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

net/publication/268591563

Evaluation of Relative Density and Shear Strength of Sands from CPT and DMT

Conference Paper  in  Geotechnical Special Publication · January 2003

DOI: 10.1061/40659(2003)7

CITATIONS READS

134 10,521

3 authors:

Michele Jamiolkowski Diego Carlo Lo Presti

Studio Geotecnico Italiano, Milano Italy Università di Pisa
86 PUBLICATIONS   2,874 CITATIONS    104 PUBLICATIONS   1,992 CITATIONS   

SEE PROFILE SEE PROFILE

Mario Manassero
Politecnico di Torino
42 PUBLICATIONS   568 CITATIONS   

SEE PROFILE

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

Assessment of Liquefaction potential in Tuscany using available mechanical CPT database View project

DEM modelling View project

All content following this page was uploaded by Diego Carlo Lo Presti on 05 December 2015.

The user has requested enhancement of the downloaded file.

 Evaluation of Relative Density and Shear Strength of Sands
from CPT and DMT

M. Jamiolkowski1, D. C. F. Lo Presti2 and M. Manassero3

Abstract

The paper summarises the experience gained by the writers in the interpretation of
the cone penetration test (CPT) and flat dilatometer test (DMT) for the assessment of
the geotechnical properties of sands. In the first part of the paper, the problem of
determining the relative density (DR) as function of the penetration test results and
ambient stress (σ’), for silica sands, is dealt with. In the second part of the paper, the
assessment of the peak angle of shearing resistance (ϕ’p) is dealt with. The attention
is given to the use of the Bolton’s (1986) strength-dilatancy theory in order to
estimate ϕ’p. Engineering correlations, based on Bolton’s (1986) work, are proposed
allowing estimation of ϕ’p as function of penetration resistance and σ’, taking into
account the compressibility and the curvilinear shear strength envelope.

Introduction

The concept of relative density (DR) suggested by Burmister (1948), despite its
intrinsic uncertainties and limitations [Tavenas and La Rochelle (1972), Tavenas
(1972), Achintya and Tang (1979)], is still extensively used in geotechnical
engineering as an index of the mechanical properties of coarse grained soils. Because
of the well-known difficulties and the high costs in retrieving good quality
undisturbed samples from sand and gravel deposits [Yoshimi et al. (1978), Hatanaka
et al. (1988), Goto et al. (1992), Yoshimi (2000)], geotechnical engineers need to

1
Professor, Department of Structural and Geotechnical Engineering, Politecnico di
Torino, Corso Duca degli Abruzzi 24, 10129 Torino Italy; phone +39 011 5644840,
[email protected]
2
Associate Professor, Department of Structural and Geotechnical Engineering,
Politecnico di Torino, Corso Duca degli Abruzzi 24, 10129 Torino Italy; phone +39
011 5644842, [email protected]
3
Associate Professor, Department of Geoenvironmental Engineering, Politecnico di
Torino, Corso Duca degli Abruzzi 24, 10129 Torino Italy; phone +39 011 5647705,
[email protected]

1
estimate the in situ DR using empirical correlations between this parameter and
penetration test results. This indirect way of evaluating DR adds further uncertainties
to those already faced when determining the relative density in the laboratory. The
first attempt to correlate the blow-count of the Standard Penetration Tests (NSPT) to
the density of sands is linked to the works by Terzaghi and Peck (1948) and Gibbs
and Holtz (1957). The continuous interest in this kind of correlation is testified by the
more recent works by Skempton (1986) and Cubrinovski and Ishihara (1999). As far
as the CPT is concerned, a pioneering work can be dated back to Schmertmann
(1976). He presented the first comprehensive correlation between the cone resistance
(qc) and DR on the basis of static Cone Penetration Tests (CPT) performed in
Calibration Chambers (CC). Such correlation relates DR to the effective overburden
stress (σ’vo) and is applicable to normally consolidated (NC) fine to medium unaged
sands. Twenty five years later, based on the results of 484 CC-CPT’s performed in
three silica sands, the writers have attempted to present similar correlations
considering the effect of CC size on the measured qc and giving appropriate
consideration to mechanically overconsolidated (OC) sands. The assessment of DR
represents the most common intermediate step in estimating the stress-strain-strength
characteristics of sands and gravels.
In the second part of the paper, we will deal with the assessment of the peak
friction angle ϕ’p of sands, making reference to the simplified Bolton’s (1986)
strength-dilatancy theory. Based on a theoretically sound framework of Rowe
(1962), the input parameters required to estimate ϕ’p are: DR, the friction angle ϕ’cv
at critical state and a parameter Q related to sand compressibility. The presentation
also includes a comprehensive discussion about the intrinsic parameters ϕ’cv, Q.

Evaluation of relative density

The first attempt to correlate the penetration resistance of the cone penetration tests
( q c ) to the density of sands dates back to the work by Schmertmann (1976). He
presented the first comprehensive correlation between q c and relative density ( D R )
on the basis of CPTs performed in the Calibration Chamber (CC). Such a correlation
is applicable to normally consolidated (NC) fine to medium, unaged, clean sands.
Schmertmann (1976) suggested a correlation between the cone resistance ( q c ) the
relative density and the vertical effective stress (σ 'vo ) , using the results of CPT's
performed on sands in the CC of the University of Florida. The analytical expression
takes the form:

( )
q c = C o ⋅ σ 'vo
C1
⋅ exp(C 2 ⋅ D R ) (1)

DR =
1  q c ⋅ σ 'vo
ln 
( ) C1

 (2)
C 2  Co 

where: C o , C1 , C 2 = empirical correlation factors

2
 Since the pioneering work by Schmertmann, many CC’s have been put into
operation in North America, Europe Australia and Japan generating a large data-base
of CPTs performed in different sands and providing a deeper insight into the merits
and limitations of this kind of large-scale laboratory test and of the empirical
correlations that can be obtained. The key points that have emerged from these
experiments can be summarised as follows:
- The analysis of the variance (Tumay 1976), performed to investigate the
relative importance of the different factors influencing the magnitude of the
q c of silica sands measured in CC tests, led to the conclusion that the relative
density ( D R ) and the consolidation stress tensor (i.e. the level of effective
stress existing in the specimen, prior penetration) are the most important
variables that influence q c . (Harman 1976, Schmertmann 1976, Garizio
1997).
- The correlation of q c vs. D R and σ 'vo holds only for NC sands. A correlation
for NC and OC deposits should refer to the effective mean in situ stress σ 'mo
instead of σ 'vo .
- Stress and strain history that can be reproduced in the laboratory play a
secondary role with the exception of increase of the horizontal effective
consolidation stress as result of the mechanical overconsolidation which
concurs to the value of the relevant stress tensor (Jamiolkowski et al. 1988).
- In the case of siliceous sands, their grain shape and crushability play a
secondary role (Robertson & Campanella 1983, Lunne et al. 1997). The
influence of grading on the penetration resistance has not been systematically
investigated. However, the use of the correlations obtained from CC
experiments leads to underestimation of D R in the case of sand deposits
containing more than 5 to 10 % of fines (Jamiolkowski et al. 1988)
- Thanks to the works by Dussealt and Morgenstern (1979) and Barton and
Palmer (1989) which have investigated the effect of geological time on
porosity, fabric and mechanical properties of coarse grained soil deposits, it is
obvious that the empirical correlations based on the results of tests performed
on laboratory reconstituted specimens, are applicable only in the case of
young, unaged NC soils. Skempton (1986) has shown a certain influence of
aging on the correlations between D R and the blow-count of the Standard
Penetration Test ( N SPT ). Analogously, it is reasonable to suppose that aging
influences the D R vs. q c correlations (see also Wride et al. 2000).
- Due to the finite dimensions of the CC, the measured cone resistance is
affected by an error in comparison to that obtainable in the case of an infinite
sand deposit with the same relative density. This phenomenon, named
chamber size effect (Schmertmann 1976, Parkin and Lunne 1982, Baldi et al.
1986, Foray 1986, Mayne and Kulhawy 1991, Tanizawa 1992, Salgado 1993)
leads, within some boundary conditions, to an underestimate or overestimate
of the field q c depending on the boundary conditions imposed on the CC
specimen during the cone penetration. The magnitude of such an
underestimation or overestimation depends on the crushability and
compressibility of the test sand, the ratio of the CC specimen diameter ( D c )

3
 to that of the cone ( d c ) , D R and confining stresses applied to the CC
specimen.
- The degree of saturation and boundary conditions imposed on the CC
specimen during the cone penetration are much less influential on the q c .

The previously illustrated considerations also apply to the penetration

resistance (q D ) as obtained from blade thrust readings in the dilatometer tests
(DMT). It is worthwhile to point out that this kind of measurement is not routinely
performed in dilatometer tests (Marchetti 1980, 1997).
In light of what has been stated above, the writers propose empirical
correlations, similar to that used by Schmertmann (1976) and based on the results of
484 CC-CPT’s and 136 CC-DMT’s that have been performed in three silica sands.
As far as DMT’s are concerned, correlations between the lateral stress index K D and
D R are also shown. The practical use of the proposed correlations is the prediction
of the relative density of granular deposits. This task can be accomplished by using
the proposed correlations and keeping in mind the intrinsic limitations of such
correlations as already discussed.

