Soil Dynamics and Earthquake Engineering 104 (2018) 33–39

Contents lists available at ScienceDirect

Soil Dynamics and Earthquake Engineering

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

Porosity estimation of unsaturated soil using Brutsaert equation MARK

a b c,⁎
Hyunwook Choo , Hwandon Jun , Hyung-Koo Yoon
a
Department of Civil Engineering, Kyung Hee University, Yongin 17104, Republic of Korea
b
Department of Civil Engineering, Seoul Nation University of Science and Technology, Seoul 139-743, Republic of Korea
c
Department of Construction and Disaster Prevention Engineering, Daejeon University, Daejeon 300-716, Republic of Korea

A R T I C L E I N F O A B S T R A C T

Keywords: Although the porosity is a crucial parameter for understanding the soil behavior under static and dynamic
Brutsaert model loading, estimating the porosity of unsaturated soil is difficult owing to the various input parameters required.
Compressional wave velocity The objective of this study is to suggest a technical method to obtain the level of porosity based on the elastic
Dynamic cone penetration test wave velocity in unsaturated soil. The Brutsaert model, which utilizes the theory of wave propagation, is applied
Energy dissipation
to modify the proposed method in terms of the porosity of unsaturated soil. The soil compressional wave velocity
Porosity
is gathered through a seismic refraction survey, and the porosity distribution at different depths is estimated. A
dynamic cone penetration test is applied to verify the converted porosity based on the compressional wave. The
two porosities estimated through the wave propagation and penetration show similar trends. Furthermore, a
special validation is performed to determine whether the energy dissipation can be ignored in the compressional
wave propagations under this experiment condition. The results of this study indicate that the suggested tech-
nique is useful for obtaining the porosity in unsaturated soil.

1. Introduction resolutions of these techniques.

A study was also conducted to investigate the characteristics of
The elastic wave velocity has been used to investigate subsurface porosity in unsaturated soils using elastic waves. Lu and Sabatier [25]
structures and obtain the design parameters when applying the theory monitored the elastic wave velocity according to changes in the water
of wave propagation in a medium. Among the various design para- potential, moisture content, and soil temperature during a two-year
meters, porosity is essential for assessing the stability of soil under static period. Shin et al. [30] used the elastic wave velocity to estimate the
and dynamic loadings, including compression, consolidation, earth- effects of soil structure, moisture content, and strength on plant growth
quakes and liquefaction [39,9]. Porosity can be obtained through a in unsaturated soil. Although these previous researches include no di-
laboratory testing of an extracted sample. However, obtaining reliable rect description of the change in porosity, the alternation of porosity
porosity data is difficult because the extraction procedure is limited to can be indirectly estimated through the measured total potential, ef-
special soil types, and extracting and moving the soil causes dis- fective stress, and moisture content. Gao et al. [17] showed the possi-
turbances that hinder the reliability [38]. To overcome the above bility of estimating the porosity in unsaturated soil using matric suction
mentioned disadvantages, the seismic wave velocity has been used to deduced through the measured elastic wave velocity, and Whalley et al.
obtain the porosity as an in-situ method. Wood [36] expressed porosity [35] demonstrated the elastic wave velocity as a function of porosity
in terms of compressional wave velocity using compressibility, a re- using Bishop's equation. Changes in elastic wave velocity in shallow
ciprocal of the elastic modulus, as an intermediary. Gassmann, whose surface soil have recently been addressed to determine seasonal and
work was translated by Berryman [8], proposed a relationship between weather effects [24], and the results of this study show that the ele-
porosity and elastic wave velocity for a low frequency range in an vation of the elastic wave velocity is caused by increased rigidity and
isotropic porous medium. Applying [10] theory of linear poroelasticity, decreased porosity of the surface soil. Previous studies have shown that
Foti et al. [16] suggested a method for estimating the porosity under the the porosity has a significant influence on the elastic wave velocity
assumptions that pore water is an undrained condition and the number under unsaturated soil conditions; however, there has been a lack of
of soil particles is infinite. Lee and Yoon [20] recently summarized methodological content to directly obtain the porosity. Therefore, in the
techniques for evaluating the porosity using a variety of assumptions, present study, we propose a practical method for estimating the por-
and assessed the sensitivities of every parameter to compare the osity of unsaturated soil using the seismic velocity.

