Soil Dynamics and Earthquake Engineering 55 (2013) 83–91

Contents lists available at ScienceDirect

Soil Dynamics and Earthquake Engineering

journal homepage: www.elsevier.com/locate/soildyn

Static and cyclic behavior of North Coast calcareous sand in Egypt

Manal Salem n, Hussein Elmamlouk, Shehab Agaiby
Faculty of Engineering, Cairo University, Giza, Egypt

art ic l e i nf o a b s t r a c t

Article history: This research studies the behavior of calcareous sand located in the North Coast of Egypt, Dabaa area,
Received 23 October 2012 under static and cyclic loading. The study is performed through a series of monotonic and cyclic
Received in revised form undrained triaxial tests conducted on two relative densities and different effective confining pressures.
30 July 2013
The cyclic tests were carried out at different cyclic stress ratios. Failure under cyclic loading was found to
Accepted 9 September 2013
Available online 8 October 2013
be governed by the gradual development of excess pore water pressure until liquefaction is reached,
rather than cumulative development of axial strain (cyclic mobility). The cyclic strength of the tested
Keywords: sand is compared with other calcareous and siliceous sands reported in the literature. The test results
Calcareous sand indicate that loose Dabaa calcareous sand has higher cyclic strength compared to other siliceous sands,
Static loading
probably due to the existence of different shapes of calcareous sand particles within the tested soil.
Cyclic loading
Relationship between cyclic resistance ratio, effective confining pressure, and relative density was
Triaxial
Liquefaction developed for the tested sand.
& 2013 Elsevier Ltd. All rights reserved.

1. Background size of the particles as well as the composition of carbonates

Calcareous sands are the result of various biological, mechan-
ical, physical, and chemical depositional environments [1,2]. There
are two main reasons behind the difference in their properties and
behavior compared to other siliceous sands subjected to similar
loading conditions. First, calcareous sands have remarkable intra-
particle void space (within the particles), which is mainly caused
by the particles being made of shells and corals that include
cavities and voids inside their bodies or on their surfaces [2–7].
Second, the irregular particle shape of this sand, which can be
found in various forms such as curved flat particles coming from
fragments of sea shells or hollow tube-shaped particles coming
from the remains of skeletons of small marine organisms resulting
in remarkable inter-particle space (between particles). Both rea-
sons lead to higher possible compressibility of calcareous sands
associated with crushing of calcareous grains when sheared.
Different particle breakage factors have been suggested in
attempts to quantify the amount of particle breakage upon
loading; these breakage factors are empirical and depend mainly
on changes in particle sizes [8,9].
Common soil classification systems (such as the Unified Soil
Classification System) do not distinguish between calcareous and
non-calcareous sands. Classification of calcareous sand/gravel has
been proposed by Hallsworth and Knox [10] to reflect the grain
n
Corresponding author. Tel.: þ 20 1001 555 454; fax: þ 20 2330 24 009.
E-mail addresses: manalasalem@eng.cu.edu.eg, manalasalem@gmail.com
(M. Salem).
forming the particles. Typically for sands, minimum void ratios
vary from 0.2 to 0.5, and maximum void ratios vary from 0.8 to 1.2
[11]. Higher void ratios have been reported for calcareous sands.
Cataño [12] presented minimum and maximum void ratios for
thirteen (13) calcareous sands, where minimum void ratios ranged
from 0.5 to 1.6, and maximum void ratios ranged from 1.1 to 2.0.
Soil specific gravity is strongly dependant on the mineralogy of the
soil particles. The specific gravity of minerals composing calcar-
eous soils like calcite is 2.75 and aragonite is 2.95 [13]. Siliceous
minerals, on the other hand, are less heavy since they typically
include quartz with a specific gravity value of about 2.65. Specific
gravities for calcareous sands typically range from 2.71 to 2.86,
while that for siliceous sands can be less than 2.65 [4,5,12,14,15].
The state of phase transformation is important while studying
undrained behavior of sands. Phase transformation is a concept
developed by Ishihara et al. [16] that refers to a transient state at
which the behavior of soil changes from contractive to dilative; the
pore water pressures begin to decrease at this point (phase
transformation point, PT) as the deviator stress increases. LaVille
[15] mentioned that the typical behavior of calcareous crushable
sand during undrained shearing is to adopt an initial contractive
tendency followed by a tendency for dilation. Hyodo et al. [6]
found that at low confining pressures, calcareous sands may dilate
without exhibiting the initial contractive behavior observed at
higher confining pressures. With the increase of effective confin-
ing pressure (s′c), both loose and dense calcareous specimens
adopted an initial contractive behavior followed by a tendency for
dilation. Thus, as s′c increases, both loose and dense sands tend to
be more contractive. In p′
q space, where p′¼(s′1 þ s′3)/2 and

0267-7261/$ - see front matter & 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.soildyn.2013.09.001
84 M. Salem et al. / Soil Dynamics and Earthquake Engineering 55 (2013) 83–91

q ¼(s′1
s′3)/2 according to MIT stress path convention, the line Table 1
drawn through the PT point from the origin is known as the phase Index properties of Dabaa calcareous sand versus other calcareous sands reported
in the literature [6,12,14,15].
transformation line [16]. Hyodo et al. [6] compared the phase
transformation stress ratio q/p′ for loose and dense Shirasu Sand Gsa D10 (mm)b Cu c emind emaxe
crushable sands and for dense and medium dense Dogs Bay
carbonate sands and found that this ratio is independent of North coast 2.79 0.15 2.40 0.75 1.04
relative density and is constant for a given soil. The previous Cabo Rojo 2.86 0.20 1.05 1.34 1.71
Playa Santa 2.75 0.16 2.75 0.80 1.22
finding was assured by LaVielle [15]. Dogs Bay 2.75 0.24 2.06 0.98 1.83
Different methods have been used to evaluate the liquefaction Ewa Plains 2.72 0.20 5.05 0.66 1.30
susceptibility of soil subjected to cyclic loading, among which
a
studying the number of cycles to failure versus Cyclic Stress Ratio Specific gravity.
b
Effective grain size.
(CSR) defined as the ratio of the applied cyclic shear stress to the c
Uniformity coefficient.
initial s′c. Researchers have investigated the effects of relative d
Minimum void ratio.
density (Dr), s′c, and soil type on the number of cycles–CSR relation- e
Maximum void ratio.
ship. Failure is commonly defined as liquefaction, which corresponds
to the number of cycles at which the pore water pressure ratio (ru)
equals 1, where the pore water pressure ratio is defined as the ratio 2.00
Standard Proctor Test
of excess pore water pressure to initial effective confining pressure
1.95 Modified Proctor Test
(ru ¼ Δu/s′c). Sometimes, large strains may occur without reaching ru
of 1.0. Accordingly, in studying the cyclic behavior of soils, failure 1.90
criterion is usually defined as the number of cycles required to cause

