An Observational Method For Consolidation Analysis of The PVD-improved Subsoil

You might also like

Download as pdf or txt
Download as pdf or txt
You are on page 1of 9

Geotextiles and Geomembranes 46 (2018) 625–633

Contents lists available at ScienceDirect

Geotextiles and Geomembranes


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

An observational method for consolidation analysis of the PVD-improved T


subsoil
Wei Guoa,b, Jian Chuc, Wen Niec,∗
a
State Key Laboratory of Hydraulic Engineering Simulation and Safety, Tianjin University, China
b
School of Civil Engineering, Tianjin University, 135 Yaguan Road, Jinnan District, Tianjin 300350, China
c
School of Civil & Environmental Engineering, Nanyang Technological University, 50 Nanyang Ave, 639798, Singapore

A R T I C LE I N FO A B S T R A C T

Keywords: Surcharge preloading together with prefabricated vertical drains (PVDs) have been widely used to improve
Geosynthetics properties of thick clay deposits. To assess the performance of soil improvement works, the average degree of
PVD consolidation needs to be estimated. A curve fitting formula is proposed in this paper to simulate the degree of
Consolidation consolidation versus a non-dimensional time factor relationship. The proposed formula fits the theoretical
Asaoka's method
consolidation solutions well with a regression coefficient R2 larger than 0.9996 and an error of less than 1.2%.
Ground improvement
Based on the formula, a modified Asaoka's observational method is proposed to predict the ultimate settlement
and calculate the coefficient of consolidation using field settlement monitoring data. The effectiveness of the
proposed observational method has been verified using some well-documented case histories. Comparisons
between the proposed method and the Asaoka's method indicate that the proposed method will give a less than
1.0% higher ultimate settlement than that by the Asaoka's method and the proposed method is able to predict the
ch value with the consideration of both vertical and horizontal flow through the ratio of time factor in horizontal
and vertical direction νhv.

1. Introduction β0
δult =
1 − β1 (2)
Preloading using fill surcharge and prefabricated vertical drains
For one-dimensional vertical drainage only, the coefficient of con-
(PVDs) is a common method adopted for the improvement of clay de-
solidation in vertical direction cv can be estimate by (Asaoka, 1978):
posit. To evaluate the degree of the soil improvement, average Degree
of Consolidation (DoC) is normally estimated. The DoC is defined as the 5 H 2 ln β1
ch =
percentage of current settlement to the ultimate primary consolidation 12 Δt (3)
settlement. The ultimate settlement is usually estimated using an ob-
where H = vertical drainage length.
servational method. The most commonly used observational methods
For one-dimensional horizontal drainage only, the coefficient of
include the Asaoka's method (Asaoka, 1978) and the hyperbolic method
consolidation in horizontal direction ch can be estimate by (Asaoka,
(Sridharan and Sreepada Rao, 1981; Tan et al., 1991; Tan, 1995; Tan
1978):
and Chew, 1996; Chung et al., 2014; Guo et al., 2018).
The Asaoka's method (Asaoka, 1978) is based on the fact that the μDe2 ln β1
ch = −
relationships between successive equally spaced settlement readings δn 8Δt (4)
and δn-1 can be expressed as a first order approximation and shown as:
In Eq. (4), μ is a factor related to drain spacing m as expressed by the
δn = β0 + β1 δn − 1 (1) following equation:
where settlement δ1, δ2, …, δn are selected from the monitoring data in m k 3 2l 2kh
μ = ln + h ln(s ) − + π
the way that δn is the settlement at time tn and the sampling interval s ks 4 3qw (5)
Δt = tn - tn-1 is constant. β0 is the intercept and β1 is the slope of the
linear trend line in the δn vs δn-1 plot. The ultimate primary con- where m = drains spacing ratio and m = De/dw; De = diameter of an
solidation settlement δult can be calculated as (Asaoka, 1978): equivalent soil cylinder influenced by vertical drain; De = 1.13S for a


Corresponding author.
E-mail address: wnie001@e.ntu.edu.sg (W. Nie).

https://doi.org/10.1016/j.geotexmem.2018.04.014
Received 16 March 2018; Received in revised form 13 April 2018; Accepted 15 April 2018
0266-1144/ © 2018 Elsevier Ltd. All rights reserved.
W. Guo et al. Geotextiles and Geomembranes 46 (2018) 625–633

square pattern and De = 1.05S for a triangular pattern; S = spacing of


the vertical drain; s = ds/dw, ds = the diameter of smear zone,
dw = equivalent diameter of the vertical drain; kh and ks = coefficient
of permeability in horizontal (or radial) direction of natural soil and
smear zone, respectively; l = drainage length; and qw = discharge ca-
pacity of PVD.
Eq. (4) is often adopted to estimate the coefficient of consolidation
in horizontal direction ch by ignoring the effect of vertical drainage. In
some cases, the vertical drainage may be significant and should not be
ignored. Then, ch estimated from Eq. (4) using field settlement mon-
itoring data will be overestimated (Edil et al., 1991; Arulrajah et al.,
2003, 2004a).
In this paper, a curve fitting formula is proposed to simulate the
degree of consolidation versus non-dimensional time factor relation-
ship. Based on this formula, a modified Asaoka's observational method
is proposed to predict the ultimate settlement and the coefficient of
consolidation using field settlement monitoring data. The effectiveness
of the proposed observational method is verified using some well-
documented case histories in Sweden, China and Singapore.

