See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/348339684

Using limit equilibrium concepts in finite element slope stability analysis

Conference Paper · November 1999

CITATIONS READS

28 89

2 authors, including:

Delwyn D. Fredlund
University of Saskatchewan
644 PUBLICATIONS   25,283 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

PhD Research Project View project

Unsaturated soils View project

All content following this page was uploaded by Delwyn D. Fredlund on 08 January 2021.

The user has requested enhancement of the downloaded file.

 Proceedingsof the InternationalSymposiumon Slopestability Engineering-IS-
shikoku'99,InvitedKeynotepaper,Matsuyama,Shikoku,Japan.pp.3 I 47-
November 8-l l, 1999
Slop StabilityEngineeing,Yagi,Yamagani&Jiango 1999Batt<ema,
Ronedam, |SBNgOsslg O7gs

Usinglimit equilibriumconcepts
in finiteelementslopestabilityanalysis

D.G.Fredlund& R.E.G.Scoular
University of Saslatchewan, Saskatoon, Sask.,Canada

ABSTRACT: This paperreviews thc developmentof finite element slope stability analysesand proposes
that sucha methodcan form a practical procedurefor solving slope stability problems.Severalstopestabitity
methodshavebeenproposedthat make useof the finite elementmerhods;tir"i" "r" summarizedin-this paper.
The proposedfinite.e-lementmethod is in a fonn th3l-g* be conveniently used in engineeringpractici. The
procedurelends itself to Presentday numerical modelling techniques.'Ihe method tri Ueen,ipA"t"a to take
advantageofrecent advancesin computertechnologyand algorithms.
The combinationof a finite element strels analysi: with-a limit equilibrium analysisprovides greatercer-
tainty and flexibility regarding the intemal distribution of stesses within the soil'mass. The nirmal forc"
a-longany selectedslip surfacecan be calculatedfrom the sfiess distribution that has been calculateJ *i.rg "
linear and non-linear stess analysis.The overall factor of safety for a slope, when the n"it. il"m"nl method
is used,can be defined as the available shearstrcngthof the soil dividedbythe resisting slear st roeth. Th.
overallfactor of safetyis a combinationof the local factorsof safetywithin the.fop". ile.etuttine-ou"ott
-L.to,
factor of safety retains the basic assumptionsinhercnt to the limii equilibrium de'finition oi tt " of
safety.
The local facton ofsafe-tyatt an exprcssionofthe stability ofthe soil massat eachpoint along the slip sur-
face' The overall factor of-safety computedusing the finite elementmethod shows good "gt""fr"ot with the
facton of safety computedusing any one of several limit equilibrium methods.fnl nnitJ element method
provides additional information-regarding the potential performanceof a slope; information not available
when using traditional limit equilibrium methods.The results indicatethat it is'important to usethe effective
shearstength characterizationof the soil when performing the slope stability anaiysis.The computedfactor
of safetyobtainedwhen using "
_toq she{ shengthcharacterizatio;of the soil, rr"y oot agreewiih the factor
of safetycomputedwhen using the finite elementstressanalysismethod.
- K9f wgrds:qlopestability analysis,finite elem€nt,enhancedmethod direct method strengthmetho4 stess
level method,factor ofsafety, Iocal factor ofsafety.

I INTRODUCTION the stressesin the soil mass.ThesesEessescan sub-

sequentlybe usedto computea factor ofsafety (Fig.
Limit equilibrium methodsof analysis have proven 1). The completestressstatefrom the finite element
to be a widely used and successfulmethod for the analysiscan be "imported" into a limit equilibrium
assessment of the stability of a slope. Limit equilib analysiswhere the normal stressand the shearstress
rium methodssum forcesand momentsrelatedto an are computedcorrespondingto any selectedslip sru-
assumedslip surface passed through a soil mass face.
(Fredhmdand Krabn, 1975; Fredhmdet al, l98l). The objective of this paper is to demonsEatea
However,thesemethodsdo not utilize the stressver- procedure for combining a finite element stress
sus stain characteristicsof the soils involved. It is analysis on a slope with the conceptsof a limiting
well known, and intuitively rmderstoodthat the sta- equilibrium method of a'alysis. The final method is
bility of a slope should be influenced by the stress called a "finite element method of slope stability
versus strain characteristics of a soil (Kondner analysis"and the results are compapd to results ob-
1963).A finite elementanalysisutilizes a stressver- tained when using conventional limit equilibrium
sus stain model for the soils involved to calculate methodof analysis.

31
 Finite ElementAnalysisfor Stresses

Limit EquilibriumAnalysis

Figue I.Illustration showing sresses that arc "rmported" from a finite element analysis into a limit equilibriurn malysis.

