CHAPTER 8

SHEAR STRENGT H O F SOI L

8.1 INTRODUCTIO N
One o f th e mos t importan t and th e mos t controversia l engineering propertie s o f soi l i s it s shea r
strength o r ability to resist slidin g along internal surfaces within a mass. The stabilit y of a cut, the
slope of an earth dam, the foundations of structures, the natural slopes of hillsides and other structures
built on soi l depend upon the shearin g resistance offere d b y the soi l alon g th e probable surface s of
slippage. There i s hardl y a proble m i n th e fiel d o f engineerin g whic h doe s no t involv e the shea r
properties o f the soil in some manner or the other.

8.2 BASI C CONCEPT O F SHEARING RESISTANCE AND

SHEARING STRENGT H
The basic concept of shearing resistance and shearing strength can be made clear by studying first
the basi c principle s o f frictio n betwee n soli d bodies . Conside r a prismati c bloc k B restin g o n a
plane surfac e MN a s show n i n Fig . 8.1 . Bloc k B is subjecte d t o th e force P n whic h act s a t right
angles to the surface MN, an d the force F a tha t acts tangentially to the plane. The norma l force P n
remains constant whereas F a gradually increases from zero to a value which will produce sliding. If
the tangentia l force F a i s relatively small, block B wil l remain a t rest, an d th e applie d horizonta l
force will be balanced by an equal and opposite force Fr on the plane of contact. This resisting force
is developed as a result of roughness characteristics of the bottom of block B and plane surface MN.
The angle 8 formed by the resultant R of the two forces F r and Pn with the normal to the plane MN
is known as the angle o f obliquity.
If th e applied horizonta l force F a i s gradually increased, th e resisting forc e F r wil l likewise
increase, alway s being equal in magnitude and opposite in direction t o the applied force. Block B
will start sliding along the plane when the force F a reaches a value which will increase the angle of
obliquity t o a certain maximu m value 8 . If block B and plane surface MN ar e made o f the same

253
 254 Chapter 8

M N

Figure 8. 1 Basi c concept o f shearin g resistanc e an d strength.

material, th e angl e 8 m i s equa l t o (ft whic h i s terme d th e angle o f friction, an d th e valu e ta n 0 i s

termed th e coefficient of friction. I f block B and plane surface MN ar e made o f dissimilar materials ,
the angl e 8 is terme d th e angle o f wall friction. Th e applie d horizonta l forc e F a o n bloc k B i s a
shearing forc e an d th e develope d forc e i s friction o r shearing resistance. The maximu m shearing
resistance whic h the materials are capable o f developing i s called th e shearing strength.
If anothe r experimen t i s conducte d o n th e sam e bloc k wit h a highe r norma l loa d P n th e
shearing force F a wil l correspondingly be greater. A series of such experiments would show that the
shearing forc e F a i s proportional t o the normal load P n, that is
F = P ta n (8.1)
If A i s th e overal l contac t are a o f bloc k B o n plan e surfac e M/V , th e relationshi p ma y b e
written as

FP
shear strength, s = —- = —- tan,
AA
or s = a tan (8.2)

8.3 TH E COULOMB EQUATIO N

The basi c concep t o f friction a s explained in Sect. 8. 2 applies t o soils whic h are purely granular in
character. Soil s whic h ar e no t purel y granula r exhibit a n additiona l strengt h whic h i s du e t o th e
cohesion betwee n th e particles. It is, therefore, still customary to separate th e shearing strength s of
such soils into two components, on e due to the cohesion betwee n the soil particles and the other due
to th e frictio n betwee n them . The fundamenta l shea r strengt h equatio n propose d b y th e Frenc h
engineer Coulom b (1776 ) is
s = c + (J ta n (8.3)
This equatio n expresse s th e assumptio n that th e cohesio n c i s independen t o f th e norma l
pressure c r acting o n the plane o f failure. At zero normal pressure , th e shea r strengt h o f the soi l i s
expressed a s
s= c (8.4)
Shear Strengt h o f Soi l 255

1
Normal pressure, a

Figure 8. 2 Coulomb' s la w

According t o Eq . (8.4) , th e cohesio n o f a soi l i s define d a s th e shearin g strengt h a t zer o

normal pressur e o n the plane of rupture.
In Coulomb' s equatio n c an d 0 are empirical parameters , th e value s o f whic h fo r an y soi l
depend upo n several factors; the most important of these are :
1. Th e pas t history of the soil .
2. Th e initial state of the soil, i.e., whether it is saturated or unsaturated.
3. Th e permeability characteristics o f the soil .
4. Th e conditions of drainage allowed to take place during the test.
Since c and 0 in Coulomb' s Eq . (8.3 ) depen d upo n man y factors , c i s terme d a s apparent
cohesion an d 0 th e angl e o f shearin g resistance . Fo r cohesionles s soi l c = 0, the n Coulomb' s
equation become s
s = a tan (8.5)
The relationshi p betwee n th e variou s parameter s o f Coulomb' s equatio n i s show n
diagrammatically in Fig. 8.2 .

8.4 METHOD S O F DETERMINING SHEA R STRENGT H

PARAMETERS
Methods
The shear strength parameters c and 0 of soils either in the undisturbed or remolded state s ma y be
determined b y any of the following methods:
1. Laboratory methods
(a) Direc t o r box shear tes t
(b) Triaxia l compression tes t
2. Field method: Van e shear tes t or by any other indirect methods

Shear Parameter s o f Soil s in-sit u

The laboratory or the field metho d that has to be chosen in a particular case depend s upo n the type
of soi l an d th e accurac y required . Whereve r th e strengt h characteristic s o f th e soi l in-sit u ar e
required, laborator y test s ma y b e use d provide d undisturbe d samples ca n b e extracte d fro m th e
 256 Chapter 8

stratum. However , soil s ar e subjec t t o disturbanc e either during sampling o r extractio n fro m th e
sampling tube s i n th e laborator y eve n thoug h soi l particle s posses s cohesion . I t i s practicall y
impossible t o obtai n undisturbe d samples o f cohesionles s soil s an d highl y pre-consolidate d cla y
soils. Sof t sensitiv e clays are nearly always remolded durin g sampling. Laboratory method s may ,
therefore, b e use d onl y i n suc h case s wher e fairl y goo d undisturbe d sample s ca n b e obtained .
Where i t is not possible to extract undisturbed samples from the natural soil stratum, any one of the
following method s ma y hav e to be use d accordin g t o convenience an d judgment :
1. Laborator y test s on remolded sample s which could at best simulat e field condition s o f th e
soil.
2. An y suitabl e field test .
The presen t tren d i s t o rel y mor e o n fiel d test s a s thes e test s hav e bee n foun d t o b e mor e
reliable tha n even th e mor e sophisticate d laboratory methods.

Shear Strengt h Parameter s o f Compacte d Fill s

The strengt h characteristic s o f fill s whic h are t o b e constructed , suc h a s eart h embankments , ar e
generally foun d i n a laboratory . Remolde d sample s simulatin g th e propose d densit y an d wate r
content o f th e fil l material s ar e mad e i n th e laborator y an d tested . However , th e strengt h
characteristics o f existin g fills ma y hav e t o b e determine d eithe r b y laborator y o r fiel d method s
keeping in view th e limitation s of each method .

8.5 SHEA R TES T APPARATU S

Direct Shea r Tes t
The origina l for m o f apparatu s for th e direc t applicatio n o f shea r forc e i s the shea r box . Th e bo x
shear test , thoug h simple i n principle, has certain shortcoming s whic h wil l b e discusse d late r on .
The apparatu s consist s of a square bras s box spli t horizontally at the leve l o f the center o f the soi l
sample, which is held between meta l grilles and porous stone s a s shown in Fig. 8.3(a) . Vertical load
is applie d t o th e sampl e a s show n i n th e figur e an d i s hel d constan t durin g a test . A graduall y
increasing horizontal load i s applied to the lower part of the box until the sample fail s in shear. Th e
shear load at failure is divided by the cross-sectional are a of the sample to give the ultimate shearing
strength. The vertica l load divided by the area o f the sample give s the applied vertica l stress <7 . The
test may be repeated wit h a few more sample s having the same initia l conditions as the first sample.
Each sampl e i s tested wit h a different vertica l load.

