Marine Clay Data
Marine Clay Data
Marine Clay Data
3.1
Introduction
The understanding of Singapore Marine Clay is very limited. The majority of the
available soil data in Singapore are from site investigation carried out during major
infrastructure projects such as Mass Rapid Transit and only the index properties,
compressibility and undrained shear strength were determined (Tan, 1983, Dames &
Moore 1983). The first effort on the detailed characterisation of Singapore Marine
Clay was undertaken by National University of Singapore and Nanyang Technological
University in the late nineties. This chapter will first give a brief description of
geological history of Singapore marine clay, followed by a review on the current
understanding on the properties of natural Singapore marine clay.
3.2
The most frequently cited sources on the geology of Singapore are Pitts (1983) and
Pitts (1992). The ensuing brief description is a summary of the existing state of
43
44
3.3
Previous
Characterisation
Studies
on
Natural
45
Authority, in preparation for the construction of the first Mass Rapid Transit (MRT)
line in Singapore.
compressibility and undrained shear strength were determined because they are of
immediate relevance to the application at hand.
The second stage was carried out at the late of 1990s by National University of
Singapore (NUS) and Nanyang Technological University (NTU). NUS carried out the
characterisation study during the construction of Singapore Art Centre (presently
called Esplanade and denoted as SAC from hereon) and the reclamation project at
Pulau Tekong (denoted as PT from hereon) and most of the data were reported in Tan
et al. (2002a and 2002b). Tan et al. (2002a and 2002b) paid attention specifically to
the effect of sample disturbance on the mechanical properties of LMC. They also
studied the small strain behaviour of LMC. The finding of these studies has provided a
reasonable database to arrive at soil models that commensurate with the increasing
sophistication involved in numerical analysis, which has become almost a standard
requirement for the geotechnical design in densely built-up Singapore. This study also
highlighted the need to obtain high quality samples to ensure reliable interpretation of
the behaviour of Singapore marine clay. On the other hand, NTU carried out some
characterisation study of UMC and LMC during the Changi Airport reclamation. A
number of conference and journal papers (e.g. Chang et al. 1997; Chu et al. 2002) were
published and mainly focused on application of in-situ tests (e.g. piezocone,
dilatometer) in the estimation of overconsolidation ratio, undrained shear strength and
consolidation properties (e.g. coefficient of consolidation). They showed that in-situ
tests provide a reasonable and cost-effective estimation of overconsolidation ratio,
undrained shear strength and consolidation properties. Some minor characterisation
46
studies of UMC and LMC were also carried out in between these two stages of major
studies (at the late of 1980s and early 1990s) such as Chang (1991) on the prediction of
overconsolidation ratio from piezocone and Todo et al. (1993) on the compressibility
of UMC and LMC.
This study only focuses on the effect of soil structure on the compressibility,
undrained behaviour of natural Singapore marine clay. Therefore, only findings of the
relevant studies will be summarized and reviewed in this chapter.
47
organic content that ranging from 0.2-3%. The organic content in LMC is fairly
uniform with about 3.5% throughout the depth.
48
noticed that YSR decrease with depth from about 4 just below the seabed to between
1.5 and 2 at the bottom of the UMC. For LMC, Figure 3.4 shows that the YSR is
about constant throughout the depth and ranging from 1.3 to 1.6. The YSR values in
UMC are consistently higher than the values reported in previous paragraph while the
YSR values in LMC are within the range reported in previous paragraph. Based on
this yield stress and YSR profile, Hanzawa and Adachi (1983) postulated that high
YSR values in UMC might be due to chemical bonding while the overconsolidation
state in LMC is probably due to secondary compression. Figure 3.4 also indicates that,
intermediate layer is moderately overconsolidated (YSR = 2.5 to 4). They attributed
this overconsolidation behaviour to dessication during the exposure of LMC during the
Last Glacial Maximum (see Section 3.2).