Experimental Data. The CPTs and DMTs have been performed in Calibration
Chambers of ENEL of Milan and the research institute ISMES of Bergamo. The
apparatus houses 1.2 m in diameter and 1.5 m in height specimen reconstituted by
means of pluvial deposition in air (Bellotti et al. 1982, 1988, Garizio 1997, Felice
1997). After deposition, samples were subject to the one-dimensional compression in
order to apply the desired consolidation stress level and stress-history. After the
consolidation stage, the penetration test (CPT or DMT) was performed, applying to
the CC specimen one of the four available boundary conditions (BC).
Most of the tests were performed under two BC’s:
- BC-1: constant axial (σ 'a ) and radial (σ 'r ) effective stresses;
- BC-3: constant axial effective stresses (σ 'a ) and zero radial strain (ε r ) ;

In addition, a limited number of CC tests were carried out using either BC-2
(axial strain ε a =0, σ 'r = constant), or BC-4 ( ε a = ε r = 0) during the penetration
stage. Table 1 indicates the percentage of tests performed under each BC’s.
All CPT’s were performed using the cylindrical Fugro-type electrical cone
tips. In most tests, the standard cone tip 35.6 mm in diameter (Lunne et al. 1997) has
been employed. A limited number of tests were also performed using cone tips

Table 1. Percentage of CPT’s and DMT’s performed under different boundary

conditions
Test BC-1 (%) BC-2 (%) BC-3 (%) BC-4 (%)
CPT’s (total
66 11 20 3
484 tests)
DMT’s (total
86 0 11 1
136 tests)

4
having diameters ( d c ) equal to 25.4, 20, 11 and 10 mm. These tests were aimed at
investigating the influence of the CC diameter ( D c ) to d c ratio ( R d ) on the q c
measured under different BC’s.
Most of the DMTs were performed using a standard dilatometer (Marchetti
1980). The probe is 14 mm thick, 95 mm wide and 220 mm high. An expandable
steel membrane, 60 mm in diameter, is located on one side of the probe; a load cell
for the measurement of the penetration resistance ( q D ) is located just above the
probe. Few tests were performed using a research dilatometer (RDMT) (Fretti et al.
1992, 1996, Bellotti et al. 1997). The main differences between standard dilatometer
and RDMT are: i) the expandable steel membrane of RDMT is equipped with strain
gauges, so that it is possible to monitor the complete expansion curve, ii) the
structure of the RDMT probe is much stiffer in comparison of the standard DMT,
even though the dimensions of the two probes are identical.
The CC tests were carried out in three well-known silica sands: fine to
medium Ticino (TS), Toyoura (TOS) and Hokksund (HS) sands. The index
properties of these sands are reported in Table 2 and Figure 1.

Data Interpretation

Size effect. The tests run with different cone sizes confirmed the well-known fact
that the penetration resistance measured in the CC is influenced by the imposed

Table 2. Index Properties of Test Sands

PARAMETER TICINO HOKKSUND TOYOURA
γ max [kN/m3] 16.67 17.24 16.13
γ min [kN/m3] 13.64 14.10 13.09
Gs [-] 2.68 2.72 2.65
Uc [-] 1.30 1.91 1.31
D 50 [mm] 0.60 0.45 0.22
Quartz [%] 30 35 90
Feldspar [%] 65 55 8
Mica [%] ∼5 ∼10 ∼3
φ'cv [°] 33 34 32

γ max and γ min = maximum and minimum dry density respectively

Gs = specific density
Uc = uniformity coefficient
D 50 = mean grain size
φ'cv = friction angle at critical state

5
 GRAIN DIAMETER (mm)
2 1 0.6 0.2 0.1 0.06 0.02 0.01 0.006 0.002
100
90
Hokksund
PERCENT FINER BY WEIGHT

80
70
60
Toyoura
50
40
30
20 Ticino
10
0
10 20 40 80 200 (ASTM sieve sizes)

Figure 1. Grain size distribution of three test sands

BC’s. Such effect is inversely proportional to the R d and decreases with increasing
the sand compressibility. Further details on such effect can be found in the already
mentioned works by Schmertmann (1976), Parkin and Lunne (1982), Baldi et al.
(1986), Foray (1986), Mayne and Kulhawy (1991), Tanizawa (1992), Salgado
(1993), Salgado et al. (1998). The measured penetration resistance appears to be
independent on the BC’s when the R d is sufficiently large, i.e. 70 to 100 for silica
sands of moderate to low compressibility respectively. Under these conditions the
penetration resistance measured in the CC matches the field value. The measured
penetration resistance were therefore corrected for chamber size effect by means of
the following empirical equation (Tanizawa 1992, Garizio 1997, Felice 1997):

[
CF = a (D R ) b ]
m
[-] (3)

where:
CF = correction factor by which the measured penetration resistance has to be
multiplied
a, b = empirical coefficients function of R d inferred from the CC performed in TS
and TOS using CPT tips having different size
m = + 1 and –1 for BC-1 and BC-3 respectively.

6
 The values of coefficients a and b for different R d ratios are given in Table 3.
The lower bound of D R = (D R ) min below which CF=1 is also reported in Table 3.
The trend of the empirical CF, yielded by Eq. 3, vs. R d is similar, although generally
lower, to what achieved by Salgado (1993) from numerical modelling.
The measured q D have not been corrected to account for CC size effect. Based on
limited CC evidence, it appears that there is no need to correct the penetration
resistance of a plate blade because the experimentally determined q D is not
influenced by the finite dimensions of the chamber (Felice 1997), even though the
reasons are not well understood.

Table 3. Coefficient a and b of Eq. 3

Dc (D R ) min
Rd = A B
dc %
100 0 0 100
60 0.412 0.221 55.8
47.2 0.166 0.457 50.8
33.6 0.090 0.624 47.4
22.1(4) 0.054 0.827 34.1
(D R ) min = D R value in percent at which CF
should be taken equal to one

Proposed correlations. The writers adopted the following equations to fit the
experimental data:
1) the same equation used by Schmertmann (1976)

C1
 σ' 
q c = C o p a   exp(C 2 D R ) (4)
 pa 

from which is it possible to obtain:

1  qc / pa 
DR = ln   (5)
(
C 2  C o σ ' / p a )
C1


where:
q c = measured cone resistance multiplied by CF of Eq. 3
σ ' = an initial effective geostatic stress component or stress invariant FL−2 [ ]
D R = relative density (as decimal)

( 4) tests performed in a smaller CC in Japan by Tanizawa (1992) in TOS

7
C o , C1 , C 2 = non dimensional empirical correlation factors, see Table 4 for CPT’s
p a = atmospheric pressure expressed in the same unit system of stress and
penetration resistance (i.e. 98.1 kPa or 1 bar etc.)

2) the equation proposed by Lancellotta (1983)

D R = A o + Bo ⋅ X (6)

where:
 q 
X = ln  ' c α 
 (σ vo ) 
A o , B o and α = empirical correlation factors (see Table 5). The parameter α is
obtained following an optimisation process which minimises the differences between
computed and measured values of the penetration resistance in terms of standard
deviation. In this case q c and σ 'vo are in kPa.

The same equations have been used in the case of DMT’s. The empirical
correlation factors obtained from DMT results are reported in Tables 5 & 6.
As to the definition of the effective stress σ ' to be introduced into Eqs. 4 and
5, the following should be taken into account:
- Zolkov and Weisman (1965) postulated that N SPT is controlled by the
horizontal in situ effective stress ( σ 'ho ).
- Similar experimental evidence emerged from CPT’s performed in CC’s
(Schmertmann 1971, 1972, Baldi et al. 1986, Houlsby and Hitchman 1988,
Mayne and Kulhawy 1991, Salgado 1993, Garizio 1997, Felice 1997). It
shows that the magnitude of the penetration resistance is much more
influenced by σ 'r than by σ 'a .
- The above statement suggests that any rational correlation between
penetration resistance and relative density should be related to the mean
(σ 'mo ) or horizontal (σ 'ho ) effective geostatic stresses rather than to the
(σ 'vo ) . The lesson learned is that the correlation between the penetration
resistance and relative density involving σ 'vo is applicable only to NC
deposits of coarse grained soils in which K o ranges from 0.4 to 0.5
remaining more or less constant with depth.

Given the above considerations, Figures 2 and 3 report the D R = f (q c , σ ' )

correlations for NC (Fig. 2) and (NC+OC) (Fig. 3) dry silica sands respectively. As
to the latter, the writers fully appreciate the extreme difficulties linked with the
estimation of σ 'ho in sand and gravel deposits. Overall, but particularly for coarse-
grained soils, the determination of σ 'ho or of the coefficient of the earth pressure at
rest ( K o ) in situ is still an unsolved problem in geotechnical engineering.