2 BACKGROI.'ND surface.The normal and shearshessesfrom an elas

tic analysiswere used to calculate an overall facto
of safety.The formulationof Kulhawy (1969) wa
Bishop (1952) noted that the stressesfrom a limit classifiedas an "EnhancedLimit StrengthMethod".
equilibrium method of analysis did not agree with A number of finite element slope stability meth
the achral stresseswithin an earth stnrcture. Other ods have been proposed and the methods can b,
researchershave confirmed this observation both categorizedas "enhancedlimit methods,'or "direc
with experimental evidence and with numerical methods",as shownin Figure 2.
modelling.Ia Rochelle(1960) estimatedthe stness Wright (1969)comparedthe factorsof safetycal
conditionsin sGepslopesusing photoelastictestson culated using the "enhancedlimit strength" methor
gelatine models. The results showed that stresses with factors of safetycalculatedusing Bishop'sSim
along a slip surfacewere over-stressedin the lower plified method(1952). A slip surfacewas selecter
portion of the slip circle. Brown and King (1966) for comparativepurposesthat had a factor ofsafet
producedcritical slip surfacesfrom a finite element of 1.0whenusingthe Bishop'sSimplifiedmethod.i
strcss analysis of slopes using a linear elastic soil was concludedthat the factors of safety determiner
model. The critical slip surfaceswere produced by by the "enhancedlimit strength" method (Kulhawf
usingthe angleof obliqwty, 0, alongthe slip surface 1969)wereapproximately3% higherthanthosede
(i.e., dequal to (45o + Q/a)). Each critical slip sur- termined applyrng Bishop's Simplified method
face representeda close approximation to an essen- Wright et al. (1973), using the "enhancedlimir
tially circularshapedslip surface. sEength"method,showedthat: l) alongone third or
Cloughand Woodward(1967)trndertooka study the slip surface,the local factors of safety are lesl
to evaluate the effect of incremental loading with than the overall factor of safety, 2) the factors or
single step loading as it related to strcssesand de- safetycalculatedby the finite elementmethodusing
formations.It was concludedthat: l) stressesand de- linear elastic material propertiesrangedfrom 0% tr
formationsin an embankmentobtainedfrom a direct 4.5% higher than those calculatedusing Bishop's
application of the gravitational body forces on the Simplified method,and 3) the factorsof safetycll-
complete structtrre were not completely accurate, culated by the finite element method using non-
and 2) changing Poisson's ratio interferes with the linear elastic material properties increased witt.
relationshipbetweensEessesand displacements,re- Poisson'sratio and arc 2%oto 8% higher than those
quiring a new analysis for each case. It was con- calculatedusingthe Bishop'sSimplifiedmethod.
cluded that "meaningfttl stability analysis can be Rcsindiz (1974) ageed with the conceptof using
madeonly d the stressdistribution within the struc- the finite elementmethodto calculatethe stability ol
ture can bepredicted reliably." a slope; however, disagreedwith points No. 2 and
Kulhawy (1969) developeda compder program No.3 of the resultsof Wright et al. (1973)because
to obtain an independentassessmentof the nonnal the factor of safety differences were too small
and shearstressdistribution along an assumedslip Res6ndizhad developeda finite elementmethod oi

32
 Finile ElementSlope StabilityMethods

Direct melhods
(finiteelementanalysisonly)

Enhancedlimit methods
(finile elementanalysiswith Load increase Strengthdecrease
a limit equilibriumanalysis) to failure to failure

Definitionof Faclor of

Strength Stress Level Streng[h& Stress Level

Kulhaury1969 Zenkiewiaet al 1975 Adikariand Commins1985

. _I (c' + dta$')M
n- zrat

Figure 2. Finite element approachesproposed in comprrting the factor ofsafety in a slope stability analysis.

slopestability analysisdefinedas an "enhancedlimit havealso beenstudiedby Martins et al. (l9gl) and

stress-level"method"n 1972(Fig. 2). This method Tan andDonald(1985).
usedthe maximum principal stressdifference of the The "enhancedlimit" slope stability methods are
soil at failure to define the factor of safety.Analyses basedon stressescalculated using a finite element
made using non-linear stressversus strain relation- analysisand combinedwith a limit equilibriumnoe
shipsled to factors of safety which in all caseswere ofanalysis alonga prescribedslip surface,to d"fin"
higher (i.e., differences as large as 3Ao/o)than con-
the factorofsafety. The prescribedslip surfaceis the
ventionalfacton of safety (e.g., Ordinary methodor
one defined by the lowest factor of safety and is
Bishop'sSimplifiedmethod).
found using a trial and error procedure.The stresses
Zier/rriewiczet al. (1975) also proposeda finite along the slip surface are computed using a finite
elementmethodof analysisto computethe factor of element analysis and can either be usea in "
safetyby using the principal stess difference in the "strength"
methodor a "stress-level"method.Farias
soil at failure to define the factor of safety. The and Naylor (1996) statedthat when using the "di-
methodis an "enhancedlimit stress- level method" rect" finite elementmethod it is, ,'not easvto obtain
(Fig. 2). Both the Resdndiz(1972) aurrd, Zienktewicz a safety factor accttrate to within the'confidence
et al. (1975)formulationsareclassifiedas "enhanced Iimits achievableby linit equilibrium methods".The
limit stress-level" methods. authorsnotedthat: l) a fine meshis required,Z) a
Naylor (1982) estdblishedtwo tJpes of finite ele- code capable of giving reliable results with ihe
mentslopestabilitymethods,a "direct,'and an "en- Mohr Coulomb elasto-plastic model for toading
hancedlimit" methodof analysis.The direct method statesclose tofailure is needed,and 3) il is tuually
useda finite elementnodal formulation to define the necessaryto carry out a set of analyseswith c, and
slip surfaceand the factor of safetydirectly from the tanQ'progressively reduced by a factor which will
analysis. The proposed "direct" slope stability becomethe safetyfactor whenfailure is eventually
methoddefinedthe factor of safety either as the in- reached.'Enhancedlimit" methodsrequireonly one
creasedload necessaryto causefailure, or as the re- finite elementanalysisto calculatefactorsofsafety
ciprocal of the reduction in the strength properties for a slope with various combinationsof c, and
requiredin order to achievefailure. Thesemethods tan6'.

