Professional Documents
Culture Documents
INTRODUCTION
Bangkok clay is well-known as a soft clay with a high
water content close to its liquid limit. It has large potential for settlement with low inherent shear strength. This
clay is classied as non- to low swelling (Horpibulsuk et
al., 2007) as per free swelling test (Prakash and Sridharan, 2004). Its swelling potential increases with depth. Besides Bangkok clay, non- to low swelling soils are generally found in many lowlands, such as Ariake bay in Japan
(El-Shafei, 2001; Modmoltin, 2002). Data provided by
Tanaka et al. (2001) indicate that some marine clays
(Pusan, Singapore, Drammen, Louiseville clays) are classied as inactive and normal clays, and are thus non- to
low swelling. Even though the clay mineralogy of many
clayey soils is primarily montmorillonite, the montmorillonite in those soils might not be the dominant
parameter controlling the soil expansivity. Other
parameters, such as other clay minerals (kaolinite, and illite, etc), the non-clay fraction (0.002 mm) and pore
medium chemistry, can also play a great role, masking
i)
ii)
iii)
iv)
240
HORPIBULSUK ET AL.
Laboratory Investigation
Since Bangkok clay is non- to low swelling, and the
swelling potential increases with depth (Horpibulsuk et
al., 2007), kaolin and Bangkok clays, with free swell ratios of 0.6 and 1.3, were used to represent non- and low
swelling clays, respectively. The free swell ratio, FSR, is
dened as the ratio of equilibrium sediment volume of
10-g oven-dried soil passing a 425 mm sieve in distilled
water (Vd) to that in carbon tetra chloride or kerosene
(Vk). Soil is classied as non-swelling and low swelling
when the free swell ratio is (less than or equal to) 1.0 and
between 1.0 and 1.5, respectively (Prakash and Sridharan, 2004). Kaolin clay was obtained from a commercial
company. Its consistency limits were liquid limit, LL
43z and plastic limit, PL of 34z. Its specic gravity was
2.78. Bangkok clay was collected from Ladkrabang district, Bangkok, Thailand at a depth of 3 to 4 meters. Its
natural water content was 85z, the liquid and plastic
limits were in the order of 89 and 30z, and the specic
gravity was 2.71. The groundwater level was at about 1.0
m from the surface. The overconsolidation ratio was 1.2
and the eective strength parameters in triaxial compression were c?0 and q?229
. The chemical properties of
the two clays and the Type I Portland cement are shown
in Table 1. The grain size distribution curves for the two
clays and the cement are shown in Fig. 1. The specic
gravity of the cement was 3.15.
The main aims of the laboratory test are to analyze the
strength development of the cement admixed Bangkok
clay using the wc/C as a prime parameter and to develop
its generalized strength prediction equation. For the rst
aim, both Bangkok and kaolin clays were passed through
241
Table 1. Chemical composition of the cement, kaolin clay, and Bangkok clay
Chemical
compounds
Portland cement
Type I
Kaolin clay
Soft Bangkok
clay
zSiO2
zAl2O3
zFe2O3
zMgO
zCaO
zNa2O
z K 2O
zSO3
20.90
4.76
3.41
1.25
65.41
0.24
0.35
2.71
59.79
31.84
1.59
3.05
0.05
63.83
21.34
8.41
1.54
0.94
0.28
2.45
1.22
Fig. 1. Grain size distribution of Bangkok and kaolin clays and Portland cement
Fig. 2.
Field Investigation
The eld study investigates the strength reduction due
to several eld factors such as non-uniformity in mixing
in-situ clay with cement, and the dierence in curing conditions between laboratory and eld improvement. Deep
mixing was performed in the Sukhaphiban 3, and Ladkrabang districts in Bangkok, and the Bangpee district in
Sumutphakarn province, Thailand. The soil proles for
the three test sites are presented in Fig. 2. It was found
that the uppermost soil of about 2 meters in thickness,
was weathered zone. The soft Bangkok clay was more
242
HORPIBULSUK ET AL.
Fig. 3.
