You are on page 1of 10

Engineering Geology 267 (2020) 105502

Contents lists available at ScienceDirect

Engineering Geology
journal homepage: www.elsevier.com/locate/enggeo

Estimation of the soil-water characteristic curve from the grain size T


distribution of coarse-grained soils

Qian Zhaia, , Harianto Rahardjob, Alfrendo Satyanagab, Guoliang Daia
a
Key Laboratory of Concrete and Prestressed Concrete Structures of Ministry of Education, Southeast University,Nanjing 210096, China
b
School of Civil and Environmental Engineering, Nanyang Technological University, Block N1, Nanyang Ave., Singapore 639798, Singapore

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

Keywords: Coarse-grained soil, such as aggregate and coarse sand, is commonly selected to construct a capillary barrier
Soil-water characteristic curve system (CBS). The barrier effect in the CBS is mainly dependent on the hydraulic properties of the coarse-grained
Air-entry value soils. It is noted that direct methods (such as experimental measurement) for the determination of the hydraulic
Grain-size distribution function properties (including the soil-water characteristic curve (SWCC) and coefficient of permeability) of soil are time-
Random distribution
consuming and costly. As a result, indirect methods for the determination of the SWCC and permeability function
of soil are commonly adopted by researchers and engineers. In this paper, a new method is proposed for the
estimation of the SWCC from the grain-size distribution (GSD) of coarse-grained soils. The proposed method is
based on the assumptions that the soil does not experience any volume change and the hysteresis of the SWCC is
ignored. A contact angle larger than zero is adopted in this study. The results computed from the proposed model
agree well with the experimental data from published studies.

1. Introduction Satyanaga et al. (2017). As is known, the accuracy of the regression


method is quite dependent on the database used for the analysis. As a
Coarse-grained soil is a cohesionless soil with a low permeability result, the correlation model proposed in the regression method is
and is commonly found in an unsaturated condition. As a result, coarse- commonly applicable to a specific condition. Different models can be
grained soil is commonly used as the barrier layer in a capillary barrier proposed for different soil types, different zones and different analyzers.
system (CBS). These systems are often used for the remediation of The fitting parameter estimation methods include the works from Rawls
rainfall-induced slope failures. The hydraulic properties of coarse- and Brakensiek (1989), Scheinost et al. (1997), Zapata et al. (2000),
grained soil are essential in the construction of a high-performance CBS. and Perera et al. (2005). In this estimation method, the fitting para-
It takes time to determine the soil-water characteristic curve (SWCC) meters are estimated from the index properties of the soil using re-
from experimental measurements. Due to the high cost and tight con- gression analyses. Again, the accuracy of the fitting parameter estima-
struction schedules, most of these time-consuming experimental tests tion method is still dependent on the database used for the analyses,
are not actually carried out. Therefore, indirect methods, such as esti- since the regression analysis is adopted. The physico-empirical methods
mation techniques using the soil index properties, are commonly include the works from Arya and Paris (1981), Tyler and Wheatcraft
adopted to estimate the SWCC for the preliminary design of the CBS. (1989), Fredlund et al., 2002, Hwang and Powers (2003), Chin et al.
In the past few decades, a number of estimation methods have been (2010), Ferraro et al. (2017) and Sweijen et al. (2017). The physico-
proposed to indirectly obtain the SWCC from the grain-size distribution empirical method adopts the assumption, as explained by Arya and
(GSD). Fredlund et al., 2002 categorized these methods into three Paris (1981), that the radius of the pores in soil can be correlated with
groups, including (i) regression analyses to estimate the water content the size of the particles. In other words, the GSD can be considered to be
at different soil suctions; (ii) estimation of the fitting parameters from analogous to the pore-size distribution.
the index properties of soil; and (iii) physico-empirical models to esti- Likos and Jaafar (2013) and Sweijen et al. (2017) indicated that the
mate the SWCC from the GSD. SWCC of soil was dependent on both the geometry and dimensions of
The regression methods for the estimation of the SWCC from the the pore space. Khoshghalb et al. (2015) stated that different pore size
GSD include works by Gupta and Larson (1979), Vereecken et al. distributions can be obtained from a single GSD for a soil with different
(1989), Schaap and Leij (1998), Meskini-Wishkaee et al. (2014) and void ratios. Sweijen et al. (2017) also indicated that the packing density