33
 Adilori and Cumrnins(1985) produccda finite mal stressat the base'of a slice is known. On the
elementmethodthat combinethe "sfength" and the other hand, limit equilibrium methods,sarting with
nstress-level" methods as defined by Kulhawy Bishop'sSimplifiedmethod(1955),have.uscdan
(1969) and ZierJriewiczet al. (1975), rEspectively estimatedfactor of safety when computing the nor-
(Fig. 2). The Adikari and Cummins(1985) method mal force at the baseofa slice. The final factor of
prodtrccd facton of safety that were between the safety is found through an ilerative proc€ss.The fi-
valuesobtainedwhen applyng the Kulhawy (1969) nite elementmethodfactor of safety is definedusing
and the Ziern/ri*vicz et al. (1975) methods. It was the normal and shearstressescomputedusing a fi-
notedthat for near-failureconditions(i.e., as defined nite elementanalysis.
by Bishop'sSirnplified method,1955),the value of Finite element numerical stress analyses have
the factor of safety calculated by the Adikari and been available for many years. The finite element
Cummins(1985) methodapproached1.0, while the method however,has not bccomepopular for slope
value ofthe factor ofsafety calculatedby the Zien- stability studies due to intense computational rc-
kiewicz et al. (1975) method remainedhigh. The quirementsand difticulties in assessingthe stress
factor of safetyby the Kulhawy (1969) methodalso versusstraincharacteristics ofthe soils. In addition,
approachedunity with the factor ofsafety being de- inexpensiveand easyto use limit equilibriummeth-
pendenton the percentageof the sftngth mobiliza- ods have provided factors of safety that appear to
tion in the component materials. The main differ- repr€sentfailure conditions in the field in most
ence in results appean related to using the stresses situations. Microcomputers now have sufficient
on the principal plane (Zienkiewicz et al. 1975) computationalcapacity to perform combined stress
rather than on the plane. By definition, failure does and limit equilibriumanalyses.As a result,it is an-
not occur on the plane of principal stressand there- ticipatedthat the latter procedurewill becomemore
forc, the Zier*iewicz et al. (1975) method(or any cornmonin engineeringpractice.
stressJevelmethod) is computing a factor of safety
that must be higher than the factors of safety pro-
3.1 Procedureusedfor thefinite elementanalysis
ducedby a "strength"method.
Duncanet al. (1996)provideda summaryof the The enhancedlimit (strength)finite elementmethod
limit equilibrium and finite element methodsthat proposedby Kulhawy (1969) was selectedas the
havebeenproposedfor slopestabilityanalyses. mostappropriatemethodfor slopestabilityanalysis.
The finite element stress-deformationsoftware.
Sigma/W(a proprietaryproductof Geo-Slopelnter-
3 SUGCESTEDSTTIDYFOR COMPARISON nationalLtd., Calgary,Albena, Canada),was modi-
BETWEEN TTIE FINITE ELEMENT AND THE fied to utilizes a searchalgorithm in order to assign
LMIT EQUILIBRJI.'MMETHODSOF SLOPE and transfer calculated finite element stressesto a
STABILITY ANALYSIS designedpoint on the slip surface (Bathe, 1982;
Krahn et al., 1996). The calculatedfinite element
The frnite element slope stability method proposed calculatedstressesare usedto calculatethe normal
in this paperis of the "enhancedlimit strength"tlpe and shear stresseson the slip surface. The latter
(Scoular,1997).The finite elementmethodusesthe strcssesare usedto calculate local facton of safety
Kulhawy (1969) definition for the factor of safety at the centerofthe baseofeach slice as well as the
combinedwith a finite elementstressanalysisof the overallfactorofsafety for the entireslip surface.
slope.Stressanalyseswerc doneusing Poisson'sra-
tios equalto 0.33 and 0.48. For eachstressanalysis,
3.2 DeJinitionoffacor of safety
the cohesionand the angleof internalfriction of the
soil were altered as the stability of the slope was The overall factorofsafety is definedin accordance
computed The selectedvaluesfor cohesion,c', were with the finite element slope stability method dc-
10,20 and 40 kPa,and for the angleof intemal fric- scribedby Kulhawy (1969),and expressed as the ra-
tion, Q',were 10,20 and30 degrees. tio of the sum of the incremental resisting shear
The finite clement slope stability method pro. strcngths, .S7,to the sum of the mobilized shear
ducesan overallfactorofsafety that is an expression forces,Sn',alongthe slip surface.
of the sability of the slopebasedon the calculated
- ES.
stresseswithin the slope. Slope stability problems Frrr"r=F. (l)
solved using the finite elementmethodhavetwo im- L"fr

portant distinctions from limit equilibrium methods.

The resisting force for each slice is calculatedin
First, the finite elementslope stability equationis termsof the shearstrength, a, at the centerof a slice
determinate; therefore, no further assumptionsarc multiplied by the base length of the slice, f. T}rc
rcguircd to completethe calculations.Second,the available resisting shear strength for a satu-
factor of safetyequationis linear, becausethe nor-
rated/rursaturatedsoil (Fredlundand Rahardjo, 1993)
canbe written as:
34
 s
t t
N1(-1,1)
I
I N r ( 1 ,1 )

tr
o
o
E
o LocalCoordinetes (r,s)
o
FiniteElement
;

GlobalCoordinates(x,y) x - Coordinate

Figurc3. Definitionofthe globalandlocal coordinatesfor a rectangularfinite element.

g, =sp={c'+(oo -u" )tanl'+(u" -u*)tar/b A common set of coordinatesis usedto identi$

1p
(2) the centerof a slice along a slip surfacewith respect
to the surroundingfinite element.The global coordi-
The mobilized shearforce, S,,, for each slice is natesfor the center of the baseare calculatedin or-
calculatedas the mobilized shear stress, 2., at the der to determine the location of the base center
centerof a slice multiplied by the baselenglh, B. within the slope, and to determinewhich elementis
S^ = t^f (3) associated with the centerofthe base.The local co-
ordinatesofthe centerofthe basearethencalculatcd
The local factor ofsafety is defined as the ratio of within the element that encompassesthe center of
the resistingshearforce, ,S7,at a point along the slip the base(Fig. 3).
surfacedivided by the mobilized shearforce, .Sr, at The global coordinatesfor the centerofthe base
the samepoint, of a slice are relatedto the globalcoordinatesof the
finite elementnodal poins through use of the shape
t* =LS , =;7
rB
(4) fimctions.
x=<N>{X} (5)
The resisting shear force, .S7,and the mobilized
shear force, S,,, tr€ both calculated using the
stressescomputedin the finite elementanalysis.The
normalstess, on, frd shearstress, rv1; @r! b€ 'im- Wherex = global x coordinatesfor the center
ported" as known values to the limit equilibrium ofthe baseofa slice;y = globaly coordinates for the
analysisand the definition of both the overall and centerofthe baseofa slice; {X} = globalx coordi-
localfactor ofsafety equationsare linear. natesfor the elementnodal points; {Y} = global y
coordinatesfor the elementnodal points; and <y'{> :
matrix of shapefirnctions.
3.3 ElementidentiJicationcorrespondingto the The shapefunctions <ly'> are defined in terms of .
baseof a slice the local coordinates(e s). Since the global coordi-
Eachelementmust be checkedto confirm that thc nates for the center of the base of a slice and the
ccnterof the baseof the slice is located within the nodes are known, the local coordinatescan be ob-
elcmentunder consideration.Then the stressescal- tained by solving Equations (5) and (6), simultane-
culated by the finite element analysis can be "im- ously. The shape finctions for a rectangular finite
ported"into the stabilityanalysis.Oncc the element elementwith fou nodesare as follows (Bathe 1982):
cmbracingthe centerof a portion along the slip sur-
tace ts located,stressvalues from the Gausspoints N ,= 1 1 t + r 1 t + s ; (7)
of the elementcan be transferrcdto the nodesbf the
clcmentand consequently to the centerofthe base.
t ne procedureis in accordancewith the mcthod de- N ,= - r X ,+ s ) (8)
scribcdby Bathe(19g2). i(l

