SOILS AND FOUNDATIONS Vol. 45, No. 5, 81-88, Oct.

2005
Japanese Geotechnical Society

ESTIMATING MODULI AND FRICTION ANGLE FROM SELF-BORING

PRESSUREMETER TESTS IN SAND

RAMESH C. GUPTAi)

ABSTRACT

Self-boring pressuremeter test (SBPMT) data has been analyzed based on cylindrical cavity expansion method and
dimensionless cylindrical cavity expansion factors (F"q) to determine angle of friction, rigidity index (/r) and modulus
of elasticity (E). Then, based on a hyperbolic model, secant modulus (Esf), secant modulus at 50 percent failure stresses
(E50),and initial modulus (E i) have been determined. Reasonable values of angle of friction, initial modulus (E1), and
secant modulus of deformation at failure (Esf)have been determined for sand deposits located in different geographical
regions. The secant modulus (E50) determined by this method compares well with unload-reload modulus, Er,
determined from unload-reload cycle of the SBPM tests.

Key words: expansion, friction angle, pressuremeter test, sand, Young's modulus (IGC: E2/E4)

be used to create a nonlinear stress-strain model or in

INTRODUCTION other words how this method can be expanded to deter-
Self-boring pressuremeter (SBPM) has been used mine secant modulus at failure (Esf), secant modulus at
widely to determine soil properties of sand. The main 50% failure stresses (E50), and initial tangent modulus
advantage of the SBPM is that it can be installed with (Ei) . In geotechnical design and practice, E50 and Ei are
minimized initial disturbance, and therefore provides frequently used to determine settlement related problems,
tests and geotechnical parameters of sands in almost un- and stress distribution under footings in working stress
disturbed conditions, and in their natural environment. conditions. In this paper, a method has been introduced
When it is expanded, it simulates the cylindrical cavity to determine, Esf, E50 and E i from SBPMT data, after
expansion from a finite radius; with this view, Gupta completing the analysis given in the companion paper
(2005) performed a finite strain analysis to determine (Gupta, 2005).
volumetric strains in the plastic zone using cylindrical
cavity expansion from finite radius so that an analysis
of SBPMT data could be performed accurately. This HYPERBOLIC NONLINEAR STRESS-STRAIN
analysis is based on Vesic's cavity expansion theory MODEL

(Vesic, 1972), which assumes a linear stress-strain Results of the triaxial compression tests and screw
relationship during expansion of cavity from beginning plate load tests have shown that the stress-strain relation-
of expansion to the stage when the state of equilibrium is ship of sand is nonlinear. It is quite difficult to determine
reached at the face of cavity. On this basis, a method of the initial modulus (E i) accurately from such tests, since
determining angle of friction (4)) and modulus of the slope of the stress-strain curve changes rapidly even at
elasticity (E) , by a technique of matching dimensional very small strains. The hyperbolic model (Kondner, 1963;
cavity expansion factor (F" q) versus circumferential Duncan and Chang, 1970; Desai and Christian, 1977),
strain (e0,9)curve obtained from a SBPM test with that which was used for triaxial compression tests, can also be
obtained from the theoretical analysis at selected values used to approximate the nonlinear stress-strain curves for
of 4) and E, has been introduced (Gupta, 2005). Here it cylindrical cavity expansion, as described below:
may be noted that F " , is equal to internal cavity pressure, At the face of cavity, prior to reaching state of
p(t), applied during a SBPM test at any instant of time, equilibrium, radial displacement (r,) and circumferential
divided by effective horizontal stress (a' h), measured strain, (e00) are given by (Baguelin, 1978):
during the SBPM test. However, real soils exhibit non-
linear stress-strain relationship even in the elastic state (1)
(Duncan and Chang, 1970; Mayne, 2001). Therefore, it
becomes necessary to examine how the above method can
i)
Ph.D., Senior Geotechnical Engineer, Structure & Bridge Div., Virginia Department of Transportation, 1401 East Broad St., Richmond, VA
23219, USA ([email protected]).
The manuscript for this paper was received for review on August 27, 2004; approved on July 20, 2005.
Written discussions on this paper should be submitted before May 1, 2006 to the Japanese Geotechnical Society, 4-38-2, Sengoku, Bunkyo-ku,
Tokyo 112-0011, Japan. Upon request the closing date may be extended one month.