— Norma l loa d
Porous ston e

Proving ring

<x><xxx><xxxp>^ Shearing
^^^^^^^^ force

Rollers

Figure 8.3(a ) Constan t rat e o f strai n shea r bo x

Shear Strengt h o f Soi l 25 7

Figure 8.3(b ) Strai n controlle d direc t shea r apparatu s (Courtesy : Soiltest )

The horizonta l loa d i s applie d a t a constan t rat e o f strain . Th e lowe r hal f o f th e bo x i s

mounted on rollers and is pushed forward at a uniform rat e by a motorized gearin g arrangement .
The upper half of the box bears against a steel proving ring, the deformation of which is shown on
the dial gauge indicatin g the shearing force. To measure the volume change during consolidation
and durin g the shearing process anothe r dial gauge is mounted to show the vertical movement of
the to p platen . The horizonta l displacement o f th e botto m o f th e bo x ma y als o b e measure d b y
another dia l gaug e whic h is no t show n in th e figure . Figur e 8.3(b ) show s a photograph o f strain
controlled direct shear test apparatus.

Procedure fo r Determinin g Shearin g Strengt h o f Soi l

In the direct shea r test, a sample o f soil is placed int o the shear box. The siz e of the box normally
used for clays and sands is 6 x 6 cm and the sample is 2 cm thick. A large box o f size 30 x 3 0 cm
with sample thickness of 1 5 cm is sometimes use d for gravelly soils .
The soil s use d fo r th e tes t ar e eithe r undisturbe d samples o r remolded . I f undisturbed , the
specimen has to be carefully trimme d and fitted into the box. If remolded samples are required, the soil
is placed into the box in layers at the required initial water content and tamped to the required dry density.
After th e specime n i s placed i n the box, an d all the other necessar y adjustment s are made, a
known norma l loa d i s applied . The n a shearing force is applied . The norma l loa d i s held constant
258 Chapte r8

throughout th e tes t bu t th e shearin g forc e i s applie d a t a constan t rat e o f strai n (whic h wil l b e
explained late r on). The shearing displacement is recorded b y a dial gauge .
Dividing the normal load and the maximum applied shearing force by the cross-sectional are a of
the specimen a t the shear plane gives respectively the unit normal pressure cran d the shearing strengt h
s at failure o f the sample. These results may be plotted o n a shearing diagra m wher e cri s the absciss a
and s th e ordinate . Th e resul t o f a singl e tes t establishe s on e poin t o n th e grap h representin g th e
Coulomb formula for shearing strength. In order to obtain sufficient point s to draw the Coulomb graph ,
additional test s mus t b e performe d o n othe r specimen s whic h ar e exac t duplicate s o f th e first . Th e
procedure i n thes e additiona l tests i s the sam e a s i n th e first, except tha t a differen t norma l stres s i s
applied each time . Normally, the plotted points of normal and shearing stresses a t failure of the various
specimens will approximate a straight line. But in the case of saturated, highly cohesive clay soils in the
undrained test, the graph of the relationship between the normal stres s and shearing strength is usually
a curved line, especially at low values of normal stress. However, i t is the usual practice to draw the best
straight line through the test points to establish the Coulomb Law. The slope of the line gives the angle
of shearing resistance an d the intercept on the ordinate gives the apparent cohesion (See . Fig . 8.2) .

Triaxial Compressio n Tes t

A diagrammatic layout of a triaxial test apparatus is shown in Fig. 8.4(a) . In the triaxial compressio n
test, thre e o r mor e identica l samples o f soi l ar e subjecte d t o uniforml y distributed flui d pressur e
around th e cylindrica l surface. Th e sampl e i s sealed i n a watertight rubber membrane . The n axia l
load i s applied t o the soil sampl e unti l it fails. Although only compressive loa d i s applied t o the soi l
sample, i t fails b y shea r o n interna l faces. I t is possible t o determin e th e shea r strengt h o f th e soi l
from th e applie d load s a t failure. Figur e 8.4(b ) gives a photograph o f a triaxial test apparatus .

Advantages an d Disadvantage s o f Direc t an d Triaxia l Shea r Test s

Direct shea r test s are generally suitable for cohesionless soil s except fin e san d an d silt whereas th e
triaxial test is suitable for all types of soils and tests. Undrained and consolidated undraine d tests on
clay sample s ca n b e mad e wit h the box-shea r apparatus . Th e advantage s o f th e triaxia l ove r th e
direct shear tes t are :
1. Th e stress distributio n across th e soil sample is more uniform in a triaxial test as compare d
to a direct shea r test .
2. Th e measuremen t of volume changes i s more accurat e i n the triaxial test .
3. Th e complete stat e of stress i s known at all stages durin g the triaxial test , whereas onl y the
stresses at failure ar e known in the direct shea r test .
4. I n th e cas e o f triaxia l shear, th e sampl e fail s alon g a plan e o n whic h th e combinatio n o f
normal stres s an d th e shea r stres s give s th e maximu m angl e o f obliquit y of th e resultan t
with th e normal , wherea s i n th e cas e o f direc t shear , th e sampl e i s sheare d onl y o n on e
plane whic h is the horizontal plane which need no t be the plane o f actual failure.
5. Por e wate r pressure s ca n b e measure d i n th e cas e o f triaxia l shea r test s wherea s i t is no t
possible i n direct shea r tests .
6. Th e triaxia l machine i s more adaptable .

Advantages o f Direc t Shea r Test s

1. Th e direct shea r machin e i s simple an d fast t o operate .
2. A thinne r soil sampl e i s used i n the direc t shea r tes t thu s facilitating drainage o f th e pore
water quickl y from a saturated specimen .
3. Direc t shea r requiremen t is much les s expensiv e a s compared t o triaxial equipment .
Shear Strengt h o f Soi l 259

Proving ring

Ram

Cell

Rubber membran e

Sample

(a) Diagrammati c layou t

Inlet Outlet

(b) Multiple x 50- E load

frame triaxia l tes t
apparatus (Courtesy :
Soiltest USA )

Figure 8.4 Triaxia l test

apparatus
 260 Chapter 8

Original sample Failure with Actual failur e

uniform strains condition

(a) Direct shear test

/— Dea d zon e

_ Stresse d
zone

Zone with
large strains

Dead zon e
(b) Triaxial shear tes t

Figure 8.5 Conditio n of sampl e durin g shearing i n direct an d triaxial shea r tests

The stress conditions across th e soil sample in the direct shear test are very complex becaus e
of th e chang e i n th e shea r are a wit h th e increas e i n shea r displacemen t a s th e tes t progresses ,
causing unequa l distributio n of shea r stresse s an d norma l stresse s ove r th e potentia l surfac e o f
sliding. Fig. 8.5(a ) shows the sample condition before and after shearin g in a direct shear box. The
final sheare d are a A,is less than the original area A.
Fig. 8.5(b) shows the stressed condition in a triaxial specimen. Because of the end restraints, dead
zones (non-stresse d zones ) triangular in section are formed at the ends wherea s th e stress distribution
across the sample midway between the dead zones may be taken as approximately uniform .

8.6 STRES S CONDITION A T A POIN T I N A SOI L MAS S

Through ever y point in a stressed bod y ther e are three planes a t right angles t o each othe r whic h are
unique as compared to all the other planes passing through the point, because they are subjected only to
normal stresse s wit h n o accompanying shearing stresse s actin g on the planes. These three plane s are
called principal planes, and the normal stresses acting on these planes are principal stresses. Ordinarily
the three principal stresses at a point differ in magnitude. They may be designated as the major principal
stress <TJ , the intermediat e principa l stress o~ 2, and th e mino r principal stress <Jy Principa l stresse s a t a
point in a stressed body are important because, once they are evaluated, the stresses o n any other plane
through the point can be determined. Many problems in foundation engineering can be approximated by
considering only two-dimensional stress conditions. The influence o f the intermediat e principal stress
(J2 o n failure may b e considered as not very significant.