35
 be usedto describethe changeof a variable within an
N ,= -r[l -s) (e)
i(l elementin terms of nodal values. The finite element
slope stability calculationsrequire that stressesat the
center of the base for each slice be within an ele-
N .: ( I + r XI - s ) (10) ment. This is achieved using the following proce-
i dure:
wher€/ ands = local coordinateswithin the element.
The localcoordinatesvary between-l and+l (Fig. {o},=<N>{F} (15)
3). A knowledgeof the local coordinatesis crucialto wherc aln = stressesat the element node; <|y'> =
identifuing the elementoverlappingthe centerof the matrix of the shapefunctions; md {F} = stressvalues
baseof a slice. By definition, an element surrounds at the Gausspoints.
the centerof the baseof a slice if the followine con- The local Gausspoint integration coordinatesarc
ditionsaremet: (0.577,0.577),however,when the local Gausspoint
For a triangular element, integration coordinatesare projected outward to the
element nodes, the local coordinates bccome
(0<r>l)and(0<s>l) (l l) (1.7320, 1.7320)Gig. 5). This projectionis carried
For a rectangularelement, out for each element and the values for the stresscs
from eachcontributing elementare averagedat each
(-l <r)l)and(-l <s>l) (12) node. Accordingly, the values of or, o, and to c,an
The centerof the baseis outsidean elementif the be computedat eachnode of the finite elementmesh.
local coordinatesare not within the above specified The nodal stresses,ar, ay, and lryy,of an elementare
ranges. The search continues until an element is transferredto the centerofthe baseofa slice along
found that satisfiestheseconditions. the slip surface.
{o}=< N >{o}, (16)
3.4 Transferof elementstressesto the centerof the
where /ol = stressesat the center of the base of a
baseofa slice
slice.
Calculated shessesare stored within the computer The stresses,a,, ay, and rry, cannow be computed
software relative to the Gausspoints of an element. at the centerofthe basefor eachslice.
SEessesmust be transferredfrom the Gausspoints of
an elementto the nodes of the element and then to
3.5 Thenormaland shear stresses
at the centerofa
the centerofthe baseofa slice.
slice
The local coordinatesof a point within a finite
elementare defined in relationship to the global co- Once the stresses,o\, oy rnd r- are known at the
ordinatesat the nodes of the element by using the center of the base for each slice, the normal stress,
shapefunctions,as per Equations(5) and (6): on, andthe mobilized shearstess, r, , can be cd-
culatedusing Equations(17) and (18), respectively
(Higdon etaL 1976):
Ix,)
x = 1 N r N 2 N 3 N 4> (t- ;) l (13) On
6r*ou or'oucos20
=------+trystnLfl
I xt I
t-l
lxt ) (r7)
o- -o.. sin20
rm = tocos2Q (18)
lt -Yt , )
where o,= total stressin the x-direction at the centcr
{l -Y} , l (14) ofthe base;o"= total stressin the y-direction at the
tY,l
t - l centerofthe base; r,n = shearstressin the x- and y-
l Y .) direction at the center of the base; and d = angle
measuredfrom the positive x-axis to the line of ap.
where r and y = global coordinate positions within
plicationofthe nonnal stress.
the elementthat are known as the centerof baseof a
The abovestepsprovide the necessaryinformation
slice (Fig. 4); X and y: global coordinateat the ele-
requiredto calculatethe stability of a slope using the
ment nodes;and lf7, N2,Nj andN4 = the shapefirnc-
finite element stresses.The calculatedvalues for the
tions definedin Equations7 to 10.
normal stress,o,r, and the mobilized shearstress,r-,
The stressesfrom a finite element analysis are
at the center of the base of a slice are entered into
stored at the Gausspoints. The shapefirnctions can
Equations(2) and (3) to give the resisting shearforce

36
 / Fictitious slice defined with
the Limit EquilibriumanalYsis

(x, y) known
(r, s) unknown

- Centerof the baseof a slice(x, y)

x_____+.

x-Coordinate

Figure4. Locationofthe centerofthe basealongthe slip surfacewithin a particularfinite element.

t
u.73m,1.73n1 t1.7320,1.7320)

O EementNodes
tr.#o.sro o.s2,6.5r4
ID
o + EementC'aussPcints

o (-0.srt,4.574 p.1n, 4.iln

o
I

(-1.7320,-1.73n) -r.7320)
(1.7320,

x - Cmrdinate

pointprojections
Figure5.Gauss ofa finiteclcnent.
to thenodes

(strength) and the mobilized shear force (actuating 4 PARAMETRIC STI.JDIESON A SIMPLE 2:I
shear),respectively. SLOPE
The local factor ofsafety is computedas the ratio
of the resisting shear force to the mobilized shear A slopeat 2 horizontal to I vertical is analyzedfor 4
force.The overall factor ofsafety is the sum ofthe conditions(Scoular,1997).The first caseis a free-
shearforce resistancevalues divided by the sum of standing slope with zero pore-water pressuresand
the actuatingshearforcesalong the slip surface. the slope is referredto as a dry slope @ig. 6). The
second case is a free-standing slope with a pie-
zometric line at thrce quaders of the slope height,
and the slopeis referredto as a wet slope(Fig. 6).

