252
Proceedings of the 18
th
International Conference on Soil Mechanics and Geotechnical Engineering, Paris 2013
Proceedings of the 18
th
International Conference on Soil Mechanics and Geotechnical Engineering, Paris 2013
2 SOIL PROPERTIES
Three soils from the region of Thrace (North-Eastern Greece)
were used in this investigation because of their swelling
characteristics. According to the properties presented in Table 1,
all three soils can be considered as clay soils since the clay
fraction (grain sizes <0.002 mm) ranges from 70% to 80%. The
P1-S2 and P2-S2 soils are classified as CH, while the P2-S1 soil
is classified as MH in accordance with the Unified Soil
Classification System. The values of maximum dry unit weight,
γ
dmax
, and optimum moisture content, w
opt
, were obtained by
conducting compaction tests with standard compaction effort.
All three soils present “high” or “very high” swelling
potential, according to the known correlations of soil index
properties with swelling characteristics (Papakyriakopoulos and
Koudoumakis 2001). One-dimensional swell tests (ASTM
D4546, Method A) were conducted using laboratory-compacted
specimens (ASTM D698) of these soils (Koudoumakis 2000).
The results obtained from specimens with initial moisture
contents similar to the optimum moisture contents of the soils
(Table 1), are presented in Table 2. It can be observed that the
swell pressure ranges from 170 kPa to 820 kPa and the free
swell ranges from 11% to 19%. These values are indicative of
the swelling potential of the soils used in this investigation.
3 EXPERIMENTAL PROCEDURES
The residual shear strength behavior of the soils was evaluated
by performing ring shear tests on remolded specimens. The use
of remolded specimens allowed the adequate control of the
specimen moisture content. The specimens were prepared with
the moisture contents shown in Table 3. At first, the optimum
moisture content, resulted from the Standard compaction test
(Table 1), was used for each soil. The following specimens were
prepared with larger moisture content, obtained for each soil as
the average value of the specimen moisture contents at the end
of the preceding tests started with optimum water content. The
specimens were placed in the cell of the ring shear apparatus by
kneading the soil with the desired moisture content evenly to fill
Table 1. Properties of soils.
Soil designation
P1-S2
P2-S1
P2-S2
Sampling depth (m)
1.5-3.0
0.0-1.6
1.6-3.0
Specific gravity
G
s
2.82
2.68
2.73
Sand (%)
11.5
5.6
3.8
Silt (%)
9.3
23.9
25.8
Grain size
analysis
Clay (%)
79.2
70.5
70.4
Liquid limit
w
L
86
82
87
Atterberg
limits
Plasticity index
I
p
63
42
55
Maximum dry unit
weight
γ
dmax
(kN/m
3
) 13.80
13.95
14.27
Compaction
characteristics Optimum moisture
content
w
opt
(%)
31.2
27.0
25.8
Table 2. Typical results of one-dimensional swell tests.
Soil designation
P1-S2
P2-S1
P2-S2
Initial moisture content
w
0
(%)
30.63 31.13 24.02 27.48
Initial void ratio
e
0
0.990 0.946 1.054 1.045
Initial degree of saturation
S
r0
(%)
87.25 88.19 61.08 71.79
Dry unit weight
γ
d
(kN/m
3
)
13.87 13.52 12.81 13.09
Swell pressure
P
s
(kPa)
820
335
170
410
Specimen height increase
Δh
(mm)
3.85
2.20 2.52
2.83
Free swell
Δh/h
0
(%)
19.23 11.60 13.28 14.17
Table 3. Testing program.
Soil
Moisture content
w
(%) Effective normal stress
σ’
n
(kPa)
31.2
25
75
250
800
P1-S2
45.8
25
50
100
200
27.0
25
75
200
600
P2-S1
47.1
25
50
100
200
25.8
25
75
200
600
P2-S2
46.4
25
50
100
200
the annular cavity between the confining rings of the cell, using
a small spatula (BS 1377 – Part 7).
Ring shear testing was based on the procedure described in
BS 1377 – Part 7. The tests were conducted using a Bromhead
ring shear apparatus (Bromhead 1979) and annular specimens of
5 mm thickness with internal and external diameters of 70 mm
and 100 mm, respectively. The specimens were consolidated for
a period of 24 hours under the effective normal stresses, σ’
n
,
presented in Table 3 and, subsequently, were sheared at a con-
stant rate of angular displacement equal to 0.048 degrees/min.
The selection of this rate of angular displacement was dictated
by the unconventional results of the consolidation stage of the
tests, described in the next section, and was based on the fact
that this rate has been found satisfactory for a large range of
soils (BS 1377 – Part 7). One of the objectives of the present
study was to investigate the residual strength behavior of the
soils for a wide range of effective normal stresses reaching or
even exceeding the values of swell pressure shown in Table 2.
Although this goal was accomplished for the specimens
prepared with the optimum moisture contents (Table 3), the use
of effective normal stresses larger than 200 kPa was not feasible
in the tests performed with soil moisture contents ranging from
46% to 47% because of excessive specimen loss during testing.
4 RESULTS AND DISCUSSION
Typical “specimen length change” – “log time” curves obtained
from the consolidation stage of ring shear tests conducted with
optimum moisture content and with moisture content ranging
from 46% to 47%, are shown in Figures 1a and 1b, respectively.
It can be observed (Figure 1a) that, in several cases, the curves
do not present the usual form attained in ordinary clays, either
due to the expansion of specimens or because the consolidation
was not completed within the predetermined period of 24 hours.
It can also be observed (Figure 1b) that, in general, the classic
type of curves appears in the test performed under the highest
effective normal stress of each test series. The overall behavior
of the three soils in the consolidation stage of the ring shear
tests is summarized in Table 4. More specifically, the tests are
divided in those exhibited specimen expansion and those
demonstrated specimen compression during consolidation. It is
evident that specimen expansion occurred almost in all series of
tests and that the interchange of expansion and compression
takes place at higher values of effective normal stress in the
tests conducted with optimum water contents. This can be
attributed to the denser condition of the specimens in these tests,
since their values of initial dry unit weight, γ
d0
, range from 1.24
gr/cm
3
to 1.59 gr/cm
3
and are larger than those (γ
d0
= 1.10
gr/cm
3
– 1.27 gr/cm
3
) in the tests conducted with moisture
contents ranging from 46% to 47%.
The residual shear stress, τ
r
, is the minimum constant value
of shear stress, determined at the end of the shearing stage of
each test and used in the drawing of the residual failure
envelopes of soils. As typically shown in Figure 2a, the residual
failure envelopes resulted from the ring shear tests conducted on
specimens with optimum water contents are curved, in
agreement with the observation of Stark and Eid (1994) that the
non-linearity of failure envelopes is significant for soils with a