Dry Density (gm/cm3)

1.85
either liquefaction (ru ¼1.0) or a specified axial strain (ε) amplitude
(ε¼5% in double amplitude [17]). 1.80
Compared to siliceous sands, particle crushing and rearrange-
1.75
ment mechanisms are more significant in calcareous sands, which
make calcareous sands more compressible than siliceous sands [2]. 1.70
Of particular note is how these mechanisms affect the pore water
1.65
pressure generation during undrained monotonic and cyclic loading
and understanding the potential for liquefaction during an earthquake. 1.60
0 2 4 6 8 10 12 14 16 18
This study aims at assessing the monotonic and cyclic behavior of
Water Content (%)
Dabaa calcareous sand through a laboratory experimental program.
Fig. 2. Standard and modified Proctor compaction curves.

2. Physical properties of tested material

The tested calcareous sand was obtained from the near surface Table 2
Elemental content and mineral composition of
of a site located at about km 135 Alex-Matrouh road, Dabaa area,
Dabaa calcareous sand.
North Coast, Egypt. The collected soil consists of irregular-shaped
grains with light tan to white color. The tested sand has a fairly Mineral Percentage (%)
uniform gradation (medium sand), with grain sizes mostly ranging
from 0.2 to 2 mm, and fines content of about 8.75%. The soil is SiO2 0.28
TiO2 0.02
classified as poorly graded sand (SP-SM) according to the Unified Al2O3 0.12
Soil Classification System (ASTM D2488); and is classified as Fe2O3 0.02
calcite-sand according to Hallsworth and Knox [10]. The grain size MnO o 0.01
distribution curve of the tested sand is plotted with other calcar- MgO 0.2
CaO 55.4
eous sands reported in the literature in Fig. 1.
Na2O o 0.01
Maximum and minimum void ratios for the Dabaa calcareous K2O 0.02
sand (ASTM D4253 and D4254, respectively) are 1.043 and 0.753, P2O5 0.06
SO3 0.12
Cl o 0.01
L.O.I.a 43.53
100
Dabaa a
90 Playa Santa L.O.I: loss on Ignition.
80 Dogs Bay

70
Cabo Rojo respectively. The specific gravity (ASTM D854) of the tested sand is
Ledge Point
Ewa Plains
2.79. Index properties of Dabaa calcareous sand are compared to
60
% Passing

other calcareous sands reported in the literature in Table 1.

50
Standard and modified proctor compaction tests were conducted
40 on the Dabaa calcareous sand according to ASTM D698 and ASTM
30 D1557, respectively. The results of both tests are plotted in Fig. 2.
20 X-Ray Diffraction analysis revealed that the sample is com-
10 posed of a major constituent of Calcite and Aragonite and a trace
0
constituent of Quartz. The mineralogy and geochemistry analysis
10 1 0.1 0.01 of the carbonate sample indicated the presence of several minerals
Diameter (mm) in the sand sample with different percentages illustrated in
Fig. 1. Grain size distribution curves of Dabaa calcareous sand versus other Table 2. A thin section was prepared for microscopic investigation
calcareous sands reported in the literature [6,7,12,14,15]. and images of the prepared section under microscopic
 M. Salem et al. / Soil Dynamics and Earthquake Engineering 55 (2013) 83–91 85

900
Dr = 40%
800

Deviator Stress, ( 1- 3) (kPa)

50 kPa
700
100 kPa
600 200 kPa
500
400
300
200
100
0
0 2 4 6 8 10 12 14 16
Axial Strain, ε (%)

1.0
0.5

Normalized Pore Water Pressure,

0.0
0 2 4 6 8 10 12 14 16
-0.5
-1.0

Δu/ 'c
-1.5
-2.0 Dr = 40%
-2.5 50 kPa
-3.0 100 kPa
200 kPa
-3.5
Axial Strain, ε (%)

Fig. 3. A section of Dabaa sand under microscopic investigation. Fig. 5. Behavior of loose calcareous sand under monotonic triaxial loading:
(a) deviator stress versus axial strain; (b) normalized pore water pressure versus
axial strain.
1.0
0.9 1000
0.8 900 Dr = 80%
Deviator Stress, ( 1- 3) (kPa)

800 50 kPa
0.7
Relative Density, Dr

100 kPa
700
0.6 200 kPa
600
0.5
500
0.4 400
0.3 300

0.2 200
Initial Dr = 75% 100
0.1
Initial Dr = 20% 0
0.0 0 2 4 6 8 10 12 14 16
0 50 100 150 200 250 300 350 400 450 Axial Strain, ε (%)
Effective Cell Pressure (kPa)
1.0
Fig. 4. Target relative density contour lines.
0.5
Normalized Pore Water Pressure,