37
 E
o
o
E
o
o
I PiezometricLine
I
t l

u 4u 60 EO 1OO
x - Coordinate (m)

Figure6. Serected
2:l free-standing
slopcwith a piczomctic rineexiting at thetoeofthe srope.

caseis a slope partially submergedin wa-

Pr titd sl.opeis simulated.bypoint loadsequalto the weight
r-er
Tm zero pore_waterpressuresin the slope (re_ ot.water on the slope.The analysesare performld
{e1-edto as dry) (Fig. D. The fourth "*" ir'" o*_ usrng potsson'sratios of 0.33 and 0.4g, and a
ltally submergedslopewith a piezometricline at one
halt Io*g't modulusequalto 20,000and ZOO,TiOO tpa.
'lhe of the.slopeheight (refenedto as wet) (Fig.7). lne results showed that the stesses changewith a
partiallysubmergedslopeis covercdwith w1rcr
"l"rg-g poisson's ratio, but ar" "orrit*t fo,
to onehalfofthe slopeheight,providingsupportfor
changesin the Young'smodulus.fhis observationis
me stope and tncreasingthe factors of safety. The
consrstent with the observations
of Matos(19g2).
cohesionof the soil was varied from l0 to 4O kp"
and the angle of internal friction was varied from l0
to 30 degreesfor eachslopet1pe. s_RESI.JLTSOF TI{E FIMTE ELEMENT SLOPE
STABILITYMETHOD
4.1 Limit equilibrium analysis
The limit equilibrium analysesare performed using Tl" l*ql factors of safety differs along the overall
slip surface (Fig. 8). Ioial factors oi-r"f"r"
Equilibrium methoa (CG), r"r"
9" F :tt_Limit *Tpl1t9d for a2:l (dry) slopewith a cohesioi"quaf
(Fredlund & kahn 1977) which provides " ,orn_ ro 4u t(pa an<l
an angle of intemal friction equai to
bined moment and force equilibrium solution. 30 degrees.While thi tocal facton-oi;;fira
L
empirical finite element interslice force function,
on fong the slip surface, the overall n rit.- lf"rn"*
911e9 an independentstess analysis (Fan et al. facton of safeg fall within the .-g" "i t " iir,rit
1986) was ued. The General Limit Equilibrium equilibrium factors of safety.
The ?itr";;"-;_
pethod along with a finite element intersiice force tween the local factors of safety for poisson;sLtio,
fimction provides a method of comparisonbetureen of 0.33 and 0.48, calculatedusing the
fi"il;t";;
the finite element basedanalysis"nd th" limit equi_ in Figure.8.m" r""toroir"f"ty
librium analysis. l1e9 f.*n-p*
TTputed by rhe limit equilibrium method and the
finite element."Fd appearto U" u.ry,i.ifrr.
fi"
resuts appearto be within-_thelimits of unccrtainty
4.2 Finite elementstressanalysis
associatcdwith slope stability calculations.ilc i_
The finite elementshessanalysiswas performed by nite elemcnt method incorporatesthe stress-;train
"switching-on" characteristicsof the soil when computins th;;h;
gravity for the fr"e-rtaodine ,LD,
and for the partially submergedslope.The toaaLf strengthand actuating shearforce oitne-rol io G
the water and the lateral support it provides to the calculationofthe factor ofsafety (Fig. 9).

38
 80

E
o
G Crest
E 4 0
o
o
---
I
\ l Water l
\- I

I roe

20 60 1(x)
x - Coordinate(m)

Figure 7. Selected 2: I partially submerged slope with a horizontal piezometric line at mid-slope.

The factor of safety results computed using the the stability coefficient, (c/7II) (Taylor, 1937),
finite elementmethod (i.e., F3 correspondingto a where 7 is the unit weight of the soil, I/ is the height
Poisson'sratio of 0.33, F4 conespondingto a Pois- of the slope,l'is the angleof intemalfriction, andc'
son'sratio of 0.48) are comparedto the factors of is the cohesion.
safetycomputedusingthe limit equilibriummethod The factors of safety are groupedaccordingto the
(GLE) and are shown in Tables I and 2. To assess soil parametersand plotted versusthe stability num-
the variationsin the factor of safetyby eachmethod ber and the stability coefficient. The greatestdiffer-
ofanalysis,the resultsare groupedaccordingto co- encein factors ofsafety is noticed at high anglesof
hesionand angle of intemal friction. The factorsof intemal friction, at low values of cohesionand at the
safetygroupedaccordingto cohesion,c', are plotted maximumvaluesof Poisson'sratio.
versusthe stability number, l(7lftan{')/cf, (Janbu' The factors of safety for the (dry) free-standing
1954).The factors of safety groupedaccordingto slope, when grouped according to cohesion and
the angle of intemal friction, /', are plotted versus plotted venus the stability number (Fig. l0) show a

S loo
ft zso
h

$ zoo
.Eorso
O

E 100
E
F
oo
s o
PoissonRatio,P=0.33
X o
20 30 40 50
(m)
x+oordinate

Figure 8. Prescntationofthe local and global faaors ofsafety for a 2:l dry slope.

39
 Table l. 2:l fiee-standingslope
Soil Parameters Dry
c'
Wet
0' GLE F3 F4 F3
kPa degree Finite tt= 0.33 tt=0.48 linire p=0.33 p=0.48
element element

20 l0 0.882 0.867 o.874 o.677

l0 20
o.634 0.647
Lt3l 1.125 l.l5l o.782 0.745 0.7s5
40 l0 1.260 1.230 1.239
1.239 0.995
20 20
0.930 0.953
t.370 1.352 1.368
.368 l n?r
r.021 0.969 0:988
l0 30 1 . 65t 1.639 t.696 t.102
40 t.077
20 1.794 1.765 1.775 I.JJ) 1.260
20 30 t.892 1.884 1.918

GlobalFacton ofSafcy
Ical F, 1P= 6.33;
Bi$op 2.3@
Janbu 2.173
GLE (F.8. firnctiar) 2.356
>' 4 (P =0.33) 23ap
,O
F, (r = O.+e; 2.339
rA
frinary 2226

o
qr

,mbuMelho4Ft=2.173

x-coordinate
Figurc9. Shearstrengthandshearfotce for a 2:l dry slopecalculatedtsing the
finire elementmethod.

2.5

2.O
.o
3 1.5
o

t r.o
6
&
+ FI(GLE)
0.5
+ Fr (F = 0.33)
- Fs (P = 0.48)
0.0

Figure 10. Facors ofsafcty versrs stability nurnbcr for a 2:l dry slope as a fimction ofcohesion.