8
 Table 4. Coefficients Co , C1 and C2 of Eqs. 4 and 5 (CPT’s)
 
C  
 σ'  1 1 qc / pa
qc = Co pa   exp(C 2 D R ) DR = ln  
 pa  C2   ' C1 
   C o  σ / p a  
   

σ' = σ' vo TS TS + TOS + HS

Co 17.74 17.68
C1 0.55 0.50
C2 2.90 3.10
R 0.90 0.89
σ 0.12 0.10
N 305 180

σ' = σ' mo TS TS + TOS + HS

Co 23.19 24.94
C1 0.56 0.46
C2 2.97 2.96
R 0.87 0.87
σ 0.10 0.10
N 299 484
R = correlation coefficient
σ = standard error
N = number of CC tests considered

Table 5. Coefficients A,B and α of eq. 6 (CPT’s & DMT’s)

 q 
D R = A o + Bo ⋅ X X = ln  c 
 (σ ' ) α 
 vo 
CC test Ao Bo α R σ N
CPT’s -1.292 0.268 0.52 0.94 7.9 456
DMT’s -1.082 0.204 0.36 0.92 6.6 100
R= Correlation coefficient
σ = standard error
N = number of CC test considered

9
 Table 6. Coefficients C o , C1 and C 2 of Eqs. 4 and 5 (DMT’s)
 
C  
 σ'  1 1 qc / pa
qc = Co pa   exp(C 2 D R ) DR = ln  
 pa  C2   ' C1 
   C o  σ / p a  
   

σ' = σ' vo TS TS + TOS + HS

Co 19.14 20.64
C1 0.62 0.52
C2 3.61 3.71
R 0.88 0.88
σ 0.11 0.10
N 57 69

σ' = σ' mo TS TS + TOS + HS

Co 26.99 26.62
C1 0.60 0.49
C2 3.75 3.80
R 0.91 0.89
σ 0.12 0.11
N 110 136
R = correlation coefficient
σ = standard error
N = number of CC tests considered

Nevertheless, the estimate of σ 'mo in coarse-grained soils is facilitated by the

following considerations:
- In NC deposits the upper limit of K o can be taken as 1- sin ϕ 'cv (i.e. the angle
of shearing resistance at critical state).
- In heavily OC sands (i.e. OCR = 15) K o is not greater than 1.0 as suggested
by the CC results (Jamiolkowski et al. 1988).

Figures 4 & 5 show Eq. 6 in the case of CPT’s and DMT’s respectively.
Figures 4 & 5 enable one to appreciate the accuracy of Eq. 6 to fit the experimental
data that are also plotted in the Figures together with the limits corresponding to
± 2σ .
Empirical correlations were also established between the lateral stress index
( K D ) and D R . K D is computed from DMT results in the following way:

10
 po − u o
KD = (7)
σ 'vo

where:
p o = lift-off pressure
u o = pore pressure, prior to penetration and expansion
σ 'vo = vertical effective stress, prior to penetration and expansion

The dependence of K D on D R is clearly shown in Figure 6 for the three

considered sands. Two different equations were used to fit the experimental data:

K D = C o (σ ' ) C1 p1a−C1 exp(C 2 D R ) (8)

K D = A exp(BD R ) (9)

Eq. 8 is similar to Eq. 4 with the only difference that, instead of considering
the penetration resistance, the lateral stress index appears in the formula. Eq. 9 is
much more crude and does not take into account the stress level prior to penetration
and expansion (σ ' ) . Moreover, D R is expressed as a fraction of one.
The accuracy of Eq. 8 is not influenced by the choice of the effective stress.

1  qc 
DR = ln 
( ) 
C 2  C 0 σ'v 0 C1 

(1 bar = 98.1 kPa)

Figure 2. Relative density of NC siliceous sands

11
  
1  qc 
DR =
( )
ln
C2  C ' C1 
σm0 
 0

(1 bar = 98.1 kPa)

Figure 3. Relative density of NC and OC siliceous sands

σ’v0 [kPa] ; DR [-] ; qc [kPa] ; σ = standard error = 7.9

100
456 experimental
data
80
DR (%)

High
60 compressibility 2σ

40

Low
20
compressibility
2σ
0
100 1000 (qc)(σ’v0)-0.5 10000

DR = -A0 + B0ln[(qc)(σ’v0)-0.5] = -1.292 + 0.268ln[(qc)(σ’v0)-0.5]

(*)qc : corrected values for taking into account the CC size

Figure 4. Experimental correlation DR-qc-σ’v0 for mainly NC sands of different

compressibility (Lancellotta, 1983; Garizio, 1997)

12
 DR = -1.082 + 0.204ln(qD/σ’v0)0.4

80 σ = 6.57 ; N = 100
Relative Density, DR %
60

40
2σ
20
2σ σ = standard error
0
10 100 0.4 1000
qD  σ'v0 
 
 p 
σ’v0 & qD in kPa pa  a 

Figure 5. Relation between wedge resistance and relative density –

Calibration chamber tests in Ticino sands (Felice, 1997)

12

Ticino
10 Toyoura
Hokksund
Lateral Stress Index KD

0
20 30 40 50 60 70 80 90 100
Relative Density DR [%]

Figure 6. KD vs DR from CC-DMT’s in three siliceous sands

13
 In Table 7 the correlation coefficients that have been obtained for the case
σ = σ (NC tests) are reported. It is worthwhile to point out that the stress exponent
' '
vo
assumes negative values, irrespective of the selected effective stress. It is also
important to notice that the use of a simpler formula. like Eq. 9 does not involve
relevant reduction in accuracy. In any case, the correlations between K D and D R are
less accurate than those between q c (or q D ) and D R . Figure 7 shows Eq. 9 ± 2 σ and
the experimental data. The empirical correlation factors used in Figure 7 are slightly
different from those reported in Table 7 because they have been obtained
disregarding those test results with a deviation from the computed value higher than
± 2σ .

Degree of saturation. All CPT’s and DMT’s used to derive the previously shown
correlations were carried out on dry specimens. Only a limited number of tests were
performed in saturated TS (Bellotti et al. 1988) showing little influence of the
saturation on the measured penetration resistance. Strictly speaking, these
correlations are applicable to dry fine to medium clean, unaged, uncemented silica
sands of low to moderate compressibility in which the static cone penetration process
corresponds essentially to a drained process.
In order to overcome, at least partially, the limitations derived from the above
specified condition the following indications can be helpful:

Table 7. Coefficients C o , C1 , C 2 A,B eqs.8 and 9 (DMT’s)

1− C
K D = C o (σ ' ) C1 p a 1 exp( C 2 D R ) K D = A exp(BD R )

Eq. 8 σ ' = σ 'vo TS TS+TOS+HS

−3
Co 5.3 ⋅ 10 6.6 ⋅ 10 −3
C1 -0.18 -0.25
C2 2.60 2.29
R 0.78 0.76
σ 12 12
N 58 73

Eq. 9 NC NC+OC
A 0.53 0.57
B 2.42 2.56
R 0.71 0.71
σ 13 13
N 73 136
R= Correlation coefficient
σ = standard error
N = number of CC test considered

14
 KD = 0.49 exp(2.72 DR)
2
σ = 1.42 ; N = 101

ln KD 1

2σ
0
2σ σ = standard error
-1
0 20 40 60 80 100
Relative Density, DR %

Figure 7. Relation between lateral stress index and relative density – Calibration
chamber tests in Ticino sands (Felice, 1997)

- The comparison between q c (dry) and q c (saturated ) resulting from CC-

CPT’s carried out on almost identical CC specimens of TS shows very small
differences. Similar conclusions have been reached by Schmertmann (1976)
comparing CC tests performed on dry and nearly saturated quartz Ottawa
sand. It is worthwhile to point out that the use of the previously shown
correlations in saturated sands leads to an underestimation of D R , whose
magnitude can be inferred from the following empirical relationship:

D R (saturated) − D R (dry) qc
100 = −1.87 + 2.32 ln (10)
D R (dry) (
σ vo p a
'
)
0.5

The above exposed formula becomes meaningless for q c /(σ 'vo p a ) 0.5 ≤ 2.24 ;
underestimation of D R , for the sands considered in this paper, ranges between 7 and
10 %.

Aging and Cementation. As far as the effects of cementation and ageing on the
penetration resistance are concerned, currently there is a lack of information able to
estimate and quantify their influence. Schmertmann (1991) has shown that the
accumulation of secondary compression in sands tends to moderately increase the
cone resistance, see for example Kulhawy and Mayne (1990).
More relevant might be the impact of even light cementation on q c (Puppala
1993, Puppala et al. 1995, Eslaamizzad 1997). It may be useful to mention that the
use of correlations similar to those here presented and established on freshly
deposited sands, in aged and/or cemented deposits leads to an overestimation of the
relative density.

15
Evaluation of shear strength

When dealing with the shear strength of non-cemented granular materials, the
friction angle, resulting from the secant slope of the failure envelope, is, in general,
the reference parameter
for both the simplified design approaches (e.g. limit equilibrium and limit analysis
methods) and the most complex multi-surface non-linear elasto-plastic work-
hardening models.
The appropriate definition of the peak friction angle ϕp and the operational
friction angle ϕop(5) referring to the simplified design approaches, appears to be even
more difficult than when more complex models of soil behaviour are used. As a
matter of fact, the operational value of ϕop, for a given boundary value problem, is a
function, among the other state parameters, of all the components of stress and strain
tensors which can be reliably assessed only through sophisticated theoretical
approaches. On the other hand, using simplified design methods, the most
appropriate operational value of ϕop, should be theoretically evaluated with reference
to the average values of the significant state parameters within the yielding volume
of soil. This kind of evaluation, using simplified design methods, is only possible in
an approximate manner and for a limited number of the recurrent boundary values
problems in Soil Mechanics.
In light of the previous considerations, in the following part of the paper, the
basic principles governing the shear strength of sandy soils, with particular reference
to the framework and the relationships proposed by Bolton (1986), are illustrated and
partially worked out to widen the possible practical applications when simplified
models of soil behaviour are adopted. In particular, an evaluation procedure to
estimate the operational friction angle from CPT is illustrated. It allows the
evaluation of operational friction angles of sands having different mineralogical
composition and/or grain size distribution, once the point resistance qc and the mean
geostatic effective stress σ’mo are known.