Corresponding author.
E-mail addresses: 101012332@seu.edu.cn (Q. Zhai), chrahardjo@ntu.edu.sg (H. Rahardjo), alfrendo@ntu.edu.sg (A. Satyanaga), daigl@seu.edu.cn (G. Dai).

https://doi.org/10.1016/j.enggeo.2020.105502
Received 29 July 2019; Received in revised form 20 January 2020; Accepted 20 January 2020
Available online 22 January 2020
0013-7952/ © 2020 Published by Elsevier B.V.
Q. Zhai, et al. Engineering Geology 267 (2020) 105502

had a significant effect on the estimation of the SWCC from the GSD. By The circles in Fig. 1(b) are further divided into different fractions by
adopting a κ-gamma pore volume distribution function for the ran- connecting the center points of the circles, as shown in Fig. 2 (a) and
domly assembled monodisperse granular materials, Ferraro et al. (b). Taking one from the divided fractions, the representative elemental
(2017) proposed an analytical derivation of the SWCC for granular triangle consists of three circles and the void in between these circles.
soils. As the soil particles are randomly distributed, the shape of the re-
When the soil is in an unsaturated condition, the soil matrix is a presentative elemental triangle varies with the changes in the radii of ri,
four-phase mixture, which includes the solid phase, water phase, air rj and rk, as shown in Fig. 2(c). As a result, the total volume of the void
phase and air-water interface (Fredlund and Rahardjo, 1993, Fredlund in the soil element can be represented by the summation of the void
et al., 2012). The air-water interface is commonly referred to as the area in all these representative elemental triangles. The void area of all
meniscus in a two-dimensional (2D) element. At the contact point be- the representative elemental triangles can be obtained by using a sta-
tween the air-water interface and a soil particle, the angle between the tistical method and considering a random positioning of the soil par-
tangent lines of the meniscus and the particle surface defines the con- ticles in the corners of the triangles.
tact angle α. The works from Goebel et al. (2004), Woche et al. (2005), The area of the void between particles in a representative elemental
Ramirez-Flores et al. (2008), Goebel et al. (2012) and Ganz et al. (2013) fraction can be obtained by using the area of triangle minus the area of
showed that the contact angle of the coarse-grained soil is not always sectors, as shown in Fig. 2(d). It is noted that the triangle is created by
equal to zero. Different SWCCs can be obtained if different contact three circles, and the three sides of the triangle are functions of the radii
angles are used in the estimation. As a result, two cases, including (i) a of soil particles. Defining the three sides of the triangle as a, b and c and
contact angle always equal to zero and (ii) a contact angle greater than the radii of the particles as ri, rj, and rk, the area of the triangle then can
zero, are considered for the estimation of the SWCC from the GSD in be calculated by using Heron's formula as follows:
this study.
s= p (p − a)(p − b)(p − c ) (1)
In this paper, a new method for the estimation of the SWCC from the
GSD is proposed for coarse-grained soils. The proposed method assumes where s is the area of triangle, a, b and c are the lengths of the sides of
that there is no soil volume change with an increase in soil suction. The the triangle, a = ri + rj, b = rj + rk, c = rk + ri, p is half of the
equations proposed in this study are derived from the geometrical re- perimeter, and p = .5(a + b + c).
lationships between the diameters of the soil particles and the radii of The area occupied by the soil particles in the representative ele-
the meniscus. Descartes' theorem of circles is adopted to determine the mental triangle can be calculated by the summation of the areas of
air-entry value of the representative elemental triangle. The estimated sectors. Defining the angle between ri and ri as α, rj and rj as β and rk and
results from the proposed method are verified with experimental data rk as γ as shown in Fig. 3, the angles α, β and γ can be calculated by
from published studies. using the theory of trigonometric functions as follows:

2. Theory 2s
α = arcsin ⎛ ⎞
⎝ bc ⎠ (2)
The theoretical background of the estimation of the SWCC from the 2s
GSD is explained in this section. In the proposed method, it is assumed β = arcsin ⎛ ⎞
⎝ ac ⎠ (3)
that the soil volume change during the variation in the soil suction is
negligible. In other words, the soil skeleton is assumed to be rigid 2s
γ = arcsin ⎛ ⎞
throughout this study; this is a reasonable assumption for coarse- ⎝ ab ⎠ (4)
grained soils with a high relative density, Dr, where Dr = (emax-e)/(emax-
Consequently, the area of the void in the representative elemental
emin) × 100%, where e is the void ratio, emax is the maximum void ratio,
triangle and the total area of voids in the soil element can be calculated
emin is the minimum void ratio. The soil particles are simplified as a
by using Eqs. (5) and (6), respectively.
series of spheres with different sizes. These spheres are randomly dis-
i
tributed in the soil matrix. The random distribution of spheres in a 3D Avoid = s − 0.5αrk2 − 0.5βri2 − 0.5γr j2 (5)
space is complex and it may not be possible for them to be represented
N
by one mathematical equation. As a result, the spheres are further Total i
simplified to circles with different radii in the 2D plane so that the Avoid = ∑ Avoid
i=1 (6)
geometrical distribution of the circles can be calculated. For the pur-
pose of simplicity, only the drying process is considered in this study, where Aivoid th
is the void area in the i elemental triangle, is the ATotal
void
which means that the hysteresis of the SWCC is ignored. In the re- total area of voids in the soil element and N is the total number of
presentative elemental triangle, the air can enter the triangle only when elemental triangles.
the radius of the meniscus is smaller than the radius of the pore, which
is tangent to the three sectors in the triangle. Once the air enters the 2.2. Grain size distribution and occurrence of the representative elemental
pore, the meniscus moves from the center towards the corner of the triangle
pores with an increase in soil suction. Consequently, the degree of sa-
turation, S, of the soil is calculated from the area of water remaining in The positions occupied by soil particles with radii of ri, rj and rk are
the representative elemental triangle. defined as position 1, position 2, and position 3, as illustrated in Fig. 4.
These three positions can be filled by a particle with any radius, and the
2.1. Simplification of the soil skeleton filling process has no sequence. In other words, soil particles of any
radius randomly fill the three positions in the representative elemental
Soil particles are simplified as a series of spheres, and these spheres triangle. Dividing the GSD into different segments, each segment is
are in contact with each other because the soil skeleton is in- represented by the effective grain size (i.e., d95, d85, −–) as shown in
compressible. To simplify the problem further, only the 2D element, as Fig. 5. In this procedure of division, each segment has the same weight
shown in Fig. 1(a), rather than the three-dimensional (3D) element is because the GSD is plotted based on the percentage of passing by mass.
considered in this study, and the spheres are represented by circles. However, the probability of a particle with radius r filling a position in
Consequently, a simplified model of the soil skeleton is illustrated in the representative elemental triangle is a function of the ratio between
Fig. 1 (b) by considering that the soil particles are randomly distributed the number of soil particles with a radius of r and the total number of all
in the soil matrix. the soil particles with different radii. In a conventional GSD curve, the

2
Q. Zhai, et al. Engineering Geology 267 (2020) 105502

(a) Section of the soil element (b) simplified model of the soil skeleton
Fig. 1. Illustration of the soil skeleton.

(a) The soil element (b) Division of the representative elemental triangle

(c) Enlarged view of divided triangles (d) selected representative elemental triangle
Fig. 2. Illustration of division of soil element into fractions.

Fig. 3. Geometrical relationship between the triangle and radii of particles.

Fig. 4. Illustration of positions of particles in a fraction.

3
Q. Zhai, et al. Engineering Geology 267 (2020) 105502

Fig. 5. Illustration of the segments of GSD.

(a) Illustration of air enter the pore, (b) Illustration of movement of meniscus
Fig. 6. Illustration of air entering the pore and movement of meniscus in the pore.

Fig. 7. Division of the triangle into sub-triangles.

(a) Unsaturated condition (b) Saturated condition


Fig. 8. Illustration of water remained in the pores.