Tan et al. (2002b) also found that a similar YSR profiles of natural Singapore
marine clay below the seabed near Pulau Tekong. Figure 3.5a shows the results
carried out by two independent commercial laboratories while Figure 3.5b shows the
results from CRS tests carried out by Port and Airport Research Institute, Japan
(PARI). Both figures include YSR values for UMC (0 to10 m), which can exceed 8 at
the top and reduce to about 4 at bottom of UMC which is higher than the data reported
by Hanzawa and Adachi (1983). Clay deposit at the depth between 10 to 14 m belongs
to intermediate layer where high sand content was found and their YSR values were
found ranging from 2 to 4. However, the reliability of these YSR values are doubtful
because the high sand content in the samples from this layer complicate the
determination of yield stress from e-log v curve. For LMC, Figure 3.5b shows that
the YSR increases very slightly with depth from 2.7 to 3.1 while Figure 3.5a indicates
a nearly constant YSR of about 2.5 in LMC. Surprisingly, the YSR in LMC remains
49
relatively higher than the values reported by Hanzawa and Adachi (1983) even up to a
depth of 40 m below the seabed. Preliminary investigation of microfabric using
Scanning Electron Microscopy (SEM) by Tan et al. (2002b) seems to suggest that
there is no clear indication of cementation. It is also simulated by Tan et al. (2002b) in
Figure 3.5b that an addition of 80 kPa to the current effective vertical overburden
would have brought the YSR to a level that is more consistent with values observed
elsewhere such as at the SAC site in Figure 3.3. Hence, Tan et al. (2002b) postulated
that the relatively high YSR at PT site is mechanically induced by erosion in this area.
The causes of overconsolidation in Singapore marine clay proposed in previous
studies (e.g. Hanzawa & Adachi, 1983; Todo et al. 1993) largely related to the
development of soil structure by processes such as cementation.
Except the
microfabric study carried out by Tan et al. (2002b), all the proposals are deduced in an
indirect way.
structuration through the comparison of yield stress and YSR profiles observed in
Singapore marine clay with that observed in the well-studied clay such as Norwegian
clay (Bjerrum, 1967). However, Perret et al. (1995) demonstrated that using YSR
alone to study the process leading to the development of overconsolidation in a clay
deposit is not enough. In addition, Hanzawa & Adachi (1983) assumed that Singapore
marine clay was not subjected to mechanical unloading. In contrast, Tan et al. (2002b)
postulated that Pulau Tekong clay might have been subjected to some degree of
mechanical unloading to explain the high YSR values after they could not identify
clear evidence of cementation through microscopic study using SEM. Since the study
by Tan et al. (2002b) is preliminary, more analysis will be carried out in this study to
identify the causes of overconsolidation in Pulau Tekong clay.
50
3.3.2.1 Estimation of Yield Stress or Yield Stress Ratio Profile from In-situ Tests
Profile of yield stress and parameters derived from yield stress such as YSR, YSD and
YSG provide a useful first indication of the causes of overconsolidation in clay (Perret
et al., 1995 and see also Section 2.8). The interpretation of yield stress profiles from
in-situ tests provide an economical alternative to oedometer test for the study of the
origin of structuration of a clay. Chang (1991) had demonstrated the determination of
YSR in Singapore marine clay by field vane test, piezocone test and dilatometer test.
According to Chang (1991), YSR could be estimated from the undrained shear
strength measured by field vane shear test, (su)FV, that normalised by the corresponding
in-situ effective overburden stress, vo based on the following two correlations:
su
'vo
(s u / 'vo ) FV
YSR =
(s u / 'vo ) NC
(3.1)
FV
1.05
(3.2)
where PI is plasticity index and (su/vo)NC is the undrained shear ratio for normally
consolidated clay. Eq. 3.1 was proposed by Mayne & Mitchell (1988) while Eq 3.2
was proposed by Chandler (1987) on the basis of a large collection of field vane and
oedometer tests results. Chang (1991) suggested that both equations appeared to
provide reasonable to slightly conservative estimates of YSR for the Singapore marine
clay.
In addition to field vane test, Chang (1991) also showed that YSR for
Singapore marine clay could be estimated from piezocone tests. He related the YSR to
the pore pressure ratio, Bq (=(u2-uo)/(qt-vo)) , where u2 and uo are the penetration pore
51
pressure measured behind the cone and the initial hydrostatic pore pressure,
respectively, qt is the cone resistance (corrected for the unequal end area effect) and
vo is the in-situ total overburden stress. He proposed the following correlation for
non-sensitive to highly sensitive clay (strength sensitivity < 8 with YSR less than 8):
YSR =
2.3B q
(3.3)
(3.7 B q 1)
This correlation was found to produce generally conservative estimates of YSR for the
Singapore marine clay (Chang, 1991).