Basic principles governing shear strength of sandy soils. The basic principles that
govern the shear strength of granular materials (i.e. the critical state concept, the
energy dissipation by particle rearrangement, the dilatancy and the dependency of the
latter on the current state parameters) and the influence of secondary factors such as
strain conditions and anisotropy, have already been clearly pointed out by
Casagrande (1936), Taylor (1948), Rowe (1962), Schofield & Wroth (1968), Bolton
(1986), Mitchell (1976), Lade & Lee (1976), Ladd et al. (1977) and Tatsuoka et al.
(1986). Looking at the framework and the related equations proposed by the
aforementioned researchers, the main components of shear strength of sands can be
split as follows:
- The pure frictional resistance between smooth surfaces of the sand mineral,
quantified by the interparticle friction angle ϕµ;
- The particle rearrangement component determining the strength increase
from ϕµ to the constant volume friction angle ϕcv;

(5) Average mobilized friction angle in correspondence of the general failure for current
geotechnical boundary value problems.

16
 - The dilation component determining the difference between the peak
strength (represented by the peak friction angle ϕp) and the steady state
strength corresponding to ϕcv.

Based on the results of triaxial (TX) and plane strain (PS) compression tests
obtained for different sands, Bolton (1986) attempted an empirical correlation to
assess the peak friction angle that takes into account the relative density (DR), the
mean effective stress at failure σ’mf, and the sand type in terms of grain size
distribution, mineralogy and grain shape.
The equation originally proposed by Bolton (1986) can be written as follows:

ϕ p − ϕ cv = m{D R [Q − ln (σ′mf )] − R} ϕp≥ϕ cv (11)

where: ϕp = peak friction angle in [°]; ϕcv = constant volume friction angle [°]; m =
coefficient equal to 3 or 5 for axisymmetric (TX) and plane strain (PS) conditions
respectively; DR = relative density; Q = particle strength parameter (reported in Table
8); σ’mf = mean effective stress at failure in [kPa]; R = coefficient that in a first
approximation is a function of (ϕcv-ϕµ) and normally it is assumed equal to 1 for
sands; ϕµ = pure friction angle between smooth surfaces of the mineral forming the
considered sand.

Table 8. Q values suggested by Bolton (1986)

GRAIN MINERAL Q
Quartz and feldspar 10
Limestone 8
Anthracite 7
Chalk 5.5

Looking in detail at Eq. 11 it can be remarked (see also Fig. 8) that it

represents a bunch of straight-lines in the plane ϕp; ln(σ’mf) converging in the point
with coordinates ln(σ’mf)=Q and ϕp= (ϕcv-m⋅R), moreover, the angular coefficient of
the straight lines is equal to (m⋅DR).
The common point of the aforementioned straight-lines is quite peculiar since
its coordinates express some intrinsic features of the considered sand not depending
on the state parameters (e.g. the typical combinations of DR and σ’mf).
In first approximation, the term (m⋅R) can represent the shear strength
component due to the grains rearrangement at constant volume strains (Rowe, 1962),
so it can be expressed by the difference between ϕcv and ϕµ.
The term Q can be expressed by the logarithmic function of an appropriate
equivalent grain yield stress (σ’c) that can be related to the grain crushing strength
for coarse materials (p’fm) as definied by Biliam (1967) and Marsal (1967) or can be

17
 ϕcv − ϕµ ≅ mR mDR ln σ’c = Q

ϕcv
~ϕµ
(1; ~ϕµ)
0
0,0001
σ'mf
Dimensionless mean effective stress at failure;
σ'c
Figure 8. Re-plotting of Bolton’s (1986) relationships referring to pure friction angle
(ϕµ) and dimensionless confining stress

referred to the threshold confining stress level at which, for a given sand, a single
value of void ratio is obtained independently from its initial relative density, fabric
and arrangement of the solid skeleton (see Figures 9 and 10).

0.90
e0=0.87
0.85
0.80
e0=0.78
0.75
Void Ratio, e

e0=0.71
0.70
0.65
e0=0.61
0.60
0.55
0.50
0.45
0.40
0.1 1 10 100 1000

Confining Pressure, σ’3 (bar) ~σ’c

(1 bar = 98.1 kPa)

Figure 9. Isotropic compression tests on Sacramento river Sand

(from Lee and Seed, 1967)

18
 Compression Curves
Sarly Point, R.1,Sand (Cs4)
Plum Island Sand (Cs2)

Void Ratio, e

~σ’c
Applied Pressure σ’v (PSI)

Figure 10. K0-compression tests for two similar sands (from Roberts, 1964)

In order to justify this simple model from a mechanical point of view it is

possible to imagine that, when the isotropic stress level is able to destroy the internal
strength of the sand grains, the shear stress increments can be only sustained by the
pure friction mobilized between smooth surfaces of the mineral forming the grains,
then without any contribution from the particle rearrangement and, of course, from
the dilatancy.
Based on the above considerations, Eq. 11 can be rewritten, referring
explicitly to the intrinsic and state parameters influencing the peak friction angle ϕp
(see also Fig. 8):

ϕ p − ϕµ = m ⋅ D R ⋅ ln (σ'c / σ′mf ) ϕp≥ϕ µ+m (12)

With reference to the silica sands considered by Bolton (1986), in order to

obtain the same results of Eq. 11, it is necessary to introduce the following
parameters into Eq. 12:
σ’c = 22026.5 (kPa);
ϕcv-ϕµ= 3° for axisymmetric conditions (TX) and
ϕcv-ϕµ= 5° for plane strain conditions (PS).
As to the value of σ’c, it is in good agreement with the expected values
observed and/or extrapolated from the isotropic compression tests of Lee & Seed
(1967) and Robertson (1964) reported in Figs. 9 and 10.
As far as the constant volume friction angle ϕcv is concerned, the
reinterpretation of the fitting parameters by Bolton (1986), under the light of the
proposed approach (Eq. 12), points out possible different values from TX and PS
conditions.
Although many authors, e.g.: Hanna et al. (1987); Schanz (1998), have
produced the experimental evidence that ϕcv(TX) and ϕcv (PS) are the same, the

19
uncertainties linked to the large strain and strain non-uniformities still leave some
open questions with respect to this problem..
Apart the practical problems of test equipment, other sources of uncertainties,
that can influence the ϕcv values, might also arise from the fitting procedures of the
experimental data carried out by Bolton (1986). Therefore, considering all the above
aspects, the difference of 2° between ϕcv values from TX and PS conditions could be
easily justified and accepted, also considering the empirical nature of Eq. 12.
In order to validate the proposed modification to the original Bolton formula,
it can be interesting to note that Eq. 12 is very similar to the one proposed by Barton
(1973) to describe the curved shear strength envelope of rock joints. Moreover, the
equation by Baligh (1975, 1976), describing the curvature of the sands failure
envelope at a given relative density, can also be re-written in the form of the above
equations.
Of course the proposed simplification of the Bolton’s (1986) formula, based
on the use of the intrinsic parameters (ϕµ, σ’c) characterizing the coarse granular
media behaviour, must be validated by further experimental data. Nevertheless, it can
be used as a reference framework for analysing the basic contributions to the shear
strength of sandy soils.
Beside the intrinsic (ϕµ, or ϕcv and σ’c) and state (DR and σ’mf) parameters,
which mainly influence the shear strength of granular materials, other aspects can
also play a significant role under some specific conditions.
Among them, the first to be mentioned is the intermediate principal stress σ’2
(see Fig. 11) that has been already introduced in an indirect way referring to PS and
TX strain conditions.

55°
Secant peak friction angle ϕp

PS
50°
σ’3 =287 kPa

45°
TX
40°

35°
DENSE LOOSE
30°
32 34 36 38 40 42 44 46
Pre-shear porosity, n %

Figure 11. Comparison of ϕp (PS) with ϕp (TX) for a siliceous sand

(Cornforth, 1964)

20
 Most of the laboratory experimental data are the output of tests in the triaxial
apparatus, whereas many geotechnical structures work in plane strain conditions. It
can be, therefore, of practical interest to transform ϕp(TX) in ϕp(PS) or vice-versa, by
means of a number of empirical formulae reported in the geotechnical literature.
Lade and Lee (1976), for siliceous sands have suggested:

ϕp(PS)= 1.5⋅ϕp(TX)-17° ϕp(TX)≥34° (13)

Schanz & Vermeer (1996), considering the results on Hostun sand as well as
those obtained by Cornforth (1964) and Leussink (1996), for siliceous sands have
proposed:

ϕp(TX)= (1/5)(3ϕp(PS)+2ϕcv) (14)

This latter equation is exactly the same as that which can be obtained by
equation (11) of Bolton (1986) evaluated for m=5 and m=3 in (PS) and (TX)
conditions respectively.
Both Eqs. 13 and 14 hold when comparing TX and PS compression tests. A
more general handling of the problem regarding the effects of σ’2 on shear strength
of granular soils must include the full range of variation of the following parameter:
b = (σ'2 −σ'3 ) / (σ'1 −σ'3 ) . However, the analysis of the changes of ϕp with variation of
b is beyond the scope of the present paper.
As documented in the last fifteen years by Tatsuoka and his co-workers
(Tatsuoka et al., 1986, 1990; Pradhan et al., 1988; Park & Tatsuoka, 1994), the peak
secant angle of shearing resistance shows a pronounced anisotropic response. As a
matter of fact, in addition to the value of parameter, b, the peak friction angle is also
affected by the angle, δ, existing between the direction of the major principal stress
at failure (σ’1f) and the direction of the bedding planes. The present ϕp anisotropy is
not usually taken into account in the interpretation of in situ tests results for strength.
The previous approaches to the assessment of the shear strength envelope of
coarse grained materials assume that there is no cohesion intercept (c’) in terms of
effective stress.
Such statement holds even in very dense and interlocked materials as recently
argued by Schofield (1998).
However, lightly cemented coarse-grained soils are noticed within natural
formations for which a c’>0 intercept is a consequence of the weak bond between the
soil grains (Nader, 1983; Bachus, 1983).
The lightly cemented soil deposits generally have unconfined compressive
strength less than 100 kPa and the c’ resulting from drained TX compression tests
falls in a range between 5 and 30 kPa.
At present there is a lack of well consolidated methods allowing to infer both
ϕ’p and c’ from in situ tests, while some possibilities can be envisaged for SBPT
(Bachus, 1983; Carter et al., 1986).
Currently the assumption of c’=0 when interpreting in situ CPT and DMT
leads to an overestimate of ϕ’p in case of lightly cemented sands.