2. Consolidation theories

The close-form expression of the Uv versus Tv relationship from the


Terzaghi's one-dimensional consolidation theory is given as (Terzaghi
et al., 1996):
m =∞
4 2 2m + 1 2 ⎤
Uv = 1 − ∑ exp ⎡−⎛ π ⎞ Tv
π2 m=0
(2m + 1) 2 ⎢
⎣ ⎝ 2 ⎠ ⎥ ⎦ (6)

where Uv is the degree of consolidation in the vertical direction; Tv is


the non-dimensional time factor for vertical direction and Tv = cvt/H2;
cv is the coefficient of consolidation in vertical direction; n is the in- Fig. 1. Results of best fitting Uhv versus Tv curves (a) best-fitting curves and (b)
errors of the obtained Uhv.
teger; H is the vertical drainage path; and t is the consolidation time.
For a pure horizontal drainage case without considering the effect of
smear and well resistance, the equation for calculating the average 3. Consolidation formula
degree of consolidation Uh based on Barron's theory was proposed by
Hansbo (1981): The Chapman-Richards model is adopted in this study to best-fit the
Uvh - Tv relationship as shown in Eq. (9). The mathematical expression
8T of the Chapman-Richards equation was derived by Richards (1959) and
Uh = 1 − exp ⎜⎛− h ⎞⎟
⎝ μ ⎠ (7) Chapman (1961) and can be written as (Ratkowsky, 1990):

where Th is the non-dimensional time factor for horizontal drainage and y = η [1 − κ exp(−μt )]λ + ε (11)
Th = ch t/De2. where η is the amplitude of the curve, ε is the offset from zero, κ, μ and λ
By assuming the time for consolidation in both vertical and hor- are rate constants and exp() is the base of the natural logarithm.
izontal flows are identical, Carrillo (1942) proposed the average degree Combining Eqs. (9)–(11) gives the proposed Uvh - Tv formula that
of consolidation for combined vertical and horizontal drainage, Uvh, to can be used to best fit the plots shown in Fig. 1a and shown as:
be calculated as:
Uvh = {1 − exp[−2(νhv + 1) Tv]} ξ (12)
Uvh = 1 − (1 − Uv )(1 − Uh) (8)
where ξ is a curve fitting constant.
Substituting Eqs. (6) and (7) into Eq. (8), we have: Similar formula has also been adopted by Guo and Chu (2017) to
n =∞ best-fit Terzaghi's one-dimensional consolidation curve. When using Eq.
8 2 2n + 1 2 (12) for the curve fitting, the value of ξ is obtained using the “Solver”
Uvh = 1 − ∑ exp ⎡−⎛ π ⎞ Tv − 2νhv Tv ⎤
π n=0
(2n + 1)2 ⎢
⎣ ⎝ 2 ⎠ ⎥
⎦ (9) function in the Microsoft Excel program by setting a target to maximize
the regression coefficient R2. The obtained curves from the curve fitting
where νhv is the ratio of time factor in horizontal and vertical direction,
functions are also shown in Fig. 1a. The errors involved in the curve
νhv = (8Th/μ)/(2Tv); and νhv can be further derived as:
fitting in Uvh are shown in Fig. 1b. The relationship between constant ξ
4 Ch H 2 and νhv is given in Fig. 2. The R2 for different νhv and ξ are also given in
νhv = Fig. 2. It can be seen that all the R2 are greater than 0.9996 and the
μ Cv De2 (10)
errors between the obtained and the close-form solutions for Uvh are less
Solving Eq. (9) gives the Uvh versus Tv curves as shown in Fig. 1a. It than 1.2%. The νhv and ξ relationship can also be best-fitted with R2 of
can be seem that Uvh is strongly influenced by the value of νhv which is 0.9972 and shown as,
dependent on μ, H/De and ch/cv. Uvh increases in value as either H/De or
ξ = −0.0247[log(vhv )]2 + 0.1834(vhv ) + 0.6572 (13)
ch/cv increases, or μ decreases. Note that the uppermost curve with
νhv = 0 is the solution for Terzaghi's one-dimensional consolidation The values of ξ are strongly influenced by νhv and increases in value
equation. Similar methods have also been used by Chai et al. (2001) as νhv increases. When νhv = 0, ξ = 0.6 which becomes the case of the
and Low (2003). Terzaghi's one-dimensional consolidation curve (Guo and Chu, 2017).