4
Q. Zhai, et al. Engineering Geology 267 (2020) 105502

Fig. 9. Illustration of the equivalent radius of the tube and the radii of the meniscus by considering different contact angles.

Fig. 10. Illustration of the SWCCs of the soil by considering different contact angles.

Fig. 11. Illustration of the GSD of soils.

y-axis is defined as the percentage of passing by mass of a certain particles with a smaller diameter is much greater than that of the soil
particle size but not the number of soil particles. As a result, the mass of particles with a larger diameter. As a result, the total number of soil
each segment needs to be converted into the number of particles. If particles with different diameters can be obtained by the summation of
there are A numbers of particles with diameter of d95, then the number the number of particles representing different segments as follows:
of particles with diameter of d85 will be (d95/d85)2A. The soil particle
with a small diameter has a higher probability of filling the corners of
the representative elemental triangle because the number of the soil

5
Q. Zhai, et al. Engineering Geology 267 (2020) 105502

Fig. 12. Comparison between the estimated results and the measured data for the coarse-grained soils.

2 2 2 2 d65 2 d65 2
⎡ ⎛ d85 ⎞ + ⎛ d75 ⎞ + ⎛ d65 ⎞ + ⎛ d55 ⎞
⎢1 + d ⎜ ⎟

d 95 ⎠
⎜ ⎟

d 95 ⎠

d 95 ⎠
⎟ ⎜

⎥ ⎟
p65 =
A ( )
d95
=
( )
d95

Atotal =A ⎢ ⎝ 95 ⎠ ⎝ ⎝ ⎝ ⎥ Atotal 2 2 2 2
⎢ 2 2 2 2 2⎥ ⎡ ⎛ d85 ⎞ + ⎛ d75 ⎞ + ⎛ d65 ⎞ + ⎛ d55 ⎞ ⎤
⎢+ ⎛
d 45 ⎞ +⎛ d35 ⎞ +⎛ d25 ⎞ +⎛ d15 ⎞ +⎛ d 5 ⎞⎥ ⎢1 + d ⎜ ⎟

d
⎜ ⎟

d

d
⎟ ⎜
⎥⎟

⎜ ⎟ ⎜ ⎟ ⎜ ⎟ ⎜ ⎟ ⎜ ⎟
⎢ ⎝ 95 ⎠ ⎝ 95 ⎠ ⎝ 95 ⎠ ⎝ 95 ⎠ ⎥
⎣ ⎝ d 95 ⎠ ⎝ d 95 ⎠ ⎝ d 95 ⎠ ⎝ d 95 ⎠ ⎝ d 95 ⎠ ⎦ (7) ⎢ d45 ⎞
2
d35 ⎞
2
d25 ⎞
2
d15 ⎞
2
d5 ⎞ ⎥
2
⎢+ ⎛⎜ ⎟+ ⎛ ⎜ + ⎛
⎟ ⎜ + ⎛ ⎟ +⎜
⎛ ⎟ ⎥ ⎜ ⎟

The probability of the particles with a diameter of d65 filling the ⎣ ⎝ d 95 ⎠ ⎝ d 95 ⎠ ⎝ d 95 ⎠ ⎝ d 95 ⎠ ⎝ d 95 ⎠ ⎦


corners of the representative elemental triangle can be calculated by (8)
dividing the number of particles with a diameter of d65 by the total
number of all the particles with different diameters as follows: It is observed that the occurrence of the representative elemental
triangle is a function of the probability that particles fill the corners of
the triangle. Therefore, Eq. (6) for the calculation of the total area of
voids in the soil element needs to be corrected by considering the
probability of the occurrence of the representative elemental triangle,