81
82 GUPTA

(2)

It may be noted that in elastic phase of cavity expan-

sion, e00is equal to-err, where err is radial strain. When
state of equilibrium has taken place at the face of cavity,
peak stresses are related by (Vesic, 1972; Gupta, 2005):
(3)
Prior to reaching a state of equilibrium at the face of
cavity, nonlinear stress-strain hyperbolic model for (a)
cylindrical cavity expansion can be represented by:

(4)

At the initial beginning, strains are small, even less

than 10' percent, such as found during wave propaga-
tion, and therefore, at that instant, the product aeo
becomes very small and can be neglected; then Eq. (4)
becomes:

(5)
(b)
Ei is the initial modulus. Substituting value of b from
Eq. (5) in Eq. (4), nonlinear stress-strain relationship Fig. 1. During cylindrical cavity expansion in medium compressibility
becomes: (MC) sand, hyperbolic model for stress-strain relationship, at
σ'o=200 kPa, (a) φ=40° and (b) φ=32°

(6)

At very large strains, the product aeoobecomes very tangent to the hyperbolic curve at this point represents

large compared to 11Ei, and therefore, at that instant, initial modulus, Ei. At Point C, peak stresses, (ƒÐrr-ƒÐƒÆƒÆ)f,

1/Ei can be neglected, then, Eq. (6) becomes: that cause Mohr circle to touch Coulomb-Mohr enve-

lope, are developed and therefore Point C represents

(7) (σrr-σ θθ)f as ordinate and coo of real soil at that instant as
abscissa. Figure 2 represents state of stresses and Mohr

Where 1 /a or (ƒÐrr-ƒÐƒÆƒÆ)asymp in Eq. (7) represents circles during cylindrical cavity expansion. In Fig. 2,

asymptotic value of the hyperbolic model. Substituting Mohr circle 1 represents the stresses at 0.5 (ƒÐrr-ƒÐƒÆƒÆ)f,

this value of 'a' in Eq. (6), and rearranging terms, hyper- whereas Mohr circle 2 represents the stresses at (ƒÐrr-ƒÐƒÆƒÆ)f,

bolic model representing the nonlinear stress-strain and therefore this circle touches Mohr-Coulomb's failure

relationship is given by: envelope. Point B in Fig. 1 represents 0.5 (arr - 600)f as

ordinate and ƒÃƒÆƒÆ value of real soil exhibiting nonlinear

(8) response as abscissa at that level of stresses. Similarly,

Point C in Fig. 1 represents (ƒÐrr-ƒÐƒÆƒÆ)f as ordinate and coo

value of real soil exhibiting nonlinear response as

Since the compressive strength or peak stresses, (arr abscissa at that level of stresses. The cavity expansion

σ θθ)f during expansion of cavity, will be achieved before analysis made in the companion paper (Gupta, 2005) uses

the hyperbolic curve becomes asymptotic i.e. before the linear Path AC for linear stress-strain relationship in

onset of achieving asymptotic value,(ƒÐrr ƒÐƒÆƒÆ)asymp, it is place of nonlinear Path ABC of real soil. Slope of Path
customary to require that (ƒÐrr ƒÐƒÆƒÆ)f is related to (arr AC represents secant modulus at failure, Esf. With this

σ θ
θ)asymp by failure factor, Rf, (Duncan and Chang, 1970): view, it is assumed that E used and determined in the

companion paper (Gupta, 2005) represents secant

(9) modulus at failure, Esf. The slope of Path AB represents

secant modulus at 50% failure stresses (E50). Path CD of

Fig. 1 represents extension of the hyperbolic curve

beyond Point C, to show that if failure conditions

MODULUS OF ELASTICITY AT VARIOUS LEVELS represented by peak stresses had not reached at Point C,
OF STRESS stress-strain relationship would have continued to be