0.0
investigation are shown in Fig. 3. The tested sand consisted of 0 2 4 6 8 10 12 14 16
-0.5
particles with different shapes. Shapes of individual particles
-1.0
varied from spherical to elongated, with mostly rounded edges
Δu/ 'c

and smooth surfaces [18]. Cementation between some particles -1.5

was observed as well. -2.0 Dr = 80%
-2.5 50 kPa
-3.0 100 kPa
3. Monotonic triaxial testing 200 kPa
-3.5
Axial Strain, ε (%)
An electro-pneumatic Cyclic Triaxial Test System was used to
conduct static (monotonic) and dynamic triaxial tests. Two con- Fig. 6. Behavior of dense calcareous sand under monotonic triaxial loading:
solidation tests were conducted at initial relative densities of 20% (a) deviator stress versus axial strain; (b) normalized pore water pressure versus
axial strain.
and 75% in the triaxial apparatus in order to determine the change
in the initial relative density due to the application of a certain
effective confining pressure. The results of these tests are pre- tamping to prepare the sample at a target relative density using
sented in Fig. 4. Using Fig. 4 and knowing the test effective undercompaction technique [19].
confining pressure and target relative density at end of consolida- Typically, the saturation process of sand specimens does not
tion stage, the initial relative density at sample preparation was require high back pressures and long periods due to the soil
determined. Sample preparation procedure is the same for the relatively high permeability. However, it has been the authors'
monotonic and cyclic triaxial tests. Samples were prepared by experience that the saturation of calcareous sand is not as easy as
86 M. Salem et al. / Soil Dynamics and Earthquake Engineering 55 (2013) 83–91

300 water pressure, axial strain at the point of phase transformation,

Dr = 40% and axial strain at zero excess pore water pressure. Samples
250 'c = 200 kPa
prepared at high void ratios (Dr ¼40%) exhibit larger contractive
q = ( 1- 3)/2 (kPa)

200 tendency compared to those prepared at low void ratios

umax (Dr ¼ 80%) as the soil structure is looser. For samples prepared
PT
150 at high void ratios (Dr ¼40%), shearing disturbs the soil structure
and soil grains tend to rearrange to a denser state of packing. For
100 samples prepared at low void ratios (Dr ¼ 80%), the initial state of
50
packing is dense and thus exhibits less contraction. Both loose
Total Stress Path
and dense states of the tested calcareous sand show no clear
Effective Stress Path
0 failure under monotonic loading, but rather a ductile behavior
0 100 200 300 400 500 600
upon shearing is apparent.
Mean Principal Stress, p, p' (kPa)
The analyses of the monotonic triaxial tests were presented in
Fig. 7. Total and effective stress paths for loose sample tested at s′c ¼ 200 kPa. p′
q space to express the total stress path (TSP) and effective
stress path (ESP) of each tested specimen, where q¼ (s′1
s′3)/2
and p′¼(s′1 þ s′3)/2 according to MIT stress path convention. At
200 phase transformation point, maximum excess pore water pressure
Dr = 40%
180 is measured, where the behavior changes from a tendency for
Dr = 80%
160 y = 0.5539x contraction to a tendency for dilation, thus the difference between
q = ( 1- 3)/2 (kPa)

R² = 0.8793
140 the effective stress and the total stress, p′ and p, is greatest, which
120 is illustrated with an example in Fig. 7.
100 From the conducted monotonic triaxial tests, the p′ and q
80 associated with phase transformation point were defined for
60 five samples and a best-fit line through these points and
40 passing through the origin defined the phase transformation
20 line (PT Line) in the p′
q space. A diagram of the PT line of
0 Dabaa calcareous sand is illustrated in Fig. 8. It is noted that the
0 50 100 150 200 250 300 350
Mean Principal Stress, p' (kPa) phase transformation points are scattered around the PT line,
irrespective of the sample relative density. This observation
Fig. 8. Phase transformation (PT) line for Dabaa calcareous sand. suggests that the phase transformation stress ratio (q/p′) for
loose and dense states is probably independent of the relative
the saturation of quartzitic sands. More time and attention is density, which agrees with the findings of Hyodo et al. [6] and
necessary, most likely due to the intra-particle void structure in LaVielle [15].
this type of sand. Thus, flushing the air outside the sample and
sample saturation are two critical steps prior to monotonic and
cyclic triaxial testing. The adopted testing approach uses a stair 4. Cyclic triaxial testing
step procedure for saturation, where the confining and back
pressures are increased simultaneously every approximately A series of stress controlled undrained cyclic triaxial tests was
40 min by 50 kPa. Then, the sample was left overnight under a conducted on the Dabaa calcareous sand to study the factors
back pressure of about 280 kPa. A minimum back pressure of
400 kPa was applied to most samples by the end of the saturation 1.2
Dr = 40%
process. Skempton's [20] pore-pressure parameter “B” of 0.98 was 1.0 CSR = 0.225
achieved for all test samples. 'c = 100 kPa
Pore Pressure Ratio, ru

The undrained behavior of the Dabaa calcareous sand under 0.8

monotonic loading was studied through a series of isotropically 0.6

consolidated compression monotonic undrained triaxial (ICU)
0.4
tests with pore water pressure measurements conducted in
accordance with ASTM D4767. Tests were carried out at three 0.2
effective confining pressures (s′c) of 50, 100, and 200 kPa, and 0.0
two relative densities (Dr) of 40 and 80%. The variation of the
deviator stress (Δsd ¼ s1
s3) and normalized excess pore water -0.2
0 20 40 60 80 100 120 140 160
pressure (Δu/s′c) with axial strain (ε) under different effective Number of Cycles
confining pressures of 50, 100, and 200 kPa for loose (Dr ¼40%)
0 20 40 60 80 100 120 140 160
and dense (Dr ¼80%) specimens is presented in Figs. 5 and 6, 1
respectively.
0
In Figs. 5 and 6, all loose samples and two of the dense
samples (tested at s′c ¼100 and 200 kPa) adopted similar initial
Axial Strain (%)