40
 Ta,ble2. 2:l partially submergedslope
Soil Parameters Dry Wet
c 6' GLE F3 F4 GLE F3 F4
kPa degree Finiteelement p= 0.33 p= 0.48 Finite p= 0.33 p= 0.48
firnction element
fimction
l0 t0 0.E45 0.843 0.827 0.649 0.635 0.64r
20 l0 1.t49 1.1t5 1.085 0.886 0.874 0.880
l0 20 1.344 1.425 1.422 1.050 1.046 1.06E
20 20 1.618 1.586 1.s75 l.3lE t.3t4 1.343
40 l0 t.721 1.722 1.691 I at', 1.296 1.316
l0 30 1.865 2.081 n.s.&* 1.482 t.505 r.530
40 20 2.297 2.3E5 2.368 1.E00 1.774 t.795
20 30 2.337 2.268 2.204 1.783 1.763 1.786
qO 30 I.OOA 2.970 2.899 2.303 2.260 2.274
rn.s.a.:no solutionachieved

2.5

)(, 1.5
(o
a
: 1.0
o
E
o
LL

0.5

0.0
0.00 0.02 0.04 0.06 0.08 0.10 0.12
StabilityCoeficent, [d( yH)l
t;igure I l. Factor ofsafety versus stability coefficient for a 2:l dry slope as firnaion ofangle ofintemat friction.

2.0

1.6

b
* 12
an
o
E o.a
6
l!

0.4

10 15 20 25
StabilityNumber,[( yHtang')/d]
ligurc 12. Factor ofsafety versusstability number as a imction ofcohesion for a2'l slope with the piczomeric
line at % ofthe
.lop hcight.

41
 2.0

1.6
.)
o
g 1.2
o

€ o.a
o
lr

0.00 0.02 0.10 o.'t2

Stabitity Coefficent, Ic't(rH ll

Figure 13. Factor of safety vems stability cocfEcient as a filrction of the angle of intemal friction for a 2:l slope with the pie-
zometric line at % ofthe slope height.

3.5

3.0

b 2-5
o
3 2.0
o
E
o
r.s
qG
i.o

0.5

0.0
0 5 1 0 1 5 2 0 2 5
StabilityNumbcr,[(tH tan0,)/c,l

Figure14.Factorof safetyvctsusstabilitynumbcrasa fimctionof cohesionfor a 2:I dry stope/, submerged

with water.

slight divergencein the factors of safety when the nite elementfactors of safety and the GeneralLimit
cohesionapproachesl0 kPa and the angle of inter- Equilibrium facton of safety when the cohesion is
nal friction approach30 degrees. 40 ud20 kPa. The difference betweenthe factorsof
The factors of safety by the finite element safety by both methods is constant at all values of
method, with a high Poisson's ratio, is greaterthan cohesionuntil the angle of intemal friction becomes
the GeneralLimit Equilibrium solution. The slight ggual to_30 degreesand cohesionbecomesequal to
divergenceis evidentwhen the factorsofsafety are l0kPa(Fig.l3).
grouped according to the angle of intemal friction The groupingofthe factorsofsafety accordinsto
and plottedversusthe stability coefficient(Fig. I l). the angle of intemal friction, ploned venus the ita-
It is also evident that at high values of cohesion, bility cocfficient (Fig. l5), shows the sarnepattem
(i.e., c'equal to 40 kPa),The factorsof safetycom- as for the (dry) free-standingslope (Fig.l0). The dif-
puted when using the General Limit Equilibrium ferencesin the results ar€ more pronouncedas the
method are greater than those from the finite ele- cohesionbecomelessthan l0 kPa.
ment methodswith either Poisson'sratio value. The factors of safety for the partiallv submersed
The factors of safety for the (wet) free-standing slopewith a piezometric line at one haliof the slipe
slope with a piezometric line at three quartersof the h.eightwere groupedby cohesionand plotted vetsus
slope height, are grotrpedaccordingto the cohesion the stability number (Fig. 16). The results show
and plotted versusthe stability number (Fig. l2). closeagreementbetweenthe GeneralLimit Equilib
The resultsshow a slight divergencebet*,eenthe fi- rium method and the finifs sts6sal mettrod. The

42
 .> 2.5
.g
3 2.0
o
b i.5
tt
o
* 1.0 + FI(GLE)
+ F r ( P= 0 ' 3 3 )
- F! (p = 0.48)

0.00 0.04 0.06 0.08 0.10 0.12

StabifltyCoefficent,le l1H )l

Figure I 5. Factor of safety versus stability coeffcient as a fimction of internal friction for a 2: I dry slope % submergedin water.

2.5

2.0

8 r' vr
6
U)

t 1.0
6
L
F3 (GLE)
0.5 Fr (P = 0'3!l)
Fc (F = 0.48)

StabilityNumber, I lHtan{')/c'l

Figure 16. Factor on safety versus stability numb€r as a firnction ofcohesion for a 2:t slope halfsubmerged with a horizontal pie-
zometric line.

samepattem of divergenceis evident as was shown terminethe stressstatein the slope.If the limit equi-
for the dry soil slope which is partially submerged librium and finite element factors of safety are
(Fig. 1a). However,the divergenceis not quite as similar for a simple slope than results from the two
extensive.The sameconrmentsapply to the factor of methodscan be interpretedin similar manners.This
safety versus the stability coefficient as shown in shrdythen setsthe stagefor using the finite element
Figure17. method for situations where the limit equilibrium
Plotting the factors of safety for the various slope methods is known to not yield satisfactory results.
conditions,(i.e., dry free-standing, wet free-standing The finite element method also producesgraphs of
and dry partially submerged),versus stabilify num- the local factors of safetythat can be combinedwith
ber on Figure 18, shows the ranking of slopesby the shearstrenglh-actuatingshearforce plots to help
factors of safety. The factor of safety can be esti- explain the bestsupportmechanismfor the slope.
matedfor a slope that is similar to one of thesecases The closeagreementbetweenthe factorsof safety
by calculatingthe stability numberand selectingthe when using the limit equilibrium method or the fi-
appropriatevalue of cohesionand angle of intemal nite elementmethod,hashistorically favoredthe use
friction. of limit equilibrium methods. Examination of the
Both the GeneralLimit Equilibrium method and the critical slip surfacesrevealsthat while the factors of
finite element method of slope stability produce safety values are close, the location of the critical
factorsof safetythat are in close agreement.The ad- slip surfacesmay be different.
vantage of the finite element method is that the
stress-stain characteristicsofthe soil are usedto de-

43
 i
o
-q 1.5
o

t r.o
6
lt
--r--- F" (GLE)
0.5 + Fr (tt = 0.33)
-- Fr (F = 0.4E)
0.0
0.00 0.04 0.08 0.10 o.12
Stability Coefficent , Ic.t( rH )l

Figure 17' Factor ofsafety versus sability coefficient as a fimction

ofthc angle ofintemal friction for a 2:l slope halfsubmerge.
with a horizontal piezometric line.