⁎
Corresponding author.
E-mail addresses: choohw@gmail.com (H. Choo), hwjun@seoultech.ac.kr (H. Jun), hyungkoo@dju.ac.kr (H.-K. Yoon).

http://dx.doi.org/10.1016/j.soildyn.2017.09.029
Received 5 July 2017; Received in revised form 12 September 2017; Accepted 29 September 2017
0267-7261/ © 2017 Elsevier Ltd. All rights reserved.
H. Choo et al. Soil Dynamics and Earthquake Engineering 104 (2018) 33–39

1
An unsaturated medium consists of one soil particle and two im- 1 2
⎡ 0.306P 3 ⎤
miscible fluids, including water and air, and thus the medium is a effeective
VP = Φ ⎢ ⎥
complicated multiphase system with three distinct phases [14,2,29]. ⎢ ρmass n ⎥
Brutsaert [13] first suggested a comprehensive model of wave propa- ⎣ ⎦ (3)
gation in unsaturated soil, which can be applied under partial water
where the amplification factor (Φ) consists of empirical constants a and
saturation [23]. This model was extended by several different re-
b (Φ = a1/2/b1/3), and Z is assumed to be 1. Eq. (3) shows that the
searchers [18,32]. The reason many researchers have selected the
porosity of unsaturated soil can be obtained using the compressional
Brutsaert model is its treatment of wave propagation in unsaturated
wave velocity, as shown in Eq. (4).
soil. The model has been applied to establish the relationship between
0.33
the elastic wave velocity and water contents of soil [15,31,7]. However, 0.306Φ 2Peffeective
n=
the Brutsaert model has a porosity parameter and can be used to esti- VP2 ρmass (4)
mate the porosity using the elastic wave velocity for unsaturated con-
ditions. Thus, the Brutsaert model was also selected for the present
study. 2.1. Sensitivity of each parameter in the Brutsaert model
This study suggests a method for evaluating the porosity based on
the Brutsaert model. First, the Brutsaert model is introduced, discussed The sensitivity of the Brutsaert model to changes in each parameter
and then rearranged in terms of the porosity. The sensitivity of every is estimated using the error norm technique, which has been widely
parameter is determined to enhance the measurement reliability. applied to verify the model [19,20,37]. The sensitivity can be estimated
Profiles of compressional waves at several depths are determined and using the ratios of the calculated porosity based on the true values to
the porosity distribution is plotted. The porosity estimated from the the predicted porosity based on various ranges of input values. The true
elastic wave velocity is compared to the porosity deduced using a the values of the compressional wave velocity, mass density of the material,
dynamic cone penetration test, and the reliability of this estimate is effective pressure and amplification factor are 137 m/s, 1530 kg/m3,
described. Finally, the low frequency range under the experiment 28,000 Pa, and 614, respectively, according to results from Adamo et al.
conditions is verified to satisfactorily fulfill of the assumptions of the [1]. The minimum and maximum values of each parameter are de-
Brutsaert model. termined by decreasing and increasing the true value of 100%, re-
spectively. Finally, the sensitivity is calculated using Eq. (5).
Porositytrue value − Porosityvarious ranges input values
2. Background theory Sensitivity =
Porositytrue value (5)
The theory of wave propagation in an unsaturated medium was first
where porositytrue value and porosityvarious ranges input values denote the
proposed by Brutsaert [13] for a three-phase porous medium including
porosity based on the true value and various ranges of input values,
solid, liquid, and gas. A random aggregation of spherical soil grains is
respectively.
assumed to idealize the contact force between particles. The liquid
The results are shown in Fig. 1, where the x-axis represents the ratio
phase is assumed to form a liquid-gas interface with the atmosphere and
of the true value to the changed values. As a result, a value of 1 means
the Kirchhoff-Helmholtz theory is applied to define the dissipation
that the true value is the same as the changed value, and values of zero
coefficients during the wave propagation.
and 2 indicate that the changed values are decreased and increased
1 100% compared to the true value, respectively. The compressional
1 2
⎡ 0.306aP 3 ⎤ wave velocity and mass density were shown to be more sensitive to the
effeective Z
VP = ⎢ 2
⎥ decreased input values and less sensitive to the increased input values.
⎢ ρmass nb 3 ⎥
⎣ ⎦ (1)