6
Q. Zhai, et al. Engineering Geology 267 (2020) 105502

Fig. 12. (continued)

as illustrated in Eq. (9). shown in Fig. 6(a). Based on Descartes' theorem of circles, the radius of
N N
r can be calculated as follows:
Total i
Avoid = ∑ Avoid = ∑ pi pj pk (s − 0.5αrk2 − 0.5βri2 − 0.5γr j2) 1 1 1 1 11 1 1 11
i=1 i=1 (9) = + + +2 + +
r ri rj rk ri r j r j rk rk ri (10)
where pi, pj, pk are the probabilities of particles with radii of ri, rj, and rk
filling the corners of the triangle. If the equivalent radius of the applied matric suction (using the
Young-Laplace equation) is larger than r, then air cannot enter the
2.3. Air-entry value of the representative elemental triangle and water triangle. On the other hand, if the equivalent radius is smaller than r, air
retention characteristic of the elemental triangle will enter the triangle and create the meniscus between soil particles as
shown in Fig. 6(b). With an increase in the matric suction, the meniscus
As shown in Fig. 6(a), air can only enter the pore when the air-water moves from the center towards the corners of the pore. The area be-
interface, which is represented by the circle in Fig. 6(a), is tangent to all tween the meniscus and the corners of the pore defines the water sto-
three of these circles. In other words, the radius of the circle is r, as rage at the applied soil suction.

7
Q. Zhai, et al. Engineering Geology 267 (2020) 105502

Fig. 12. (continued)

Table 1
Effective particle sizes of soils.
Effective sizes (mm) Medium sand Fine sand Fine RCA Coarse RAP Fine RAP Coarse RCA

d95 3.08 0.69 11.16 13.40 6.45 24.21


d85 2.02 0.51 5.28 12.68 4.16 19.37
d75 1.66 0.44 3.73 12.01 3.28 16.89
d65 1.38 0.38 2.68 11.16 2.49 15.10
d55 1.12 0.33 1.97 10.76 1.97 13.66
d45 0.90 0.28 1.44 9.82 1.58 12.67
d35 0.71 0.25 1.06 8.96 1.27 11.61
d25 0.54 0.22 0.66 8.18 0.96 10.51
d15 0.37 0.19 0.29 7.33 0.64 8.94
d5 0.19 0.16 0.08 6.57 0.25 7.24