3.0

2.5

g 2.0
o
U)
o 1.5
o
()
o
lJ-
1.0

3/4 Piezometnc,C = 10 kpa

StabilityNumber, [( yH tang,ltc,l
Figure 18' Factor of safety vcrsrs sability nunber as a ftrndion of
cohesion fot a 2:l slope, evaluated for dry, piezometic and
submergedconditions.

6 ANALYSIS FOR TI{E LOCATION OF TTIE showedthe deepestslip surface.For the partialll.
CRITICAL SLIP SI.JRFACE
ylmerged slope,the finite elementmethodwith a
Poisson'sratio equalto 0.4g,showed" .o*ia"r"Uty
shallowerslip surface.
The location of the critical circle changesdepend.ing
on the sitrution Fiog qotfr"{. The biggesrchange
in location of critical slip surface was-experiencJd 7 CONCLUSION
for the (wet) free-standingsloae (Figs. 19 and 20)
andthe (wet) supportedslope(Figs. 2l and,22\. The finite elementmethodof slopestabiliw is a vi-
In general,the finite elementmethodslip surfaces ablemethodof analysisthat is now availabiefor en_
go deeperthan the limit equilibrium slip surfacesfor gingenng
.practice.The use of the finite "lem.nt
the (wet) free-standing slope. The partially sub. method yelds morc detailed information on tt.
mergedslopesshow that the limit equilibrium slip stressstrte in the soil than is availablefrom .onln"n-
surfacesgo deeperthan the finite element method tional limit equilibrium methods.This information
slip surfaces.For the free- standingslope, the frnite can assists.engineers in the design of slopes and
elementmethodwith a Poisson'sratio equalto 0.4g sloperetainingstructures.

4
 Cohesion = 40 kPa, 0'= 30", Piezometric litr- 3t4 of the ralay up the slope
Method X y R FactorofSafety
GLE (EE. Function) 58.50 56.00 37.88 ,t.741
F" (p = 0.3€t) 57.50 49.50 34.69 1.627
Fg (p = 0.48) 57.50 53.00 37.83 1.661
Fs (p = 0.33)
6
-? 50
.g
P
? ? 4 0 Fs (p = 0.48)
d
i 3 0

10 20 30 40 50 60 70 80 90 loo .t10
x - Coordinate (m)
Figure 19. Location of the critical slip surface for a slope with a piczometric line where
the soil properties are c, = 40Wa and ('=
300.

Coh.esion= 40 kPa, 0,= 3Oo,piezometricline 3/4 of lhe way up lhe slope

Method x y R Fjctorofs;feiy
GLE (F.E.Function) 63.50 S9.OO 39.56 1.102
F, (p = 0.33) 63.00 S9.OO 41.54 1.076
=
F. (p 0.48) 61.50 S9.SO 42.gg 1.100

Fs (F = 0.48)
Fs(rr= 0.33)

E
-? 40
o
6
.g
p GLE (F.E.tunction)
6 3 0
o
o
I

10 20 30 40 50 60 70 80 90 1oo 110
x - Coordinate (m)
Figure 20' Location of the critical slip surface for a slope with
a piezometric line where the factors of safety are closest to I .0.

The valueof Poisson'sratio can affect the calcu- the finite elementmethod to slope stability prob,
lation of the factor of safety as well as the location lems, a u"tt"t *a"ot"nding is required regarding
of the slip surface.with anjncreasingapplicationof oe etrectoiFoi.*nt otio andthe overall deforma-
tion modelon the stabilityof slopes.
45
 Cohesion = 40 kPa, 0'= 30o, half submeqed slope
Method /\ Y R FaciorofSafety
GLE (F.E. Function) 58.fi) 58.50 40.20 2.303
F, (p = 0.33) 52.50 50.50 31.76 2.259
F. (p = 0.48) 51.50 51.50 30.94 2.273

E
o
G
c
E
o
()o
I

10 20 30 40 50 60 70 80 90 100 110
x - Coordinate (m)

Figure 2l ' Location ofthe critical slip surfacefor a halfsubmerged slope where the soit propertiesare c,= 40kpa and,
O'= 31,.

Cohesion = 40 kPa, 0,= 30o, half submerged stop-

Method X Y R Factor of Safety
G L E ( F . E .F u n c t i o n ) 6 2 . 5 0 63.50 44.11 1.050
F, (p = 0.33) 54.00 51.50 31.97 1.046
F. (p = 0.48) 53.00 56.50 34.69- 1.068
G L E ( F . E .f u n c t i o n )

Fs (tr = 0.48)

E Fr (P = 0.33)
o
r0
.c
E
o
o
o
I

.10 20 30 40 50 60 70 80 90 100 1 1o
x - C o o r d i n a t e( m )

Figute22.Locationof thecritical slip srrfacefor a submerged

slopewherethefactorsof safetyareclosesrto 1.0.

The finite element stress.analy-sisprovides input boundary conditions are being used and that a rca-
information for the calculation of the stability of a sonable stressdeformation riodel is beine used.
slope. Further research must be undertaken on the With this arisurianc€,soil structures can be dner de-
stress analysis in order to ensure that the proper sigpedtoaccountforavari€tyofstressconditions.