where VP is the compressional wave velocity. The empirical parameters

a and b depend on the granular properties of the material. In addition,
Peffective, ρmass, and n are the effective pressure under the unsaturated
conditions, the mass density of the material and the porosity, respec-
tively. The interstitial effects between the total density and effective
pressure, Z, are expressed mathematically in Eq. (2).

5
3 3
⎧ 30.75[(1 − S ) Bgas + SBliquid] 2 b ⎫
1+ 1
⎨ 2
Peffective
⎬
Z= ⎩ ⎭
3
⎧ 46.12[(1 − S ) Bgas + SBliquid] 2 b ⎫
1+ 1
⎨ 2
Peffective
⎬
⎩ ⎭ (2)

where Bgas and Bliquid are the bulk moduli of the gas and liquid, re-
spectively and parameter S indicates the saturation of the medium. The
remaining parameters are the same as those used in Eq. (1). According
to a study by Shin et al. [31], the value of Z is almost unity under
unsaturated conditions, and Z is known to have less of an effect in Eq.
(1) at a shallow depth [11,15]. Furthermore, Adamo et al. [1] also
suggested that Z is equal to 1 because this value is insensitive to Fig. 1. Sensitivity of the Brutsaert model. The ratio of input values is the ratio between
changes in saturation, porosity, and depth of wave propagation for a the true value and changed values, which were decreased and increased in comparison to
range of values. Thus, Eq. (1) can be rearranged into Eq. (3) with an the true value. The order of the relative effects of the parameters in the Brutsaert model:
compressional wave velocity ≈ mass density > effective stress ≈ amplification factor.
added amplification factor (Φ).

34
H. Choo et al. Soil Dynamics and Earthquake Engineering 104 (2018) 33–39

Fig. 2. Profiles of field tests (elastic wave velocity

and DCP) and soil sampling.

On the other hand, the sensitivities of the amplification factor and ef- Table 1
fective stress were relatively low irrespective of the true value. Eq. (4) Properties of extracted materials.
has the greatest effect on the compressional wave velocity and mass
D10(mm) D30(mm) D60(mm) Cc Cu USCS emax emin Gs
density, and thus more accurate input values are needed to obtain the
porosity under unsaturated conditions using a high-resolution elastic S1 0.08 0.5 1.2 2.6 15 SW 1.07 0.60 2.62
wave. S2 0.08 0.3 1.1 1.0 13 SW 1.00 0.68 2.64
S3 0.08 0.3 1.0 1.1 12 SW 1.05 0.54 2.63
Average 0.08 0.36 1.1 1.5 13 – 1.04 0.62 2.63
3. Field test
*D10, D30, and D60 denote the diameters at passing percentages of 10%, 30%, and 60%,
3.1. Site description respectively. Cc and Cu are coefficients of curvature and uniformity. e and Gs are the void
ratio and specific gravity.