Dilatometer (DMT) results were also used for estimating the YSR for clays.
The commonly use correlation between the YSR and horizontal stress index (KD) is:
YSR = (0.5K D )
(3.4)
where KD = (po-uo)/vo with po is the contact pressure after corrected for membrane
stiffness, uo is the initial hydrostatic pore pressure and vo is in-situ effective
overburden stress. Marchetti (1980) recommended =1.56 for uncemented natural
clays. However, Chang (1991) proposed value of 0.84 for the Singapore marine
clay.
A recent theoretical study carried out by Cao et al. (1996) and Cao (1997)
based on the modified Cam clay concept, the following equations were proposed for
the estimation of YSR from field vane test:
3 (s u ) FV
YSR = 2 '
vo M cos 'ps
1/
(3.5)
52
qt u2
YSR = 2
'vo (1 + 0.67M )
1/
(3.6)
(3.7)
1/
3 (p o vo )
YSR = 2
d 'vo M (ln I r + 1)
1/
(3.8)
where is the plastic volumetric strain ratio with typical value of 0.75 (Cao et al.,
1996), M is the slope of critical state line, ps is the effective friction angle in the plane
strain compression condition, Ir is the rigidity index, is the strain rate correction
factor for piezocone which can be taken as 1.64 (Cao et al. 1996) and d is the strain
rate correction factor for flat dilatometer which can be taken as 1.57 (Cao, 1997).
Chang et al. (2001) showed that these equations provided a reasonable estimation of
YSR of Singapore marine clay from Changi site. However, these equations involve
parameters such as M and ps that are not readily known without conducting detailed
experimental tests. Hence, these equations are difficult to use without a detailed
knowledge about a soil.
According to the comprehensive analytical study carried out Aubeny (1992)
using an advanced elastoplastic anisotropic soil model MIT-E3 (Whittle 1987, Whittle
and Kavvadas 1994), normalized cone tip resistance (Qt) is one of the piezocone
parameters exhibiting a higher sensitivity to the YSR. In addition, the correlation
53
between YSR and Qt is one of the most common proposals found in the literature (e.g.,
Mayne & Holtz 1988, Chen & Mayne 1996, Demers & Leroueil 2002):
YSR = k
q t vo
= kQ t
'vo
(3.9)
where k is a constant, qt is the corrected cone tip resistance, and vo and vo are the insitu total and effective overburben stresses, respectively. Powell et al. (1988) found
that the k value for some clays from United Kingdom varies between 0.2 and 0.5 for
nonfissured clays and between 0.9 and 2.2 for heavily overconsolidated fissured clays.
Lutenegger & Kabir (1988) also obtained scattered results for marine clays in New
York with a mean k value of 0.3. Based on a large worldwide database, Chen &
Mayne (1996) obtained a k value of 0.32 with a relatively low coefficient of
determination (r2 = 0.67). The degree of variability increases further when fissured
clays (r2 = 0.47) are included. For sensitive clays in Quebec, Demers & Leroueil
(2002) found a k value of 0.29 with a high coefficient of determination (r2 = 0.99).
This degree of accuracy is exceptional, given the spatial scope of the calibration.
Nevertheless, Demers & Leroueil (2002) concluded that their correlations and others
found in literature are not accurate enough for final designs, which concurred with
sentiments expressed by Lunne et al. (2002).