Hyperbolic model as defined by Eq. (8) is shown by represented by hyperbolic path CD.

curve ABCD in Fig. 1. At Point A, eoe equals zero and the It may be noted that the laboratory tests have shown
 ESTIMATING MODULI AND FRICTION ANGLE 83

(13a)
or

(13b)

It may be noted that E50 has been considered to be

approximately equal to modulus determined from
unload-reload cycle during SBPM tests (Wroth, 1984). It
can also be seen that relationships between Esf and Ei and
between Esf and E50 or between E50 and Ei are dependent
on the value of failure factor Rf. These relationships can
only be used when the value of Rf is known.
Fig. 2. During cylindrical cavity expansion, stress path, states of
stresses and Mohr-Coulomb envelope
Esf, Ei and FAILURE FACTOR, Rf
As shown below, it is possible to estimate Esf using the
existing correlations and Ei using wave equation theory
that at any state of stress, strain is not dependent upon and then value of Rf can be estimated using Eqs. (11b) or
the stress path (i.e. linear or nonlinear) between the two (13). For normally consolidated sands (Schmertmann,
points, but depends on stresses at the initial such as A and 1978; Mayne, 2001; Baldi et al. , 1989), secant modulus
the subsequent point of the stress path such as B or C, (Esf) or constrained modulus (M) can be estimated based
under consideration (Lambe and Whitman, 1969; Vesic, on correlations developed between angle of friction,
1972). relative density (DR), cone tip resistance (q.), effective
As shown below, using the above concept, values of E50 overburden stress WO, effective octahedral stress (a'oct),
and Ei can be determined by using hyperbolic model as and atmospheric pressure (Pa) as given below:
defined by Eq. (8). Using E = Es f, when the state of
equilibrium is reached at the face of cavity or at the
interface of elastic and plastic zone, such as at point C in for high compressibility sand (HC) (14a)
Fig. 1, circumferential stress can be written as (Gupta,
2005):
for medium compressibility sand (MC) (14b)
(10)
for low compressibility sand (LC) (14c)
Where cep denotes circumferential strain at the inter-

face of the elastic and plastic zones or at the face of cavity From Schmertmann (1978)
when the state of equilibrium is reached for the first time.

For Point C of Fig. 1, the expression as given in Eq. (10) (15)

for goo i.e. for eop, and the expression as given in Eq. (3)

for (ƒÐrr-ƒÐƒÆƒÆ)i.e. (arr 600)f can be substituted in Eq. (9); From Mayne (2001)
and after rearranging terms, Esf or Rf are given by:
(16)
(11a)
From Baldi et al. (1989)
or Using Eqs. (15), (14a), (14b), and (14c), it can be
shown that terms qc and Dr can be eliminated from
(11b) Eq. (16), to express M in terms of .95,Pa, a'o and 6'0,, by
the following:
Similarly, the relationship between Ei and E50 can also

be determined. For Point B of Fig. 1, eee at 0.5 (ƒÐ1-ƒÐ3)f

is given by (Gupta, 2005):

(17a)
(12)

As stated before, E50, represented by linear Path AB of

Fig. 1, is secant modulus at half of the value of (arr
σ θθ)f. Substituting (ƒÐrr-ƒÐƒÆƒÆ)= 0.5 (arr - 0-00)f and the

above value of e000.50 in Eq. (9) and arranging terms, (17b)

following expressions for E50 in terms of Ei or Esf are
obtained:
84 GUPTA

(17c)

Where: qc, E, M, P a, - 'o, U'oct are in kPa. Relative

density, DR is in percentage. v is Poisson's ratio. y, is in
kN/m3. Vsis in m/s. Depth, z is in m. E and Mare related
by well-known equation:

(18)
(a)

Initial modulus (Ei) can be estimated based on wave

propagation theory. Baldi et al. (1989) performed exten-
sive tests on sand and established a correlation between
shear wave velocity ( Vs), cone tip resistance (q,), and
effective overburden stress (0-'0), as shown in Eq. (19).
Similarly, Mayne (2001) performed numerous tests to
establish a correlation between shear wave velocity ( Vs)
and total unit weight (yt) as shown in Eq. (20).