626
W. Guo et al. Geotextiles and Geomembranes 46 (2018) 625–633

settlements δn and δn-1 as,


1 1
δnξ = α + βδnξ− 1 (19)
where α is the intercept and β is the slope of the straight line in the δn1/ξ
vs δn-11/ξ plot which can be expressed as,
1
ξ
α = (1 − β ) δult (20)

c
β = exp ⎡−2(νhv + 1) v2 Δt ⎤
⎣ H ⎦ (21)
Deriving from Eq. (20) gives the ultimate settlement δult as,
ξ
α ⎞
δult = ⎜⎛ ⎟

⎝ 1 − β⎠ (22)
Fig. 2. Effect of νhv on the constants of curve fitting ξ Deriving from Eq. (21) gives the coefficient of consolidation in
vertical direction cv as:
When νhv = ∞, ξ = 1.0 which represents the case where the effect of H 2 ln β 1
vertical drainage is negligible. When νhv = 100, ξ = 0.924. This value cv = −
2Δt 1 + νhv (23)
may set the boundary for the contribution of vertical consolidation to
the combined degree of consolidation Uvh to be considered. Combining or the coefficient of consolidation in horizontal direction ch as:
Eqs. (10) and (12), we have: μDe2 ln β νhv
ch = −
ξ 8Δt 1 + νhv (24)
8T
Uvh = ⎡1 − exp ⎜⎛− h − 2Tv ⎟⎞ ⎤
⎢ ⎥
⎣ ⎝ μ ⎠⎦ (14) Eqs. (19)–(24) can now be used as an observational model to predict
δult, cv and ch based on monitored settlement data. The procedure in-
Eq. (14) is another equation to calculate the average degree of volves five steps as outlined below. This procedure is similar to that
consolidation of soil due to combined vertical and horizontal drainage. used in the Asaoka's method, except using (δn)1/ξ instead of δn for the
Substituting Uvh = δ/δult into Eqs. (12) and (14), we have the following plot. It can be seen that the two methods become the same when ξ = 1.
δ - t relationships, respectively: Therefore, the proposed method can be considered as a generalization
δ = δult [1 − exp( −2(νhv + 1) Tv )]ξ (15) of the Asaoka's method.

or Step (1) calculate the ratio of time factor in horizontal and vertical
ξ directions νhv using Eq. (10), and the ratio ch/cv is an input;
8T
δ = δult ⎡1 − exp ⎜⎛− h − 2Tv ⎞⎟ ⎤ Step (2) identify the value of ξ from Fig. 2 or calculate ξ using Eq.
⎢ ⎥
⎣ ⎝ μ ⎠⎦ (16) (13);
It should be pointed out that Eq. (14) was derived based on excess Step (3) select settlement δ1, δ2, …, δn from a monitored settlement
pore water pressure distribution and the Uvh calculated is different from curve, where δn is the settlement at time tn and the sampling interval
Uvh = δ/δult for elasto-plastic soil. As pointed out by Chai and Zhou Δt = tn - tn-1 is constant;
(2018), the maximum difference is about 8.6% occurred when Uvh is Step (4) plot (δn)1/ξ versus (δn-1)1/ξ curve and use linear regression
about 40% and the difference becomes smaller when Uvh is getting to get the slope of the line β and its intercept with the vertical axis α;
greater. Thus, Eqs. (15) and (16) are good estimations for practical Step (5) calculate the ultimate settlement δult and coefficient of
applications where the range of degree of consolidation interested is consolidation in vertical direction cv and horizontal direction ch
well above 40%. Similar approach has also been adopted in other ob- using Eqs. (22)–(24), respectively.
servational methods proposed by Asaoka (1978); Sridharan and
Sreepada Rao (1981); Tan et al. (1991); Tan (1995); Tan and Chew As the proposed method uses linear regression to best-fit (δn)1/ξ
(1996); Chung et al. (2014). versus (δn-1)1/ξ curve, the predicted results are affected by the sampling
interval. To evaluate the effect of sampling interval, a parameter N90
4. Proposed observational method was defined as the number of samples to achieve 90% degree of con-
solidation (Guo and Chu, 2017). Combining the definition of N90 and
Eq. (15) can be further processed as an observational method to Eq. (12) gives,
predict the ultimate primary consolidation settlement δult, the coeffi- ξ
cients of consolidation in vertical direction cv and in horizontal direc- ⎡1 − exp ⎛−2(νhv + 1) c v N90 Δt ⎞ ⎤ = 90%
H 2
⎣ ⎝ ⎠⎦ (25)
tion ch using monitored settlement data. The same procedure as used in
the Asaoka's method can be applied too, see Eqs. (1)–(4). This involves Combining Eqs. (21) and (25) gives the expression of N90 as,
selecting settlement δ1, δ2, …, δn from the settlement curve at a con-
ln(1 − 0.91/ ξ )
stant sampling interval Δt = tn – tn-1 and substituting δn, δn-1 and their N90 =
ln β (26)
responding time tn, tn-1 to Eq. (15):
ξ
c
δn = δult ⎡1 − exp ⎛−2(νhv + 1) v2 tn ⎞ ⎤ 5. Applications
⎣ ⎝ H ⎠⎦ (17)
ξ The effectiveness of the proposed procedure is examined by ap-
c
δn − 1 = δult ⎡1 − exp ⎛−2(νhv + 1) v2 tn − 1⎞ ⎤ plying the method to six well documented case histories. The first four
⎣ ⎝ H ⎠⎦ (18)
cases are for PVDs with fill surcharge and the last two are for PVDs with
Combining Eqs. (17) and (18) gives the relationship between the vacuum preloading.