Due to the availability of data in this study, the YSR profile will only be
estimated from piezocone test with Eq. 3.9. The wide range of reported k value in the
literature led to a need to develop a site specific correlation that is suitable for
Singapore marine clay. A site specific calibration will be carried out to examine how
accurate are the YSR-Qt correlation for the estimation of yield stress ratio profiles for
the clay deposit below the seabed near Pulau Tekong in Chapter 6. The yield stress
54
profile will be obtained by multiplying Eq. 3.9 with the corresponding in-situ effective
overburden stress, vo
3.3.3 Compressibility
Tan (1983) reported that Singapore marine clay has a compression index Cc of 0.7 to
1.3 for upper marine clay and 0.45 to 0.95 for the lower marine clay, whereas the
recompression index ranges from 0.083Cc to 0.3Cc with an average of 0.18 for both the
UMC and LMC. Dames and Moore (1983) also reported the similar average values
for both UMC and LMC. Tan (1983) proposed a correlation between Cc and initial
void ratio (measured at the start of oedometer test), eo for Singapore marine clay (both
UMC and LMC) based on 233 oedometer tests:
Cc = 0.344 (eo + 0.51)
(3.10)
Another correlation was proposed by Dames & Moore (1983) for design purposes in
Singapore:
Cc = 0.54 (eo - 0.15)
(3.11)
Todo et al. (1993) observed that the e-log v curve of the Singapore marine
clay after yield stress is non-linear. They attributed this behaviour to the postulation
that Singapore marine clay is cemented. However, the cementation agent was not
revealed. Tan et al. (2002b) also observed similar behaviour in soil from SAC and PT
site (Figure 3.6). However, no clear evidence of cementation was observed through
the investigation of microfabric using SEM. In addition, they also found that this
55
behaviour is only noticeable when sampling disturbance is minimized and the more
refined CRS test is used.
Tan et al. (2002b) described the non-linear behaviour in compression curve
after vy with two compressibility indices which are Cc1 (termed as Ccmax from hereon)
and Cc2. The parameter Ccmax is the compression index immediately after vy and Cc2
is the usual value along the normally consolidated line well after vy (>1000 kPa). In
Chapter 5, it will be shown that Cc2 reported by Tan et al. (2002b) is comparable with
compression index for reconstituted clay, Cc* as defined by Burland (1990).
Therefore, to ensure the consistency of symbol for the discussions in subsequent
chapters, Cc2 used by Tan et al. (2002b) will be replaced by Cc*. Results on Ccmax and
Cc* from the complete series of one-dimensional compression tests carried out by Tan
et al. (2002b) on LMC from SAC site are summarized in Table 3.2 and Figure 3.7. As
shown in Figure 3.7, the effect of sample quality on compressibility is most obvious in
CRS tests. It could be noticed that for samples retrieved using a local sampler (BH3),
the measured ratio of Ccmax/Cc* are significantly lower than Ccmax/Cc* measured from
samples retrieved using the Japanese sampler. It is also shown in Figure 3.7, for a
given sampling method, CRS tests also tend to produce higher ratios of Ccmax/Cc* than
standard oedometer tests.
In addition, Tan et al. (2002b) also compared the measured Ccmax and Cc* from
CRS tests on LMC from SAC and PT site retrieved using Japanese sampler with Eq.
3.10 and Eq. 3.11. They found that these equations act as a lower bound for the Ccmax
data in Figure 3.8, but is an upper bound for the Cc* data in Figure 3.9. Hence, they
concluded that these equations could be potentially unconservative for compressibility
56
calculations when Ccmax is operative, but is acceptable when the compression is well
beyond vy.
From the previous paragraphs, the effect of sample quality and test type on
compressibility was well-studied. Sample quality and test type have a pronounced
effect on the measured compressibility curve immediately after vy. The reduction of
measured compressibility caused by sample disturbance is believed due to the partial
destructuration resulting from strain induced on soil samples during sampling (Hight et
al., 2002).
Singapore marine clay is structured and some in-depth study is needed to further verify
this postulation and to investigate the origin of this soil structure. In addition, if this
progressive destructuration of soil structure could be quantified, the non-linear
compression curve after vy could be predicted (Liu & Carter, 2000).
generally influenced by test type, strain rate and overconsolidation ratio (Kulhawy &
Mayne, 1990). Based on 500 tests data from UU and UCT tests, Tan (1983) reported
57
that su/vo for UMC ranging from 0.18 to 0.41 while su/vo for LMC ranging from
0.25 to 0.41 (Table 3.1). Dames & Moore (1983) also reported a similar range
between 0.22 and 0.40 for both UMC and LMC from another independent set of data
obtained from UU tests. They attributed this wide range of su/vo to error in the
estimation of vo, sample disturbance and overconsolidation state of Singapore marine
clay.
Tan et al. (2002b) showed that su from UCT tests on LMC from SAC site is
sensitive to sampling quality. For samples retrieved using the Japanese sampler, su is
generally higher than those determined from samples retrieved using local sampler
(Figure 3.10a). The average difference is roughly 30%. However, with isotropic
(CIU) or anisotropic (CKoU) reconsolidation to in-situ stresses, the difference in shear
strengths arising from samples of differing quality and the variability with depth
reduced very significantly (Figure 3.10b).
observations to the nearly strength isotropy of LMC from SAC site. Tanaka et al.