A Two-Dimensiona l Demonstratio n o f th e Existenc e o f Principa l Plane s

Consider th e body (Fig . 8.6(a)) is subjected to a system of forces suc h as Fr F 2 F 3 an d F 4 whos e
magnitudes and lines of action are known.
Shear Strengt h o f Soi l 261

dx
(c)

Figure 8.6 Stres s at a point i n a body i n two dimensiona l spac e

Consider a small prismati c elemen t P . The stresse s actin g on this element i n the direction s
parallel to the arbitrarily chosen axe s x and y are shown in Fig. 8.6(b) .
Consider a plane AA through the element, making an angle a with the jc-axis. The equilibrium
condition o f th e elemen t ma y be analyze d b y considerin g th e stresse s actin g o n the face s o f th e
triangle ECD (shaded) which is shown to an enlarged scale in Fig. 8.6(c) . The normal and shearing
stresses o n the faces o f the triangle are also shown.
The unit stress in compression and in shear on the face ED are designated as crand T respectively.
Expressions fo r c r an d T may b e obtaine d b y applyin g th e principle s o f static s fo r th e
equilibrium condition of the body. The sum of all the forces in the jc-direction is
<Jxdx ta n a + T dx+ rdx se c a cos a - crdx sec a sin a = 0 (8.6)
The sum of all the forces i n the y-direction is
cr dx + TX d x tan a - T dx sec a sin a - crdx se c a cos a = 0 (8.7)
Solving Eqs. (8.6 ) an d (8.7 ) for cran d T , we have

aV+GX a -G J
— Ho i — cos2
•* a + T ™? sm2 a (8.8)

T = —|CT
fj \ yV •
- c r *r )/ - sin2a-irvv cos2 a (8.9)
By definition, a principal plane is one on which the shearing stress is equal to zero. Therefore,
when i is made equa l to zero in Eq. (8.9), the orientation of the principal plane s is defined by the
relationship

2i,
tan2a = (8.10)
262 Chapter 8

Equation (8.10 ) indicates that there are two principal planes through the point P in Fig. 8.6(a )
and tha t the y ar e a t right angle s t o each other . By differentiating Eq. (8.8 ) wit h respect t o a , an d
equating to zero, w e have

— y= - a..y sin 2a + ar sin 2a + 2t _. cos 2a = 0

da

or tan 2a = (8.11)
a -G X

Equation (8.11 ) indicate s th e orientatio n o f th e plane s o n whic h th e norma l stresse s e r are

maximum and minimum. This orientatio n coincides wit h Eq. (8.10). Therefore, i t follows that the
principal planes ar e also planes on which the normal stresses ar e maximum and minimum.

8.7 STRES S CONDITION S I N SOI L DURIN G TRIAXIA L

COMPRESSION TES T
In triaxia l compressio n tes t a cylindrica l specime n i s subjecte d t o a constan t all-roun d flui d
pressure whic h is the minor principal stres s O" 3 since the shear stress on the surface i s zero. The two
ends ar e subjected t o axial stres s whic h is the major principal stres s o r The stress conditio n in the
specimen goe s o n changin g wit h the increas e o f the majo r principa l stres s cr r I t i s of interes t t o
analyze the state o f stress alon g inclined sections passin g through the sampl e a t any stress leve l (J l
since failur e occurs alon g incline d surfaces .
Consider th e cylindrica l specime n o f soi l i n Fig . 8.7(a ) whic h i s subjecte d t o principa l
stresses <7 { an d <7 3 (<7 2 = <T 3).
Now CD, a horizontal plane, is called a principal plane since it is normal to the principal stres s
<TJ an d th e shea r stres s i s zero o n thi s plane. EF i s the othe r principa l plan e o n whic h the principal
stress <7 3 acts. AA i s the inclined section o n which the state of stress is required to be analyzed.
Consider a s befor e a smal l pris m o f soi l show n shade d i n Fig . 8.7(a ) an d th e sam e t o a n
enlarged scal e i n Fig. 8.7(b) . All the stresses actin g on the prism are shown. The equilibrium of the
prism require s
Horizontal force s = cr 3 sin a dl - a sin a dl + T cos adl = (8.12)

-D
A/
E

(a) (b )

Figure 8. 7 Stres s conditio n i n a triaxial compressio n tes t specime n

Shear Strengt h o f Soi l 26 3

£ Vertical forces = o{ cos a dl - a cos a dl - i sin a dl - 0 (8.13 )

Solving Eqs. (8.12 ) and (8.13) w e have

<7, + <7 , <7 , — (7-,

cr = — - + — -cos2 « (8.14 )
2 2
1
r = -(cr1 -<J 3 )sin2« (8.15 )

Let the resultant of <rand Tmake an angle 8 with the normal to the inclined plane. One should
remember that when ens less than 90°, the shear stress Ti s positive, and the angle S is also positive.
Eqs. (8.14 ) an d (8.15 ) ma y b e obtaine d directl y fro m th e genera l Eqs . (8.8) and (8.9)
respectively by substituting the following:
cr = < 7 . , < T = ( T , a n d T= 0

8.8 RELATIONSHI P BETWEE N TH E PRINCIPA L STRESSE S AN D

COHESION c
If the shearing resistance s of a soil depends on both friction an d cohesion, slidin g failure occurs in
accordance wit h the Coulomb Eq . (8.3), that is, when
T = s = c + c r t a n0 (8.16 )
Substituting for the values of erand rfrom Eqs. (8.14) and (8.15) into Eqs. (8.16) and solving
for <7 j w e obtai n
c + <7 3 tan </>
~
= <r j, +-"'v-'^'v-cos 5 ^ tftan^
(8.17 )

The plane with the least resistance to shearing along it will correspond t o the minimum value
of <7 j whic h can produce failure in accordance wit h Eq. (8.17). o l wil l be at a minimum when the
denominator i n the second membe r o f the equation is at a maximum, that is, when
d
— — (sin a cos a - cos z a tan <z>) = 0
da
Differentiating, an d simplifying, w e obtain (writing a - a c)
«, = 45° + 0/2 (8.18 )
Substituting for a i n Eq. (8.17) and simplifying, we have
CTj = CT 3 tan 2 (45° + 0/2) + 2c ta n (45 ° + 0/2) (8.19 )

or (T =l v3N0 + 2cN(8.20 )

where A^ = tan2 (45° + 0/2) is called th e flow value.

If the cohesion c = 0, we have

°i = °IN* (8.21 )
If 0 = 0, we have
<T = < T + 2c (8.22 )
264 Chapte r8

If the sides of the cylindrical specimen ar e no t acted o n by the horizontal pressur e <7 3, the loa d
required t o caus e failur e i s calle d th e unconfme d compressiv e strengt h qu. I t i s obviou s tha t a n
unconfmed compressio n tes t can be performed onl y o n a cohesive soil . According to Eq. (8.20) , th e
unconfmed compressiv e strengt h q i s equal to

u = a — 2r N
<T 8 71\
i y « -\] </> f (o.Zj)
If 0 = 0, then qu = 2c (8.24a )
or the shear strength

s = c = — (8.24b )

Eq. (8.24b ) show s on e o f th e simples t way s o f determinin g the shea r strengt h o f cohesiv e
soils.

8.9 MOH R CIRCL E O F STRESS

Squaring Eqs. (8.8 ) an d (8.9 ) an d adding, we have

i2 / _ ^ x 2

+ ^ = I" 2j + *ly (8.25 )

Now, Eq. (8.25) is the equation of a circle whose cente r ha s coordinate s

and whos e radius is — i/(c7y - c r ) -

2v v '
The coordinates of points on the circle represent the normal and shearing stresses on inclined
planes at a given point. The circl e is called th e Mohr circle o f stress, after Mohr ( 1 900), who firs t
recognized thi s usefu l relationship . Mohr's metho d provide s a convenien t graphica l metho d fo r
determining
I . Th e normal and shearin g stress on any plane through a point in a stressed body .
2. Th e orientation of the principal planes if the normal and shear stresses on the surface of the
prismatic elemen t (Fig . 8.6 ) ar e known . Th e relationship s ar e vali d regardles s o f th e
mechanical propertie s o f th e material s sinc e onl y th e consideration s o f equilibriu m ar e
involved.
If th e surface s o f th e elemen t ar e themselve s principa l planes , th e equatio n fo r th e Moh r
circle o f stress may be writte n as

T + oy - - =
- - y -- ( 8.26)

The center of the circle has coordinates T - 0 , and o= (a{ + (T3)/2, and its radius is (<J l - (T 3)/2.
Again from Mohr's diagram, the normal and shearing stresses o n any plane passing through a point
in a stressed bod y (Fig. 8.7 ) may be determined if the principal stresses cr l an d (J 3 are known. Since
<7j an d O" 3 are alway s known in a cylindrical compression test , Mohr's diagram i s a very useful too l
to analyze stresses o n failure planes.
Shear Strengt h o f Soi l 265

8.10 MOH R CIRCL E O F STRESS WHE N A PRISMATI C ELEMEN T

IS SUBJECTE D T O NORMA L AN D SHEA R STRESSE S
Consider first the case of a prismatic element subjected to normal and shear stresses a s in Fig. 8.8(a).