The landslide selected as the testing site provided unsaturated

study, the elastic refraction method was applied to obtain the com-
conditions for this study. The site is on Geohwa Mountain in South
pressional wave velocity.
Korea, a location where a debris-flow occurred approximately 5 years
The total length used to conduct the elastic wave survey was 90 m,
ago and that consists of many different streams. Among these streams,
and geophones were installed at 2 m intervals to increase the resolution
one main stream, which is 90 m in length, was selected for the field test
at shallow depths. Fig. 2 outlines the elastic wave survey. A sledge-
because of the deposited volume of the materials.
hammer was selected to generate artificial vibrations, and subsequent
The materials in the landslide were extracted using a sampler. The
impactions were conducted in three different locations near E1, E20,
positions of the samples, labeled S1, S2, and S3, were separated at
and E45 to obtain high quality signals, as shown in Fig. 2.
horizontal intervals of approximately 25 m, shown in Fig. 2. The sam-
ples obtained were used to estimate the material properties through
experiments including sieve test [3], minimum and maximum void 3.3. Dynamic cone penetration test
ratio [4], and specific gravity [6]. Fig. 3 shows the results of the sieve
test, where the average coefficients of the curvature (Cc) and uni- A dynamic cone penetration (DCP) test measures the penetrated
formity (Cu) are approximately 1.5 and 13, respectively. The extracted depth using a penetrometer, and this study utilized a dropped hammer
materials are classified as well graded sands (SW) based on the unified (78.8 N) with a constant height of 575 mm [5]. The penetrated depth
soil classification system (USCS) using the estimated coefficients. The per blow is called the dynamic cone penetration index (DPI), which is
ranges of the maximum and minimum void ratios are 1.04 and 0.62, useful for understanding the strength of a material. Material properties
respectively, and the average specific gravity is 2.63. Detailed material including the porosity, elastic modulus, dry unit weight and friction
properties are provided in Table 1. angle were characterized using a DCP test [12,26,27]. Thus, in the
study, the DCP test was selected for determining the reference value for
a comparison with the porosity estimated using the elastic wave velo-
3.2. Elastic wave survey
city.
The DCP tests were carried out in the same stream where the elastic
The elastic wave survey captures the wave profiles propagated into
wave survey was conducted. As shown in Fig. 2, the DPI was gathered
the medium through an induced artificial impaction on the surface. The
at 10 m intervals. and thus the experiments were carried out ten times.
wave profiles obtained can be converted into the elastic wave velocity,
which is used to characterize the medium because the elastic wave is
4. Results
propagated into individual grains of the medium. Therefore, in the

4.1. Elastic wave velocity

The waveforms were gathered using the seismic refraction method,

and the travel time-distance curve was determined using the
SeisImager-Poltrefa program (OYO Co.). Fig. 4 shows the distribution of
the analyzed elastic wave velocity as a function of elevation based on
the geographic information system. The test site consists of four layers
based on the ranges of the compressional wave velocity: 0–0.7 km/s
(colluvial soil), 0.8–1.2 km/s (weathering soil), 1.3–1.9 km/s (weath-
ering rock) and over 2.0 km/s (soft rock). The thicknesses of the soil
and weathering rock are approximately 1 and 3 m, respectively and
exhibit similar trends, whereas the soft rock has much greater thick-
nesses at distances of around approximately 40 and 60 m.

4.2. Dynamic cone penetration index (DPI)

Fig. 3. Sieve test results. Note that Cc and Cu are the coefficients of curvature and uni-
formity, respectively.
The measured penetration depth per blow was converted into the

35
H. Choo et al. Soil Dynamics and Earthquake Engineering 104 (2018) 33–39

Fig. 4. Measured elastic wave velocity profiles.

dynamic cone penetration index (DPI), and the values obtained at ten
sites are plotted in Fig. 5. The average initial DPI is approximately
227 mm/blow, and positions D1, D7 and D9 are where the initial DPI is
below the average value, indicating that the surface may have loosened.
A high stiffness of the surface soil may be predicted at positions D4 and
D10 owing to the high initial DPI (approximately greater than 300 mm/
blow). The final penetrated depths of the DCP at positions D1, D2, D3, Fig. 6. Distribution of measured elastic wave velocity for different depths.