243
Fig. 6. Compression behavior of cement admixed Bangkok clay samples for wc/C12 after 28 days of curing
down. The lower the wc/C, the greater the strength and
the yield stress. This nding is the same as that reported
by Horpibulsuk et al. (2005) and Miura et al. (2001) for
the cement admixed Ariake clay. As such, it is possible to
adopt the wc/C and Abrams' law to analyze the strength
development in the cement admixed Bangkok clay.
Figures 7 and 8 show an analysis of laboratory strength
development in the cement admixed kaolin and Bangkok
clays for dierent curing times. It is found that the wc/C
can be applied to analyze the strength development of cement admixed clay with dierent cement content and
water content at a specic curing time. The test data are
represented well by a power function in the form:
qu
A
(wc/C )B
(1)
HORPIBULSUK ET AL.
244
varies in a narrow band between 1.25 and 1.31, irrespective of the swelling potential and curing time considered.
The parameter B can thus be taken as a constant for both
cement admixed clays (non- to low-swelling clays) for 7 to
120 days of curing. When the parameter B is 1.27, the
result is the following relation:
1.27
(2)
where q(w /C) is the strength estimated at a clay-water/cement ratio of (wc/C )1, and q(w /C) is the strength value at
clay-water/cement ratio of (wc/C )2.
At a particular wc/C, the strength development with
time is controlled only by the value of A since B is regarded as constant. Even though the parameter A depends on clay type, the rate of strength development with
time is identical for various admixed clays since it is
predominantly inuenced by the hydration process (Horpibulsuk et al., 2003). As such, it is possible to generalize
the strength development using the 28-day strength, q28 of
cement admixed clays as a reference value (vide Fig. 9). A
linear regression analysis gives the following relationship
c
qD
0.0390.283 ln D
q28
(3)
{ }
q(w /C)
q(w /C)
c
28
(wc/C )28
(wc/C )D
1.27
(0.0390.283 ln D)
(4)
where q(w /C) is the strength of the cement admixed Bangkok clay to be estimated at clay-water/cement ratio of (wc
/C ) after D days of curing, and q(w /C) is the strength of
the cement admixed Bangkok clay at clay-water/cement
ratio of (wc/C ) after 28 days of curing.
Tables 2 to 4 show that the predicted strengths are in
c
28
12
245
Trial mix for determination of input of cement for Sukraphiban 3 district for dry mixing method
Claywater/cement
ratio,
wc/C
Laboratory
strength,
qul (kPa)
Predicted
strength,
qup (kPa)
106.7
106.7
106.7
5.09
5.09
5.09
873
1167
1453
857
1142
Reference
1.83
2.17
0
24.44
24.44
24.44
106.7
106.7
106.7
4.37
4.37
4.37
1008
1378
1745
1042
1389
1735
3.40
0.78
0.54
7
14
28
27.93
27.93
27.93
106.7
106.7
106.7
3.82
3.82
3.82
1262
1628
1969
1235
1645
2056
2.13
1.10
4.42
250
7
14
28
34.92
34.92
34.92
106.7
106.7
106.7
3.06
3.06
3.06
1596
1996
2250
1639
2185
2730
2.70
9.42
21.34
150
7
14
28
21.45
21.45
21.45
114.5
114.5
114.5
5.34
5.34
5.34
803
985
1246
807
1076
1344
0.53
9.17
7.92
175
7
14
28
25.03
25.03
25.03
114.5
114.5
114.5
4.58
4.58
4.58
992
1181
1463
982
1308
1635
1.05
10.83
11.74
200
7
14
28
28.6
28.6
28.6
114.5
114.5
114.5
4.00
4.00
4.00
1224
1446
1694
1163
1550
1937
4.98
7.19
14.33
250
7
14
28
35.75
35.75
35.75
114.5
114.5
114.5
3.20
3.20
3.20
1676
1932
2168
1544
2058
2572
7.86
6.52
18.62
150
7
14
28
16.12
16.12
16.12
66.6
66.6
66.6
4.13
4.13
4.13
1022
1246
1549
1118
1490
1862
9.36
19.60
20.15
175
7
14
18.81
18.81
66.6
66.6
3.54
3.54
1209
1429
1360
1812
12.50
26.76
Input of
cement, C?