Sign Conventio n
1. Compressiv e stresse s ar e positive and tensile stresses ar e negative.
2. Shea r stresse s ar e considered a s positive if the y give a clockwise momen t abou t a point
above the stressed plane as shown in Fig. 8.8(b), otherwise negative.
The norma l stresse s ar e take n a s absciss a an d th e shea r stresse s a s ordinates . I t i s
xy d th e shea r stres s r ( xyT = yx T ) acting o n th e surfac e of
assumed th e norma l stresses c rx ,y c r an
the elemen t ar e known . Tw o point s P l an d P 2 ma y no w b e plotte d i n Fig . 8.8(b) , whos e
coordinates ar e

If the points P} an d P2 are joined, the line intersects the abscissa at point C whose coordinates
are [(0,+op/2,0].

Minor principal
plane >ai

(a) A prismatic element subjected to normal and shear stresses

(ax + a y)/2

+ ve

(b) Mohr circle of stress

Figure 8.8 Moh r stres s circl e fo r a general cas e

266 Chapte r8

Point O is the origin of coordinates fo r the center of the Mohr circl e o f stress. With center C
a circle ma y no w be constructed with radius

This circl e whic h passe s throug h points P l an d P 2 i s calle d th e Mohr circle o f stress. The
Mohr circl e intersect s the abscissa a t two points E and F . The major and minor principal stresse s
are o l ( = OF) an d cr 3 ( = OE) respectively .

Determination o f Norma l an d Shea r Stresse s o n Plan e A A [Fig . 8.8(a) ]

Point P { o n the circle of stress i n Fig. 8 . 8(b) represents th e state of stress on the vertical plane of the
prismatic element ; similarl y poin t P 2 represent s th e stat e o f stres s o n th e horizonta l plan e o f th e
element. If from poin t P{ a line is drawn parallel to the vertical plane, it intersects the circle at point PQ
and i f from th e point P 2 o n th e circle , a lin e is drawn parallel t o th e horizontal plane, this line als o
intersects th e circle at point P Q . The point PQ s o obtained i s called th e origin o f planes or the pole. If
from the pole P Q a line is drawn parallel to the plane AA in Fig. 8.8(a) to intersect the circle at point P3
(Fig. 8.8(b)) then the coordinates o f the point give the normal stress cran d the shear stres s To n plan e
AA as expressed b y equations 8.8 and 8.9 respectively. This indicates that a line drawn from the pole P Q
at an y angl e a t o the cr-axi s intersect s the circle a t coordinates tha t represent th e norma l an d shea r
stresses o n the plane inclined at the same angle to the abscissa.

Major an d Mino r Principa l Planes

The orientations o f the principal planes may be obtained b y joining point P Q to the points E and F
in Fig 8.8(b) . P Q F is the direction o f the major principal plane on which the major principal stres s
dj acts ; similarl y P Q E is the directio n o f the mino r principal plane o n whic h the mino r principa l
stress <7 3 acts. It is clear from the Mohr diagram that the two planes P Q E and P Q F intersect at a right
angle, i.e., angle EP Q F = 90° .

8.1 1 MOH R CIRCL E O F STRES S FO R A CYLINDRICA L SPECIME N

COMPRESSION TES T
Consider the case of a cylindrical specimen o f soil subjected to normal stresses <7 j and <J 3 whic h are
the major and minor principal stresses respectivel y (Fig . 8.9 )
From Eqs . (8.14 ) and (8.15) , we may write

22
(8.27)
/ O /-*^T \

Again Eq. (8.27 ) is the equation of a circle whos e center ha s coordinate s

<7, + CT , (7 , — (J-.
<J = — - -and T =0 and whose radius is
-
22
A circle wit h radius (o { - cr 3)/2 with its center C on the abscissa a t a distance o f (a l + cr 3)/2
may b e constructe d a s show n in Fig . 8.9 . Thi s i s the Moh r circl e o f stress . Th e majo r an d mino r
principal stresses are show n in the figur e wherei n cr , = OF an d <7 3 = OE .
From Fig. 8.8 , w e ca n writ e equations for cf j an d <7 3 and T max as follow s

±
Shear Strengt h o f Soi l 267

.A

Figure 8. 9 Moh r stres s circl e fo r a cylindrical specime n

(8.29)

where T max is the maximum shear stress equal to the radius of the Mohr circle.
The origin of planes or the pole P Q (Fig. 8.9) may be obtained as before by drawing lines fro m
points E and F parallel to planes on which the minor and major principal stresses act . In this case,
the pole P O lies on the abscissa and coincides with the point E.
The normal stress < J and shear stress T on any arbitrary plane AA makin g an angle a wit h the
major principa l plane may be determined a s follows.
From th e pole P0 draw a line PQ Pl paralle l t o the plane AA (Fig. 8.9) . The coordinates o f the
point Pl giv e the stresses c r and i. From the stress circle we may write

= 2a

cr, + cr, cr, - cr .

- (8.30)

Normal stress a

0° 15 ° 30 ° 45 ° 60 ° 75 ° 90 °
Angle of inclination of plane, a ^

Figure 8.1 0 Variatio n o f cran d r with a

268 Chapte r8

(j, -cr ,
r= 3 sin2 # (8.31 )

Equations (8.30) an d (8.31) are the same a s Eqs. (8.14 ) an d (8.15) respectively .
It is of interest to study the variation of the magnitudes of normal and shear stresse s with the
inclination o f the plane.
Eqs. (8.30 ) an d (8.31 ) ar e plotted with a a s the abscissa show n in Fig. 8.10 . The following
facts ar e clear fro m thes e curves:
1. Th e greates t an d leas t principa l stresse s ar e respectivel y th e maximu m an d minimu m
normal stresse s on any plane throug h th e point i n question .
2. Th e maximu m shear stress occur s o n planes at 45° to the principal planes .

8.12 MOHR-COULOM B FAILURE THEOR Y

Various theories relatin g to the stress condition i n engineering material s at the time of failure are
available i n th e engineerin g literature . Eac h o f thes e theorie s ma y explai n satisfactoril y th e
actions o f certain kind s o f materials a t th e tim e they fail , bu t n o one o f them i s applicabl e t o all
materials. Th e failur e o f a soi l mas s i s mor e nearl y i n accordanc e wit h th e tenet s o f th e Moh r
theory o f failur e tha n wit h thos e o f an y othe r theor y an d th e interpretatio n o f th e triaxia l
compression tes t depends t o a large extent on this fact. The Mohr theory is based on the postulate
that a material wil l fail whe n the shearing stres s o n the plane alon g whic h the failure is presume d
to occur is a unique function o f the normal stress acting on that plane. The material fail s along the
plane only when the angle between th e resultant of the shearing stres s an d the normal stres s i s a
maximum, tha t is , wher e th e combinatio n o f norma l an d shearin g stresse s produce s th e
maximum obliquity angle 8.
According t o Coulomb's Law , the condition of failure i s that the shear stres s

