained
condtionsresult when loads are applied to a massof
soil faster than the soil can drain. In the laboratory,
undrainedconditionsareachievedby loadingtestspecimens so rapidly that they cannotdrain, or by sealing
them in impermeablemembranes.(As noted previously, it is pref'erableto control drainagethrough the
use of impermeablemembranesratherthan very hi-eh
ratesof loading,to avoid high strainratesthat are not
representative
of lield conditions.)
Imaginethat the direct sheartest specimenshownin
Figure3.1 reacheda drainedconditionunderthe load
of 200 N and was then loadedto failure by increasing
Z rapidly.As shownby Eq. (3.11),the effectivestress
on the horizontalplaneat equilibriumunderthe 200N load. befbrethe shearload was increased,would be
20.I kPa. The pore pressurebeforethe shearload was
increased
would be 0.5 kPa.as shownby Eq. (3.4).
As the shearload 7 was appliedwithout allowing
tirne fbr drainage,the pore pressurewould increase,
under the
becausethe clay is normally consolidated
stress.As the shearload I is in20.1 kPa ef1-ective
creased,
the pore pressurewithin a specimenof a typclay undertheseconditions
ical normallyconsolidated
would increaseby aboutl2 kPa,and the efl'ectivenormal stresson the tailure plane at tailure (oi) would

DRAINED AND UNDRAINED SHEAR STRENCTHS

decreaseby the sameamount.The effectivestresson

rhe lailure planeat failure would thus be equalto 20.1
kPa - 12 kPa, or about 8 kPa. The undrainedshear
:trengthof the clay would thus be about4.6 kpa:
.r: c'* o'tanS':0

+ (8.0X0.58):4.6kpa

(3.1s)
Figure 3.4 shows the stresspaths and shear strengths
lbr drained and undrained failure of the direct shear
test specimen. The drained stress path is vertical, corresponding to an increase in shear stress and constant
:ftective normal stress on the horizontal plane. The
,rndrained stresspath curves to the left, as the increase
in shear stress is accompanied by a decrease in eff-ectire normal stressdue to the increasein pore pressure.
As is typical fbr normally consolidated clays, the
undrained strength is lower than the drained strength.
This is due to the fact that the pore pressure increases
.rnd the effective stress decreases during undrained
.hear. For very heavily overconsolidated clays, the re',erse is true: The undrained strength is greater
than the
,lrained strength, because pore pressure decreasesand
:tfective stress increases during undrained shear.
\trength Envelopes
itrength envelopesfor soils are developed by perfbrmng strength tests on soils using a range of pressures
,,nd plotting the results on a Mohr stressdiagram, as
.jrown in Figure 3.5. Both eff-ectivestress and total
.tfess strength envelopes can be developed. The
.trength envelopesshown in Figure 3.5 are represen.rtlve of the results of tests on undisturbedspecimens
i clay. all trimmed fiom the same undisturbed sample
.nd therefore all having the same preconsolidation

Undrained ,
stresspath-/
10
15
20
Effeclivestress- 6' (kPa)

a
a

E
a
(6
c)
c

Totalstress(undrained)

c'
I

Effectivestressor total stress- o' or o

Figure 3.5 Drained and undrained strength envelopes for

saturated clav.

pressure. The effective stress envelope represents the

fundamental behavior of the clay, because the strength
of the clay is controlled by effective stressand density.
The total stress envelope reflects the pore pressures
that develop during undrained shear as well as the fundamental behavior in terms of efl'ective stresses.
EfTective stress strength envelopes fbr clays consist
of two parts. At high stressesthe clay is normally consolidated, and the high-pressure part of the envelope
extends back through the origin. At low stressesthe
clay is overconsolidated.The strengthenvelopein this
range of pressuresdoes not extend through the origin.
The values of the effective stress shear strength parameters c' and @' depend on whether the clay is normally consolidatedor overconsolidated.If the clay is
testedin a range of pressureswhere it is normally consolidated, c'' is zero and g' is constant.If the clay is
tested in a range of pressureswhere it is overconsolidated, c' is greater than zero and $' is smaller than the
normally consolidatedvalue. Becausethe values of r''
and @' that characterize the strength of the clay depend
on the magnitude of the stressesin relation to the preconsolidationpressure.it is important that the range of
stressesused in laboratory strength tests should correspond to the range of stressesinvolved in the problem being analyzed.
The total stressenvelope is horizontal. representing
shear strength that is constant and independentof the
magnitude of the total stressused in the test. This behavior is characterizedby these relationshios:

25
.t//

(3.16a'1

(b,,- 0

(3.r6b)

{t :

I igure 3.4 Drained a n d u n d r a i n e d s t r e s s p a t h s a n d s h e a r

:.'ncths.

26

sorl

MECHANICS PRINCTPLES

where c: is the total stress cohesion intercept, s,, the

undrained shear strength, and @,,the total stressfiiction
angle.
The shear strength is the same for all values of total
normal stress because the clay is saturated and undrained. Increasing or decreasing the total normal
stressresults only in a change in pore pressurethat is
equal in n-ragnitudeand opposite in sign to the change
in normal stress.Thus. the effective stressis constant.
and becausethe eff-ectivestressis constant the strength
is constant becausestrength is controlled by etfective
strcss. Although strength is controlled by ef-fbctive
stress.it is more convenientfbr some purposesto Llse
the total stress envelope and the corresponding total
stress parameters. Use of efI'ective and total stress parameters in stability analysesis discussedlater in the
chapter.
If the clay were only partly saturated,the undrained
strength envelope wor"rldnot be horizontal. Instead, it
would be inclined and shapedas shown in Figure 3.6.
As the total normal stressincreases.the strength also
increases.becausechangesin total stressdo not cause
equal increasesin pore pressure.As the total stress
applied to a partly saturated specimen is increased,
both the pore pressure and the eff'ective stressincrease.
This occurs because.with both water and air in the
voids of the clay, the pore fluid (the mixture of water
and air) is not incompressible,and only part of the
added total stress is carried by the pore fluid. The balance is carried by the soil skeleton,which results in an
increase in eflective stress.
How much of a change in total stress is borne by
the change in pore pressure and how much by change
in effective stress depends on the degree of saturation
of the clay. At degrees of saturation in the range of
10% and lower, the change in pore pressureis negli-

Effectivestress
(drained
and

gible and virtually all of the change in total stress is

reflected in change in efI'ective stress. At degrees of
saturation approaching l00Vo, the opposite is true: Virtually all of the change in total stress is reflected in
change in pore pressure, and the change in efl'ective
stressis negligible. This behavior is responsiblefor the
curvature of the total stressenvelope in Figure 3.6: The
degree of saturation increases as the total confining
pressure increases. Therefore, at low values of total
stress. where the degree of saturation is lower, the envelopc is steeper because changes in e1I-ectivestress
are a larger portion of changes in total stress. At high
values of total stress,where the degreeof saturationis
higher, the envelope is flatter because changes in effective stress are a smaller portion of changes in total
StTCSS.

Recapitulation
. Sheor strensthis dellned as the maximum shear
stressthat the soil can withstand.
. The strength of soil is controlled by efl-ective
stresses.
whetherfailure occursunderdrainedor
undrainedconditions.
. Drained strengthis the strengthcorespondingto
failure with no changein effective stresson the
tailureplane.
. Undruinedstrengthis the strengthcorresponding
to tailurewith no changein watercontent.
. EfTectivestressstrengthenvelopesrepresentfundamentalbehavior.becausestrengthis controlled
by eftectivestressand density.
. Total stress strength envelopesreflect the pore
pressuresthat developduring undrainedshear,as
well as fundamentalbebaviorin termsof ef-tbctive
SITCSS.