(2000) reported that the ratio between the compression undrained shear strength and
extension undrained shear strength for LMC from SAC is ranging from 0.9 to 1.1. In
addition, it could be noticed in Figure 3.10b that the shear strength profile obtained
from UCT tests is highly variable as compared to that obtained from CIU and CKoU.
Nevertheless, when compared the data from UCT tests to the range of su/vo reported
by Tan (1983) and Dames & Moore (1983), it can be seen that using an su/vo of 0.40
is a gross overestimation, whereas the lower bound value of 0.22 seems more
appropriate (Tan et al. 200b).
58
Based on 303 FVT test results, Tan (1983) found that the su/vo ranges from
0.19 to 0.42 for both UMC and LMC. Dames & Moore (1983) also reported the
almost similar range (from 0.22 to 0.42) for su/vo obtained from LVT and FVT tests.
Early work by Skempton (1957) suggested that the undrained shear strength ratio of
normally consolidated clays (su/vo) is a linear function of PI for vane shear test
(Figure 3.11).
This
observation for DST is consistent with that reported by Mesri (1989) (su/vy = 0.22),
but very low for FVT in comparison with Bjerrum (1973). They attributed the equality
of su/vy for FVT and DST to the strength isotropy as discussed above.
It is important to recognize that the sampling process causes two different
disturbances, namely, the loss of residual effective stress and soil structure (Tanaka,
2000; Tan et al. 2002a). Reconsolidation to in-situ stresses are used to negate the
effect of loss of residual effective stress. Since it was shown that reconsolidation to insitu stresses can recover significant portion of strength loss due sample disturbance,
this implied that the lower su measured by UCT or UU test is primarily due to the loss
of residual effective stress and sample disturbance induced during sampling is likely to
have resulted in small scale of destructuration (Tan et al. 2002a). However, the degree
of destructuration induced by that the sampling disturbance could not be assessed
59
because comparison with su from ideal sample that is free from sample disturbance is
not possible.
Based on 303 test data obtained from FVT, Tan (1983) reported that the strength
sensitivity, St of UMC and LMC lies between 1.5 to 6 and 3 to 5, respectively. Dames
& Moore (1983) also reported the almost similar range (between 2.5 to 7.5) from both
UMC and LMC determined by FVT.
Skempton & Northey (1952), Singapore marine clay can be classified as medium
sensitive to sensitive clay. As mentioned in Section 2.5.1, St gives an indication of the
effect of soil structure on the enhanced resistance to undrained shearing of intact soil
as compared to the remoulded soil at the same water content. Therefore, these data
clearly reveal that natural Singapore marine clay is structured. However, systematic
study on the origin of soil structure and its effect of soil structure on mechanical
behaviour of natural Singapore marine clay is still lacking.
60
same test to give Eu/su. Based on 13 test data obtained in conjunction with the
laboratory testing works for the MRT Detailed Geotechnical Study, Tan (1983) found
that most of the Eu/su data lie between 170 to 300. This range falls within the right
zone given in the chart for Eu/su as a function of plasticity index and yield stress ratio
by Duncan & Buchignani (1976) (Figure 3.13). However, it should be reminded that
Figure 3.13 is derived based on data from direct simple shear test reported by Ladd et
al. (1977) following the SHANSEP testing procedure which is generally known to
eliminate the contribution of soil structure to the stiffness of soil (Burland, 1990).
It is now recognised that laboratory measurement of soil stiffness using
external LVDT is typically far smaller than local strain measurement with Halls effect
strain transducer (Tan et al., 2002a). In addition, it is also recognised that Eu is also
strain dependent where Eu decreases with increasing shear strain (Hight & Leroueil,
2002). Therefore, Eu obtained from the unload-reload cycle at about one-third the
peak deviator stress might experience a higher strain level than those induced by
engineering structures such as foundation of building and hence the modulus may to be
underestimated. This highly likely be one of the main reasons contributing to the
discrepancy between the stiffness determined using conventional triaxial apparatus and
those back analysed stiffness from field data. However, due to the availability of data,
Eu was estimated as secant modulus at stress level of about one-third the peak deviator
stress with external strain measurement in UCT and UU will used in this study as a
qualitative indication of the effect of soil structure on the stiffness of natural Singapore
marine clay.