T ^ c + atan^ (8.32 )
In Fig 8. 1 l(b) M QN an d M QNl ar e the lines that satisfy Coulomb' s conditio n o f failure. If the
stress a t a given point withi n a cylindrical specimen unde r triaxia l compression i s represented b y
Mohr circl e 1 , it ma y b e note d tha t every plan e through this point ha s a shearin g stres s whic h is
smaller than the shearing strength.
For example, if the plane AA in Fig. 8.1 l(a) is the assumed failur e plane, the normal and shear
stresses on this plane at any intermediate stage of loading are represented b y point b on Mohr circl e
1 where the line PQb is parallel to the plane AA. The shearing stres s o n this plane i s ab which is less
than the shearing strengt h a c at the same norma l stress Oa . Under thi s stress conditio n there is no
possibility o f failure . O n th e othe r han d i t woul d no t b e possibl e t o appl y th e stres s conditio n
represented b y Mohr stres s circle 2 to this sample because i t is not possible fo r shearing stresse s t o
be greate r tha n th e shearin g strength . At th e norma l stres s Of, th e shearin g stres s o n plan e AA is
shown t o b e fh whic h is greate r tha n the shea r strengt h o f th e material s fg whic h i s no t possible .
Mohr circl e 3 i n th e figur e i s tangent to th e shea r strengt h lin e M QN an d M QNj a t points e an d e {
respectively. O n th e sam e plan e AA at normal stres s Od , the shearin g stres s d e i s the sam e a s th e
shearing strengt h de . Failur e i s therefor e imminen t o n plan e A A a t th e norma l stres s O d an d
shearing stres s de . The equation for the shearing stres s d e is
s = de - de'+ e'e = c + crtan 0 (8.33 )
where 0 is the slope o f the line MQN whic h is the maximum angle o f obliquity on the failure plane.
The valu e of the obliquity angle can neve r exceed <5 m = 0, the angle of shearing resistance , withou t
the occurrence of failure. The shear strength line MQN whic h is tangent to Mohr circle 3 is called th e
Shear Strengt h o f Soi l 269

'i /
Rupture
plane Moh r
envelope N

Mohr circle of
rupture
(b)

Figure 8.11 Diagra m presenting Mohr's theory o f ruptur e

Mohr envelope o r line of rupture. The Mohr envelope may be assumed as a straight line although it
is curved under certain conditions. The Mohr circle which is tangential to the shear strength line is
called th e Mohr circle o f rupture. Thu s th e Moh r envelop e constitute s a shea r diagra m an d i s a
graph o f th e Coulom b equatio n fo r shearin g stress . Thi s i s calle d th e Mohr-Coulomb Failure
Theory. Th e principal objective of a triaxial compression tes t is to establish the Mohr envelope fo r
the soil being tested. The cohesion and the angle of shearing resistance can be determined from this
envelope. Whe n th e cohesion o f th e soi l i s zero, tha t is, when the soi l i s cohesionless, th e Moh r
envelope passes throug h the origin.

8.13 MOH R DIAGRA M FO R TRIAXIAL COMPRESSIO N TES T A T

FAILURE
Consider a cylindrica l specime n o f soi l possessin g bot h cohesio n an d frictio n i s subjecte d t o a
conventional triaxia l compressio n test . I n th e conventiona l tes t th e latera l pressur e cr 3 i s hel d
constant an d th e vertica l pressur e <T J i s increase d a t a constan t rat e o f stres s o r strai n unti l th e
sample fails . I f cr l i s th e pea k valu e o f th e vertica l pressur e a t whic h th e sampl e fails , th e tw o
principal stresse s tha t ar e t o b e use d fo r plottin g th e Moh r circl e o f ruptur e ar e cr 3 an d o r I n
Fig. 8.12 the values of cr { an d <7 3 are plotted on the er-axi s and a circle i s drawn with (o ^ - cr 3) as
diameter. The center of the circle lies at a distance of (<j { + cr3)/2 from th e origin. As per Eq. (8.18),
the soil fails alon g a plane which makes an angle a, = 45° + 0/2 with the major principal plane. In
Fig. 8.1 2 the two lines P QPl an d P QP2 (wher e P Q i s the origin of planes) ar e the conjugate rupture
planes. The two lines M QN an d M QN^ draw n tangential to the rupture circle a t points P l an d P 2 are
called Mohr envelopes. If the Mohr envelope can be drawn by some other means, the orientation of
the failure planes may be determined .
The result s o f analysi s o f triaxia l compressio n test s a s explaine d i n Sect . 8. 8 ar e no w
presented i n a graphical for m in Fig. 8.12 . The various information that can be obtained fro m th e
figure include s
1. Th e angle of shearing resistance 0 = the slope of the Mohr envelope .
270 Chapter 8

Aa Mohr envelope
(a, - a3)/2

Figure 8.1 2 Moh r diagra m fo r triaxia l tes t a t failur e fo r c- 0 soi l

Rupture
plane Rupture
plane

0 = 0

T
I
c

0 C

(a) c = 0 (b) 0 = 0

Figure 8.1 3 Moh r diagra m fo r soil s wit h c = 0 and =0

2. Th e apparent cohesio n c = the intercept of the Mohr envelop e o n the T-axis .

3. Th e inclination of the rupture plane = a .
4. Th e angle between th e conjugate planes = 2a.
If th e soi l i s cohesionless wit h c = 0 the Mohr envelopes pas s throug h the origin, an d i f the
soil is purely cohesive wit h 0 = 0 the Mohr envelope is parallel to the abscissa. The Mohr envelope s
for thes e tw o types o f soils are shown in Fig. 8.13.

8.14 MOH R DIAGRA M FO R A DIREC T SHEA R TES T A T FAILUR E

In a direct shea r tes t the sample i s sheared alon g a horizontal plane. This indicate s that the failur e
plane is horizontal. The normal stress don thi s plane is the external vertical load divided by the area
of th e sample . Th e shea r stres s a t failur e i s th e externa l latera l loa d divide d b y th e are a o f th e
sample.
Point Pj on the stress diagram in Fig. 8.14 represents the stress condition on the failure plane.
The coordinate s o f the point are
normal stress = <7 , shea r stress i- s.
Shear Strengt h o f Soi l 271

Minor
Plane o f rupture

0
t Majo r principal
plane

Figure 8.1 4 Moh r diagra m fo r a direct shea r test a t failur e

If i t i s assume d tha t th e Moh r envelop e i s a straigh t lin e passin g throug h th e origi n

(for cohesionles s soi l o r normall y consolidate d clays) , i t follow s tha t th e maximu m
obliquity 8 m occurs o n the failure plane an d 8 m = 0. Therefore th e lin e OP { mus t be tangent
to th e Moh r circle , an d th e circl e ma y be constructe d a s follows :
Draw Pj C norma l t o OP r Poin t C whic h i s th e intersectio n poin t o f th e norma l wit h th e
abscissa i s the cente r o f th e circle . CP { i s th e radiu s of th e circle . Th e Moh r circl e ma y no w b e
constructed whic h gives the majo r and mino r principal stresses cr { an d <7 3 respectively .
Since th e failur e i s o n th e horizonta l plane , th e origi n o f plane s P Q ma y b e obtaine d b y
drawing a horizontal lin e through P { givin g P Q. P QF an d P QE giv e the directions o f the majo r and
minor principal planes respectively .

Example 8. 1
What is the shearing strengt h of soil along a horizontal plane at a depth of 4 m in a deposit o f sand
having the following properties :
Angle of internal friction, 0 = 35°
Dry uni t weight, y d - 1 7 kN/m 3
Specific gravity , Gs = 2.7.
Assume the ground water table i s at a depth of 2.5 m from th e ground surface. Also fin d th e
change in shear strengt h when the water table rises to the ground surface .