21
Evaluation of the peak friction angle from CPT and DMT. Nowadays the CPT and
DMT are among the tools most commonly used in design to evaluate ϕp. With this
respect two basic different approaches can be envisaged (Jamiolkowski & Lo Presti,
2000):

A. The first approach can apply to both CPT and DMT results and refers to the
use of the existing bearing capacity theories (e.g. Durgunoglu & Mitchell,
1975; Janbu & Senneset, 1974; Vesic, 1975, 1977; Salgado, 1993, Salgado
et al., 1997). In this case, the ultimate bearing capacity is measured (i.e. the
cone resistance qc in case of CPT and the wedge resistance qD in case of
DMT), therefore, the bearing capacity formula is used to estimate ϕp. The
summary of the input data required when using these approaches is shown in
Table 9. Further details can be found in the works by Mitchell & Keaveny
(1986) and Yu & Mitchell (1998).

B. The second approach consists of the in situ evaluation of DR from the results
of the considered penetration tests. Once the DR has been assessed, the
estimation of ϕp can be carried out by using correlations ϕp= f(DR, grading)
like the one proposed by Schmertmann (1978), see Fig. 12 or, in a more
refined manner, by means of an iterative use of Bolton’s (1986) stress
dilatancy Eqs. (11) and/or (12).

The main features and the use of the methods belonging to the groups A and
B are summarized in the paper of Jamiolkowski & Lo Presti (2000).
As a matter of fact, the methods for estimating ϕp by the bearing capacity
theories, in spite of the more elegant initial approach, require rather complex input
data and /or are affected by important limitations and approximations of the original
theoretical models so that, most of the procedures, practically applicable, must turn
to “calibrating” coefficients and/or “operational” parameters that reintroduce
empiricism into the initial equations (see for example Mitchell & Keaveny, 1986).
For such reasons the methods of group B, passing from the empirical
evaluation of DR by CPT and DMT before the final assessment of ϕp at different
confining stress levels, can still be considered more robust, reliable, and useful for
practical applications within the current geotechnical design practice.
Referring to the ϕp assessment by the methods of the group B, once DR has
been evaluated by means of one of the approaches outlined in the first section of this
paper, an estimate of the ϕp(TX) can be attempted with reference to Fig. 12 by
selecting the line appropriate for the gradation curve for the soil layer in question.
The main limitation which arises from the use of the functions of Fig. 12 is that ϕp
only refers to triaxial and direct shear (DS) tests carried out at a confining stress
range of 50<σoct<350 kPa and at a normal stress range of 80<σ’vc<400 kPa
respectively. Therefore, it is not possible to take into account the influence of
parameters such as: different confining stress level at failure, strain conditions and
last but not least sands characterized by different compressibility and mineralogy
from the tested ones.
In order to overcome these limitations, Eqs. 11 or 12 can be used to assess
ϕp(TX) or ϕp(PS) following the procedure outlined in Fig. 13. This approach has

22
 Table 9. Friction angle from cone resistance – possible approaches

STRESS-STRAIN STRENGTH STRESS

REFERENCE COMPRESSIBILITY OTHER INPUT REQUIRED
RELATIONSHIP ENVELOPE TENSOR
SCHMERTMANN NO ASSUMPTION DR
(1978) EMPIRICAL NO LINEAR σ’v0 or σ’m0 GRADING CURVE
BEEN ET AL.
(1987) NO ASSUMPTION YES (IMPLICITLY) LINEAR σ’ m0 ξ = f(NCR)0
STRENGTH-
BOLTON DILATANCY Q, ϕ’cv
YES (IMPLICITLY) CURVILINEAR σ’mf
(1986) THEORY SEMI-
EMPIRICAL
JANBU AND
SENNESET RIGID PLASTIC EMPIRICALLY LINEAR σ’v0 sq
(1974)
DURGUNOGLU
AND MITCHELL RIGID PLASTIC NO LINEAR σ’v0 sq, K0, δ
(1975)
ELASTIC
VESIC
PERFECTLY YES LINEAR σ’m0 or σ’h0 sq, K0, G
(1972, 1977)
PLASTIC
SALGADO ELASTIC NON
YES CURVILINEAR σ’h0 sq, ϕ’cv, K0, DR, Q, G(γ, σ’m)
(1993) LINEAR PLASTIC
The table is not exhaustive, for similar approaches see also Robertson & Campanella (1983), Jamiolkowski et al. (1988), Trofimenkov
(1974)

23
 42°
LT
EL -SI
R AV AND
-S

Secant Peak Friction Angle ϕp

G
O RM AVEL
IF R
UN ED G ND D
A D E SA SAN
38° R S
LG A R IU M
1 WE L C O ED D
RM D M AN D
FO ADE S
I
UN G R AN
I UM E S
L L E D IN
36° 2 WE M
M DF D
E N
R D SA ϕp=a+bDR
I FO RA NE
G
UN LL FI Curve a b
E M
W O R (n°) (°) (°)
3 IF 1 37.0 0.08
32° UN
2 34.5 0.10
3 31.5 0.12
4 4 28.0 0.14

28°
20 40 60 80 100
Relative density, DR %

Figure 12. Correlation ϕp = f(DR Grading) (Schmertmann, 1978)

been validated for a number of siliceous sands by Bellotti et al. (1989), Jamiolkowski
(1990), Yoon (1991) and others. This method is equally conditioned by the reliability
of the DR best estimate as that involving the use of Fig. 12, but has the advantage to
be able to take into account the stress level relevant to the considered boundary value
problem via the introduction of an appropriate value of σ’mf. The higher rationality of
this approach has, however, the limitation of a more elaborated input involving the
knowledge and/or the assumption, in addition to DR, of two intrinsic (ϕcv or ϕµ and Q
or σ’c) and one state (σ’mf) parameter.
As far as Q=lnσ’c is concerned, Table 8 and 10 report respectively the values
of Q as suggested by Bolton (1986) and those resulting for a number of sands
inferred from triaxial compression tests. Table 10 associates with each value of Q
also the corresponding ϕcv. In the light of the preliminary considerations related to
the possible difference of 2° existing between ϕcv(TX) and ϕcv(PS), in the authors
opinion, the value reported in Table 10 should be considered as ϕcv(TX). Moreover it
may be worthy to point out that the values of Q displayed apply to grains having
dimensions corresponding to those of fine to medium sands. As documented by Lee
(1992) for coarse sands and gravel particles constituted by the same mineral, Q tends
to decrease with increasing the equivalent grains diameter.
Regarding the estimate of σ’mf corresponding to the boundary value problem
of practical interest, the issue is far from being solved in a rigorous manner. At
present, only the following rules of thumb can be suggested to the readers:

24
 For shallow foundations (De Beer, 1967):
q + 3σ 'v 0
σ 'ff ≅ lim (1 − sin ϕ op ) (15)
4

qlim : limit bearing capacity of the considered shallow foundation;

ϕ’op : operational friction angle for qlim evaluation.