D4, and D10 are almost 1 m, whereas positions D5, D6, D7, and D8
demonstrate relatively small penetrated depths (approximately less between particles. The velocity finally converges to 4 km/s, indicating
than 1 m) owing to deposits of stiff soil. However, position D10 shows a bedrock. The measured compressional wave velocity is converted into
high penetration depth (approximately 2 m) where loose soil was de- the porosity for these depths using Eq. (4). The compressional wave
posited. velocity is shown in Fig. 6 (measured values). The values of the mass
density, effective pressure, and amplification factor in Eq. (4) are
4.3. Porosity 1530 kg/m3 (measured value), 28,000 Pa [1] and 614 [1], respectively.
Fig. 7 shows the porosity distribution, based on the compressional wave
The elastic wave velocity was chosen at positions E1, E5, E10, E15, velocity, has a significant range of 0.85 to 0.004. Fig. 7 shows almost
E20, E25, E30, E35, E40, and E45, which are the same positions used the opposite trend as in Fig. 6 because the compressional wave velocity
for the DCP tests. The velocity as a function of depth is plotted in Fig. 6, is highly affected by Eq. (4), as shown in Fig. 1. Note that Fig. 7 shows
which shows that the elastic wave velocity increases as the depth in- that it is possible to obtain the porosity in unsaturated soil using the
creases owing to the growth of the effective stress and the contact force compressional wave velocity.

Fig. 5. Measured DPI by a dynamic cone penetrometer at positions of: (a) D1; (b) D2; (c) D3; (d) D4; (e) D5; (f) D6; (g) D7; (h) D8; (i) D9; (j) D10. The locations are shown in Fig. 2.

36
H. Choo et al. Soil Dynamics and Earthquake Engineering 104 (2018) 33–39

elastic wave velocity, respectively. Fig. 8 shows that there are two
different dispersed relationships which ranges A and B based on a
porosity of under and over 0.6, respectively. The coefficients of de-
termination are 0.7 and 0.2 for ranges A and B, respectively, and thus,
the relationship of range A is higher than that of range B. The results
shows that Eq. (4) is appropriate for medium to dense soil under a
porosity of approximately less than 0.6.

5. Discussion

The results indicate that the Brutsaert model is suitable for esti-
mating the porosity in unsaturated soil, and thus, a detailed verification
of the measured compressional wave velocity is performed: First, the
reliability of the measured compressional wave velocity is reviewed
through a comparison between the elastic moduli estimated using the
compressional wave velocity and the DPI. Second, the validation of
Brutsaert model is examined under these experiment conditions with
consideration of the energy dissipation.

5.1. Elastic modulus

The relationship between the seismic wave velocity and the DPI was
Fig. 7. Converted porosity profiles for different depths. estimated to verify the measured values using the exponential function
between the Young's modulus (E) and DPI, as suggested by Mohammadi
Lee et al. [21] suggested an equation to obtain the porosity using the et al. [26].
DPI, and thus the deduced porosity based on velocity is compared with
E = 55.033 ⋅ DPI−0.5459 (7)
the porosity estimated through the DPI for verification. In the case of
the porosity determined from Eq. (4), the mass density is a secondary A seismic wave propagates in an elastic medium without causing
factor affecting the porosity. However, the mass density at the soil any disturbances or altering the on-going processes [28]; thus, the
surface is used to determine the porosity because the extraction of seismic wave velocity is related to the elastic modulus including the
deeper samples is limited. Thus, 20 porosities determined from the Young's modulus (E), constraint modulus (M), shear modulus (G) and
surface soil were selected for comparison with the DPI. The results are bulk modulus (B) [33]. The compressional wave velocity (VP), which is
compared in Fig. 8, which shows that the two different porosities have a obtained through a seismic survey, can be converted into a constraint
linear relationship with a coefficient of determination of 0.4, as in- modulus (M) with the mass density (ρ) because particle motion is
dicated in Eq. (6) below. parallel to the direction of the wave propagation.

nDPI = 0.302nVelocity + 0.3 (R2 = 0.4) (6) M = VP2 ⋅ ρ (8)

where nDPI and nvelocity denote the porosities based on the DPI and The constraint modulus consists of shear and bulk moduli as

Fig. 8. Comparison of porosities based on elastic wave velocity

and DPI.