3.4
61
Concluding Remarks
Strength sensitivity (Tan, 1983 and Dames & Moore, 1983), non-linear virgin
compression curves (Todo et al., 1993; Tan et al. 2002b) and the effect of sample
disturbance on the mechanical properties of Singapore marine clay (Tan et al., 2002a
and 2002b), are the indication of the existence of soil structure in Singapore marine
clay. However, at present, no special attention was paid on the degree of soil structure
and the significance of its effect on the mechanical behaviour of natural Singapore
marine clay. Similarly, no systematic study was carried out on the investigation of the
possible causes of overconsolidation in natural Singapore marine clay.
In order to provided a better understanding of the behaviour of natural
Singapore marine clay, the degree of soil structure and the significance of its effect on
the mechanical behaviour of natural Singapore marine clay will be characterised using
the frameworks discussed in Chapter 2.
Table 3.1 Properties for Singapore Marine Clay (after Tan 1983)
Geotechnical Properties
879
757
314
293
568
466
Liquidity Index
0.65-1
0.6
Activity
Su/vo
0.18-0.41
0.25-0.41
Sensitivity
1.5-6
3-5
3
21.5
7
21.5
1-1.5
1-1.5
Compression Index, Cc
0.7-1.3
0.45-0.95
Recompression Index, Cr
0.083-0.3Cc
0.083-0.3Cc
0.5-5.0
0.5-2.0
2-3 cv
2-3 cv
Coefficient of permeability, kv x
10-8 (cm/sec)
1-20
Eu/su
170-300
170-300
Ko
0.52-0.72
0.52-0.72
62
Table 3.2 Difference in Ccmax and Cc* for different samplers and test methods.
(after Tan et al., 2002b)
(a) BH1 (retrieved using Japanese Sampler) using CRS
Sample no.
Ccmax
Cc* #
Ccmax/Cc*
BH1-UDP22
0.56
0.44
1.27
BH1-UDP32
0.96
0.57
1.68
BH1-UPD45
1.3
0.53
2.45
BH1-UDP52
1.42
0.57
2.49
BH1-UDP62
1.80
0.67
2.69
BH1-UDP72
1.07
0.58
1.85
BH1-UDP82
1.14
0.57
2.00
BH1-UDP92
1.42
0.55
2.58
Ccmax
Cc* #
Ccmax/Cc*
BH1-UDP46
0.90
0.66
1.36
BH1-UDP64
1.23
0.90
1.37
BH1-UDP84
0.86
0.73
1.17
Ccmax
Cc*#
Ccmax/Cc*
BH3-UDP22
0.40
0.40
1.00
BH3-UDP32
0.62
0.52
1.19
BH3-UDP44
0.66
0.58
1.13
BH3-UDP52
1.01
0.57
1.77
BH3-UDP64
0.81
0.57
1.42
BH3-UDP73
0.97
0.68
1.42
BH3-UDP83
0.74
0.60
1.23
BH3-UDP92
0.94
0.58
1.62
BH3-UDP102
0.74
0.61
1.21
63
Table 3.2 Difference in Ccmax and Cc* for different samplers and test methods.
(Tan et al., 2002b) (continued)
(d) BH3 (retrieved using Local Sampler) using oedometer
Sample no.
Ccmax
Cc*#
Ccmax/Cc*
BH3-UDP25
0.70
0.56
1.25
BH3-UDP35
0.70
0.67
1.04
BH3-UDP45
0.66
0.66
1.00
BH3-UDP55
0.70
0.65
1.08
BH3-UDP65
0.92
0.71
1.30
BH3-UDP75
1.01
0.81
1.25
BH3-UDP85
0.88
0.78
1.13
BH3-UDP95
0.74
0.62
1.19
BH3-UDP105
0.69
0.69
1.00
64
65
25
Non-deposition
0
-25
Maximum thickness of
marine clays in Singapore
Kenney (1964)
-50
-75
6000
4000
2000
Figure 3.1 Present understanding of the age of the Kallang Formation (after Tan et al.,
2002b).