For deep foundations and penetration tests (Fleming et al., 1992):

σ 'ff ≅ σ 'v 0 ⋅ N q (16)

Nq : dimensionless bearing capacity factor for deep foundation, function

of ϕ’op.

qc or qD σ’ and ϕ’ ➀

➀ σ’v0 or σ’0 & ϕµ or ϕcv

DR = f(σ’, qc) ➁ Function of the
boundary value
problem considered

σ’mf ➁ ➂ ➂ Can require iterative

approach

IRD = DR(Q-lnσ’mf)-R or IC = DRln(σ’c/σ’mf)

ϕp = ϕcv + mID or ϕp = ϕµ + mIC

• m ➔ function of imposed stress system: 3 for TXT; 5 for PST

• Q, σ’c ➔ functions of grain strength and size, see: Bolton (1986), Yoon
(1991), Lee (1992), Garizio (1997)
• R ➔ fitting parameter taken equal to 1 and function of (ϕcv - ϕµ)

Figure 13. Friction angle of sand from penetration tests Bolton’s (1986)
stress-dilatancy theory

25
 Table 10. Q-values of different uniform sands (*)

SAND MINERALOGY Q ϕcv REFERENCE

TICINO SILICEOUS (**) 10.8 34.6
TOYOURA QUARTZ 9.8 32 Jamiolkowski et al., 1988
HOKKSUND SILICEOUS 9.2 34
MOL QUARTZ 10 31.6 Yoon, 1991
FINES 0% 9.8 30
QUARTZ
FINES 5% 10.9 32.3 Salgado et al. 1997
OTTAWA FINES 10% 10.8 32.9 Salgado et al. 2000
FINES 15% 10 33.1
FINES 20% 9.9 33.5
ANTWERPIAN QUARTZ & 7.8 to 8.3 31.5 Yoon, 1991
GLAUCONITE
KENYA CALCAREOUS 9.5 40.2
Jamiolkowski et al., 1988
QUIOU CALCAREOUS 7.5 41.7
(*) inferred from TX compression tests
(**) i.e.: containing a comparable amount of quartz and feldspar grains

26
 The link between σ’ff and σ’mf for compression loading stress path is given
by:

 3 − sinϕ0 p 
σ'mf ≅ σ'ff ⋅   (17)
 3 cos 2 ϕ0 
 p 

As already pointed out, the method, based on the Bolton (1986) theory,
sketched in Fig. 13 has been mainly validated referring to silica sands. An attempt
has been carried out, by the writers, to extend the aforementioned procedure, for the
assessment of DR and then ϕp from qc values, to sands characterized by different
mineralogical compositions that results mainly in a different deformability under
isotropic stress increments and in different values of ϕcv and ϕµ.
As a first step, the relationships qc - DR, reported in Fig. 4, has been
considered. These correlations have been worked out by Lancellotta (1983) and
subsequently have been revised by Garizio (1997) for taking into account the
influence of CC dimensions, geometry and boundary conditions and referring to a
much more extended data base.
To account for the influence of different sand compressibilities, the fitting
parameters Ao and Bo of the equation in Fig. 4 have been evaluated for the average,
upper and lower bond respectively of the data set displayed in the same figure.
This data set is mostly related to different normally-consolidated (NC) and lightly
over-consolidated (OC) sands for which, as previously mentioned, DR may be related
to qc through σ’vo.
In order to be able to include, within the proposed procedure, the OC sands,
as a first attempt, the equation of Figure 4 has been modified as reported in Table 11,
i.e. instead of the geostatic vertical stress (σ’vo) the relative density has been
expressed as a function of the mean geostatic effective stress [σ’mo=σ’vo(1+2K0)/3)]
and the Ao value has then been transformed by using the following equation:

1 − 2K 0 ( NC)
A' 0 = A 0 − B 0 ln (18)
3

where K0(NC) is the coefficient of horizontal pressure at rest for NC sands assumed,
in a first approximation, equal to 0.4.

The proposed modification of the equation of Figure 4 into the one of Table
11 does not change the results of the original correlations in the case of NC sands but
via the K0 values allows, in principle, to extend the proposed procedure to the OC
sands.
The obtained values of A’o and Bo for the different mineralogical
compositions are reported in Table 11 together with the values of Q characterizing
the same kind of sands as suggested by Bolton (see also Table 8).
Substituting the equation reported in Table 11 into Eqs. 11 or 12 and using
the parameters given in Table 11 and the values of ϕµ or ϕcv within the ranges of

27
 Table 11. Parameters A’0, B0 and Q used in Figs. 14, 15 and 16 for different sands

MINERALOGICAL COMPRESSIBILITY EXAMPLES IN THE A’0 B0 Q

COMPOSITION LITERATURE
QUARTZ

MONTEREY
SANDS

OTTAWA -1.506 0.268 10.0

QUARTZ LOW TOYOURA
SYDNEY
SANDS
CALCAREOUS SILICA

FELDSPAR
TICINO
QUARTZ MEDIUM -1.360 0.268 9.5
HOKKSUND
MICA

QUIOU
SANDS

KENYA
SANDSTONE
HIGH BASS STRAIT
MICA -1.214 0.268 8.5
ANTWERPIAN
CHATTAHOOCHEE

 
'  qc 
D R = A 0 + B 0 ln ; q c , σ 'mo in kPa
 σ' 
 m0 

28
 Table 12. Range of ϕcv e ϕµ values for sands of different mineralogical compositions

ϕµ ϕcv
Quartz sands 25° ÷ 30° 30° ÷ 34°
Silica sands 27° ÷ 32° 32° ÷ 36°
Calcareous sands 32° ÷ 38° 36° ÷ 42°

Table 12, the peak friction angle ϕp(TX) or ϕp(PS) can be evaluated for the considered sand.
Figures 14, 15 and 16 show ϕp(TX) trends, for silica, calcareous and quartz sands
respectively, evaluated by the proposed procedure and referred to a mean confining stress at failure
σ’mf equal to the geostatic mean effective stress σ’mo. In the same figures the adopted values of ϕcv
and Q are also reported.
The values of ϕp(TX) obtained from the aforementioned figures have been compared with the
estimation of the same parameter at the same confining stress levels carried out in different ways by
Durgunoglu & Mitchell (1975), Robertson & Campanella (1983) and Chen & Juang (1996), The
comparison results have pointed out a very good agreement for all the mineralogical compositions
of the considered sands.
Once the set of intrinsic and state parameters, characterising the mechanical behaviour of the
considered sand, has been assessed, it is possible to evaluate the ϕop(TX) and/or the ϕop(PS) values
for different confining stress levels at failure by using the Bolton (1986) Eqs. 11 or 12 and referring
to Eqs. 15, 16 and 17 for some typical boundary value problems. However, since σ’mf is a function
of qlim that in turn depends on ϕop, a trial and error procedure must be adopted.

ϕcv = ϕµ + 3° = 34° ; Q = lnσ’c = 9.5 ; σ’c in kPa

50000
ϕp = 46°
ϕp = 48° ϕp @ σ'mf=σ'm0
40000 ϕp = 44° σ'm0=σ'v0(1+2K0)/3
Point resistance, qc (kPa)

30000 ϕp = 42°

ϕp = 40°
20000
ϕp= 38°
10000 ϕp = 36°

ϕp = 34°
0
0 50 100 150 200 250 300
Geostatic mean effective stress, σ'm0 (kPa)

Figure 14. Peak friction angle from CPT for silica sands using Bolton (1986) theory
29
 ϕcv = ϕµ + 3° = 38° ; Q = lnσ’c = 8.5 ; σ’c in kPa
50000
ϕp =52°
ϕp = 50° ϕp = 46° ϕp @ σ'mf=σ'm0
40000 ϕp = 48° σ'm0=σ'v0(1+2K0)/3
Point resistance, qc (kPa)

ϕp = 44°
30000

ϕp = 42°
20000

10000 ϕp = 40°

ϕp = 38°
0
0 50 100 150 200 250 300
Geostatic mean effective stress, σ'm0 (kPa)

Figure 15. Peak friction angle from CPT for calcareous sands using Bolton (1986) theory

50000
ϕp = 46° ϕp = 42°
ϕp = 40°
ϕp = 48°
ϕp = 44°
40000
Point resistance, qc (kPa)

ϕp @ σ'mf=σ'm0
σ'm0=σ'v0(1+2K0)/3
30000 ϕp = 38°

20000 ϕp = 36°

10000 ϕp = 34°

0
0 50 100 150 200 250 300
Geostatic mean effective stress, σ'm0 (kPa)

Figure 16. Peak friction angle from CPT for quartz sands using Bolton (1986) theory

30
 For the estimation of qlim and Nq of Eqs. 15 and 16 it is recommended to adopt
ϕop=(ϕp+ϕcv)/2 so as to take also into account, beyond the effect of the curvature of the strength
envelope, the influence, on the mobilized average friction angle, of the strain level in
correspondence of the limit pressure of the considered boundary value problem (i.e. progressive
failure effect).
Finally, it is worthy to recall that the proposed procedure for the estimation of ϕp and ϕop
allows also to consider the over-consolidation ratio (OCR) of the considered sand via the Ko value.
Unfortunately, at the present state of the art, the quantification of the OCR degree and then of Ko,
within coarse grained materials, is still very unreliable.

Closing remarks

1. Eq. 5, referring to the empirical coefficient C0, C1 and C2, shown in Table 4, allows the
estimation of DR in deposits of unaged, uncemented silica sands of low to moderate
compressibility.
2. The use of the above equations involves the effective overburden stress σ’vo in case of NC sands
but requires the estimation of the mean geostatic stress σ’mo for OC deposits.
3. An alternative approach for the evaluation of DR by Lancellotta (1983) and later reworked by
Garizio (1997) is displayed in Fig. 4. This correlation, using the available data-base for both NC
and OC sands makes reference to σ’vo.
4. Making reference to the above mentioned DR = f ( q c , σ’) correlation, an iterative approach
based on the Bolton’s strength-dilatancy theory is proposed to estimate ϕ’p on the basis of the
CPT results.
5. The proposed approach, at least in principle, allows evaluation, for a given sand, of ϕ’p taking
into account: sand compressibility, curvilinear nature of the shear strength envelope and
imposed strain conditions.
6. Figures from 14 to 16 display the correlations ϕ’p = (qc,σ’mo) for three qualitatively defined
classes of sand compressibilities assuming σ’mf = σ’mo. For σ’mf different than the σ’mo iterative
procedure outlined in Fig. 13 should be followed.
7. A comparison between ϕ’p = f(qc,σ‘) yielded by Fig. 14 and ϕ‘p, obtained from drained
compression loading CKoD triaxial tests, suggests that the proposed procedure tends to
underestimate ϕ’p by 1° to 1°½, see Jamiolkowski (1990).
8. On the basis of the aforementioned approaches, an iterative procedure has been suggested for
the estimation of the operational friction angle ϕ‘op referred to some current limit equilibrium
boundary value problem. This approach allows also to take into account, in a first
approximation, the progressive failure aspect.
9. This paper gives also engineering correlations between Marchetti’s (1980) dilatometer lateral
stress index KD, DR and ϕ’p respectively as inferred from CC tests on siliceous sands.