Solution
The effectiv e vertica l stres s a t the plane of interest is
<r'=2.50xy d + l.SO x y b
Given y d = 17 kN/m3 and G s = 2.7

We haver, = 17- = — X9.81

9A9
or lie = 26.5 - 1 7 = 9.49 o r e = —— = 0.56

3
Therefore, Y b = *9.81 =10.7 kN/m
l+e 1 + 0.56
272 Chapter 8

Hence c / = 2.5 x 1 7 + 1. 5 x 10. 7 = 58.55 kN/m 2

Hence, the shearing strength of the sand is
5 = (/ tan 0 = 58.55 x tan 35° = 41 kN/m 2
If th e wate r tabl e rise s t o th e groun d surfac e i.e. , b y a heigh t o f 2. 5 m , th e chang e i n th e
effective stres s will be ,
Ao" = y d x 2.5 -Yb* 2. 5 = 1 7 x 2.5 - 10. 7 x 2.5 = 15.7 5 kN/m 2 (negative)
Hence th e decrease in shear strength will be,
= Ac/ ta n 35 ° = 15.7 5 x 0.7 0 = 1 1 kN/m 2

Example 8. 2
Direct shea r test s wer e conducte d o n a dr y sand . The siz e o f th e sample s use d fo r th e test s wa s
2 in. x 2 in. x 0.75 in . The tes t results obtained are given below:

Test No . Norma l loa d Norma l stres s a Shea r forc e Shea r stres s

2 2
(Ib) (Ib/ft )a t failur e (Ib ) (Ib/ft )

11 5 54 01 2 43 2
22 0 72 01 8 64 8
33 0 108 02 3 82 8
46 0 216 04 7 169 2
5 12 0 432 09 3 334 8
Determine th e shear strength parameters c and 0.

/
4000-

//
3000-

c/f
C/3

£ 200 0 -
C/3 y
/
j3

1000-

^L
S^ A ^" 7 8 °

1000 2000 3000 4000 5000

2
Normal stress , a Ib/ft

Figure Ex . 8. 2
Shear Strengt h o f Soi l 273

Solution
The failure shea r stresses r^ a s obtained from th e tests are plotted against the normal stresses a , i n
Figure Ex 8.2. The shear parameters fro m th e graph are: c = 0, 0 = 37.8° .

Example 8. 3
A direct shear test, when conducted on a remolded sample of sand, gave the following observations
at th e tim e o f failure : Norma l loa d = 288 N; shea r loa d = 17 3 N. The cros s sectiona l are a o f the
sample = 36 cm2.
Determine: (i ) the angle of internal friction, (ii ) the magnitude and direction o f the principal
stresses i n the zone of failure.

Solution
Such problem s ca n b e solve d i n tw o ways , namel y graphicall y an d analytically . The analytica l
solution has been lef t a s an exercise fo r the students.
Graphical Solutio n
173
(i) Shea r stres s T =
= 4.8 N/cm 2 = 48 k N / m2
36
288
Normal stress a = —= 8.0 N / cm2 = 80 k N / m2
36

We know one point on the Mohr envelope. Plot point A (Fig. Ex. 8.3) with coordinates 1 -
48 kN/m2, and o= 80 kN/m2. Since cohesion c = 0 for sand, the Mohr envelope OM passes
through the origin . The slop e o f OM gives the angl e of internal friction (j) = 31° .
(ii) I n Fig. Ex. 8.3, draw line AC norma l to the envelope O M cutting the absciss a a t point C .
With C as center, and AC as radius, draw Mohr circle C l which cuts the abscissa a t points B
and D, which gives

120

80 Mohr circle C\

Major principa l plane

40 C2

40 F 8 0C 12 0 160 200
a, kN/m 2

Figure Ex . 8. 3
274 Chapte r8

major principa l stress = O B = (J l = 163.5 kN/m 2

minor principal stress = OD = <J 3 = 53.5 kN/m 2
Now, ZACB = 2cc = twic e th e angl e betwee n th e failur e plan e an d th e majo r principal
plane. Measurement gives
2a= 121 ° or a- 60.5 °
Since i n a direc t shea r tes t th e failur e plane i s horizontal , th e angl e mad e b y th e majo r
principal plan e wit h th e horizonta l wil l b e 60.5° . Th e mino r principa l plan e shoul d b e
drawn at a right angle to the major principal plane.
The direction s o f th e principa l planes ma y als o b e foun d by locatin g th e pol e P o. P o i s
obtained b y drawing a horizontal line from poin t A which is parallel to the failure plane in
the direct shear test. Now PE an d P(D giv e the directions of the major and minor principal
planes respectively.

8.15 EFFECTIV E STRESSES

So far, the discussio n ha s bee n base d o n consideratio n o f tota l stresses . I t i s t o b e note d tha t th e
strength an d deformatio n characteristic s o f a soi l ca n b e understoo d bette r b y visualizin g it a s a
compressible skeleto n o f solid particles enclosing voids. The void s may completely b e fille d with
water o r partl y wit h water and air . Shea r stresse s ar e t o b e carrie d onl y b y th e skeleto n o f soli d
particles. However , th e tota l norma l stresse s o n an y plan e are , i n general , th e su m o f tw o
components.
Total norma l stres s = component of stress carried b y soli d particle s
+ pressure in the fluid i n the voi d space .
This visualizatio n of th e distributio n o f stresse s betwee n soli d an d flui d ha s tw o important
consequences:
1. Whe n a specimen of soil is subjected to external pressure, the volume change of the specimen
is not due to the total normal stress but due to the difference between the total normal stres s
and the pressure of the fluid in the void space. The pressure in the fluid is the pore pressure u.
The difference which is called the effective stres s d ma y now be expressed a s
tf = cr-u (8.34 )
2. Th e shea r strengt h o f soils , a s o f al l granula r materials , i s largel y determine d b y th e
frictional force s arisin g during slip at th e contact s betwee n th e soi l particles . Thes e ar e
clearly a functio n o f th e componen t of norma l stres s carrie d b y th e soli d skeleto n rathe r
than o f th e tota l norma l stress . Fo r practica l purpose s th e shea r strengt h equatio n o f
Coulomb i s given by th e expression

s = c' + (o - U ) tan </)' = c' + a' tan </)' (8.35 )

where c' = apparent cohesion i n terms of effective stresse s
0' =angle of shearing resistance i n terms of effective stresse s
<7 = total norma l pressure to the plan e considere d
u = pore pressure.
The effectiv e stres s parameter s c' an d 0 ' of a give n sampl e o f soi l ma y b e determine d
provided th e pore pressure u developed durin g the shea r tes t i s measured . Th e por e pressur e u is
developed whe n th e testin g o f th e soi l i s don e unde r undraine d conditions . However , i f fre e
Shear Strengt h o f Soi l 27 5

drainage take s plac e durin g testing, there wil l not b e an y developmen t o f pore pressure . I n suc h
cases, th e total stresses themselve s are effective stresses .

8.16 SHEA R STRENGT H EQUATIO N I N TERM S O F EFFECTIV E

PRINCIPAL STRESSE S
The principal stresses ma y be expressed eithe r as total stresses o r as effective stresse s i f the values
of pore pressur e ar e known.
If u is the pore pressur e develope d durin g a triaxial test, we may write as befor e

o = o, -u

where a j and <5' 3 are the effective principa l stresses. The equation for shear strength in terms
of effective stresse s i s

<7,' — <7o G<— (J-, <J, — (T-.

s = — si n 2a = — si n 2a = — ; co s 0 (8.37 )
2 2 2
where 2a= 90° + 0'
Coulomb's equatio n in terms of effective stresse s i s
s = c''+ (<7-u) tan 0'

(7, — (J~.
Therefore, — cos<z> ' = c' + (er-u) tan0'

Since, cr =
22

we hav e — co s (/)' = c' H— l- ta n <f)'

+ — cos(9 0 + 0') ta n 0' - u tan 0'

Simplifying

<7,. —
- cr-. , , . . . cr , + or, . , O" , -1
-wsn
2 2 2
1c ' cos$)' + (<73 — «) sin^'
or

where (ci j - cr 3) indicate s th e maximum deviato r stres s a t failure . E q (8.38 ) ma y also b e

expressed i n a different for m as follows by considering effectiv e principa l stresse s

1, , c' cos^' + a. sin^'

— l(<j, 3 -cr,f ) , = - - -
2 1-si n

or —
276 Chapte r8

Simplifying, w e have

(o[ -o' 3)f = (o{ + o'3 ) sin (/)' + 2c' cos 0' (8.39 )