37
H. Choo et al. Soil Dynamics and Earthquake Engineering 104 (2018) 33–39

hydraulic conductivity under saturated conditions (kis) based on Van

Genuchten's model [34].
1 1
ki = kis S 2 [1 − (1 − S m )m]2 (15)
where S is the saturation and m is a dimensionless parameter. The value
of m is deduced from the Van Genuchten curve shape (VGS).
1
m=1−
VGS (16)
The range of output frequencies of the measured signals is ap-
proximately 30–50 Hz, and the input frequency in Eq. (14) has an
average value of 40 Hz. The hydraulic conductivity under standard
conditions is estimated as 0.01 mm/s based on the Hazen equation, and
the average saturation is calculated as 10% using measured values,
which range from 3% to 27%, based on the time domain reflectometry.
The value of VGS was assumed to be 0.6 in the study by Van Genuchten
[34] and the dimensionless parameter (m) was calculated to be 0.62.
Finally, the viscosity factor deduced using the input values is 0.02,
which is less than the reference value of unity, and thus the experi-
Fig. 9. Relationship between Young's modulus (E) and DPI. The subplot shows a change mental conditions satisfy the assumption of a low frequency range
of E in the range of 0 MPa~0.6 MPa. without a dissipation of energy. Note that the measured value in this
study is fully satisfied when applying the Brutsaert model.
follows.
5.3. Advantage and limitation of suggested methodology
4
VP2⋅ρ = B + ⋅G
3 (9) The compressional wave velocity shows highly affecting factors in
Eq. (9) can be rearranged in terms of the Young's modulus, as shown Brutsaert model and the Fig. 1 show the sensitivity increases in an in-
in Eq. (11) using Eq. (10). verse relationship with the growth of compressional wave velocity. The
compressional wave velocity depends on effective stress of materials
4
B = M− ⋅G and thus, the velocity increase when the soil particles are highly com-
3 (10)
bined. Note that the compressional wave velocities are approximately
9BG 9MG−12G 2 ρ (9VP2 VS2−12VS4 ) 60–230 m/s for soils and 2200–7000 m/s for rocks [28]. Given that the
E= = =
3B + G 3M −4G + G ρ (3VP2−4VS2 + VS2) (11) reference compressional wave velocity is 137 m/s in this study, the
sensitivity is small in soil and rocks because the velocities of the ma-
where VS is the shear wave velocity. terials are generally larger than the reference velocity, however when
Finally, the compressional wave velocity is deduced in terms of the the velocity is smaller than reference velocity, the correct value should
Young's modulus because the compressional wave velocity is generally be considered. The small compressional wave velocity means that the
0.6 times faster than the shear wave velocity [22]. effective stress of the soil is weak, which is called as loose soils. And
E ≈ 0.9VP2 (12) thus, the limitation is the same as the results of the porosity comparison
based on DPI and elastic wave velocity (Fig. 8), showing that the sug-
The relationship between the Young's modulus and the DPI, as gested model is more reliable in medium and dense soil.
measured in this study, is plotted in Fig. 9. Although the converted
Young's modulus shows a smaller range than in previous studies [26] 6. Conclusion
owing to the different characteristics at this site, the relationship can
still be described using an exponential function as in Eq. (13). In this study, the porosity in unsaturated soils was obtained using
E = 0.3736 ⋅ DPI−0.284 (13) the compressional wave velocity, and the converted porosity was
compared to the DPI using a dynamic cone penetrometer test to verify
This relationship demonstrates that the measured compressional
reliability. The detailed results of this study are as follows.
wave velocity and the DPI have a reliable relationship for this particular
site. Furthermore, additional support of the appropriateness of this re-
lationship is suggested through the low range of values (0–0.6 MPa) of • The Brutsaert model was applied to deduce the porosity, and the
sensitivity of every parameter in this model was estimated. The
the Young's modulus based on the low compressional wave velocity.
compressional wave velocity is a highly sensitive parameter in this
model, indicating that the selected model is suitable for evaluating
5.2. Validation of the Brutsaert model
the porosity based on the elastic wave velocity.