. Total stressstrengthenvelopesfor saturatedclays

are horizontal,correspondingto c : s,,,6,, - 0.
Total stressenvelopesfor partly saturatedclays
are not horizontal,and @,,is greaterthan zero.

BASIC REQUIREMENTSFOR SLOPE STABILITY

ANALYSES

o
o
o
c
a

E f f e c t i v es t r e s so r t o t a ls t r e s s- a ' o r o

l'igure 3.6 Draincd and undrained strength envelopcs firr

oartlv saturatcdclav.

Whether slope stability analysesare perfbrmed fbr

drainedconditionsor undrainedconditions,the most
basicrequirementis that equrlibriummustbe satisfied
All body fbrces(weights;,
in terms of total stresses.
and all externalloads,includingthose due to water
must be inpressures
acting on externalboundaries,
providetwo usecludedin the analysis.Theseanalyses
ful results: ( l) the total normal stresson the shear

BASIC REQUIREMENTS FoR SLoPE STAB]I-ITY ANAIJSES

\Lrrface and (2) the shear stress required for equilibrium.

The foctor of ,safetl,fbr the shear surface is the ratio
of the shear strength of the soil divided by the shear
;tress required for equilibrium. The normal stresses
along the slip surface are needed to evaluate the shear
stren-eth: Except fbr soils with @ : 0, the shear
strength depends on the normal stress on the potential
plane of failure.
In efl-ectivestress analyses,the pore pressuresalong
the shear surface are subtracted from the total stresses
to determine efl-ectivenormal stresses.which are used
to evaluate shear strengths. Therefilre, to perfbrm ef't'ectivestressanalyses,it is necessaryto know (or to
estimate) the pore pressuresat every point along the
shear surface. These pore pressures can be evaluated
rvith relatively good accuracy fbr drained conditions,
rvhere their values are determined by hydrostatic or
steady seepage boundary conditions. Pore pressures
can seldom be evaluated accurately for undrained
condtions, where their values are determined by the
fesponseof the soil to external loads.
In total stressanalyses,pore pressuresare not subtracted fiom the total stresses,because shear strengths
are related to total stresses.Therefbre. it is not necessary to evaluate and subtract pore pressuresto perfbrm
total stressanalyses.Total stressanalysesare applicable only to undrainedconditions.Thc basic premise of
Iotol stressanctlt,sisis this: The pore pressuresdue to
LrndrainedIoading are deterrnineciby the behavior of
the soil. For a given value of total stresson the potential failure plane. there is a unique value of pore pressure and therefbre a unique value of efl'ective stress.
Thus, although it is true that shear strength is really
controlled by el1'ectivestress,it is possible fbr the untlrained condition to relate shear strength to total norrnal stress.becauseeffbctive stressand total stressare
Lrniquelyrelated for the undrained condition. Clearly,
this line of reasoning does not apply to drained conditions. where pore pressuresare controlled by hydraulic boundary conditions rather than the response
o l ' t h e s o i l t o e x t e r n a ll o a d s .
.{nalyses of Drained Conditions
Druined contlitions are those where chanses in load
are slow enough,ur where they have been in place long
cnough, so that all ol the soils reach a stateof equilibrium and no excess pore pressuresare caused by the
loads. In drained conditiclns pore pressuresare contlolled by hydraulic boundary conditions. The water
n'ithin the soil may be static. or it ntay be seeping
steadily.with no change in the seepagcover tinte and
no increaseor decreasein the amount of water within