627
W. Guo et al. Geotextiles and Geomembranes 46 (2018) 625–633

Table 2
Comparisons of ch predicted by proposed and Asaoka's observational methods.
Case Measured Asaoka's νhv Proposed Difference
ch (m2/s) method ch method ch between
(m2/s) (m2/s) proposed and
Asaoka's method
(%)

SkaEdeby-V13 0.22 0.274 38.9 0.248 −9.6


Orebro-G11 0.50 0.503 172 0.473 −6.1
CTE-SP2 1.50 1.403 12 1.123 −20.0
CA-PVD 2.51 0.955 2508 0.949 −0.6

Tianjin R-S1 3.40 4.038 1205 3.960 −1.92


Tianjin OT-S2 3.50 3.309 1205 3.239 −2.12

Fig. 3. Application of the proposed method to SkaEdeby-V13 case history (a)


monitored and predicted time-settlement curves, and (b) plots using proposed
method.

Table 1
Comparisons of δult predicted by proposed and Asaoka's observational methods.
Cases Observed Asaoka's νhv Proposed Difference
δult (m) method method δult between
δult (m) (m) proposed and
Asaoka's
methods (%)

SkaEdeby-V13 0.780 0.778 38.9 0.783 0.63


Orebro-G11 1.013 1.076 172 1.087 0.99
CTE-SP2 1.474 1.484 12 1.498 0.94
CA-PVD 2.295 2.354 2508 2.359 0.20
Fig. 4. Application of the proposed method to Orebro-G11 case history (a)
Tianjin R-S1 0.977 0.971 1205 0.974 0.31 monitored and predicted time-settlement curves, and (b) plots using proposed
Tianjin OT-S2 0.930 0.984 1205 0.987 0.30 method.

628
W. Guo et al. Geotextiles and Geomembranes 46 (2018) 625–633

Fig. 5. Application of the proposed method to CTE-SP2 case history (a) mon- Fig. 6. Application of the proposed method to CA-PVD case history (a) mon-
itored and predicted time-settlement curves, and (b) plots using proposed itored and predicted time-settlement curves, and (b) plots using proposed
method. method.

5.1. PVDs with fill surcharge analyze the monitored time-settlement curve at SkaEdeby-V13. The
settlement plots were selected from the monitoring data with the
5.1.1. Skå-Edeby test site, Sweden sampling interval Δt of 0.1 year. The time factor ratio in this case is
The first case is the Skå-Edeby test site for an airport construction in calculated as νhv = 38.9 using Eq. (10). The curve-fitting constant ξ of
Stockholm, Sweden, as presented by Hansbo (1960) and Holtz and 0.89 was obtained from Fig. 2. Linear regression of the plots indicates
Broms (1972). In this case, 27 kPa fill surcharge was used together with the slope of 0.9451 and intercept of 0.0417 which gave the N90 of 38.8
sand drains (dw = 0.18 m) with a triangular spacing of 1.5 m to con- and the predicted ultimate settlement δult of 0.783 m. The observed
solidate the normally consolidated soft clay. The clay stratum was ultimate settlement was 0.780 m. Using Asaoka's method as shown in
about 10-m-deep with drainage provided from both top and bottom. Eq. (2), the ultimate settlement would be 0.778 m. The ch and cv values
The monitored settlement data at Skå-Edeby V13 is selected and shown obtained from Fig. 3b were 0.247 m2/yr and 0.177 m2/yr, respectively.
in Fig. 3a. Oedometer tests indicate the average cv of 0.158 m2/yr. The These values were comparable with the average cv and ch given by
best-fitting pore pressure dissipation curve gives the average ch of Oedometer tests and the back calculated data. A comparison of ob-
0.22 m2/yr (Holtz and Broms, 1972). tained δult and ch values are given in Table 1 and Table 2, respectively. It
The curve-fitting using the proposed method is also shown in should be pointed out that same sampling interval are used in the two
Fig. 3a. It can be seen good agreement is achieved. Fig. 3b shows the observational methods (Δt = 1 yr).
results of using proposed method as the observational method to

629
W. Guo et al. Geotextiles and Geomembranes 46 (2018) 625–633

Fig. 7. Application of the proposed method to TianjinR-S1 case history (a) Fig. 8. Application of the proposed method to TianjinOT-S2 case history (a)
monitored and predicted time-settlement curves, and (b) plots using proposed monitored and predicted time-settlement curves, and (b) plots using proposed
method. method.