Wave propagation leads to relative motion between phases, and • Detailed porosity profiles as a function of depth were determined,
and porosities based on the compressional wave velocity and the
thus the Brutsaert [13] model is recommended at sufficiently low fre-
DPI were compared. Two different ranges of the dispersed re-
quency ranges when neglecting the dissipation of energy. The inertia-
lationship were shown.
viscosity factor (β) is used to estimate whether dissipation has occurred
[1]. If β is less than 1, the condition of low frequency is satisfied. • To verify the measured compressional wave velocity, a relationship
between the compressional wave velocity and the Young's modulus
5.13fs ki was proposed. The energy dissipation was also identified to confirm
β= <1 the low frequency range of the measured seismic wave.
nS (14)
where fs, ki, n and S are the frequency of the output signal, the hy-
draulic conductivity under unsaturated conditions, the porosity, and
the saturation, respectively. The value of ki can be obtained using the

38
H. Choo et al. Soil Dynamics and Earthquake Engineering 104 (2018) 33–39

Acknowledgement saturated elastic porous media. J Appl Phys 1986;60(9):3045–55.

[19] Lee JS, Fernandez AL, Santamarina JC. S-wave velocity tomography: Small-scale
laboratory application; 2005.
This work was supported by the National Research Foundation of [20] Lee JS, Yoon HK. Porosity estimation based on seismic wave velocity at shallow
Korea (NRF) grant funded by the Korea government (MSIP) (NRF- depths. J Appl Geophys 2014;105:185–90.
[21] Lee C, Kim KS, Woo W, Lee W. Soil Stiffness Gauge (SSG) and Dynamic Cone
2017R1A2B4008157). Penetrometer (DCP) tests for estimating engineering properties of weathered sandy
soils in Korea. Eng Geol 2014;169:91–9.
References [22] Lillie RJ. Whole earth geophysics: an introductory textbook for geologists and
geophysicists (No. 550.8 LIL); 1999.
[23] Lo WC, Sposito G. Acoustic waves in unsaturated soils. Water Resour Res
[1] Adamo F, Andria G, Attivissimo F, Giaquinto N. An acoustic method for soil 2013;49(9):5674–84.
moisture measurement. IEEE Trans Instrum Meas 2004;53(4):891–8. [24] Lu Z. Feasibility of using a seismic surface wave method to study seasonal and
[2] Arora A, Tomar SK. The effect of inertial coupling on seismic reflection amplitudes. weather effects on shallow surface soils. J Environ Eng Geophys 2014;19(2):71–85.
Geophys Prospect 2008;56(5):643–54. [25] Lu Z, Sabatier JM. Effects of soil water potential and moisture content on sound
[3] ASTM. C136. Standard test method for sieve analysis of fine and coarse aggregates; speed. Soil Sci Soc Am J 2009;73(5):1614–25.
1984. [26] Mohammadi SD, Nikoudel MR, Rahimi H, Khamehchiyan M. Application of the
[4] ASTM. D4254. Standard test methods for minimum index density and unit weight of Dynamic Cone Penetrometer (DCP) for determination of the engineering para-
soils and calculation of relative density; 2016. meters of sandy soils. Eng Geol 2008;101(3):195–203.
[5] ASTM. D6951. Standard test method for use of the dynamic cone penetrometer in [27] Salgado R, Yoon S. Dynamic cone penetration test (DCPT) for subgrade assessment.
shallow pavement applications; 2015. Jt Transp Res Program 2003:73.
[6] ASTM. D854. Standard test methods for specific gravity of soil solids by water [28] Santamarina JC, Klein A, Fam MA. Soils and waves: particulate materials behavior,
pycnometer; 2014. characterization and process monitoring. J Soils Sediment 2001;1(2). [130-130].