(kg/m3)
Curing
time,
(days)
Cement
content,
C (z)
150
7
14
28
20.95
20.95
20.95
175
7
14
28
200
Clay water
content,
w c (z )
1
n
i 1
|qupqul|
100)
qul
|qupqul|
100
qul
z
8.24z
Remark: for dry mixing, the natural water content, wn and the clay water content, wc are the same.
HORPIBULSUK ET AL.
246
Table 3.
Depth
(m.)
3
Trial mix for determination of input of cement for Sukraphiban 3 district for wet mixing method
Input of
cement, C?
(kg/m3)
Curing
time,
(days)
Cement
content,
C (z)
Water
content,
w c (z )
Claywater/cement
ratio,
wc/C
Laboratory
strength,
qul (kPa)
Predicted
strength,
qup (kPa)
175
7
14
28
27.05
27.05
27.05
122.62
122.62
122.62
4.53
4.53
4.53
631
758
1053
596
794
992
5.56
4.77
5.75
185
7
14
28
28.78
28.78
28.78
124.00
124.00
124.00
4.31
4.31
4.31
703
874
1121
635
847
1058
9.61
3.11
5.60
200
7
14
28
31.40
31.40
31.40
126.07
126.07
126.07
4.02
4.02
4.02
793
920
1175
695
926
1158
12.35
0.69
1.49
175
7
14
28
24.63
24.63
24.63
108.33
108.33
108.33
4.40
4.40
4.40
565
722
999
619
825
1031
9.60
14.30
3.22
185
7
14
28
26.20
26.20
26.20
109.59
109.59
109.59
4.18
4.18
4.18
634
773
1119
660
879
Reference
4.08
13.76
0
200
7
14
28
28.58
28.58
28.58
111.48
111.48
111.48
3.90
3.90
3.90
668
918
1236
721
961
1201
7.95
4.68
2.85
175
7
14
28
20.87
20.87
20.87
89.60
89.60
89.60
4.29
4.29
4.29
552
658
807
638
851
1063
15.64
29.28
31.73
185
7
14
28
22.19
22.19
22.19
90.67
90.67
90.67
4.09
4.09
4.09
638
775
940
680
906
1132
6.52
16.86
20.40
200
7
14
28
24.19
24.19
24.19
92.28
92.28
92.28
3.82
3.82
3.82
705
900
1090
742
988
1235
5.19
9.81
13.30
1
n
|qupqul|
100)
qul
i 1
|qupqul|
100
qul
z
9.56z
Remark: W/C1.0
column before the onset of the hydration (prior to cementation bonds formation) (Miura et al., 1998; Shen and
Miura, 1999; Shen et al., 2003; Shen et al., 2008). This
consolidation results in a signicant reduction in the eld
water content, as experimentally observed by Consoli et
al. (2006) and Rotta et al. (2003).
Figures 12 and 13 show the strength development in
the dry and the wet mixing columns at Sukhaphiban 3
district. Figure 14 shows the strength development in the
wet mixing columns at Ladkrabang district. For the same
input of content in kg per cubic meter of clay for the
whole depth, the weight of cement per dry weight of clay,
C, decreases with depth since water content decreases
(unit weight increases) with depth. The decrease in both
wc and C yield almost the same wc/C value, and therefore
practically the same laboratory strength, qul, for the
whole improved depth. The laboratory strengths shown
in Figs. 12 to 14 are thus the average values from dierent
depths. It is found that the eld strength is between 0.6
and 1.5 times the laboratory strength for the wet mixing
and between 0.6 and 1.0 times the laboratory strength for
the dry mixing. It must be kept in mind that the quf/qul
12
247
Trial mix for determination of input of cement for Ladkrabang district for wet mixing method
Input of
cement, C?