5.1.2. Örebro test site, Sweden 5.1.3. Central Expressway - Braddel Road Interchange, Singapore
The second case was a test site for a motorway constructed through The third case is a construction site for the Central Expressway and
Örebro, Sweden (Hansbo et al., 1981). The settlement data monitored Braddel Road Interchange in Singapore (Tan et al., 1985), where a
at Orebro-G11 is selected as shown Fig. 4a. The fill surcharge load 90 kPa fill surcharge was used together with PVDs (dw = 0.06 m)
applied was 40 kPa. Geodrains (dw = 0.065 m) were installed at trian- spaced at 1.3 m in a triangular grid. The soil was 6 m thick, doubly
gular spacing of 1.1 m to consolidate the 7-m-deep one-way drained drained soft organic clay. The average cv and ch of the organic clay are
light overconsolidated soft clay. The field ch value was not determined. approximately 1.0 and 1.5 m2/yr, respectively. The monitored settle-
The ch/cv ratio of 2.5 was assumed in the calculation (Tan, 1995). ment data at CTE-SP2 (Tan and Chew, 1996) is selected and shown
The curve-fitting using the proposed method is shown in Fig. 4a. Fig. 5a. The prediction using the proposed method is shown in Fig. 5b.
The ultimate settlement predicted using Fig. 4b was 1.087 m which was Comparisons of the ultimate settlement and ch predicted by the pro-
in good agreement with the settlement estimated based on the oed- posed method as well the Asaoka method are made in Tables 1 and 2.
ometer data. The cv and ch obtained from the proposed method was
0.189 m2/yr and 0.473 m2/yr, respectively. The cv value matched that
from oedometer tests which varied from 0.06 to 1.0 m2/yr with the 5.1.4. Changi east land reclamation project, Singapore
average value of 0.2 m2/yr. The δult and ch predicted from the Asaoka's The fourth case is the Changi East Reclamation Project, Singapore
method are also given and compared in Tables 1 and 2, respectively. (Choa, 1995; Bo et al., 2005; Chung et al., 2009; Chu et al., 2009). PVD
drains (93 × 4 mm) was installed to approximately 43 m with a square

630
W. Guo et al. Geotextiles and Geomembranes 46 (2018) 625–633

Fig. 9. Effects of νhv on the difference of the predicted results from the proposed
and Asaoka's methods.

drain spacing of 1.5 m cv varied from 0.5 to 2.0 m2/yr as measured by


oedometer tests. ch varied from 2.0 to 5.0 m2/yr as determined by the
dissipation test (CPTu) before the soil improvement (Chu et al., 2002;
Arulrajah et al., 2004b, 2004c; 2006; Bo et al., 2005). The ch value
generally increases with depth and is approximately 2.5 times higher
than that of cv. The monitored settlement data at CA-PVD is selected
and shown in Fig. 6a. The ultimate settlement and ch predicted using
the proposed method as shown in Fig. 6b as well using the Asaoka
method are compared with the field data in Tables 1 and 2.

5.2. PVDs with vacuum preloading

5.2.1. Road construction, China


The first case was at a construction site for a 364.5-m-long and 51-
m-wide road section leading to a container terminal at Tianjin Port,
China (Yan and Chu, 2003). Vacuum preloading of 80 kPa with PVDs
was adopted to improve the soft clay deposit. The PVDs
(dw = 67.5 mm) were installed at 1.0 m in a square grid to consolidate a
20 m thick, singly drained silty clay layer. The monitored settlement
data at section-1 (Tianjin R-S1) is selected in the study and shown in
Fig. 7a. The δult and ch were predicted using the proposed method in
Fig. 7b and compared with the field monitoring data as well those
Fig. 10. Effect of sampling interval Δt on the estimated (a) ultimate settlement
predicted using the Asaoka's method as shown in Tables 1 and 2, re- and (b) horizontal coefficient of consolidation.
spectively.

square grid. The applied total surcharge was 120 kPa including 80 kPa
5.2.2. Oil storage station, China
vacuum pressure and the 2 m thick surcharge materials and 0.3 m thick
The second case was at a construction site for a section of an oil
sand blanket. The monitored settlement data at section 2 (Tianjin OT-
storage station near the coast of Tianjin, China (Chu et al., 2000). The
S2) is selected and shown in Fig. 8a. Predictions of δult and ch using the
soil was very soft, had a high water content, and was still undergoing
proposed method are shown in Fig. 8b and comparisons of the predicted
consolidation with its ch in the range of 3.5–14.8 m2/yr. The PVDs
values and field data as well as the predictions from the Asaoka's
(dw = 67.5 mm) were installed to 20 m deep at a spacing of 1.0 m in a