(19)
From Baldi et al. (1989)

(20) (b)
From Mayne (2001)
Using Eq. (15), an equation for qc in terms of q and a'0 Fig. 3. For high compressibility sand (HC) at effective overburden
stress (7'0) of 100 and 200 kPa, (a) Ratios E50/Ei, Esf/Ei,Rf versus
can be obtained and substituted in Eq. (19) and after
φ curves and (b) Ei,E50,Esf versus φ curves
rearranging terms, the relationship for Vs in terms of 4),
Pa, and u'o becomes:

(21)

Finally, after determining Vs and yt, initial modulus

(Ei) and shear modulus (Gmax)are calculated using follow-
ina well-known relatinnchip.

(22)

Where g is gravitational acceleration constant and is

equal to 9.8 m/s2. Using above equations, values of Ei,
Esf and E50were calculated at a'0 of 100 and 200 kPa, for (a)
various values of 0, and the Ei, Esf and E50 versus 4
relationships are obtained as shown in Figs. 3, 4 and 5. It
may also be noted that typical values of Esf or E0 at an
arbitrary stress of 100 kPa range from 10 to 25 MPa for
loose sand, 20 to 60 MPa for medium dense sand and 60
to 100 MPa for dense sand (Coduto, 1994). It can be
shown that these typical values of E0 or Esf and the values
of Est-calculated from above equations are about in the
same range, see Figs. 3, 4 and 5.
Values of failure factor, Rf, which is defined by
Eq. (1 lb), can now be estimated because values of Ei and
Esf can be determined using Eqs. (22) and (18). In Figs. 3, (b)
4, and 5, values of Rf at u'o of 100 and 200 kPa, for
values of 4) have been shown. In these figures, estimated Fig. 4. For medium compressibility sand (MC) at effective overburden
values of Rf are found to vary from 0.5 for very dense stress (6'0) of 100 and 200 kPa, (a) Ratios E50/E;, Esf/Ei, Rf versus
sand to about 0.85 for loose sands. Analyzing the triaxial φ curves and (b) Ei,E50,Esf versus φ curves
 ESTIMATING MODULI AND FRICTION ANGLE 85

(a)

Fig. 6. When circumferential strain at the face of cavity is 10%,

relationship of cylindrical cavity expansion factor, F" ,, with angle
of friction and effective overburden stress

where q' = ath.

In this way, for normally consolidated sands, Ir
becomes dependent only on 4 and o-'0. Therefore, at
(b) selected values of 4 and 0-'0, Ir can be calculated, and then
value of F" q can be calculated at coo= 10%, using the
Fig. 5. For low compressibility sand (LC) at effective overburden procedure explained in the companion paper, Gupta
stress (a'0) of 100 and 200 kPa, (a) Ratios E50/Ei,Esf/Ei Rf versus (2005). The F", versus curve at 6'0 = 50, 100 and 200
φCUrVeS and(b)Ei, E50, E,f VerSUSφCUrVeS kPa, when coois equal to 10% is shown in Fig. 6.