46
 ACKNOWLEDGEMENTS t"cshchinsky, D. 1990. Slope sabitity analysis: Gcncralizod
L*,nal of Geotechnicot Engineering. ASCE
l9gq9acy
Thc authorswant to acknowledgethe initial dis- I l6(5): E5l-867.
cussionsregardingthe potential for using a finite Marrins, J.8., EB. Rcis & A.C. Maros I9gl. New mcthods of
arulysis for srability of slopcs. proc. lhth Int. ('onf Soil
clemcnt slope stability method rhat were held with Mech. Found. Eng. 3: 463467.
hofl Wong lki Sin of NanyangTechnologicalUni- Matos. A,C. 1982, The numcrical influcncc of thc poisson ntio
venity, Singapore.Thesediscussionsformed the ba- on the safay factor. Proceed@ of the 4th laennatiotul
sis for the study of this topic. The assistanceof Dr. Conference on Numerical Methds in Geo-Mechonis l:
FangshcngShuai, Ms. Noshin Tadenadehand Ms. 207-2tt.
Brigine Boldt-Lrppin in assemblingthis manuscript Morgcnsrern.N.R. & V.E. Price t965. The analysisof rhe sa-
bifity ofgeneral slip surfaces.Geotechniquel5(l ): 79-93.
is also acknowledged. The authorsare also gnateful Nayfor, D.J. 1982. Finite elemcntsand slope srabiliry.Numeri-
to Gco-Slopelntemational, Calgary, for the modifi- col Methods in Geomechanics. D. Rcidcl publisiring Corn-
cations made to their software in order thst this pany.
studycould be readilyperformed. Rescndia D. 1972. Accuracy of cmbankmcnt deformations.
Proceedings, ASCE Speciolty Conference on perlormonce
of Earth and Earth-Supported Structures, purdul Univer-
siry. West Lafayene. Indiana t2-14 June. l(part l): gt7_
REFERENCES 836.
Resendiz D. 1974. Accuacy of equitibrium slope sability
Adikari. G.S.N. & P.J. Cunmins I9E5. An ctfcctive stcss anzlysis. Journal ofthe Soil Mechaniq and Foundotion Di
slopc sability ana.lysis mcthod for dams. proc. llth Int. virjoz. ASCE 100(GT8): 967-970.
Conl Soil Mech. Foand. Eng.2:713-718. Scoular.RE.G. 1997.Lirnit cquilibrium slope sability analysis
Bathc, KJ. 19E2. Finite clemcnt procedures in cnginecring uing a suess analysis. M.Sc. Thesis, Ilniversii. of ial.-
analysis: 200-233. hentice..Hall. lcotchewan, Saskat oon, Conada.
Bisbo,p,A.W. 1952. The stability of earth dams. ph.D. Thesis. Ts& C.P. & I.B. Donald 19E5. Finirc elananr calculation of
Univcrsity ofLondon dam $abifity. Proc. I lth Int. Conf. Soil Mech. Found. Eng.,
Bishop, A.W. 1955. The use-of the slip circle in the subility San Francisco.Califomia 4: 2CAlr-2O44.
uulysis of slopes.Geotechnique5(l ): 7-17. Taylor, D.W. 1937. Srabiliry of canh slopcs. Journal of rhe
Bror.,g C. B. & I. P. King l!)62. Automatic embanfrmerr Boston Sociery ofCivil Engineers )Q(lV(3): 337-3g6.
amlysis cguilibrium and instabiliry conditions. Journal of Wrigfit, S. G. 1969. A study of slope srability and the
Soil Mechanics and Foandation Division, ISCE-93 (SM4i undrained shear strength ofclay shalcs. Ph. D. thesis, tJni-
2W-2r9 versity ofCalifornia ar Berrlley.
DuncanJ.Ir,t1996.Srarc-of-rhc.art: Srabitityanddeformation Wright,.S.G., F. Kulhany, & J.M. Duncan t973. Accuracy
y!y_1is.__J_ournal of GeorechnicalEngineering, ASCE of
equrlrbnum sfope stabiliry analysis. ASCE. Journat of-Soil
122(7):577-597. Mech anicsFoundation Division 99(SMl 0): 7g3_79l. -
Fao, K, D.G. Fr€dlund& G.W. Wilson 19g6.An intcrslicc Zienkiewicz,O.C., C. Humpheson
A nW. tewis f Sii. Asso_
forcefimction for limit equilibriumslopestabiliry"rof"rir. crated and non-associatcd visco_ptasticity and plasticitv
CanadianGeotechnical in
Journal 23e\ig7 -Z%.' soil Mechanics. Gdotechnique ZSiq: All<Sg.
Farias.M.M. & D.J. Naylor 1996.SaGi analysisusinglinitc
clements"Infogeo94 Sa6paulo.Brazil.
Frcdlund. D.G.A J. Knhn 1977.Comparison of slopestability
methodsof analysis.CanadionGiotechnique iq!), qZS-
439-
Frcdlun4D.G., J. Krahn& D.E. pu&hl 198t. Thc rclationship
bctuccn timir cquilibriumslopesability methods.ir"". i7
I cnth-Intemationol Conferenceon Soil Mechonicsontt
_ t.oundatiorlsEngineerizg,Stockholm,Swedcn3: 4@416.
t'tedfun4 D.G. & H. Ratrardjo1993. Soitmechotis for un-
_.. sanraed soils.New york: JohnWitey & Sons.
Higdon a.. E.H. Ohlscn,W.B. Stiles, l'4,. Wecsea W.F. Ri_
W-1976. Mechanicsof Materials.New york: JohnWiley
& Sons.
Janbu.N. t954.Stabiliryanalysisofslopeswith dimensionless
_- paraneters.HarvardSoil McchanicsSedcs(46).
Kondncr,RL. I 963. HyAerbolicsoess-strainrefrnse: cohe-
-
sive soils Journal ofthe SoitMechania and'Foundariorc
Division. ASCEE9(SM!):I t5-t43.
Kratr\ J., L. Lrm & D.G. Frcdlund 1996.The use of finire
elunant computedporc-natcrpnessurcs in a slopestability
arufysis.Landslides,Senneset(editor) 2: l27j:l2gl. Rot-
terdam:Balkema.
Kulhawy.F.H. 19,69.Finiteelementanalysisof thebehaviorof
embankments. phD. Thesis,the Universiryof Califomia.at
Berkley.Cslifomia.U,S.A.
La Rochellc.P. I 960.Theshorttermsability of slopesin Lon-
donclay. Ph.D.Thesis.Universityoflondon, London UK.

47

View publication stats