Table 3
Effects of sampling interval on the proposed and Asaoka's observational methods.
Δt (yr) Asaoka's Method Proposed Method Difference

β0 β1 δult ch j90 α β δult ch N90 δult ch


(m) (m2/yr) (m) (m2/yr) (%) (%)

0.2 0.250 0.772 1.096 0.461 9 0.246 0.778 1.100 0.445 9 0.38 −3.60
0.1 0.136 0.876 1.092 0.472 17 0.132 0.881 1.098 0.450 18 0.61 −4.76
0.05 0.073 0.932 1.076 0.503 33 0.070 0.936 1.094 0.466 34 1.61 −7.22
0.025 0.038 0.965 1.073 0.514 64 0.036 0.967 1.083 0.480 67 0.88 −6.64
0.0125 0.019 0.982 1.073 0.509 129 0.018 0.983 1.087 0.475 134 1.26 −6.72

631
W. Guo et al. Geotextiles and Geomembranes 46 (2018) 625–633

Table 4
Effects of data range on the proposed and Asaoka's observational methods.
U Asaoka's Method Proposed Method Difference
(%)
β0 β1 δult ch j90 α β δult ch N90 Δδult Δch
(m) (m2/yr) (m) (m2/yr) (%) (%)

31.0 0.102 0.792 0.492 1.663 10 0.095 0.804 0.502 1.544 10 2.10 −7.15
42.7 0.088 0.875 0.697 0.956 17 0.081 0.886 0.726 0.856 19 4.12 −10.43
49.7 0.082 0.899 0.812 0.758 22 0.076 0.909 0.850 0.673 24 4.67 −11.23
60.5 0.074 0.930 1.057 0.515 32 0.069 0.939 1.128 0.445 36 6.64 −13.68
70.2 0.073 0.932 1.078 0.500 33 0.069 0.939 1.124 0.446 36 4.26 −10.87
80.8 0.072 0.934 1.097 0.487 34 0.069 0.939 1.126 0.445 36 2.65 −8.66
90.6 0.073 0.932 1.080 0.500 33 0.070 0.936 1.094 0.466 34 1.28 −6.65

method are given in Tables 1 and 2, respectively.

6. Discussions and limitations

The comparisons made in Table 1 indicate that the ultimate con-


solidation settlement δult predicted using the proposed method is con-
sistently higher than that using the Asaoka's method although the dif-
ferences are generally rather small. In other words, either the Asaoka or
the proposed method can be used to predict the ultimate consolidation
settlement given the interval is selected properly. On the other hand,
the difference in the ch predicted using the two methods can be relative
large depending on the time factor ratio νhv. The differences in the δult
and ch predicted using the predicted and the Asaoka's methods (as given
in Tables 1 and 2) are plotted in Fig. 9. It can be seen that the difference
in the prediction of ch is dependent on the νvh value. When νvh is less
than 100, the effect of vertical drainage becomes relatively significant.
The proposed method is able to consider both the vertical and hor-
izontal drainage more conveniently and thus give a better prediction
when the effect of vertical drainage needs to be taken into considera-
tion.
The accuracy of the values of δult and ch predicted either by the
proposed or the Asaoka's method is dependent on the selection of the
sample interval. The monitored data at Orebro-G11 is used to in-
vestigate the effect of sample interval Δt on the predictions of the
proposed method. The predicted results of using proposed method are
summarized in Table 3 and plotted in Fig. 10. It can be seen that the
selection of Δt does influence the predicted results. The value of δult
decreases and ch increases as the number of samples to achieve U90, N90,
increases. The predicted results become constant when N90 is greater
than 30. Fig. 10 also shows that the δult predicted from the proposed
method is always higher and ch always lower than that from the
Asaoka's method.
It should be pointed that the values of δult and ch predicted by either
the proposed or the Asaoka's method will only be reliable when the
settlement data for a soil layers having a degree of consolidation of no
less than 60%. Using the settlement data for the Orebro-G11 case as an
example, the settlement data within different degree of consolidation U
ranges (31, 42.7, 49.7%, …) were used to predict the values of δult and
ch using the proposed and the Asaoka's method respectively. The degree
of consolidation was determined as the ratio of settlement at con-
solidation time t to the ultimate primary consolidation settlement of
1.088 m which is predicted in Fig. 4b. The predicted data are tabulated
in Table 4. The values of δult and ch are also plotted versus degree of
consolidation in Fig. 11. It is apparent in Fig. 11 that the predicted δult
and ch values are only reliable when the settlement data used are col-
Fig. 11. Effect of data range on the estimated (a) ultimate settlement and (b)
lected for a soil layer having achieved 60% or more degree of con-
horizontal coefficient of consolidation.
solidation.