21

the soil. If these conditions prevail in all the soils at a

site, or if the conditions at a site can reasonably be
approximated by these conditions, a drained analysis
is appropriate. A droined anall-sis is perfbrmed using:
. Total unit weights
. EfTective stress shear strength parameters
. Pore pressuresdetermined from hydrostatic water
levels or steady seepageanalyses
Analyses of Undrained Conditions
Undrained conditions are those where changes in loads
occur more rapidly than water can flow in or out of
the soil. The pore pressures are controlled by the behavior of the soil in responseto changes in external
loads. If these conditions prevail in the soils at a site,
or if the conditions at a site can reasonably be approximated by these conditions, an undrained analysis is
appropriate. An undruined unul\si.v is performed usin-t:
. Total unit weights
. Total stress shear strensth Darameters
How Long Does Drainage Take?
As discussed earlier, the di1l-erencebetween undrained
and drained conditions is time. The drainage characteristics of the soil mass, and its size, detennine how
long will be requrredfor transition fiom an undrained
to a drained condition. As shown by Eq. (3.1):

,D.
t u , , -4 -

(3.11)

L7,

where /,,,,is the tin-rerequired to reach 99c/cof drainage

equilibrium, D the length of the drainagepath, and r',,
the coefficient of consolidation.
Values of c. fbr clays vary fiom about 1.0 crn:/h
(10 firlyr) to about 100 times this valr.re.Values of r',,
tbr silts are on the order of 100 times the values fbr
clays. and values of c,. fbr sands are on the order ol'
100 times the valuesfor silts, and higher.Thcsc typical
values can be used to develop surne rou_ehideas o1'the
lengths of time required to achieve drained conditions
in soils in the lleld.
Drainage path lengths are related to layer thicknesses.They are half the layer thicknessfbr layers that
are bounded on both sides by l.norc permeablc soils.
and they are equal to the layer thicknessfor layers that
are drained only on one side. Lenses or layers of silt
or sand within clay layers provide internal drainage.
reducin-ethe drainagepath length to half ol'the thickness between internal drainace lavers.

28

sorr- MECHANTcsPRINCTPLES

Values of /.,.,calculated using Eq. (3.17) are shown

in Figure 3.7. For rnost practical conditions, many
years or tens of years are required to reach drainage
equilibrium in clay layers. and it is usually necessary
to corrsiderundrainedconditions in clays. On the other
hand, sand and -9ra'',elslayers almost always reach
drainage equilibrium quickly, and only drained conditions need be consideredfbr these materials.Silts fall
in between sandsand clays, and it is often difficult to
anticipate whether silt layers are better approximated
as dritined or undrained.When there is doubt whethcr
a l a y e r w i l l b e d r a i n e do r u n d r a i n c d ,t h e a n s w e r i s t o
analyze both conditrons, to cover the range of possibilities.

analysis of the clay fbundation. There is no problem

with perfbrming a single analysis in which the embankment is considered to be drained and is treated in
terms of effective stresses.and in which the fbundation
is consideredto be undrained and is treated in terms
of total stresses.
As discussed earlier, equilibrir-rmin terms of total
stresses must be satisfied tor both total and efl-ective
stressanalyses.The only difl'erencesbetweentotal and
efl'ective stressanalysesrelate to the strength pararncters that arc uscd and whether pore pressuresare specified. In the case o1' short-terrn analysis ol' a sand
embankmcnt on a clay fbundation, thc strength of thc
sand would be charircterized in terms ol' ell'ective
s t r e s s e s( b y a v a l u e o l ' r f ' l b r
t h e s a n d ) ,a n d t h e
strcngthof the clay would bc charactcrizcdin tcrrls of
t o t a l s t r c s s c s( b y v a l u c so 1 ' s , ,- r ' v a r y i n g w i t h d c p t h ,
wilh d-l,,: 0 lbr a saturatedclay).
Porc pressureswould be specifiedlirr thc sand rf thc
water table was above the top of the clay or if there
was seepagethlough the embankrnent.but pore pressureswould not be specilieclfirr tlre clay. There woLrld.
<lf course. be pore pressLrres
in the clay. However, bec a u s ct h c s t r c n g t ho 1 ' t h c c l a y i s r c l a t c dt o t o t a l s t r c s s ,
it would be unnecccsaryto specify these nonzero valucs. Becausc lnost computcr programs subtract porc
p r e s s u r e sw h e n t h e y a r e s p e c i f r e d .s p e c i l y i n g p ( ) r e