[7] Attivissimo F, Cannazza G, Cataldo A, De Benedetto E, Fabbiano L. Enhancement [29] Sawada S, Tsukamoto Y, Ishihara K. Residual deformation characteristics of par-
and metrological characterization of an accurate and low-cost method based on tially saturated sandy soils subjected to seismic excitation. Soil Dyn Earthq Eng
seismic wave propagation for soil moisture evaluation. IEEE Trans Instrum Meas 2006;26(2):175–82.
2010;59(5):1216–23. [30] Shin HC, Taherzadeh S, Attenborough K, Whalley WR, Watts CW. Non‐invasive
[8] Berryman JG. Origin of Gassmann's equations. Geophysics 1999;64(5):1627–9. characterization of pore‐related and elastic properties of soils in linear Biot–Stoll
[9] Bian H, Shahrour I. Numerical model for unsaturated sandy soils under cyclic theory using acoustic‐to‐seismic coupling. Eur J Soil Sci 2013;64(3):308–23.
loading: application to liquefaction. Soil Dyn Earthq Eng 2009;29(2):237–44. [31] Shin HC, Whalley WR, Attenborough K, Taherzadeh S. On the theory of Brutsaert
[10] Biot MA. Theory of propagation of elastic waves in a fluid‐saturated porous solid. I. about elastic wave speeds in unsaturated soils. Soil Tillage Res 2016;156:155–65.
Low‐frequency range. J Acoust Soc Am 1956;28(2):168–78. [32] Tuncay K, Yavuz Corapcioglu M. Body waves in poroelastic media saturated by two
[11] Blum A, Flammer I, Friedli T, Germann P. Acoustic tomography applied to water immiscible fluids. J Geophys Res: Solid Earth 1996;101(B11):25149–59.
flow in unsaturated soils. Vadose Zone J 2004;3(1):288–99. [33] Uhlemann S, Hagedorn S, Dashwood B, Maurer H, Gunn D, Dijkstra T, Chambers J.
[12] Brough M, Stirling A, Ghataora G, Madelin K. Evaluation of railway trackbed and Landslide characterization using P-and S-wave seismic refraction tomography—the
formation: a case study. NDT E Int 2003;36(3):145–56. importance of elastic moduli. J Appl Geophys 2016;134:64–76.
[13] Brutsaert W. The propagation of elastic waves in unconsolidated unsaturated [34] Van Genuchten MT. A closed-form equation for predicting the hydraulic con-
granular mediums. J Geophys Res 1964;69(2):243–57. ductivity of unsaturated soils. Soil Sci Soc Am J 1980;44(5):892–8.
[14] Conte E, Cosentini RM, Troncone A. Shear and dilatational wave velocities for [35] Whalley WR, Jenkins M, Attenborough K. The velocity of shear waves in un-
unsaturated soils. Soil Dyn Earthq Eng 2009;29(6):946–52. saturated soil. Soil Tillage Res 2012;125:30–7.
[15] Flammer I, Blum A, Leiser A, Germann P. Acoustic assessment of flow patterns in [36] Wood AB. A textbook of sound. London: G. Bell and Sons Ltd.; 1949.
unsaturated soil. J Appl Geophys 2001;46(2):115–28. [37] Yoon HK, Lee JS. Field velocity resistivity probe for estimating stiffness and void
[16] Foti S, Lai CG, Lancellotta R. Porosity of fluid-saturated porous media from mea- ratio. Soil Dyn Earthq Eng 2010;30(12):1540–9.
sured seismic wave velocities. Géotechnique 2002;52:359–73. [38] Yoon HK, Jung SH, Lee JS. Characterisation of subsurface spatial variability using a
[17] Gao W, Watts CW, Ren T, Shin HC, Taherzadeh S, Attenborough K, Jenkins M, cone resistivity penetrometer. Soil Dyn Earthq Eng 2011;31(7):1064–71.
Whalley WR. Estimating penetrometer resistance and matric potential from the [39] Zhang M, Wang X, Yang G, Xie L. Solution of dynamic Green's function for un-
velocities of shear and compression waves. Soil Sci Soc Am J 2013;77(3):721–8. saturated soil under internal excitation. Soil Dyn Earthq Eng 2014;64:63–84.
[18] Garg SK, Nayfeh AH. Compressional wave propagation in liquid and/or gas

39