7. Conclusions

The average degree of consolidation is often used as a design spe-


cification to evaluate the degree of the soil improvement achieved. A

632
W. Guo et al. Geotextiles and Geomembranes 46 (2018) 625–633

formula is proposed in this paper to simulate the degree of consolida- and back-analysis by Asaoka and the hyperbolic methods. Aust. GeoMech. 38 (2),
tion versus time factor relationship for consolidation with combined 29–37.
Arulrajah, A., Nikraz, H., Bo, M.W., 2004a. Observational method of assessing improve-
vertical and horizontal flow. The curve given by this formula fits well ment of marine clay. ICE- Ground Improv 8 (4), 151–169.
with the theoretical solutions given by Terzaghi, Hansbo and Carrillo Arulrajah, A., Bo, M.W., Nikraz, H., Hashim, R., 2004b. Piezocone dissipation testing of
with the regression coefficient R2 of larger than 0.9996 and an error of Singapore marine clay at Changi. Geotech. Eng. J. SEAGS AGSSEA 35 (3), 119–126.
Arulrajah, A., Nikraz, H., Bo, M.W., 2004c. Factors affecting field instrumentation as-
less than 1.2%. sessment of marine clay treated with prefabricated vertical drains. Geotex.
A modified Asaoka's observational method is proposed to predict Geomembr 22 (5), 415–437.
the ultimate settlement and the coefficient of consolidation using set- Arulrajah, A., Nikraz, H., Bo, M.W., Hashim, R., 2006. In-Situ pore water pressure dis-
sipation testing of marine clay under reclamation fills. Geotech. Geol. Eng. 24 (1),
tlement data monitored in the field. The effectiveness of this proposed 29–43.
method was verified using several well-documented case histories. Asaoka, A., 1978. Observational procedure of settlement prediction. Soils Found. 18 (4),
Comparisons between the proposed method and the Asaoka's method 87–101.
Bo, M.W., Chu, J., Choa, V., 2005. The Changi east reclamation project in Singapore. In:
indicate that the proposed method will give a less than 1.0% higher
Indraratna, B., Chu, J. (Eds.), Ground Improvement – Case Histories. Elsevier,
ultimate settlement than that by the Asaoka's method. The main dif- Elsevier Geo-Engineering Book Series, Ch. 9. Amsterdam, the Netherlands, pp.
ference is in the ch value. The maximum difference in ch can be 20% for 247–278.
the analysed case histories. The proposed method is able to predict the Carrillo, N., 1942. Simple two and three dimensional cases in the theory of consolidation
of soils. J. Math. Phys. 21, 1–5.
ch value with the consideration of both vertical and horizontal flow Chai, J.C., Shen, S., Miura, N., Bergado, D.T., 2001. Simple method of modeling PVD-
through the ratio of time factor νhv. When νhv is less than 100, the effect improved subsoil. J. Geotech. Geoenviron. Eng. 127 (11), 965–972.
of vertical drainage becomes relatively significant. Chai, J., Zhou, Y., 2018. Method for considering the effect of nonuniform consolidation.
Int. J. GeoMech. 18 (2), 04017151.
This study also pointed out that the prediction using either the Chapman, D.G., 1961. Statistical problems in population dynamics of exploited fisheries
proposed or the Asaoka's method will only be reliable when the number populations. In: Neyman, J. (Ed.), Proceedings of the 4th Berkeley symposium on
of samples to achieve 90% degree of consolidation, N90, is greater than mathematical statistics and prob- ability, Berkeley, CA, USA. University of California
Press, Berkeley, CA, USA, pp. 153–186.
30 and the settlement data are collected for a soil layer having achieved Choa, V., 1995. Changi east reclamation project. In: Proceedings of the International
at least 60% degree of consolidation. Symposium on Compression and Consolidation of Clayey Soils, IS-hiroshima'95.
Hiroshima, Japan, pp. 1005–1017.
Chu, J., Bo, M.W., Arulrajah, A., 2009. Soil improvement works for an offshore land
Notations reclamation project. Proc. ICE – Geotec. Engng. 162 (1), 21–32.
Chu, J., Bo, M., Chang, M., Choa, V., 2002. Consolidation and permeability properties of
ch coefficient of consolidation in horizontal (or radial) direction Singapore marine clay. J. Geotech. Geoenviron. Eng. 128 (9), 724–732.
Chu, J., Yan, S.W., Yang, H., 2000. Soil improvement by vacuum preloading method for
cv coefficient of consolidation in vertical direction
an oil storage station. Geotechnique 50 (6), 625–632.