Short-Term Analyses
Shrtrl ternr ref'ers to conclitions dr-rringrlr lirllowing
construction-thc time irlrnediately fbllowing the
changc in load. For exanrple. if constructins a sand
e r n b a n k n r e not n a c l a v l i r u n d a t i o nt a k e s t w o r r o n t h s ,
the short-terrnconclitionfirr the enrbankntcntwoulcl bc
t h e e n c l o f c o n s t r u c t i o n .r l r t w o l n ( ) n t h s .W i t h i n t l t i s
p c r i o d o l ' l i r n c . i t w o u l d b e a r e a s o n a b l ea p p l o x i n t a t i o n
that no drainage wor-rldoccur in the clay firundation,
whcrcas thc sand crnbankrnentwould bc I'ully clraincd.
F o r t h i s c o n d i t i o n i t w o u l d b c l o g i c a l t o p e r f i r r ma
d r a i n e da n a l v s i so l ' t h e e r n b a n k r n e nat n d a n u n d r a i n e d

D r a i n a g ep a t h l e n g t h- m
10

Years
10,000
1,000

10

f-q2+

'100
Months

10 100
1

3.0

Days
100

"*49

1 0 Ho u r s
100
1

I ^.,iut'
---f-

a$"

Drainagepath length- ft
Figure 3.7 Time required fbr' drainage of' soil deposits {1.,.,based on Terzagh's theory ol'
consolidation).

BASIC REQUIREMENTS FOR ST,OPE STABILITY ANALYSES

pfessuresfor soils that are being treatedas undrained

.-anresultin errors.Therefore,for soils that are treated
pore pressuresshouldbe set
rn terms of total stresses,
rr) zero,eventhough,in fact, they are not zero. (In the
particularcaseof $ : 0, no enor will result if pore
pressuresare not specifiedas zero, becausestrengths
.rre independentof normal stress,and misevaluating
normal stress does not result in strengthsthat are
nrongly evaluated.)
Externalwaterpressuresactingon the surfaceof the
tbundationor the embankmentwould be specifiedfor
hoth materials,becauseexternalwater pressuresare a
romponentof total stress,and they must be included
to satisfyequilibrium in terms of total stress.
Long-TermAnalyses
.\fier a periodof time. the clay foundationwould reach
.i drainedconCition,and the analysisfbr this condition
ri ould be perfbrmedas discussedearlierunder "Anali:es of Drained Conditions,"becauselong term and
.lruined conditions cany exactly the same meaning.
Both of thesetermsrefer to the conditionwheredrain:rveequilibriumhas beenreachedand thereare no exJesspore pressuresdue to externalloads.
For the long-termcondition,both the sandembanknrent and clay foundationwould be characterizedin
ic'rrnSof eff-ectivestresses.
Pore pressures,determined
irom hydrostaticwater levels or steadyseepageanali ses.would be specifiedfor both materials.External
',\ater pressureson the surfaceof the foundation or
.'rrbankmentwould be specihedfor both materials,as
.rlrvays;thesemust be includedto satisfyequilibrium
ln (ermsof lotal stress.