References

Achintya, H. & Tang, W.H. (1979) “Uncertainty Analysis of Relative Density” Journal of
Geotechnical Engineering Division, ASCE, 105(GT7), 899-904.

Bachus R.C. (1983) “An investigation of the strength deformation response of naturally occuring
lightly cemented sands” Ph.D. Thesis, Civil Engineering Stanford University, CA.

31
Baldi, G., Bellotti, R., Ghionna, V., Jamiolkowski, M. & Pasqualini, E. (1986) “Interpretation of
CPT's and CPTU's. 2nd Part: Drained Penetration” Proceeding 4th International Geotechnical
Seminar, Singapore, pp.143-156.

Baligh M.M. (1975). “Theory of deep static cone penetration resistance” Research Report R75-56,
MIT, Cambridge, Mass.

Baligh M.M. (1976) “Cavity Expansion in Sands with Curved Envelopes” Journal of Geotechnical
Engineering Division, ASCE, 102(GT11), 1131-1146.

Barton N.R. (1973). “Review of a new shear strength criteria for rock joints” Engineering Geology,
7, 287-332.

Barton M.E. & S.N. Palmer (1989) The Relative Density of Geologically Aged British Fine and
Fine-medium Samds. Quaterly Journal of Engineering Geology. 1: 49-58.

Been K., Jefferies M.G., Crooks, J. H. & Rothenburg L. (1987) “The cone penetration test in sands
II: general inference of state” Geotechnique, 37(3), 285-299.

Bellotti, R., Bizzi, G. & Ghionna, V.N. (1982) “Design, Construction and Use of Calibration
Chamber” Proc. 2nd ESOPT, Amsterdam, Vol.2, pp.439-446.

Bellotti, R., Crippa, V., Pedroni, S. & Ghionna, V.N. (1988) “Saturation of Sand Specimen for
Calibration Chamber Tests” Proc. ISOPT-1, Orlando, Vol.2, pp.661-672.

Bellotti R., Ghionna V.N., Jamiolkowski M., & Lancellotta R. (1989) “Shear strength from CPT”
Proc. XII ICSMFE, Vol 1, pp. 165-170, Rio de Janeiro.

Bellotti R., Benoit J., Fretti C. & Jamiolkowski M. (1997) “Stiffness of Toyoura sand from
dilatometer Tests” Journal of Geotechnical and Geoenvironmental Engineering, ASCE, 123(9),
836-846.

Biliam J. (1967). “Some aspects of the behaviour of granular materials at high pressure” Proc.
Roscoe Memorial, Stress Strain Behaviour of Soils, Cambridge University.

Bolton M.D. (1986) “The strength and dilatancy of sands” Geotechnique, 36(1), 65-78

Brinch Hansen J. (1970) “A revised and extended formula for bearing capacity”. Danish
Geotechnical Institute Bulletin, 11.

Burmister, D.M. (1948) “The Importance and Practical Use of the Relative Density in Soil
Mechanics” ASTM Proc. 48, 1249-1268.

Carter J.P., Booker J.R. & Yeung S.K. (1986) “Cavity expansion in cohesive frictional soils”
Geotechnique, 36(3), 349-353.

Casagrande A. (1936) “Characteristics of cohesionless soils affecting the stability of slopes and
earth fills” Journal of Boston Society of Civil Engineers., Jan.: 257-276.

Chen J.W. & Juang C.H. (1996) “Determination of drained friction angle of sands from CPT”
Journal of Geotechnical Engineering Division, ASCE, 122(5), 374-381.

32
Cornforth D.H. (1964) “Some experiments on the influence of strain conditions on hte strength of
sands” Geotechnique, 14(2), 143-167.

Cubrinovski, M. & Ishihara, K. (1999) “Empirical Correlation between SPT N-value and Relative
Density of Sandy Soils” Soils and Foundations, 39(5), 61-71.

De Beer E.E. (1967) “Bearing capacity and settlement of shallow foundations in sand”. Proc.,
Bearing capacity and settlement of foundations, Duke University, Durham, N.C.

Durgunoglu H.T. & Mitchell J.K. (1975). “Static penetration resistance of soils, I-Analysis, II-
Evaluation of the Theory and implications for practice”. ASCE Spec. Conference In situ
measurements of soil properties, Vol. 1, Raleigh, NC.

Dussealt M.B. & Morgenstern N.R. (1979) “Locked Sands” Quarterly Journal of Engineering
Geology. 2: 117-131

Eslaamizaad, S. & Robertson, P.K. (1996) “Estimation of In-Situ Lateral Stress and Stress History
in Sands” Proceeding 49th Canadian Geotechnical Conference, St. John's, Newfoundland, pp.439-
448.

Fleming W.G.K., Weltman A.J., Randolph M.F. & Elson W.K. (1992) “Piling engineering” Blakie
Academic and Professional, Glasgow.

Foray, P. (1986) “First Results in the I.M.G. Calibration Chamber: the Boundary Condition
Problem” 1st International Seminar on Research Involving Validation of In-Situ Devices in Large
Calibration Chambers, Milano, unpublished.

Felice A. (1997). Determinazione dei Parametri Geotecnici e in Particolare di Ko da Prove

Dilatometriche. M.Sc. Thesis, Department of Structural Engineering, Politecnico di Torino.

Fretti C., Lo Presti D.C.F. & Salgado R. (1992) "The Research Dilatometer: in Situ and Calibration
Chamber Test Results" Rivista Italiana di Geotecnica, Vol. XXVI, n. 4, pp 237-243.

Fretti C., Froio F., Jamiolkowski M., Lo Presti D.C.F., Olteanu A. & Pedroni S. (1996)
"Dilatometer Tests in Calibration Chamber: Stiffness and Ko Assessment" VIII Conferenza
Nazionale Rumena di Geotecnica, Iasi, Romania, 25-27 September 1996. Vol. I, pp. 85-94.

Garizio G.M. (1997) “Determinazione dei parametri geotecnici e in particolare di K0 da prove

penetrometriche” M.Sc. Department of Structural Engineering, Politecnico di Torino.

Gibbs, H.J. & Holtz, W.G. (1957 “Research on Determining the Density of Sands by Spoon
Penetration Testing” Proceedings 4th ICSMFE, Vol.1, pp.35-39, London.

Goto, S., Suzuki, Y., Nishio, S. & Oh Oka, H. (1992) “Mechanical Properties of Undisturbed Tone-
River Gravel obtained by In-Situ Freezing Method” Soils and Foundations 32(3), 15-25.

Hanna A.M., Massoud N. & Youssef H. (1987) “Prediction of plane-strain angles of shear
resistance from triaxial test results” Proc. Prediction and Performance in Geotechnical
Engineering, Univ. of Calgary, pp. 369-376.

33
Harman, D.H. (1976) “A Statistical Study of Static Cone Bearing Capacity, Vertical Effective
Stress and Relative Density of Dry and Saturated Fine Sands in Large Triaxial Test Chamber”
M.Sc. Dissertation, University of Florida, Gainesville, Fla.

Hatanaka, M., Suzuki, Y., Kawasaki, T. & Endo, M. (1988) “Cyclic Undrained Shear Propeties of
High Quality Undisturbed Tokyo Gravel” Soils and Foundations, 28(4), 57-68.

Houlsby, G.T. & Hitchman, R. (1988) “Calibration Chamber Tests of Cone Penetrometer in Sand”
Geotechnique, 38(1), 39-44.

Jamiolkowski M. (1990) “Shear strength of cohesionless soils from CPT” De Mello Volume, pp.
191-204, Editoria Edgar Blucker, San Paulo.

Jamiolkowski M., Ghionna V.N., Lancellotta R. & Pasqualini E. (1988). “New correlations of
penetration tests for design practice”. Proc., Penetration Testing 1988, ISOPT 1, Orlando, Florida,
J. De Ruiter ed., Vol. 1 pp: 263-296

Jamiolkowski M. & Lo Presti D. (2000). “Shear strength of coarse grained soils from in situ tests. A
compendium” Proc., 4th International Geotechnical Engineering Conference, Cairo, Egypt.

Janbu N. & Senneset K. (1974) “Effective stress interpretation of in situ static penetration tests”
Proc., ESOPT I, Stockholm. Vol. 2.2, pp. 181-194.

Kulhawy, F.H. & Mayne, P.H. (1990) “Manual on Estimating Soil Properties for Foundation
Design” Electric Power Research Institute, EPRI.

Ladd C.C., Foott R., Ishihara K., Schlosser F. & Poulos H.G. (1977) “Stress-deformation and
strength characteristics” S.O.A. Report, 9th ICSMFE, Tokyo, 2: 421-494.

Lade P.V. & Lee K.L. (1976). “Engineering properties of soils” Report UCLA-ENG-7652,
University of Califirnia at Los Angeles.