BASIS OF ANALYZING SBPMT DATA

compression tests, Duncan and Chang (1970) had found
the value of Rf varying between 0.6 and 0.9 for dense to As explained above, at selected values of 4 and Ir,
loose sands. dimensionless factor (F" q) can be calculated for various
values of circumferential strain (e00). Internal pressure,
p(t), applied during SBPM test and corresponding in-
RELATIONSHIP BETWEEN F" , and 4 crease in radius of cavity, which occurs at applied internal
Dimensionless cavity expansion factor (F" 0 is a varia- pressure, p(t), is utilized to develop F" q versus coocurve.
ble and relationship between F" a and (/)is dependent on F", is equal to p(t)lo-'h. a'h is measured during a SBPM
Ir and also on the value of circumferential strain (coo), see test. coois equal to [{R(t)-Ri} R(t) is radius of the
companion paper (Gupta, 2005). To find a relationship cavity at time t, when internal pressure is p(t). Ri is the
between F" q and 4), an effort was made to estimate F" , initial radius of the cavity, i.e. initial radius of the SBPM
versus 4 curve at a selected value of gee, say at 10%, and probe prior to application of the internal pressure just
also at three selected values of 6'0, for normally consoli- above the value of in-situ ash. In this way, F" q versus co
dated sands. For selected type of sand (LC, MC, or HC), curve obtained from SBPMT data is compared with
the constrained modulus, M, can be calculated using theoretical F" q versus coo curve developed at selected
Eqs. (17a) or (17b) or (17c), and the secant modulus, Esf values of Ir and (/). Several trials are made, and the value
can be calculated using Eq. (18). It can be seen that the of Ir and that produces a theoretical F" qversus coocurve
value of Esf determined as above, is dependent only on closely matching with actual F" qversus coocurve obtained
variables, 0, 0-'0, and aoct. As previously stated, the value from SBPMT is selected. It may be mentioned that be-
of Esf has been assumed equal to Young's modulus of cause there are two variables, and Ir, it becomes difficult
elasticity, E. When value of E based on these equations is to find an optimum set of 4)and Ir. This has to be decided
substituted in the expression for rigidity index, Ir, which based on judgment. Author first determined F", value at
is defined by Eq. (23), it can be seen that Ir also becomes εθθ=l0% from the SBPMT data and then at that deter-
dependent on variables, 0, 0-'0, and a'oct. o-'oct is equal to mined value of F" q and o-'0 at the depth where SBPM
σ,0(l+2K0)/3,and K0 for normally consolidated sand is tests was conducted, an approximate value of cb was
equal to (1-sin 4)). determined from Fig. 6. Using this value of 0, and select-
ed values of Ir, theoretical F" q versus eo curve were deter-
(23) mined and matched with actual F" q versus eoe curve,
obtained from SBPM test. Generally, it was found that
86 GUPTA

optimum value of 0 was within •} 0.5 degrees of 4 deter- Eqs. (23) and (13b), and 11(a), respectively as shown in

mined from Fig. 6. In this way, an optimum set of 0 and Table 1. Value of E50 compares well with value of Er

Ir was determined. From known values of 4) and /r, the determined from unload/reload cycles. The value of Est.
value of E can be determined from Eq. (23). Poisson's compares well with typical values of Esf for dense sands.
ratio, v, is generally assumed as 0.3 for SBPM tests. The value of 4 selected as 38•‹ matches with that esti-

As already mentioned that the value of E so determined mated by Bruzzi et al. (1986).

is assumed equal to secant modulus, Esf, at failure, i.e.

peak stresses at the state of equilibrium. Once, the value Shotter Paper Mill Sand

of Esf is determined, an effort is made to determine the Houlsby et al. (1986) performed SBPM tests at a depth

value of E50 and E. With this objective, an approximate of 5.8 m in uniformly graded fine to coarse sand at the

value of Rf is estimated from Figs. 3, or 4 or 5, based on site of Shotter Paper Mill, North Wales, U.K. The

estimated value of 4) and known value of (7'0. Knowing ground water level was at a depth of 3.8 m. Houlsby et al.
the value of Rf, the value of E50 can be determined from (1986) stated the value of y, o-'0, K0, and 4 as 18 kN/m3,
Eq. (13b) and value of El from Eq. (11a). On this basis, 84.8 kPa, 0.9 and 38 degrees, respectively. A good match

SBPMT data of several sites were analyzed to determine between measured curve and theoretical curve derived at

values of 0, Esf, and E50, as described hereunder. Ir = 380, K0 = 0.9 and 4) = 38•‹ was obtained as shown in

EXAMPLES

SBPM tests in sands have been performed worldwide.