8.17 STRESS-CONTROLLE D AN D STRAIN-CONTROLLE D TEST S

Direct shear tests or triaxial compression test s may be carried ou t by applying stresses o r strains at
a particularly known rate. When the stress is applied at a constant rate it is called a stress-controlled
test an d whe n th e strai n i s applie d a t a constan t rat e i t i s calle d a strain-controlled test. Th e
difference betwee n the two types of tests may be explained with respect t o box shea r tests .
In the stress-controlled test [Fig. 8.15(a) ] the lateral load F a whic h induces shear is gradually
increased unti l complete failur e occurs. Thi s ca n b e don e b y placin g weight s o n a hange r o r by
filling a counterweighte d bucke t o f origina l weigh t W a t a constan t rate . Th e shearin g
displacements ar e measured by means of a dial gauge G as a function o f the increasing load F . The
shearing stress a t any shearing displacement, is

where A i s th e cros s sectiona l are a o f th e sample . A typica l shap e o f a stress-strai n curv e o f th e

stress-controlled tes t is shown in Fig. 8.15(a) .
A typical arrangement of a box-shear test apparatus for the strain-controlled tes t is shown in
Fig. 8.15(b) . Th e shearin g displacements ar e induce d and controlle d i n suc h a manne r tha t they
occur at a constant fixed rate. This can be achieved by turning the wheel either by hand or by means
of any electrically operate d motor s o that horizontal motio n is induced throug h th e worm gea r B .
The dial gauge G gives the desired constant rate of displacement. The bottom o f box C is mounted
on frictionless rollers D . The shearing resistance offere d to this displacement b y the soil sampl e is
measured b y the proving ring E. The stress-strain curves for this type of test have the shape shown
in Fig. 8.15(b) .
Both stress-controlled an d strain-controlled types of test are used in connection wit h all the
direct triaxia l and unconfine d soi l shea r tests . Strain-controlle d test s ar e easier t o perfor m an d
have the advantage of readily giving not only the peak resistance a s in Fig. 8.1 5 (b ) but also the
ultimate resistance whic h is lower than the peak suc h as point b in the same figure, whereas th e
stress controlle d gives only the peak values but not the smaller values after the peak is achieved .
The stress-controlle d test is preferred onl y in some special problem s connecte d wit h research.

8.18 TYPE S O F LABORATOR Y TEST S

The laborator y tests on soils may be on
1. Undisturbe d samples, o r
2. Remolde d samples .
Further, the tests may be conducted on soils that are :
1 . Full y saturated, or
2. Partiall y saturated.
The typ e of test to be adopted depend s upo n how best w e can simulate the field conditions.
Generally speaking , th e various shear tests for soils may be classified as follows :
Shear Strengt h o f Soi l 277

Dial
gauge

Displacement
\_-

(b) Strain controlled

Figure 8.1 5 Stres s an d strai n controlle d bo x shea r tests

1. Unconsolidated-Undraine d Test s (UU )

The samples ar e subjected to an applied pressur e unde r conditions in which drainage i s prevented,
and then sheared under conditions o f no drainage.

2. Consolidated-Undraine d or Quick Test s (CD)

The sample s ar e allowe d t o consolidat e unde r a n applie d pressur e an d the n sheare d unde r
conditions of no drainage .

3. Consolidated-Draine d o r Slow Test s (CD)

The sample s ar e consolidated a s in the previous test, but the shearing i s carried out slowly under
conditions of no excess pressure i n the pore space .
The drainage conditio n of a sample i s generally th e deciding factor in choosing a particular
type o f tes t i n th e laboratory . Th e purpos e o f carryin g ou t a particula r tes t i s t o simulat e fiel d
conditions a s fa r a s possible . Becaus e o f th e hig h permeabilit y o f sand , consolidatio n occur s
relatively rapidl y an d i s usuall y completed durin g the applicatio n o f th e load . Test s o n san d ar e
therefore generall y carried out under drained conditions (drained or slow test) .
For soils othe r tha n sands the choice o f test conditions depend s upo n the purpose fo r which
the shear strengt h i s required. The guiding principle i s that drainage conditions o f the test should
conform a s closely as possible t o the conditions unde r which the soils wil l be stressed i n the field .
Undrained o r quick test s ar e generall y use d fo r foundation s on clay soils , sinc e durin g the
period of construction only a small amount of consolidation will have taken place and consequently
the moisture content wil l have undergone little change. For clay slopes o r cuts undrained tests are
used both for design an d for the investigation of failures.
Consolidated-undrained test s are used where changes i n moisture content are expected t o take
place due to consolidation before the soil is fully loaded. An important example is the condition known
as "sudden drawdown" such as that occurs in an earth dam behind which the water level is lowered at
278 Chapte r8

a faster rate than at which the material of the dam can consolidate. In the consolidated-undrained test s
used in this type of problem, the consolidation pressure s ar e chosen to represent the initial conditions
of the soil, and the shearing loads correspond to the stresses calle d into play by the action of sudden
drawdown.
As already stated , drained tests ar e always used i n problems relatin g to sandy soils. In clay
soils drained tests are sometimes used in investigating the stability of an earth dam, an embankment
or a retaining wall after a considerable interval of time has passed .
Very fine sand, silts and silty sands also have poor drainage qualities. Saturated soils of these
categories ar e likely to fail i n the field unde r conditions similar to those under which consolidated
quick tests are made.

Shearing Tes t Apparatu s fo r th e Variou s Type s o f Test s

The various types of shear tests mentioned earlier may be carried ou t either by the box shear test or
the triaxial compression test apparatus. Tests that may be made by the two types of apparatus are :

Box Shea r Tes t Apparatu s

1. Undraine d and consolidated- undraine d tests on clay samples only.
2. Draine d o r Slow tests on any soil.
The bo x shea r test apparatus is not suite d fo r undrained or consolidated-undrained test s on
samples other than clay samples, because the other soils are so permeable that even a rapid increase
of the stresses i n the sample may cause at least a noticeable chang e of the water content.

Triaxial Compressio n Tes t Apparatu s

All types of tests can conveniently be carried ou t in this apparatus.

8.19 SHEARIN G STRENGT H TEST S O N SAN D

Shear tests on sand may be made when the sand is either in a dry state or in a saturated state. No test
shall b e mad e whe n th e soi l i s in a moist stat e a s this state exist s onl y due t o apparen t cohesio n
between particle s whic h woul d be destroye d whe n i t i s saturated . Th e result s o f shea r test s o n
saturated samples are almost identical with those on the same sand at equal relative density in a dry
state except tha t the angle 0 is likely to be 1 or 2 degrees smaller for the saturated sand .
The usua l typ e o f tes t use d fo r coars e t o mediu m san d i s th e slo w shea r test . However ,
consolidated undraine d tests ma y be conducted on fine sands , sand y silts etc. whic h do not allow
free drainag e unde r changed stress conditions. If the equilibrium of a large body o f saturated fin e
sand i n a n embankmen t is disturbe d by rapi d drawdown of th e surfac e o f a n adjoinin g body of
water, the change i n water content of the fil l lag s behind the change i n stress .
In all the shearing tests on sand, only the remolded samples are used as it is not practicable to
obtain undisturbed samples. The soil samples are to be made approximately to the same dry density
as it exists in-situ and tested either by direct shear or triaxial compression tests .
Tests on soils are generally carried out by the strain-controlled type apparatus. The principal
advantage o f this type of test on dens e san d is tha t its peak-shear resistance, a s well a s the shea r
resistances smalle r than the peak, can be observed an d plotted.