8
Q. Zhai, et al. Engineering Geology 267 (2020) 105502

To calculate the water area between the meniscus and the corner of Table 1.
the pore, the representative elemental triangle is further divided into The effective particle sizes shown in Table 1 are used to estimate the
three sub-triangles. The water area in each sub-triangle is defined by SWCC of these soils using the spreadsheet created in this study. The
the radii of the soil particles and the radius of the meniscus. The radius estimated results are illustrated in Fig. 12.
of the meniscus, r, is calculated from the applied suction using the Fig. 12 indicates that the estimated results from the GSD of the
Young-Laplace equation. Consequently, the water area in a sub-triangle different sandy soils agree well with the experimental data if a contact
can be calculated by using the same theories as explained in Section angle greater than zero is considered. If the contact angle is assumed to
2.1. be equal to zero, then the estimated results may overestimate the de-
When the soil suction of ψ is applied to the soil sample, some tri- gree of saturation at a given matric suction. In addition, it is interesting
angles are in the unsaturated condition, as shown in Fig. 8(a), and some to observe that soils with the same GSD have different SWCCs as in the
are in the fully saturated condition, as shown in Fig. 8(b). As a result, case of the coarse RCA shown in Fig. 12 (f). It is noted that the dif-
the degree of saturation, S, can be obtained from Eq. (11) as follows: ference between the two SWCCs is mainly due to the coating on the
Am aggregates. Coating the soil particles with oil or wax changes the con-
S= tact angles between the air-water interface and the soil particles.
Atotal (11)
Rahardjo et al. (2017) indicated that coating the aggregates with oil or
where Am is the total water area in all the representative elemental wax increased the contact angle between the meniscus and soil parti-
triangles at a suction of ψm, and Atotal is the total water area in all the cles. The results shown in Fig. 12 (f) provide the experimental evidence
representative elemental triangles at a fully saturated condition. for the observation in this study, that is, a smaller AEV can be obtained
At a suction of ψm, the equivalent radius of the meniscus is calcu- if a higher contact angle is used. It should be noted that the deviation
lated as rm. If rm is greater than the radius of the biggest circle, r, as between the estimated results and the measured data may not be due
shown in Fig. 6(a), then air cannot enter this triangle, and the water purely to the variation in the contact angles. The different particle
area in this triangle can be calculated from Eq. (5). On the other hand, if shapes and the variation in the void ratios may also lead to different
rm is less than r, then air has entered the triangle. The water area in this SWCCs. However, there were no data or valuable information reported
triangle needs to be calculated from the summation of three sub-tri- on these two factors for these sandy soils, and it is difficult to quantify
angles as shown in Fig. 7 using Eq. (5) by replacing ri, rj or rk with rm. their effects on the variation in the SWCC. The contact angles estimated
Consequently, an electronic spreadsheet incorporating Eq. (5) can be for the aggregates are between [30°, 75°] as shown in Fig. 12, which is
created to estimate the SWCC from a given GSD. consistent with the experimental data reported by Geobel et al. (2004)
It should be noted that the soil-water characteristic curve explained and Ramirez-Flores et al. (2008).
in this section only considers the capillary water. The works of Plaster Fig. 12 also indicates that the estimated results for coarse RCA un-
(2009) and Zhai et al. (2019a) indicated that there were three forms of derestimate the degree of saturation at a given matric suction even for
water including gravity water, capillary water and hygroscopic water in the case of the contact angle being equal to zero. As explained in
the soil matrix. The capillary water dominates in sandy soil while the Section 2.3, only the capillary water is considered, and the adsorbed
hygroscopic water dominates in clayey soil. Therefore, the theories water, which is attached around the soil particles, is not considered in
explained in this study are only applicable to sandy soil. this study. In a normal coarse-grained soil, the capillary water dom-
inates the soil matrix, and the proposed method is applicable to the
2.4. Capillary model and relationship between radius of meniscus and normal coarse-grained soil. However, the coarse RCA is manufactured
diameter of the capillary tube from the crushing of recycled concrete, and cement grout is attached
around the aggregates. Water can be adsorbed around the aggregates
To explain the water flow in soil, the capillary model (Fredlund due to the attached cement grout. As a result, the measured degree of
et al., 2012, Zhai and Rahardjo, 2015, Zhai et al., 2019b) is commonly saturation may be higher than the estimated results from the proposed
used to convert the applied matric suction into the equivalent pore size method. Once the coarse RCA is coated with oil or wax, the cement
or the radius of the capillary tube, d. As shown in Fig. 9, if the contact grout attached on the aggregates is separated with water, and the ad-
angle α is assumed to be equal to zero, then the radius of the meniscus r sorbed water due to the cement grout is eliminated. As a result, the
will be equal to the radius of the tube d. On the other hand, if the results from the proposed method agree well with the experimental
contact angle α is assumed to be greater than zero, then r will be greater data for the coarse RCA coated with oil and wax, as shown in Fig. 12(f).
than d. As a result, a smaller air-entry value (AEV) can be obtained if a
higher contact angle is adopted in the calculation. 4. Conclusions
Consequently, the SWCC of the soil with a higher contact angle α2
can be horizontally shifted from the SWCC of the soil with a smaller In this study, a new method for the estimation of the SWCC from
contact angle α1. The suction ψ2 on the curve with contact angle α2 and GSD is proposed for coarse-grained soils. In the proposed method, a
suction ψ1 on the curve with contact angle α1, as shown in Fig. 10, can rigid soil skeleton is assumed, soil particles are simplified as spheres in
be correlated by the Eq. (12) as follows: a 2D setting and no hysteresis of the SWCC is considered. The equations
ψ1 cos α1 proposed in this study are derived from the geometrical relationships
= between both the radii of the soil particles and the meniscus. Descartes'
ψ2 cos α2 (12)
theorem of circles is adopted for the determination of the air-entry
value of the representative elemental triangles. In the derivation, a
3. Verification of the proposed method contact angle greater than zero is considered for the estimation of the
SWCC from the GSD. The comparisons between the estimated results
As only the capillary water is considered in this study, sandy soil from the proposed method and experimental data indicate that the
including medium sand and fine sand from Yang et al. (2004); fine proposed method can provide good estimations of the SWCC for a
recycled concrete aggregate (RCA), coarse reclaimed asphalt pavement coarse-grained soil. The contact angles presented in this study are
(RAP) and fine RAP from Rahardjo et al. (2013); and coarse RCA, coarse consistent with the experimental results reported by Geobel et al.
RCA coated with oil, coarse RCA coated with wax from Rahardjo et al. (2004) and Ramirez-Flores et al. (2008). However, the exact values of
(2017) were selected for the verification of the proposed method. The the contact angle for different types of soil require further research. As
grain-size distribution data of these soils are illustrated in Fig. 11. The only the capillary water is considered in this study, the proposed
effective grain sizes of d95, d85 to d5 of these soils are illustrated in method is only applicable to the soil in which the capillary water