-1
contractive behavior associated with the generation of positive
pore water pressure (Δu), which continues until Δumax is reached -2
at phase transformation point (PT). As the axial strain increases
-3
beyond PT point, the excess pore water pressure decreases
indicating dilation of tested samples. Similar behavior was -4
reported by Hyodo et al. [6] and LaVielle [15]. Note that the
dense sample tested at s′c ¼50 kPa exhibited dilative behavior -5
with almost no initial contraction. The contractive behavior can Fig. 9. Behavior of loose calcareous sand subjected to cyclic loading: (a) pore
be quantified using several approaches: maximum excess pore pressure ratio versus number of cycles; (b) axial strain versus number of cycles.
 M. Salem et al. / Soil Dynamics and Earthquake Engineering 55 (2013) 83–91 87

affecting the cyclic behavior of this sand. Three main variables
were investigated: sample relative density, effective confining
pressure, and applied cyclic shear stress. The cyclic shear stress
normalized by the initial effective confining pressure is expressed
as a Cyclic Stress Ratio (CSR). Each specimen was prepared at a
specified initial relative density in order to reach the target relative
density (Dr) of either 40% (loose) or 80% (dense) after consolida-
tion. Effective confining pressures of 50, 100, and 200 kPa were
applied to the specimens, and then the samples were axially
loaded at cyclic stress ratios of 0.15, 0.225, and 0.30 for specimens
with Dr ¼40%; and 0.20, 0.30, and 0.50 for specimens with
Dr ¼80%. The frequency of the cyclic loading used in this study
was 0.30 Hz and the loading function was sinusoidal wave [21].
It is noted that an average seating stress of 7 kPa was applied in all
tests. The seating stress was part of the first compression cycle
applied to the specimen. The compression and extension loading
cycles were equal and symmetric around zero value except for the
first quarter cycle.
During cyclic loading, excess pore water pressure is generated.
The ratio of the excess pore water pressure (Δu) to the initial
60
Dr = 40%
50
CSR = 0.225
40 'c = 100 kPa
30
q = ( 1- 3)/2 (kPa)
effective confining pressure (s′c) is defined as the pore pressure
ratio, ru (ru ¼Δu/s′c). If the excess pore water pressure increases
until it reaches the value of the initial effective confining pressure,
the specimen fails as the effective stress drops to zero defining the
onset of liquefaction at which ru ¼1.0. Sometimes, large strains
may occur without reaching zero effective stress. Accordingly,
in studying the cyclic behavior of soils, failure criterion is usually
defined as the number of cycles required to cause either liquefac-
tion (ru ¼1.0) or a specified double amplitude axial strain (ε ¼ 5%).
For the calcareous sand tested herein (Dr ¼40% and Dr ¼80%), the
soil reached ru of 1.0 at relatively low strains (less than 5% in
double amplitude). The number of cycles required to reach
liquefaction ranged from two (2) cycles (Dr ¼80%, s′c ¼200 kPa,
CSR ¼0.5) to 547 cycles (Dr ¼ 40%, s′c ¼50 kPa, CSR ¼0.225).
Typical response of tested loose Dabaa calcareous sand speci-
men subject to cyclic loading is presented in Fig. 9. The specimen
was initially prepared at a relative density of 38% and consolidated
under an effective confining pressure of 100 kPa to reach a target
relative density of 40% before applying the axial load. The specimen
was axially loaded with a CSR of 0.225. During the first 140 cycles,
though the pore water pressure gradually increased, the axial accu-
mulative strain remained relatively small (nearly 0.5% at ru of 0.5).
After cycle number 140, the pore water pressure increased signifi-
PT Line
cantly until it reached the value of the effective initial confining
pressure (i.e., ru ¼ 1.0), thus reaching liquefaction at cycle number 152.
This failure is typical for contractive calcareous and siliceous loose
sands [15,22].

20 The effective stress path of this specimen is shown in Fig. 10.

The phase transformation (PT) line, as determined from all con-
10
ducted monotonic compression triaxial tests, is plotted for refer-
0 ence. The PT line obtained from monotonic compression triaxial
0 20 40 60 80 100 120
- 10 loading corresponded well to the PT points observed in the
- 20 compression side of the effective stress path from cyclic loading.
Similar observation was noted by Hyodo et al. [6]. It is worth
- 30
mentioning that phase transformation occurs under both com-
- 40 pression and extension loading. Data reported in the literature [6]
Mean Principal Stress, p' (kPa)
from monotonic compression and extension triaxial tests as well
Fig. 10. Effective stress path of loose Dabaa calcareous sand. as cyclic triaxial tests on sands showed that the behavior in
compression and extension was not symmetric with extension
being weaker than compression. In this study, only monotonic
1.2 compression triaxial tests were conducted; therefore, the PT line
Dr = 80% was limited only to the compression side of the effective stress
1.0 CSR = 0.20
path of the cyclic loading. In Fig. 10, the sample transitioned from
Pore Pressure Ratio, ru

'c = 100 kPa

0.8 contractive to dilative behavior as it crossed the PT line in the
0.6 compression side of the effective stress path. As the axial load
reversed from compression to extension, two phase transforma-
0.4
tions were observed. The sample first dilated as the excess pore
0.2 water pressure decreased and p′ increased. Then, the sample
0.0

-0.2 60
0 100 200 300 400 500 Dr = 80%
50 CSR = 0.20
Number of Cycles PT Line
40 'c = 100 kPa
0 100 200 300 400 500
1
30
q = ( 1- 3)/2 (kPa)

0 20
Axial Strain (%)