Direct Shea r Tes t

Only th e draine d o r th e slo w shea r test s o n san d ma y b e carrie d ou t b y usin g th e bo x shea r tes t
apparatus. The box i s filled wit h san d to the required density. The sampl e i s sheared at a constant
Shear Strengt h o f Soi l 279

vertical pressure a. The shear stresses are calculated at various displacements of the shear box. The
test is repeated with different pressure s <7 .
If th e sampl e consist s o f loos e sand , th e shearin g stres s increase s wit h increasin g
displacement until failure occurs. If the sand is dense, the shear failure of the sample is preceded by
a decreas e o f th e shearin g stres s fro m a peak valu e to a n ultimate value (also know n as residual
value) lower than the peak value.
Typical stress-strain curves for loose and dense sands are shown in Fig. 8.16(a).
The shear stress of a dense sand increases from 0 to a peak value represented b y point a, and
then gradually decreases and reaches an ultimate value represented b y point b. The sample of sand
in a dense state is closely packed and the number of contact points between the particles ar e more
than in the loose state. The soil grains are in an interlocked state. As the sample is subjected to shear
stress, th e stres s ha s t o overcom e th e resistanc e offere d b y th e interlocke d arrangemen t o f th e
particles. Experimenta l evidence indicates that a significant percen t of the peak strengt h is due to
the interlocking of the grains. In the process o f shearing one grain tries to slide over the other and
the voi d rati o o f th e sampl e whic h i s th e lowes t a t th e commencemen t o f th e tes t reache s th e
maximum value at point a, in the Fig 8.16(a). The shear stress also reaches th e maximum value at
this level . An y furthe r increas e o f strai n beyon d thi s poin t i s associate d wit h a progressiv e
disintegration o f th e structur e of th e san d resulting in a decrease i n th e shea r stress . Experienc e
shows that the change i n void ratio due to shear depends on both the vertical loa d an d the relative
density of the sand. At very low vertical pressure, the void ratio at failure is larger and at very high
pressure it is smaller than the initial void ratio, whatever the relative density of the sand may be. At

Peak value
Dense san d
b ultimat e value

Displacement

(a) Shear stres s vs displacement

Dense san d

Loose sand
0
Normal stress, a

(b) Volume change (c) Shear strength vs normal stres s

Figure 8.1 6 Direc t shea r test o n sand

 280 Chapte r8

Table 8. 1 Typica l value s o f 0 and (j) u fo r granula r soil s

Types o f soi l 0 de g 0udeg

Sand: rounded grains

Loose 28 to 30
Medium 30 to 35 26 to 30
Dense 35 to 38
Sand: angular grains
Loose 30 to 35
Medium 35 to 40 30 to 35
Dense 40 to 45
Sandy gravel 34 to 48 33 to 36

intermediate value s of pressure, the shearing force causes a decrease in the void ratio of loose sand
and a n increase i n the voi d rati o o f dense sand . Fi g 8.16(b ) show s ho w the volum e of dense sand
decreases u p t o a certai n valu e o f horizonta l displacemen t an d wit h furthe r displacemen t th e
volume increases , wherea s i n th e cas e o f loose san d th e volum e continues to decreas e wit h a n
increase i n th e displacement . I n saturate d san d a decrease o f th e voi d rati o i s associate d wit h an
expulsion of pore water, and an increase with an absorption of water. The expansion of a soil due to
shear a t a constan t valu e o f vertica l pressure i s calle d dilatancy. A t som e intermediat e stat e o r
degree of density in the process of shear, the shear displacement does not bring about any change in
volume, that is, density. The densit y of sand at which no change in volume is brought about upon
the applicatio n o f shea r strain s i s calle d th e critical density. Th e porosit y an d voi d rati o
corresponding t o th e critica l densit y ar e calle d th e critical porosity an d th e critical void ratio
respectively.
By plottin g the shea r strength s corresponding t o th e stat e o f failur e i n th e differen t shea r test s
against the normal pressure a straight line is obtained for loose sand and a slightly curved line for dense
sand [Fig . 8.16(c)] . However , fo r al l practica l purposes , th e curvatur e fo r th e dens e san d ca n b e
disregarded an d an average line may be drawn. The slopes of the lines give the corresponding angles of
friction 0 of the sand. The general equation for the lines may be written as
s = <J tan (f)
For a given sand , the angle 0 increases wit h increasin g relativ e density . For loose san d i t is
roughly equal to the angle of repose, defined as the angle between the horizontal and the slope of a
heap produced b y pouring clean dry sand from a small height. The angle of friction varie s with the
shape of the grains. Sand samples containing well graded angula r grains give higher values of 0 as
compared t o uniformly graded san d with rounded grains. The angl e of frictio n </ > for dens e san d at
peak shea r stres s i s higher than that at ultimate shear stress . Table 8.1 gives some typical values of
0 (at peak) an d 0 M (at ultimate).

Triaxial Compressio n Test

Reconstructed san d samples at the required density are used for the tests. The procedure of making
samples shoul d b e studie d separatel y (refe r to an y book o n Soi l Testing). Test s o n san d ma y b e
conducted eithe r in a saturated state or in a dry state. Slow or consolidated undraine d tests may be
carried ou t as required.

Drained o r Slo w Test s

At leas t thre e identica l sample s havin g the sam e initia l conditions ar e t o be used . Fo r slo w test s
under saturated conditions the drainage valve should always be kept open. Each sampl e shoul d be
Shear Strengt h o f Soi l 281

'jv:,:^-V^
: v•••:..-y<; A
-• v' -i .* • '' - '' • " '
• ••*». • '- x ' ' ' « ' "
• •> " "•.. , \." ' • ••> '

(a) Dense san d (b ) Loose san d

Figure 8.1 7 Typica l shape s of dens e an d loos e sand s a t failur e

Strain

(a) Stress-strain curves for three samples at dense state

Mohr
envelope

(b) Mohr envelop e

Figure 8.1 8 Moh r envelop e fo r dens e san d

282 Chapte r8

tested unde r differen t constan t all-round pressures fo r example, 1 , 2 and 3 kg/cm2. Each sampl e i s
sheared t o failure by increasing the vertical load a t a sufficiently slo w rat e t o prevent any build up
of exces s por e pressures .
At an y stag e o f loadin g th e majo r principa l stres s i s th e all-roun d pressur e <7 3 plu s th e
intensity of deviator stress (o { - cr 3). The actually applied stresse s ar e the effective stresses i n a slow
test, tha t is <7 } = a\ and O" 3 = <r' 3, Dense sample s fai l alon g a clearly define d ruptur e plane wherea s
loose sand samples fai l alon g many planes which result in a symmetrical bulging of the sample. The
compressive strengt h o f a sampl e i s define d a s th e differenc e betwee n th e majo r an d mino r
principal stresses a t failure (G I - <T 3),,. Typical shapes of dense and loose sand samples a t failure are
shown i n Fig. 8.17 .
Typical stress-strai n curve s for three sample s i n a dense stat e an d the Mohr circle s fo r these
samples a t peak strengt h are shown in Fig. 8.18 .
If the experiment i s properly carrie d ou t there will be one common tangen t to all these thre e
circles an d this will pass throug h the origin. This indicates that the Mohr envelop e i s a straight line
for san d and the sand has no cohesion. The angle made by the envelope wit h the a-axis is called th e
angle o f interna l friction . The failur e planes fo r eac h o f thes e sample s ar e show n i n Fig. 8.18(b) .
Each o f them mak e a n angle a wit h the horizontal which is approximately equal t o
a = 45° + 0/2
From Fig . 8.18(b ) an expression for the angle of internal friction ma y be writte n as

- (J 3 (T j / <7 3 - 1
(840)
{Q
-™ }

Example 8. 4
Determine th e magnitude of the deviator stress if a sample of the same sand with the same void ratio
as given in Ex. 8. 3 was tested i n a triaxial apparatus with a confining pressur e o f 60 kN/m 2.

Solution
In th e case of a triaxial test on an identical sampl e of san d as given in Ex. 8.3 , us e the sam e Moh r
envelope O M (Fig . Ex . 8.3) . No w th e poin t F o n th e absciss a give s th e confinin g pressur e
<73 = 60 kN/m 2 . A Mohr circl e C 2 may no w be drawn passing through point F and tangentia l t o the
Mohr envelop e OM . Th e poin t E gives the major principal stress <J } fo r th e triaxia l test.
Now cr j = O E = 188 kN/m 2 , <7 3 = 60 kN/m 2
Therefore a l - <7 3 = 188 - 6 0 = 128 kN/m2 = deviator stres s

Example 8. 5
A consolidated draine d triaxia l test was conducted on a granular soil. At failure cr'/o ^ = 4.0. Th e
effective mino r principa l stres s a t failur e wa s 10 0 kN/m 2 . Comput e 0 ' an d th e principa l stres s
difference a t failure.

Solution

-j -1 4- 1
sin<z$' = —; ~ ==~ 0.6 o r 6' - 37°
cr,7cr3 +1+ 414 + 1

The principa l stres s differenc e at failure is