Some published test data/results of SBPM tests per-

formed by Bruzzi et al. (1986), Houlsby and Clark (1986),

Robertson and Hughes (1986) and Wroth (1984), is

analyzed, using the method described in this paper.

Po River Sand

SBPM test data of Bruzzi et al. (1986) was analyzed

(a)
and the analyzed results for a SBPM test performed at a

depth of 17.9 m in Po river medium to coarse sand, are

shown in Fig. 7(a) and Table 1. The depth to water table

was 0.5 m. The values of YT, a'o, 4), and shear modulus

(Gur) estimated from unload-reload cycles by Bruzzi et al.

(1986) are shown in this table. The coefficient of earth

pressure at rest (K0) was stated to be equal to 0.65 •} 0.18.

SBPMT data consisted of a curve of internal cavity

pressure, p(t), versus radial strain. Dividing p(t) by q',

F"q versus 800 curve was obtained as shown in Fig. 7(a).

An excellent match was found between measured data

and the theoretical curve derived using the above method (b)
when values of L, 0, and K0 were selected as 300, 38•‹ and
0.725, respectively. Based on values of (/) and 6'0, approx- Fig. 7. Comparison of F", versus goocurves determined from SBPMT
imate value of Rf as 0.68 was selected from Fig. 3. The data with theoretical curve for (a) Po River Sand (Bruzzi et al.,
values of Esf, E50 and El were then calculated using 1986) and (b) Shotter Paper Mill Sand (Houlsby et al., 1986)

Table 1. From SBPM tests in sands, 4 and Esf determined using finite strain analysis for cavity expansion (Gupta, 2005) and then, E50 and E, deter-
mined using hyperbolic model

Note: The values of y, o-'0, Ko, Gur and Er are from Bruzzi et al. (1986), Wroth (1984), Houlsby et al. (1986), and Robertson and Hughes (1986).
 ESTIMATING MODULI AND FRICTION ANGLE 87

(a)
(a)

(b)

Fig. 8. Comparison of theoretical curves with F" q versus e00 curves

determined from SBPMT performed in sand at McDonald's Farm (b)
(Robertson and Hughes, 1986) at depths (a) 9.3 m and (b) 10.9 m
Fig. 9. (a) Comparison of theoretical curves with F" q versus Coocurves
determined from SBPMT performed in Zeebrugh sand (Wroth,
1984) and (b) Comparison E50 estimated by theoretical method with
Fig. 7(b). An approximate value of Rf as 0.6 was
Er from SBPMT
estimated from Fig. 5. As shown in Table 1, the value of
Er calculated from the unload-reload cycle compares well

with E50, estimated based on Esf and Rf. 0 of 38•‹ was also

determined by Houlsby et al. (1986). curve derived at Ir = 500, K0 = 0.52 and 0 = 39•‹ was ob-

tained as shown in Fig. 9(a). As shown on Table 1, the

Dense Sand at McDonald's Farm value of Er from unload-reload cycle compares well with

Robertson and Campanella (1988) and Robertson and E50 estimated by the above method.

Hughes (1986) have performed cone penetration tests and Figure 9(b) shows that Er estimated from unload-

SBPM tests in dense sand at McDonald's Farm site. The reload cycles compare well with E50 for Po river sand,

cone tip resistance from 9 to 12 meters varies widely from Zeebrughe sand, Shotter paper mill sand and

15 bars to more than 200 bars, showing either the McDonald's Farm sand. These examples show that the

variability of the density of the sand deposit or the SBPM data can be analyzed to determine reasonable

values of 4) and modulus at various stress levels (Es f , E50

presence of very thin sand-witched layers of saturated
cohesive material. Two SBMT tests were performed at and Ei), when volumetric strains in the plastic zone are

depths of 9.3 and 10.9 m (Robertson and Hughes, 1986). calculated by finite strain method for solving cylindrical

A good match between measured curve and theoretical cavity expansion.