9
Q. Zhai, et al. Engineering Geology 267 (2020) 105502

dominates. moisture retention curve from soil particle size distribution and bulk density data
using a packing density scaling factor. Hydrol. Earth Syst. Sci. 18, 4053–4063.
Perera, Y.Y., Zapata, Z.E., Houston, W.N., Houston, S.L., 2005. Prediction of the soil-
Declaration of Competing Interest water characteristic curve based on grain-size distribution and index properties.
Proceed. Geo-Front. Austin Texas UAS January 24–26.
The authors declare that they have no known competing financia- Plaster, E.J., 2009. Soil Science and Management. Delmar, Clifton.
Rahardjo, H., Satyanaga, A., Leong, E.C., Wang, J.Y., 2013. Unsaturated properties of
linterestsor personal relationships that could have appeared to influ- recycled concrete aggregate and reclaimed asphalt pavement. J. Eng. Geol., April
ence the work reported in this paper. 161, 44–54. https://doi.org/10.1016/j.enggeo.2013.04.008.
Rahardjo, H., Prasad, A., Satyanaga, A., Mohammad, H., Leong, E.C., Wang, C.L., Wong,
L.H., 2017. 1-D Infiltration Behavior of Two-layered Recycled Concrete Aggregates
Acknowledgement using Hydrophobic Materials in a Column Apparatus. ASCE J. Mater. Civil Eng. 29 n
8, August 1, 2017.
The first author would like to acknowledge the financial supports he Ramirez-Flores, J.C., Woche, S.K., Bachmann, J., Goebel, M.O., Hallett, P.D., 2008.
Comparing capillary rise contact angles of soil aggregates and homogenized soil.
received from the National Natural Science Foundation of China (No.
Geoderma 146, 336–343.
51878160), the National Key Research and development program of Rawls, W.J., Brakensiek, D.L., 1989. Estimation of soil-water retention and hydraulic
China (No. 2017YFC00703408). properties. In: Morel-Seytoux, H.J. (Ed.), Unsaturated Flow in Hydrologic Modeling
Theory and Practice. Kluwer Academic Publishers, Beltsville, Md, pp. 275–300.
Satyanaga, A., Rahardjo, H., Zhai, Q., 2017. Estimation of Unimodal Water Characteristic
References Curve for Gap-graded Soil. Soils Found. 57, 789–801.
Schaap, M.G., Leij, F.J., 1998. Database-related accuracy and uncertainty of pedo-transfer
Arya, L.M., Paris, J.S., 1981. A physico-empirical model to predict the soil moisture functions. Soil Sci. 163, 765–779.
characteristic particle-size distribution and bulk density data. Soil Sci. Soc. Am. 45, Scheinost, A.C., Sinowski, W., Auerswald, K., 1997. Regionalization of soil-water reten-
1023–1030. tion curves in a highly variable soilscape, I. Developing a new pedo-transfer function.
Chin, K.B., Leong, E.C., Rahardjo, H., 2010. A simplified method to estimate the soil- Geoderma 78, 129–143.
water characteristic curve. Can. Geotech. J. 47 (12), 1382–1400. Sweijen, T., Aslannejad, H., Hassanizadeh, S.M., 2017. Capillary pressure–saturation re-
Ferraro, A., Sufian, A., Russell, A.R., 2017. Analytical derivation of water retention for lationships for porous granular materials: pore morphology method vs. pore unit
random monodisperse granular media. Acta Geotech. 12 (6), 1319–1328. assembly method. Adv. Water Resour. 107, 22–31.
Fredlund, D.G., Rahardjo, H., 1993. Soil Mechanics for Unsaturated Soils. Wiley, New Tyler, S.W., Wheatcraft, S.W., 1989. Application of fractal mathematics to soil water