-1 10

0
-2
0 20 40 60 80 100 120
-10
-3
-20
-4 -30

-5 -40
Mean Principal Stress, p' (kPa)
Fig. 11. Behavior of dense calcareous sand subjected to cyclic loading: (a) pore
pressure ratio versus number of cycles; (b) axial strain versus number of cycles. Fig. 12. Effective stress path of dense Dabaa calcareous sand.
88 M. Salem et al. / Soil Dynamics and Earthquake Engineering 55 (2013) 83–91

contracted as the excess pore water pressure increased and p′ reversed from dilative to contractive upon extension, the stress
decreased. Hyodo et al. [6] observed similar behavior in Hiroshima path for the loose samples showed decreased q and p′ values,
Masado and Ube Masado sands tested in extension monotonic causing the stress path to proceed to the left towards the origin
triaxial tests. Similar behavior can be observed as well for Playa and occurrence of flow liquefaction with total strength loss.
Santa calcareous sand tested under cyclic triaxial loading by Typical response of dense Dabaa calcareous sand specimen
LaVielle [15]. Hyodo et al. [6] attributed the further change in subject to cyclic loading is shown in Fig. 11. This specimen is
behavior from dilative to contractive upon extension to the prepared initially at a relative density of 78% and consolidated
concurrently observed necking of the samples, which was under an effective confining pressure of 100 kPa to reach a target
observed in this study as well. Generally, as the sample behavior relative density of 80% before applying the axial load. The specimen
was axially loaded with a CSR of 0.20. The behavior of the dense
sand specimen showed a gradual increase in the pore water pressure
0.40 during the first 430 cycles, while the axial accumulative strain
0.35
remained relatively small (nearly 0.5% at ru of 0.5). After cycle
number 430, the pore water pressure increased significantly until
Cyclic Stress Ratio, CSR

0.30 it reached the value of the initial effective confining pressure,

reaching liquefaction (ru ¼1.0) at cycle number 453. Dabaa dense
0.25
sand showed initial contractive behavior under cyclic loading (which
0.20 was also shown under monotonic loading) leading to liquefaction
rather than strain-defined failure. LaVielle [15] reported a similar
0.15
behavior for Playa Santa calcareous dense sands (Dr ¼ 60%) tested
Dr = 40%
0.10 under cyclic triaxial loading.
σc' = 50 kPa
The effective stress path of this specimen in response to cyclic
0.05 σc' = 100 kPa
loading is shown in Fig. 12. The phase transformation (PT) line is
σc' = 200 kPa
0.00 plotted for reference. The cyclic stress path for the tested dense
1 10 100 1000
specimen showed a similar behavior to the loose specimen shown
Number of Cycles to reach ru = 1.0 in Fig. 10. Similar to loose samples, as the sample behavior
Fig. 13. Cycles to liquefaction (ru ¼1.0) versus cyclic stress ratio at various effective reversed from dilative to contractive upon extension, q and p′
confining pressures for loose sand (Dr ¼ 40%). values decreased, which led the stress path to the left towards the
origin and liquefaction occurred.

0.6
Dr = 80% 4.1. Effect of cyclic stress ratio

0.5 σc': 50 kPa

The effect of cyclic stress ratio on liquefaction susceptibility is
Cyclic Stress Ratio, CSR

σc': 100 kPa

presented in Figs. 13 and 14 for loose and dense sands, respec-
0.4 σc': 200 kPa
tively. The number of cycles required to cause liquefaction (ru ¼1.0)
is plotted versus CSR, at various effective confining pressure
0.3
values. The results of both loose and dense specimens indicated
that for specimens tested at same relative density and effective
0.2
confining pressure, the number of stress cycles required to reach
liquefaction (ru ¼ 1.0) decreased as CSR increased. Similar trends
0.1 have been reported in previous researches for clean siliceous sands
[22–27]; and calcareous sands [6,7,15,28,29].
0
1 10 100 1000
Number of Cycles to reach ru = 1.0 4.2. Effect of effective confining pressure
Fig. 14. Cycles to liquefaction (ru ¼1.0) versus cyclic stress ratio at various effective
confining pressures for dense sand (Dr ¼ 80%). The effect of effective confining pressure on liquefaction
susceptibility is illustrated in Figs. 15 and 16 for loose and dense

250 250
Dr = 80%
Effective Confining Pressure (kPa)
Effective Confining Pressure (kPa)

CSR: 0.20
200 200 CSR: 0.30
CSR: 0.50

150 150

100 100

Dr = 40%
50 CSR: 0.15 50
CSR: 0.225
CSR: 0.30
0 0
1 10 100 1000 1 10 100 1000
Number of Cycles to reach ru =1.0 Number of Cycles to reach ru=1.0

Fig. 15. Cycles to reach ru ¼1.0 versus effective confining pressure at various cyclic Fig. 16. Cycles to reach ru ¼ 1.0 versus effective confining pressure at various cyclic
stress ratios for loose sand (Dr ¼ 40%). stress ratios for dense sand (Dr ¼ 80%).
 M. Salem et al. / Soil Dynamics and Earthquake Engineering 55 (2013) 83–91 89

0.6 0.45
'c = 100 kPa Dr = 40%

0.5 Dr = 40% Dr = 80%

Cyclic Resistance Ratio, CRR

0.40
Dr = 80%
Cyclic Stress Ratio, CSR

0.4 y = 1.315x-0.292
0.35 R² = 0.9994
0.3

0.30 y = 1.1179x-0.281
0.2 R² = 0.9815

0.1 0.25

0 0.20
1 10 100 1000 0 50 100 150 200 250
Number of Cycles to reach ru = 1.0 Effective Confining Pressure (kPa)

Fig. 17. Cycles to reach ru ¼ 1.0 versus cyclic stress ratio at various relative densities Fig. 18. Variation of cyclic resistance ratio of loose and dense Dabaa calcareous
under confining pressure of 100 kPa. sand with effective confining pressure at various relative densities.