curve derived at Ir = 500, K0 =1.1 and 4) = 40•‹ for the test

at 9.3 m and at I, = 250, Ko= 0.83 and 4) = 35•‹ for test at

CONCLUSIONS
10.9 m was obtained as shown in Figs. 8(a) and 8(b). The

value E50 estimated by the above method, compares well Using cylindrical cavity expansion analysis and es-

with Er determined from a reload /unload cycle per- timating volumetric strains by finite strain method, the

formed at 10.9 m depth, see Table 1. value of the angle of friction and modulus of deforma-

tion at various stress levels can be estimated with reasona-

Zeebrughe Sand ble accuracy as demonstrated by the analysis of several

Wroth (1984) performed a SBPM test in sand at a SBPM tests in sand performed in various geographical

depth of 10.6 m. The ground water table was at a depth regions. The modulus of deformation (Esf), determined

of 4.3 m. a was measured as 107 kPa and 60 was esti- from analysis of SBPMT data was found to lie in the

mated as 160 kPa and on that basis K0 was determined as range of typical values for that type of sand. The hyper-

0.46. The value of 4) was determined as 39•‹ and the value bolic model for nonlinear stress-strain relationship can

of Gur was estimated as 31 MPa, see Wroth (1984). A be successfully used to determine modulus at 50% failure

stresses and also initial tangent modulus by using the

good match between measured curve and theoretical
88 GUPTA

σh= total horizontal stress;

procedure described in this paper. It is considered that
effective horizontal stress;
unload/reload modulus determined from SBPMT is σ'h=

σ'o= effective vertical stress;

related to modulus at 50% failure stresses (Wroth, 1984). a'm= mean effective horizontal stress in elastic zone;
Modulus at 50% failure stresses determined by using the σ'mp= mean effective horizontal stress in plastic zone;
procedure described in this paper matched well with σm= mean horizontal stress in elastic zone;
unload/reload modulus determined from SBPM tests σ'oct= effective octahedral stress in elastic zone;

conducted in sand at several sites, demonstrating the σoct= total octahedral stress;
σ'rp= effective radial stress at interface of elastic and plastic
reliability and accuracy of this method.
zones;
σ'θp= effective circumferential stress at interface of elastic and
plastic zones;
NOTATION (σrr-σ θθ)a,ymp= asymptotic value of (ƒÐrr-ƒÐƒÆƒÆ) in hyperbolic model;