York. retention estimations. Soil Sci. Soc. Am. J. 53 (4), 987–996.
Fredlund, D.G., Rahardjo, H., Fredlund, M.D., 2012. Unsaturated Soil Mechanics in Vereecken, H., Maes, J., Feyen, J., Darius, P., 1989. Estimating the soil moisture retention
Engineering Practice. Wiley, New York. characteristic from texture, bulk density, and carbon content. Soil Sci. 148 (6),
Fredlund, M.D., Wilson, G., Fredlund, D.G., 2002. Use of the grain-size distribution for 389–403.
estimation of the soil-water characteristic curve. Can. Geotech. J. 39, 1103–1117. Woche, S.K., Goebel, M.O., Kirkham, M.B., Horton, R., Van Der Ploeg, R.R., Bachmann, J.,
Ganz, C., Bachmann, J., Lamparter, A., Woche, S.K., Duijnisveld, W.H.M., Goebel, M.O., 2005. Contact angle of soils as affected by depth, texture, and land management. Eur.
2013. Specific processes during in situ infiltration into a sandy soil with low-level J. Soil Sci. 56, 239–251 April 2005.
water repellency. J. Hydrol. 484, 45–54. Yang, H., Rahardjo, H., Leong, E.C., Fredlund, D.G., 2004. Factors affecting drying and
Goebel, M.O., Bachmann, J., Woche, S.K., Fischer, W.R., Horton, R., 2004. Water po- wetting soil-water characteristic curves of sandy soils. Can. Geotechn. J. October 41
tential and aggregate size effects on contact angle and surface energy. Soil Sci. Soc. (5), 908–920. https://doi.org/10.1139/t04-042.
Am. J. 68, 383–393. Zapata, C.E., Houston, W.N., Houston, S.L., Walsh, K.D., 2000. Soil-water characteristic
Goebel, M.O., Woche, S.K., Bachmann, J., 2012. Quantitative analysis of liquid pene- curve variability. In: Shackelford, C.D., Houston, S.L., Chang, N.-Y. (Eds.), Advances
tration kinetics and slaking of aggregates as related to solid–liquid interfacial prop- in Unsaturated Geotechnics (GSP 99), Proceedings of the GeoDenver Conference,
erties. J. Hydrol. 442–443, 63–74. Denver, Colo., 5–8 August 2000. American Society of Civil Engineers, Reston, Va, pp.
Gupta, S.C., Larson, W.E., 1979. Estimating soil-water retention characteristics from 84–124.
particle size distribution, organic matter percent, and bulk density. Water Resour. Zhai, Q., Rahardjo, H., 2015. Estimation of permeability function from the Soil-Water
Res. 15 (6), 1633–1635. Characteristic Curve. Eng. Geol. 199, 148–156.
Hwang, S., Powers, S.E., 2003. Using particle-size distribution models to estimate soil Zhai, Q., Rahardjo, H., Satyanaga, A., Dai, G.L., 2019a. Estimation of unsaturated shear
hydraulic properties. Soil Sci. Soc. Am. J. 67, 1103–1112. strength from soil-water characteristic curve. Acta Geotech. 14 (6), 1977–1990.
Khoshghalb, A., Pasha, A.Y., Khalili, N., 2015. A fractal model for volume change de- https://doi.org/10.1007/s11440-019-00785-y.
pendency of the water retention curve. Géotechnique 65 (2), 141–146. Zhai, Q., Rahardjo, H., Satyanaga, A., Priono, Dai, G.L., 2019b. Role of pore-size dis-
Likos, W.J., Jaafar, R., 2013. Pore-scale model for water retention and fluid partitioning tribution function on the water follow in soil. J. Zhejiang University-SCIENCE A 20
of partially saturated granular soil. J. Geotech. Geoenviron. 139 (5), 724–737. (1), 10–20.
Meskini-Wishkaee, F., Mohammadi, M.H., Vanclooster, M., 2014. Predicting the soil

10

You might also like