De diameter of an equivalent soil cylinder influenced by vertical Chung, S.G., Kweon, H.J., Jang, W.Y., 2014. Observational method for field performance
drains of prefabricated vertical drains. Geotex. Geomembr 42 (4), 405–416.
ds diameter of smear zone Chung, S., Lee, N., Kim, S., 2009. Hyperbolic method for prediction of prefabricated
vertical drains performance. J. Geotech. Geoenviron. Eng. 135 (10), 1519–1528.
dw equivalent diameter of the vertical drain Edil, T.B., Fox, P.J., Lan, L.T., 1991. Observational procedure for settlement of peat. In:
H vertical drainage length Proceedings International Conference on Geotechnical Engineering for Coastal
j90 number of samples to achieve U90 in the Asaoka's method Development, Geo-coast’91, Yokohama, Japan, pp. 165–170.
Guo, W., Chu, J., 2017. New observational method for prediction of one-dimensional
kh coefficient of permeability in horizontal (or radial) direction consolidation settlement. Geotechnique 67 (6), 516–522.
of natural soil Guo, W., Chu, J., Nie, W., 2018. Design chart for the modified hyperbolic method. Soils
ks coefficient of permeability in smear zone Found. https://doi.org/10.1016/j.sandf.2018.02.014.
Hansbo, S., 1960. Consolidation of Clay with Special Reference to the Influence of
l drainage length Vertical Sand Drains. Swedish Geotechnical Institute, Stockholm, Sweden.
m drains spacing ratio m = De/dw Hansbo, S., 1981. Consolidation of fine-grained soils by prefabricated drains. In:
n integer, n = 0, 1 … Proceedings of 10th International Conference on Soil Mechanics and Foundation
Engineering, Stockholm, vol. 3. Balkema, Rotterdam, pp. 677–682.
N90 number of samples to achieve U90 in the proposed method
Hansbo, S., Jamiolkowski, M., Kok, L., 1981. Consolidation by vertical drains.
qw discharge capacity of PVD Geotechnique 3 (1), 45–66.
S spacing of the vertical drain Holtz, R.D., Broms, B., 1972. Long-term loading tests at Ska-edeby, Sweden. In:
Proceedings ASCE Specialty Conference on Earth and Earth Supported Structures.
t consolidation time
Purdue Lafayette Indiana, pp. 435–464.
Th non-dimensional horizontal (radial) time factor Low, B.K., 2003. Theories, computations, and design procedures involving vertical drains.
Tv non-dimensional vertical time factor In: Bo, M.W., Chu, J., Low, B.K., Choa, V. (Eds.), Soil Improvement: Prefabricated
Uh degree of horizontal consolidation, % Vertical Drain Techniques. Thomson Learning, Singapore, pp. 5–56 Ch. 2.
Ratkowsky, D.A., 1990. Handbook of Nonlinear Regression Models. Marcel Dekker, New
Uv degree of vertical consolidation, % York, NY, USA.
Uvh average degree of consolidation for combined vertical and Richards, F.J., 1959. A flexible growth function for empirical use. J. Expl. Bot. 10 (2),
horizontal consolidation 290–300.
Sridharan, A., Sreepada Rao, A., 1981. Rectangular hyperbolic fitting method for one-
β0 intercept in Asaoka's plot dimensional consolidation. Geotech. Test J. 4 (4), 161–168.
β1 slope in Asaoka's plot Tan, S.A., 1995. Validation of hyperbolic method for settlement in clays with vertical
δ consolidation settlement δ1, δ2, …, δn settlement at time t1, drains. Soils Found. 35 (1), 101–113.
Tan, S.A., Chew, S.H., 1996. Comparison of the hyperbolic and Asaoka observational
t2, …, tn method of monitoring consolidation with vertical drains. Soils Found. 36 (3), 31–42.
Δch difference of ch obtained from two methods Tan, S.B., Tan, S.L., Yang, K.S., Chin, Y.K., 1985. Soil improvement methods in Singapore.
Δt sampling interval In: Proceeding of 3rd International Geotechnical Seminar at Nanyang Technological
Institute, Singapore, pp. 249–272.
δult ultimate consolidation settlement
Tan, T.S., Inoue, T., Lee, S.L., 1991. Hyperbolic method for consolidation analysis. J.
μ factor to account for drains spacing Geotech. Geoenviron. Eng. 117 (1), 1723–1737.
νhv ratio of time factor in horizontal and vertical direction Terzaghi, K., Peck, R.B., Mesri, G., 1996. Soil Mechanics in Engineering Practice, third ed.
Wiley-Interscience, John Wiley & Sons, New York, NY.
ξ curve fitting constant
Yan, S.W., Chu, J., 2003. Soil improvement for a road using the vacuum preloading
method. Proc. ICE - Ground Improv. 7 (4), 165–172.
References

Arulrajah, A., Nikraz, H., Bo, M.W., 2003. Factors affecting field settlement assessment

633

You might also like