Lancellotta R. (1983) “Analisi di Affidabilità in Ingegneria Geotecnica” Atti dell’Istituto di Scienza

delle Costruzioni, No 625, Politecnico di Torino.

Lee D.M. (1992) “The angles of friction of granular fills” Ph.D. Thesis, Cambridge University.

Lee K.L. & Seed H.B. (1967) “Drained strength characteristics of sands”. Journal of Soil
Mechanics and Foundation Division, ASCE, Vol. 93(SM6), Proc. Paper 5561, 117-141.

Leussink H., Wittke W. & Weseloh K. (1966) “Unterschiede im scherverhalten rollinger erdstoffe
und kugelschuttungen im dreiaxial und biaxialversuch“ Veroffentlichungen des Institut fur
bodenmechanik und felsmechanik, Universitat Karlsruhe.

Lunne, T., Robertson, P.K. & Powell, J.J.M. (1997) “Cone Penetration Testing in Geotechnical
Practice” Blackie Academic and Professional, London.

Marchetti S. (1980) “In situ tests by Flat Dilatometer” Journal of the Geotechnical Engineering
Division, ASCE, Vol. 106(GT3), 299-321.

34
Marchetti, S. (1997) “The Flat Dilatometer: Design, Applications” Proceeding 3rd International
Geo-technical Engineering Conference, Cairo, pp.421-428.

Marsal R.J. (1967) “Large scale testing of rockfill materials” Journal of Soil Mechanics Foundation
Division, ASCE, 93(SM2), 27-43.

Mayne, P.W. & Kulhawy, F.H. (1991) “Calibration Chamber Data Base and Boundary Effects
Correction on CPT Data” Proceedings International Symposium on Calibration Chamber Testing,
Potsdam, NY, pp.257-264.

Mitchell J.K. (1976). “Fundamentals of Soil Behaviour”. J. Wiley & Sons, New York.

Mitchell J.K. & Keaveny J.M. (1986) “Determining sand strength by penetrometers” Proc. ASCE
GSP No. 6 Use of In Situ Tests in Geotechnical Engineering. Blacksburg, Va, pp. 823-929.

Nader S.R. (1983) “Static and dynamic behaviour of cemented sands“ Ph.D. Thesis, Civil
Engineering Stanford University, CA

Park C.S. & Tatsuoka F. (1994) “Anisotropic strength and deformation of sands in plane strain
compression“ Proc. XIII ICSMFE New Delhi, India, Balkema, Rotterdam, pp. 1-4.

Parkin, A.K. & Lunne, T. (1982) “Boundary Effects in the Laboratory Calibration of a Cone
Penetrometer in Sand” Proc. 2nd ESOPT. Amsterdam, Vol.2, pp.761-768.

Pradhan T.B.S. & Tatsuoka F., Horii N. (1988) ”Simple shear testing on sands in a torsional shear
apparatus” Soil and Foundations 28(2), 95-112.

Puppala, A.J. (1993) “Effect of Cementation on Cone Resistance in Sands: A Calibration Chamber
Study”. Ph. D. Thesis, Louisiana State University, Baton Rouge.

Puppala, A.J., Acar, Y.B. & Tumay, M.T. (1995) “CPT in Very Weakly Cemented Sand”
Proceedings International Symposium on Cone Penetration Testing, CPT '95, Vol.2, pp.269-276.

Roberts J.E. (1964) “Sand compression as a factor in oil field subsidence”. Ph.D. Thesis,
Massachusetts Institute of Technology.

Robertson P.K. & Campanella R.G. (1983) “Interpretation of cone penetration tests. Part I: Sand”
Canadian Geotechnical Journal, 20(4), 718-733.

Rowe P.W. (1962). “The stress-dilatancy realtion for static equilibrium of an assembly of particles
in contact” Proc. Royal Society London.

Salgado R. (1993). “Analysis of penetration resistance in sands”. Ph.D. Thesis, University of

California, Berkeley.

Salgado R., Mitchell J.K. & Jamiolkowski M. (1997). “Cavity expansion and penetration resistance
in sand” Journal of Geotechnical Engineering Division, ASCE, 123(4), 344-354

Salgado R., Mitchell J.K. & Jamiolkowski M. (1998). “Calibration Chamber Size effects on
Penetration resistance in Sand” Journal of Geotechnical Engineering Division, ASCE, 124(9), 878-
888

35
Salgado R., Bandini P. & Karim A. (2000) “Shear Strength and Stiffness of Silty Sand” Journal of
Geotechnical Engineering Division, ASCE, 126(5), 451-462.

Schanz T. & Vermeer P.A. (1996) “Angles of friction and dilatancy of sands“ Geotechnique, 46(1),
145-151.

Schanz T: (1998). “Zur modellierun des mechanischen verhaltens von reibungsmaterilaen“ Institute
fur Geotechnik, Universitat Stuttgart, Mitteilung 45, Edit. P.A. Vermeer.

Schmertmann J.H. (1971) “Discussions and Conference Record – Discussion Session 1”

Proceedings of the Fourth Panamerican Conference on Soil Mechanics and Foundation
Engineering, San Juan, Puerto Rico, June 1971, ASCE, Vol. III

Schmertmann J.H. (1972) “Effects of in situ lateral stress on friction – cone penetrometer data in
sands” Verhandelingen Fugro, Sonder Symposium 1972, pp: 37-39.

Schmertmann, J.H. (1976) “An Updated Correlation between Relative Density DR and Fugro-Type
Electric Cone Bearing, qc.” Contract Report DACW 39-76 M 6646 WES, Vicksburg, Miss., 1976

Schmertmann J.H. (1978) „Guidelines for cone penetration test performance and design” US Dept.
of Transportation, FHWA, R.78-209. Washington D.C.

Schmertmann, J.H. (1991) “The Mechanical Ageing of Soils” Journal of Geotechnical Engineering
Division, ASCE, 117(9), 1288-1330.

Schofield A. (1998). “Mohr Coulomb error correction”. Ground Engineering, 31(8), 30-32.

Schofield A.N. & Wroth C.P. (1968) “Critical state soil mechanics” McGraw-Hill, London.

Skempton, A.W. (1986) “Standard Penetration Test Procedures and the Effects in Sands of
Overburden Pressure, Relative Density, Particle Size and Overconsolidation” Geotechnique, 36(3),
425-447.

Tanizawa, F. (1992) “Correlations between Cone Resistance and Mechanical Properties of Uniform
Clean Sand” Internal Report, ENEL-CRIS, Milano.

Tatsuoka F., Sonoda F., Hara K., Fukishima S. & Pradhan T.B.S. (1986) ”Failure and deformation
of sands in torsional shear” Soil and Foundations 26(4), 79-97.

Tavenas, F.A. (1972) “Difficulties in the Use of Relative Density as a Soil Parameter” ASTM, STP
523, Selig and Ladd Editors, pp.478-483.

Tavenas, F. & La Rochelle, P. (1972) “Accuracy of Relative Density Measurements”

Geotechnique, 22(4), 549-562.

Taylor D.W. (1948) ”Fundamentals of soil mechanics”. J. Wiley & Sons, New York.

Terzaghi, K. & Peck, R.B. (1948) “Soil Mechanics in Engineering Practice” J. Wiley & Sons New
York.

36
 Trofimenkov J.B. (1974) “Penetration Testing in USSR” State of the Art Report, Proc., ESOPT I,
Stockholm. Vol. 1, pp. 147-154.

Tumay M. (1976) “Cone Bearing vs. Relative Density Correlation in Cohesionless Soils” Dozen
Dissertation. Bogazici Univ. Istanbul.

Vesic A.S. (1972) “Expansion of cavities in infinite soil mass” Journal of Soil Mechanics
Foundation Division, ASCE, 98(SM3), 265-290.

Vesic A.S. (1975) “Bearing capacity of shallow foundations” Foundation Engineering Handbook,
Winterkorn and Fang eds., Van Nostrand Reinhold.

Vesic A.S. (1977) “Design of pile foundations” National Cooperative Highway Research
programs. Synthesis of Highway Practice 42, Transportation Research Board, Washington D.C.

Wride C.E., Robertson P.K., Biggar K. W., Campanella R.G., Hofmann B.A., Hughes J.M.O.,
Küpper A. & Woeller D.J. (2000) “Interpretation of in situ test results from the CANLEX sites”
Canadian Geotechnical Journal, 37(3), 505-529.

Yoon Y. (1991). “Static and dynamic behaviour of crushable and non-crushable sands” Ph.D.
Thesis, Ghent, University.

Yoshimi, Y. (2000) “A Frozen Sample that did not Melt” Proceedings of Conference on
Development in Geotechnical Engineering. GEOTECH-YEAR 2000, pp.293-296, Bangkok.

Yoshimi, Y., Hatanaka, M. & Oh-Oka, H. (1978) “Undisturbed Sampling of Saturated Sands by
Freezing” Soils and Foundations, 18(3), 59-73.

Yu H.S. & Mitchell J.K., (1998) “Analysis of cone resistance: review of methods” Journal of
Geotechnical Engineering Division ASCE, 124(2), 140-149.

Zolkov, I. & Weisman, G. (1965) “Engineering Properties of Dune and Beach Sands and the
Influence Stress History” Proc. 6th ICSMFE, Montreal, Vol.I, pp.134-138.

37

View publication stats