0.45
sands, respectively. The number of cycles required to cause Dr = 40%

liquefaction (ru ¼1.0) is plotted versus effective confining pressure Dr = 80%

Cyclic Resistance Ratio, CRR

0.40
(s′c), at various values of CSR. For a sample prepared at a specific
relative density and subjected to a specific CSR, the number of
0.35
cycles required to cause liquefaction increased as the effective y = 0.6418x-0.287
confining pressure decreased. This may be attributed to the trend R² = 0.9913

observed in the monotonic triaxial testes, in which the samples 0.30

became more contractive as the confining pressure increased. This
behavior was observed for loose and dense sands.
Various opinions have been reported in the literature regarding
the effect of effective confining pressure on the number of cycles
to cause liquefaction for calcareous and siliceous sands. Stedman
[30] observed that in cyclic loading of siliceous sands, increasing
the effective confining pressure generally decreased the resistance
to liquefaction; however, in the loosest states, the increase in
confining pressure had little effect on resistance to liquefaction.
Furthermore, Hyodo et al. [6] found that in case of dense Shirasu
calcareous sands, the effect of the effective confining pressure on
number of cycles to failure is unclear. However, for Dogs Bay
carbonate sand [6] and Cabo Rojo calcareous sand [29], the cyclic
strength increased as the confining pressure decreased, which is
similar to the behavior of Dabaa calcareous sand. The effect of
effective confining pressure on the cyclic resistance of dense
siliceous sand under triaxial loading was also investigated by
Castro and Poulos [31], where they showed that the cyclic stress
ratio required to cause failure decreased with increasing effective
confining pressure, which is in agreement with the behavior of
Dabaa calcareous sand. Yunoki et al. [32] and Ishihara et al. [33]
have indicated that there was no effect of confining pressure on
the behavior of silica sands for relative densities ranging from 50
to 70%. Moreover, Finn et al. [34] mentioned that at a given void
ratio, the cyclic stress ratio and number of stress cycles to reach
liquefaction in siliceous sands are uniquely related, independent of
the consolidation stress. However, the behavior of tested Dabaa
calcareous sand contradicts with what has been reported by Seed
and Lee [23] and Peacock and Seed [35] for clean siliceous sand, as
they found that as the confining pressure increased the number of
cycles to failure increased. It is worth mentioning that Seed
and Lee [23] found their observations contradictory to their
expectations in reference to the observed trends in monotonic
tests as well as the confining pressure–critical void ratio relation-
ship defined for their tested soil.
4.3. Effect of relative density
The effect of relative density on liquefaction susceptibility is
studied for samples tested at the same cyclic stress ratio and
0.25
0.20
0 5 10 15 20 25 30
Effective Confining Pressure, 'c
Dr0.55
Fig. 19. Variation of cyclic resistance ratio of Dabaa calcareous sand with effective
confining pressure normalized to relative density.
effective confining pressure. The higher the relative density, the
higher the number of cycles required to cause liquefaction as
illustrated in the example shown in Fig. 17. The effect of relative
density on cyclic resistance of Dabaa calcareous sand is in agree-
ment with previous researches conducted on calcareous sands
[6,15,29]; and on siliceous sands [22,23,25,34].
5. Evaluation of crushability during triaxial testing
Calcareous sands consist mainly of carbonate sediments that
are characterized with their possible crushability potential, and
thus needs to be evaluated. Hardin's [8] method was used to
quantify the crushing in the tested sand upon testing in both
monotonic and cyclic triaxial tests. Hardin [8] based his measure-
ments on the changes in the grain-size distribution curve before
and after the test. He defined the breakage potential (Bp) as the
area between the original grain size distribution curve of the soil
and the No. 200 sieve size, representing the total possible change
in the soil gradation. He also defined the total breakage (Bt) as the
area between the original and final grain size distribution curves.
Hardin [8] then defined the relative breakage factor (Br) as the (Bt/Bp).
The calculated values of Hardin's relative breakage factor (Br) ranged
from 0.01 (1.09%) to 0.02 (2.23%). Similar low breakage factors are
reported by Hardin [8] for basalt tested at similar effective confining
pressures. Higher relative breakage factor of 0.27 (27%) was observed
by Hardin [8] for basalt tested at higher effective confining pressure
of 3.4 MPa. It is worth noting that there is no significant difference in
the amount of observed crushing between samples subjected to
90 M. Salem et al. / Soil Dynamics and Earthquake Engineering 55 (2013) 83–91

monotonic and cyclic loading in this study. It was concluded that 7. Comparison of cyclic behavior with other sands in literature
crushing during testing the Dabaa calcareous sand was insignificant
for the range of effective confining pressures and axial loads applied 7.1. Comparison with siliceous sands
herein. Similar observation was noted for Cabo Rojo calcareous
sand [29]. The obtained test results for the Dabaa sand were compared to
other results reported in the literature for the more commonly
studied siliceous sands. For most cases, the general behavior of the
Dabaa calcareous sand followed the behavior adopted by siliceous
sands. Generally, under the same effective confining pressure and
6. Correlation between CRR and other factors initial relative density, the number of cycles required to reach
liquefaction (ru ¼1.0) increased as the value of CSR decreased as
In this experimental study, the Cyclic Resistance Ratio (CRR) is shown in Fig. 20.
defined as the CSR required to cause liquefaction for loose or dense As previously mentioned, calcareous sands have large void
specimens in 20 loading cycles to represent an earthquake of ratios and are well known for being more compressible than
magnitude 7.5 [17]. For both loose and dense calcareous sands, the siliceous sands. Therefore, it was initially thought that this
value of CRR at each effective confining pressure was determined. compressible soil skeleton of calcareous soils would result in a
The variation of CRR with effective confining pressure is shown in more contractive soil matrix, and thus result in calcareous sands
Fig. 18. For both loose and dense samples, similar trends were being more susceptible to liquefaction than siliceous sands. This
observed for the variation of CRR with effective confining pressure. research and others have found that this is not true for all relative
Results shown in Fig. 18 were used to perform further linear densities. The results plotted in Fig. 20 indicate that loose Dabaa
regression analyses to develop a correlation between CRR and the calcareous sand has greater cyclic strength and is less susceptible
change in relative density and effective confining pressure. The to liquefaction when compared to siliceous sands at same relative
result of the regression analysis is illustrated in Fig. 19 and Eq. (1). density. Increased cyclic strength is likely to be due to the
!0:287 existence of different shapes of calcareous sand particles within
D0:55 the tested soil, which provides more stable interlocking soil fabric
CRR ¼ 0:6418 r
ð1Þ
sc0 resistant to liquefaction [5–7,15,28,36]. The effect of particles
interlocking decreased as the relative density increased as the
where s′c is the effective confining pressure in kPa and Dr is the cyclic resistance of dense Dabaa calcareous sand was less than
relative density in percent. those reported for siliceous sand as shown in Fig. 21.