Failure
Progressive
One of the fundamentalassumptionsof limit equilibrium analysesis that the strengthof the soil can be
mobilized over a wide range of strains,as shown by
the curve labeled"ductile" in Figure 3.3. This implicit
assumptionarisesfrom the fact that limit equilibrium
analysesprovide no information regardingdeformations or strains.
Progressivefailure is a strongpossibilityin the case
of excavatedslopes in overconsolidatedclays and
shales, particularly stifT-fissuredclays and shales.
These materialshave brittle stress-straincharacteristics, and they contain high horizontal stresses,often
higher than the vertical stress.When an excavationis
madein stiff fissuredclay or shale,the excavatedslope
reboundshorizontally,as shown in Figure 3.8. Finite
elementstudiesby Duncanand Dunlop(1969),(1910)
showedthat shearstressesare very high at the toe of
the slope,and there is a tendencyfbr failure to begin
at the toe and progressback beneaththe crest. as
shownin Figure3.8.
Immediatelyafter excavationof the slope (at time
/r), the stresses
at point A mightjust havereachedthe
peakof the stress-displacement
curve,and the stresses
at pointsB and C would be lower.With time, the slope
would continueto reboundinto the cut. due to a delayed responseto the unloadingfiom the excavation,
and possiblyalso due to swellingof the clay as its
water content increasesfollowing the reduction in
stress.At a later time (r.), therefore,the displacements
at A, B, and C would all be larger,as shownin Figure
3.8. The shearstressat point A would deceaseas it
moved beyond the peak, and the shear stressesat

Excavatedslope
O v e r c o n s o l i d a t ec dl a v

1'::i:>
-\
B

a
F
a

d
0)
c
</)

D i s p l a c e m e n- tA x
Figure 3.8

l--_

29

z#

P o t e n t i asl l i ps u r f a c e

o
6
0)

a
o
o)

6
o
a

0)
s

(u

T i m et 3

u)
D i s p l a c e m e n- tA x

D i s p l a c e m e n- tA x

Mechanism of progressive failure of an excavated slope in overconsolidated clay

30

sott- MECHANICS PRINCIPLES

pointsB and C would increase.At a later time (r',),the

displacementat point B would be largeenoughso that
the shear stress there would fall below the peak'
Through this process, progressively,lailure would
spreadaroundthe slip surface,without evermobilizing
the peak shear strengthsimultaneouslyat all points
along the slip surface.
Becauseprogressivefailure can occur for soils with
brittle stress-strain characteristics,peak strengths
shoulclnot be usedfor thesesoils in limit equilibrium
fbr brittlesoilscan lead
usrngpeakstrengths
analyses;
of stabilassessment
unconservative
and
to inaccurate
with
slopes
experience
5,
in
Chapter
ity. As discussed
clays'
fissured
particularly
clays,
in overconsolidated
for
appropriate
are
strength,s
shows that fullv softened
ncll
have
slickensides
where
thesematerialsin cases
developed,and residuol strengthsare appropriatein
havedeveloped.
conditionswhereslickensides

RecaPitulation
. Equilibrium must be satisfiedin terms of total
stressfor all slopestabilityanalyses.
. ln el'fectivestress analyses,pore pressuresare
subtractedfrom total stressesto evaluatethe ef'fective stresseson the shearsurface.
. In total stressanalyses,pore pressuresare not
subtracted.Shear strengthsare related to total
SITCSSCS.

, The basic premiseof total stressanalysesis that

there is a uniquerelationshipbetweentotal stress
and effective stress.This is true only for undrainedconditions.
' Total stressanalysesare not applicableto drained
conditions.
' The time requiredlbr drainageof soil layersvaries from minutesfor sandsand gravelsto tens or
hundredsof yearslbr claYs.
r In short-termconditions,soils that drain slowly
may best be characterizedas undrained.while
soils that drain more quickly are best characterized as drained.Analysesof such conditionscan
be performed by using effective stressstrength
parameterstor the drained soils and total stress
strengthparametersfbr the undrainedsoils.
. When eff'ective stress strength parametersare
used, pore pressuresdeterminedfiom hydraulic
bounclaryconditions are specified.When total
stressstrengthparametersare used,no pore pressuresare specified.
. An implicit assumption
of limit equilibriumanalyses is that the soils exhibit ductile stress-strain
behavior.Peak strengthsshould not be used for
materialssuch as stiff fissuredclays and shales
which have brittle stress-straincharacteristics'
becauseprogressivetailure can occur in these
materials.Using peak strengthscan result in
evaluationsof stainaccurateand unconservative
bilitv.