(kg/m3)
Curing
time,
(days)
Cement
content,
C (z)
Water
content,
w c (z )
Claywater/cement
ratio,
wc/C
Laboratory
strength,
qul (kPa)
Predicted
strength,
qup (kPa)
150
7
14
28
21.78
21.78
21.78
110.54
110.54
110.54
5.08
5.08
5.08
359
500
669
395
526
Reference
10.01
5.19
0
175
7
14
28
25.81
25.81
25.81
113.83
113.83
113.83
4.41
4.41
4.41
453
614
865
471
628
785
4.03
2.26
9.19
200
7
14
28
29.95
29.95
29.95
117.12
117.12
117.12
3.91
3.91
3.91
593
770
1010
549
732
915
7.31
4.96
9.47
250
7
14
28
38.57
38.57
38.57
123.70
123.70
123.70
3.21
3.21
3.21
762
940
1197
707
942
1177
7.24
0.14
1.66
150
7
14
28
21.16
21.16
21.16
100.34
100.34
100.34
4.74
4.74
4.74
439
556
758
430
573
716
2.07
3.09
5.53
175
7
14
28
25.08
25.08
25.08
103.53
103.53
103.53
4.13
4.13
4.13
517
673
928
513
684
854
0.75
1.51
7.94
200
7
14
28
29.12
29.12
29.12
106.72
106.72
106.72
3.67
3.67
3.67
596
808
1061
596
795
993
0.04
1.68
6.35
250
7
14
28
37.52
37.52
37.52
113.10
113.10
113.10
3.01
3.01
3.01
743
959
1261
764
1019
1273
2.92
6.25
0.96
150
7
14
28
18.69
18.69
18.69
95.66
95.66
95.66
5.12
5.12
5.12
463
601
792
390
520
650
15.72
13.41
17.89
175
7
14
28
22.13
22.13
22.13
98.51
98.51
98.51
4.45
4.45
4.45
544
712
946
466
621
776
14.32
12.74
18.00
200
7
14
28
25.65
25.65
25.65
101.35
101.35
101.35
3.95
3.95
3.95
646
836
1101
542
722
903
16.10
13.58
18.00
250
7
14
28
32.97
32.97
32.97
107.04
107.04
107.04
3.25
3.25
3.25
788
1051
1348
696
927
1159
11.73
11.78
14.03
1
n
i 1
|qupqul|
100)
qul
|qupqul|
100
qul
z
7.72z
Remark: W/C1.0
248
HORPIBULSUK ET AL.
Fig. 14.
Fig. 11. Field curing water content of wet mixing columns at Sukhaphiban 3 district
Fig. 15. Eect of water to cement ratio, W/C, on the strength development of wet mixing columns
Fig. 12. Strength prole of dry mixing columns at Sukhaphiban 3 district
in this case.
Figure 17 shows the strength development of the high
strength columns (C220 kg/m3) compared with that of
the low strength column (C220 kg/m3) for the same
W/C and installation rate (PRWR1.0 m/min). It is
clearly noted that the strength of all the wet mixing
249
Fig. 18. Suggested procedure of wet mixing method for soft Bangkok
clay
columns made up from dierent cement contents is practically the same. In other words, the strength is irrespective of the cement content for such a high installation
rate. This result is the same as that reported by Horpibulsuk et al. (2004c) for improvement with high cement contents with installation rates close to and higher than 1.0
m/min. As such, it is not advantageous or economical to
opt for the high cement content columns particularly at
high installation rate. Horpibulsuk et al. (2004c) concluded that an enhancement of strength does occur when the
installation rate is less than 0.7 m/min. The higher the cement content, the greater the reduction in water due to
larger hydration, making mixing dicult.
From this study and the work of Horpibulsuk et al.
(2004c), wet mixing is suitable for executing low strength
columns since the fast installation rate can be adopted
and the quf/qul is high. For high strength columns, high
input of cement is required with the very low installation
rate. The quf/qul value is about 0.330.67, which is much
lower than that of low strength columns (Horpibulsuk et
al., 2004c). As such, for a particular design (dead and
live) load on the soft Bangkok clay, it might be more economical to improve soft ground with many low strength
Based on the laboratory and eld study, our suggestions for improving deep mixing for the soft Bangkok
clay are summarized and presented (vide Fig. 18). The
method below is for low strength columns (qul(28days)
1500 kPa) by the wet mixing method.
Determination of input of cement compensating for eld
factors
1. From the soil prole and the design load on the soft
ground, determine the diameter, length, spacing
and eld strength at the service time of the deep
mixing column.