Figure 3.2 Changes in the volume of water in the oceans of the world, expressed as
metres of equivalent sea-level change from the last interglacial (LIG) to the present.
(after Bird et al., 2003)
66
14
14
16
16
18
BH1-CRS
18
BH1-OED
Depth (m)
20
BH3-CRS
BH3-OED
22
20
22
'vo
24
24
26
26
28
28
30
100
30
200
300
400
0.8
1.2
1.6
2.0
2.4
Figure 3.3 Yield stress and YSR for SAC Site (after Tan et al., 2002b)
(Samples from BH1 were retrieved with Japanese sampler while samples from BH3
were retrieved with local sampler)
67
Figure 3.4 Yield stress and yield stress ratio profile (after Hanzawa & Adachi, 1983)
(a)
(b)
Figure 3.5 (a) YSR data from two independent commercial testing facilities for PT
(b) Measured YSR from PARI and hypothetical addition of 80 kPa overburden to recreate YSR values before erosion. (after Tan et al., 2002b)
68
2.0
BH1(Jap)-CRS
1.8
BH1(Jap)-OED
BH3 (Local)-CRS
1.6
Void Ratio
BH3(Local)-OED
1.4
1.2
1.0
0.8
0.6
10
100
1000
10000
Figure 3.6 Typical compressibility behaviour of samples retrieved using two different
samplers and subjected to CRS and oedometer tests (sampling depth between 21.5 m
and 22.5 m) at SAC site. (after Tan et al., 2002b)
14
16
BH1-CRS
BH1-OED
Depth (m)
18
BH3-CRS
BH3-OED
20
22
24
26
28
30
0.0
0.4
0.8
1.2
1.6
*
C1/C
CCcmax
/CC2c
2.0
2.4
2.8
Figure 3.7 Variation of Ccmax/Cc* with depth for different samples subjected to
oedometer and CRS tests. (after Tan et al., 2002b)
69
2.0
PT-Upper
PT-Intermediate
PT-Lower
SAC
Dames & Moore (1983)
1.2
Tan (1983)
1.6
0.8
0.4
0.0
0.0
0.5
1.0
1.5
2.0
2.5
Figure 3.8 Ccmax versus in-situ void ratio for clays from the two test sites (after Tan et
al., 2002b)
1.6
PT-Upper
PT-Intermediate
PT-Lower
1.2
SAC
C
CcC2
0.8
0.4
0.0
0.0
0.5
1.0
1.5
In-situ void ratio (e0)
2.0
2.5
Figure 3.9 Cc* versus in-situ void ratio for clays from the two test sites (after Tan et
al., 2002b)
70
14
14
16
16
18
18
0.40'vo
0.22'vo
22
CIU (Japan)
CIU (Local)
CKoU (Japan)
CKoU (Local)
22
24
24
26
26
UCT (Japan)
28
UCT (Local)
20
Depth (m)
Depth (m)
20
UCT (Japan)
28
0.30'vo
UCT (Local)
30
30
0
20
40
60
80
100
120
20
40
60
80
100
(a)
(b)
120
Figure 3.10 (a) Effect of sampling method on undrained shear strength for SAC clay
(b) Comparison of undrained shear strength using different recompression methods.
(after Tan et al., 2002b)
0.6
Singapore (SAC)
Singapore (PT)
su (field vane)/p'c
su(field vane)/vy
0.5
Skempton (1957)
Drammen
Ariake
0.4
Bangkok
Louiseville
0.3
Bjerrum (1973)
0.2
0.1
0.0
0
20
40
60
Plasticity Index (PI)
80
100
Figure 3.11 Undrained shear strength ratio (su/vy) versus plasticity index for different
clays. (after Tan et al. 2002b)
71
14
UCT (Japan-SAC)
UCT (Local-SAC)
16
UCT (PT)
DST (PT)
Depth (m)
20
22
24
26
0.21vy
p'c(PT)
(PT)
0.21
28
p'vyc (SAC)
0.21
(SAC)
30
0
40
80
120
160
Eu/su
Figure 3.12 Undrained shear strengths from PT and SAC Sites versus prediction based
on vy (after Tan et al. 2002b)
Figure 3.13 Generallized undrained modulus ratio versus yield stress ratio and
plasticity index (after Duncan & Buchignani, 1976)