0.35 7.2. Comparison with other calcareous sands in literature

0.30
Calcareous sands generally followed similar trends regarding
Cyclic Stress Ratio, CSR

0.25 the behavior under cyclic undrained triaxial loading [15,29]. The
variation in cyclic behavior of Cabo Rojo and Playa Santa calcareous
0.20 sands tested at similar relative densities and effective confining
pressures is compared to Dabaa sand in Fig. 22. Index properties of
0.15 the three soils are summarized in Table 1. The dense Cabo Rajo sand
'c = 100 kPa showed higher cyclic resistance when compared to the dense Dabaa
0.10
sand; opposite behavior was observed when comparing the two
Dabaa Calcareous Sand, Dr = 40%
0.05 sands at loose state. The reversed observed behavior at the two
Siliceous Sand (A), Dr = 40% (Hussein, 2008 [22])
Siliceous Sand (B), Dr = 40% (Evans & Zhou, 1995 [26])
densities could not be related to differences in grain-size distribu-
0.00 tion, void ratios, specific gravity, etc. On the other side, Playa Santa
1 10 100 1000
and Dabaa sands (Dr ¼40%) have similar grain-size distributions,
Number of Cycles to reach ru = 1.0
void ratios, and specific gravities; yet, the cyclic resistance of the
Fig. 20. Comparison between loose Dabaa calcareous sand and other siliceous Dabaa sand was higher than that of Playa Santa sand. The variations
sands under effective confining pressure of 100 kPa. in the existent marine remains, constituent mineralogy, shape and

0.8
0.7 Dabaa Calcareous Sand, Dr = 40%
Dabaa Calcareous Sand, Dr = 80%
0.7
Cabo Rojo Calcareous Sand, Dr = 45% (Sandoval & Pando, 2012 [29])
0.6
Cabo Rojo Calcareous Sand, Dr = 80% (Sandoval & Pando, 2012 [29])
0.6 Playa Santa Calcareous Sand, Dr = 40% (LaVielle, 2008 [15])
Cyclic Stress Ratio, CSR

Cyclic Stress Ratio, CSR

0.5
'c = 100 kPa
0.5
0.4
0.4
0.3
0.3

0.2
'c = 100 kPa 0.2

0.1 Dabaa Calcareous Sand, Dr = 80% 0.1

Siliceous Sand (A), Dr = 80% (Hussein, 2008 [22])
0.0 0.0
1 10 100 1000 1 10 100 1000
Number of Cycles to reach ru =1.0 Number of Cycles to reach ru = 1.0