(σrr-σ θθ)f= Peak stresses at which Mohr circle touches Mohr-

The following symbols are used in this paper:
Coulomb failure envelope, or failure stresses;
a, b= Hyperbolic model parameter;
(σrr-σ θρ)05f= 50% of failure stresses;
DR = Relative density in percent; △= average volumetric strain in plastic region, dimension-
G= gravitational acceleration constant; less;
E= Young's Modulus of Elasticity;
φ= angle of shearing resistance of soil.
Esf = Secant Modulus at failure or at state of equilibrium;
E50= Secant Modulus at half value of failure stresses;
Ei = Initial modulus of elasticity;
REFERENCES
Er =Unload /reload modulus from SBPMT;
E0 = Modulus of Elasticity at a' 0=100 kPa; 1) Baguelin, F., Jezequel, J-F. and Shields, D. H. (1978): The pres-
G= shear modulus; suremeter and foundation engineering, Trans. Tech. Publications,
Gmax= Initial or maximum shear modulus; Clausthal, Germany, 617.
Ko = coefficient of earth pressure at rest; 2) Baldi, G., Bellotti, R., Ghionna, V. N., Jamiolkowski, M. and
hw= height of ground water table above test location; LoPresti, D. C. F. (1989): Modulus of sands from CPTs and
Jr = rigidity index; DMTs, Proc. 12th ICSMFE, Rio de Janeiro, Balkema/Rotterdam,
M= Constrained modulus; 1, 165-170.
Rf = failure factor; 3) Bruzzi, D., Ghionna, V., Jamiolkowski, M., Lancelotta, R. and
Ri = initial radius of cavity Manfredini, G. (1986): Self-boring pressuremeter tests in Po river
rp = radius of plastic zone at time t; sand, ASTM Special Technical Publication 950, The Pressuremeter
R(t) = radius of cavity at time t; and its Marine Applications: Second International Symposium,
△R= change in radius of Cavity, {R(t)— Ri}; 57-74.
r(t) = radial distance of particle from axis of cavity in its 4) Coduto, D. P. (1994): Foundation Design, Principles and Practices,
displaced position; Prentice Hall, Englewood, NJ, 796.
p(t) = internal cavity pressure at time t; 5) Desai, C. S. and Christian, J. T. (1977): Numerical Methods in Ge-
q= total isotropic ground stress, equal to o-h for cylindrical otechnical Engineering, McGraw-Hill Book Company, New York,
cavity; 783.
q' = effective isotropic ground stress, equal to or'h for cylin- 6) Duncan, J. M. and Chang, C. Y. (1970): Nonlinear analysis of
drical cavity; stress strain in soils, J. Geotech. Engrg., ASCE, 96 (5), 1629-1653.
F" q= Dimensionless cylindrical cavity factor at time t; 7) Gupta, R. C. (2005): Finite strain analysis for self-boring
z= depth from ground surface; pressuremeter tests in sand, Soils and Foundations, 45(5), 73-79.
qc = cone penetration resistance; 8) Houlsby, G. T., Clark, B. G. and Wroth, C. P. (1986): Analysis of
vs = shear modulus; the unloading of a pressuremeter in sand, ASTM Special Technical
Pa = atmospheric pressure, equal to 101.3 kPa; Publication 950, The Pressuremeter and its Marine Applications:
v= Poisson's ratio; 2nd International Symposium, 245-262.
γw= unit weight of water; 9) Kondner, R. L. (1963): Hyperbolic stress-strain response: Cohesive
γt= total unit weight of soil; soils, J. Soil Mech. Found. Div., ASCE, 89 (SM1), 115-143.
εrr= radial strain; 10) Lambe, T. W. and Whitman, R. V. (1969): Soil Mechanics, John
εθθ= circumferential strain in cylindrical coordinates; Wiley and Sons, New York, 553.
εθp= circumferential strain at interface of elastic and plastic 11) Mayne, P. W. (2001): Stress-strain-strength-flow parameters from
zones; enhanced in-situ tests, Proc. Inter. Conf. In-situ Measurements of
εθθro-5f)= circumferential strain at 50% failure stresses; Soil Properties and Case Histories, Bali, Indonesia, 27-48.
εv= volumetric strain; 12) Robertson, P. K. and Hughes, M. 0. (1986): Determination of
εvl= volumetric strain in Path AC; properties of sand from self-boring pressuremeter tests, ASTM
εv2= volumetric strain in Path C'C; Special Technical Publication 950, The Pressuremeter and its
dV= volume of an element of width dr and unit height; Marine Applications: Second International Symposium, 283-302.
dr= width of an element at radial distance r(t); 13) Robertson, P. K. and Campanella, R. G. (1988): Guidelines for
ζr= radial displacement of particle at any radial distance; using CPT, CPTU and Marchetti DMT for Geotechnical Design,
ζ,c= radial displacement of particle at face of cavity; FHWA-PA-87-023 + 84-24, FHWA, Washington, D. C.
ζ,p= radial displacement of particle at interface of elastic 14) Schmertmann, J. H. (1978): Guidelines for Cone Penetration Test:
and plastic zones; Performance and Design, FHWA-TS-209, FHWA, Washington,
σrr= total radial stress; D. C.
σ θθ= total circumferential stress in cylindrical coordinates; 15) Vesic, A. S. (1972): Expansion of cavities in infinite soil Mass, J.
σ'rr= effective radial stress; Soil Mech. Found. Div., ASCE, 98 (3), 265-290.
σ'θθ= effective circumferential stress in cylindrical coor- 16) Wroth, C. P. (1984): The interpretation of in situ soil tests,
dinates; Geotechnique, 34 (4), 449-489.
σ'zz= normal stress in vertical direction;