2. From the target eld strength at service time (open
square symbol), estimate the target eld strength at
7 days of curing (Target quf(7days)) (cross symbol),
which can be approximated using Eq. (3).
3. Determine the laboratory strength at 7 days of curing (qul(7days)), using the eld strength reduction of
1.7 (quf/qul0.6) (black circle symbol).
4. Determine the cement content to attain the laboratory strength at 7 days of curing (qul(7days)) and service time. This task can simply be done using Eq. (4).
Field execution and examination of eld strength
5. Execute the deep mixing column with a W/C of 1.0
and an installation rate (PR and WR) of about 1.0
m/min.
6. Core the soil-cement samples from a selected
column at dierent depths to determine the eld
strength at 7 days of curing for every section (every
3000 column).
7. If quf(7days)0.6 qul(7days), then the tested section meets
the requirement.
8. In the case that quf(7days)0.6 qul(7days),
8.1 If the actual quf(7days) is slightly lower than the
target quf(7days) (black square), the service time of
the section should be postponed to increase curing time.
8.2 If the actual quf(7days) is much lower than target
HORPIBULSUK ET AL.
250
ACKNOWLEDGEMENT
The authors would like to acknowledge the nancial
support provided by the Suranaree University of Technology. The eld test results of the soil-cement columns
obtained from the Bureau of Materials Analysis and Inspection, Department of Highways, Thailand are appreciated.
LIST OF NOTATIONS
Aempirical constant
Bempirical constant
Ccement content in terms of percentage
C?input of cement in terms of kg/m3
quunconned compressive strength
qufeld strength of the column
quf(7days)eld strength of the column at 7 days of curing
qullaboratory strength
qul(28days)laboratory 28-day strength
q(w /C) strength to be estimated at clay-water/cement
ratio of (wc/C )1
q(w /C) strength value at clay-water/cement ratio of
(wc/C )2
q(w /C) strength of the cement admixed Bangkok clay
to be estimated at clay-water/cement ratio after D days of curing
q(w /C) strength of the cement admixed Bangkok clay
clay-water/cement ratio after 28 days of curing
wnnatural water content
w(lab)3dayslaboratory 3-day curing water content
w(lab)28dayslaboratory 28-day curing water content
wc/Cclay-water/cement ratio
W/Cwater to cement ratio
FSfactor of safety
c
28
REFERENCES
1) Abrams, D. A. (1918): Design of concrete mixtures, Structural
Materials Research Material, Bulletin 1, Lewis Institute, Chicago.
2) Buttereld, R. (1979): A natural compression law for soils (an advance on e-log p?), Geotechnique, 29(4), 469480.
3) Consoli, N. C., Rotta, G. V. and Prjetto, P. D. M. (2006): Yielding-compressibility-strength relationship for an articially cemented soil cured under stress, Geotechnique, 56(1), 6972.
4) El-Shafei, A. (2001): Geo-environmental and geotechnical assessments of nitrate pollute soil, Ph.D. Dissertation, Saga University,
Saga, Japan.
5) Hashizume, H., Okochi, Y., Dong, J., Horii, N., Toyosawa, Y.
and Tamate, S. (1998): Study on the behavior of soft ground improved using Deep Mixing Method, Proceedings of International
Conference on Centrifuge, 98, 851856.
6) Horpibulsuk, S. (2001): Analysis and Assessment of Engineering
Behavior of Cement Stabilized Clays, Ph.D. Dissertation, Saga
University, Saga, Japan.
7) Horpibulsuk, S. and Miura, N. (2001): A new approach for
studying behavior of cement stabilized clays, Proc. 15th International Conference on Soil Mechanics and Geotechnical Engineering
(ISSMGE), Istanbul, Turkey, 3, 17591762.
8) Horpibulsuk, S., Miura, N. and Nagaraj, T. S. (2003): Assessment
9)
10)
11)
12)
13)
14)
15)
16)
17)
18)
19)
20)
21)
22)
23)
24)
25)
26)
27)
28)
29)
30)
31)
32)
33)
34)
35)
36)
251