Fig. 21. Comparison between dense Dabaa calcareous sand and other siliceous Fig. 22. Comparison between Dabaa calcareous sand and other calcareous sands
sand under effective confining pressure of 100 kPa. under effective confining pressure of 100 kPa.
 M. Salem et al. / Soil Dynamics and Earthquake Engineering 55 (2013) 83–91
texture of soil grains, particles crushability, and intra-particle voids
may explain the variation in liquefaction susceptibility of different
calcareous sands.
8. Conclusions
1. Both loose and dense states of the Dabaa calcareous sand
exhibit similar initial contractive behavior upon shearing. This
contractive behavior is more apparent as the effective confining
pressure increases and for looser state.
2. Both loose and dense states of the Dabaa calcareous sand show
no clear failure under monotonic loading, but rather a ductile
behavior upon shearing is apparent.
3. Failure under cyclic loading in both loose and dense states of
the Dabaa calcareous sand is governed by the gradual devel-
opment of excess pore water pressure until liquefaction is
reached, rather than cumulative development of axial strain
without liquefaction (cyclic mobility).
4. For the same initial relative density and effective confining
pressure, the number of cycles to reach liquefaction decreases
as the cyclic stress ratio increases.
5. For the same initial relative density and cyclic stress ratio, the
number of cycles to reach liquefaction decreases as the effec-
tive confining pressure increases.
6. For the same effective confining pressure and cyclic stress ratio,
the number of cycles to reach liquefaction increases as the soil
relative density increases.
7. As the confining pressure increases, the effect of the initial
relative density on the cyclic behavior diminishes.
8. Dabaa calcareous sand showed less susceptibility to liquefac-
tion compared to siliceous sands subjected to similar loading
conditions at relative density of 40%. This may be attributed to
the existence of different shapes of calcareous sand particles
within the tested soil providing more stable interlocking soil
fabric resistant to liquefaction. The effect of particles interlocking
decreased as the relative density increased.
9. A soil-specific correlation relating the cyclic resistance ratio,
soil relative density, and effective confining pressure is devel-
oped for the Dabaa calcareous sand.
References
[1] Fookes PG. The geology of carbonate soils and rocks and their characterization
and description. In: Proceedings of the international conference on calcareous
sediments, Perth. Vol. 2. Rotterdam: Balkema; 1988, 787–806.
[2] Coop MR, Airey DW. Carbonate sands. In: Tan TS, Phoon KK, Hight DW,
Leroueil S, editors. Characterization and engineering properties of natural
soils. Balkema; 2002. p. 1049–86.
[3] Datta M, Gulhati SK, Rao GV. Engineering behavior of carbonate soils of India
and some observations on classification of such soils. Geotechnical Properties,
Behavior and Performance of Calcareous Soils, ASTM STP 1982;777:113–40.
[4] Golightly CR, Hyde AFL. Some fundamental properties of carbonates sands.
Engineering for Calcareous Sediments, Vol. 1. Perth: Balkema; 69–78.
[5] Hyodo M, Aramaki N, Itoh M, Hyde AFL. Cyclic strength and deformation of
crushable carbonate sand. Soil Dynamics and Earthquake Engineering 1996;15
(5):331–6.
[6] Hyodo M, Hyde AFL, Aramaki N. Liquefaction of crushable soils. Geotechnique
1998;48(4):527–43.
[7] Sharma SS, Ismail MA. Monotonic and cyclic behavior of two calcareous soils
of different origins. Journal of Geotechnical and Geoenvironmental Engineering,
ASCE 2006;132(12):1581–91.
91
[8] Hardin BO. Crushing of soil particles. Journal of Geotechnical Engineering,
ASCE 1985;111(10):1177–92.
[9] Lade PV, Yamamuro JA, Bopp PA. Significance of particle crushing in granular
materials. Journal of Geotechnical Engineering, ASCE 1996;122(4):309–16.
[10] Hallsworth CR, Knox RWO’B. Classification of sediments and sedimentary
rocks. BGS Rock classification scheme. Vol. 3. British Geological Survey
Research Report, RR 99-03; 1999.
[11] Kullhawy FH, Mayne PH. Manual on estimating soil properties for foundation
design. Electric Power Research Institute, EPRI; August, 1990.
[12] Cataño Arango J. Stress strain behavior and dynamic properties of Cabo Rojo
calcareous sands. M.Sc. thesis, University of Puerto Rico, Mayaguez; 2006.
[13] Hurlbut CS. Dana's manual of mineralogy. 18th edition. NY: John Wiley & Sons,
Inc.; 1971.
[14] Morioka BT. Evaluation of the static and cyclic strength properties of calcar-
eous sand using cone penetrometer tests. Ph.D. dissertation. University of
Hawaii, Manoa; 1999.
[15] LaVielle TH. Liquefaction susceptibility of uncemented calcareous sands from
Puerto Rico by cyclic triaxial testing. M.Sc. thesis, Virginia Polytechnic Institute
& State University, Blacksburg; 2008.
[16] Ishihara K, Tatsuoka F, Yasuda S. Undrained deformation and liquefaction of
sand under cyclic stress. Soils Found 1975;15(1):29–44.
[17] Ishihara K. Liquefaction and flow failure during earthquakes. Geotechnique
1993;43(3):351–415.
[18] Mitchell JK, Soga K. Fundamentals of soil behavior. 3rd edition. New Jersey:
Wiley; 2005.
[19] Ladd RS. Preparing test specimens using undercompaction. Geotechnical
Testing Journal 1978;1(1):16–23.
[20] Skempton AW. The pore-pressure coefficients A and B. Geotechnique 1954;4
(4):143–7.
[21] Agaiby SS. Behavior of calcareous sands of northern coast in Egypt under cyclic
triaxial loading. M.Sc. thesis, Cairo University, Egypt; 2011.
[22] Hussein MS. Undrained cyclic behavior of gravelly sand soil. M.Sc. thesis, Cairo
University, Egypt; 2008.
[23] Seed HB, Lee KL. Liquefaction of saturated sands during cyclic loading. Journal
of the Soil Mechanics and Foundations Division 1966;92(SM6):105–34.
[24] Silver L, Chan K, Ladd S, Lee L, Tiedemann A, Townsend C, Valera E, Wilson H.
Cyclic triaxial strength of standard test sand. Journal of Geotechnical Engi-
neering, ASCE 1976;102(GT5):511–23.
[25] Lin PS, Chang CW, Chang WJ. Characterization of liquefaction resistance in
gravelly soil: Large hammer penetration test and shear wave velocity
approach. Soil Dynamics and Earthquake Engineering 2004;24:675–87.
[26] Evans MD, Zhou S. Liquefaction of sand-gravel composites. Journal of Geo-
technical Engineering, ASCE 1995;121(3):287–98.
[27] Toki S, Tatsuoka F, Miura S, Yoshimi Y, Yasuda S, Makihara Y. Cyclic undrained
triaxial strength of sand by a cooperative test program. Soils Found 1986;26
(3):117–28.
[28] Morioka BT, Nicholson PG. Evaluation of the liquefaction potential of calcar-
eous sand. In: Proceedings of the International Offshore and Polar Engineering
Conference, Seattle, WA, USA; 2000, 494–500.
[29] Sandoval EA, Pando MA. Experimental assessment of the liquefaction resis-
tance of calcareous biogenous sands. Earth Science Research Journal 2012;
16(1).
[30] Stedman JD. Effects of confining pressure and static shear on liquefaction
resistance of Fraser River sand. B.A.Sc. thesis, The University of British
Columbia, Canada; 1997.
[31] Castro G, Poulos J. Factors affecting liquefaction and cyclic mobility. Journal of
Geotechnical Engineering, ASCE 1977;103(6):501–16.
[32] Yunoki Y, Ishihara K, Seki M, Yoneda Y. The effect of the initial effective
confining pressure on cyclic triaxial shear behaviour of dense sand. In:
Proceedings of the 17th Japanese National Soil Mechanics Research Meeting,
JSSMFE; 1982, 1649–1652.
[33] Ishihara K, Kikuchi Y, Tsutsumi K. Cyclic deformation behaviour of sands under
low confining pressure. In: Proceedings of the 18th Japanese National Soil
Mechanics Research Meeting, JSSMFE; 1983, 353–354.
[34] Finn WD, Pickering DJ, Bransby PL. Sand liquefaction in triaxial and simple
shear tests. Journal of Geotechnical Engineering, ASCE 1971;97(4):639–59.
[35] Peacock WH, Seed HB. Sand liquefaction during cyclic loading simple shear
conditions. Journal of Geotechnical Engineering, ASCE, 94 (SM3); 689–708.
[36] Kaggwa WS, Poulos HG. Comparison of the behaviour of dense carbonate
sediments and silica sand in cyclic triaxial tests. Research Report No. R611,
University of Sydney, School of Civil and Mining Engineering, 1990.