See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/309889831

Heat conduction in two and three-phase media with solid spherical particles
of the same diameter

Article  in  International Journal of Thermal Sciences · November 2016

DOI: 10.1016/j.ijthermalsci.2016.10.022

CITATIONS READS

3 36

2 authors:

Sandra Corasaniti Fabio Gori

University of Rome Tor Vergata University of Rome Tor Vergata
25 PUBLICATIONS   250 CITATIONS    75 PUBLICATIONS   811 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Thermal diffusivity of sintered steels View project

All content following this page was uploaded by Sandra Corasaniti on 21 December 2018.

The user has requested enhancement of the downloaded file.

 International Journal of Thermal Sciences 112 (2017) 460e469

Contents lists available at ScienceDirect

International Journal of Thermal Sciences

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

Heat conduction in two and three-phase media with solid spherical

particles of the same diameter
S. Corasaniti, F. Gori*
University of Rome Tor Vergata, Department of Industrial Engineering, Via del Politecnico 1, 00133 Rome, Italy

a r t i c l e i n f o a b s t r a c t

Article history: Heat conduction in two and three-phase media, composed of solid spherical particles of the same
Received 27 January 2016 diameter, is investigated theoretically and experimentally. The theoretical model has no empirical
Received in revised form constants and is based on the solution of the Fourier heat conduction equation, under the thermal
14 October 2016
assumption of parallel heat fluxes, in porous medium with porosity greater than 0.4764. The medium can
Accepted 14 October 2016
be two-phase, fully saturated with water or air, or three-phase, partially saturated with water and air. The
Available online 10 November 2016
problem is investigated experimentally with the thermal probe, a vertical cylinder inserted into the glass
beads of the diameter of 3 mm. The thermal probe, made in laboratory, has a diameter of 1.5 mm and
Keywords:
Two and three-phase media
length 150 mm, and contains an electric heater and a temperature sensor. The perfect line source theory
Solid spherical particle with same diameter in the transient regime is employed to measure the thermal conductivity of the water saturated glass
Effective thermal conductivity beads. The theoretical results for the two-phase media, glass beads and water, are in good agreement
Thermal probe with the experiments. The results of the theoretical model for the three-phase media, glass beads with
Theoretical model water and air, are in good agreement with the experimental data of the literature, and are compared to
Parallel heat fluxes assumption several theoretical models of the literature.
© 2016 Elsevier Masson SAS. All rights reserved.

1. Introduction according to the two thermal assumptions, are the upper and lower
bounds on the effective thermal conductivity of all normally
Heat transfer by conduction in multiple-phase media is an distributed mixtures, [1,2]. The difficulty in the solution of the
important problem in several different applications, e.g. agricul- Fourier equation arises when the knowledge of the shape, size,
ture, geology and geophysics, thermal insulation of buildings, location and thermal properties of each particle is uncertain, or
geothermal engineering, filtration processes, ground water pollu- when shape and location are very complicated. For this reason the
tion and heat storage systems, cooling of nuclear reactors, catalytic use of the electro-magnetic equations, specifically the Ohm law, has
reactors, heat exchangers, high-voltage underground cables, received a great attention, and the approach of the so-called elec-
radioactive waste storage facilities below the ground. trical analogy is resumed by the Maxwell approach [3], for suffi-
The solution of this kind of problem has been carried out ciently dilute dispersion of the spherical particles, later on
employing several approaches which are described in the extended to ellipsoidal particles and generalized to the case of n
following. Since the heat conduction equation has been discovered dispersed phases embedded in one continuous phase, [4]. Also the
first, the first kind of approach has been the direct solution of the electrical analogy presents some limitations and the third approach
basic heat transfer conduction equation, the Fourier law. For an to be mentioned is the empirical one. A limited number of papers is
idealized two phase media composed of a cubic solid and a fluid possible to mention here in order to review briefly the attention
around, the simplest solution of the Fourier equation is to assume a that each of the three approaches has received.
one-dimensional heat transfer through the unit cell, under the Starting the review from the electrical analogy, the extension of
hypothesis of temperature or heat flux distribution [1]. The effec- the Maxwell approach [3], sees the empirical proposal of an
tive thermal conductivities of the two phase media, calculated expression for the electrical permittivity of a composite, consisting
of any number of randomly mixed components [5], the application
to the physics of plant environment [6], and to three-phase mixture
in spherical coordinates, [7]. The Maxwell approach, valid for a solid
* Corresponding author. fraction tending to zero, or for porosity tending to 1, was extended
E-mail address: [email protected] (F. Gori).

http://dx.doi.org/10.1016/j.ijthermalsci.2016.10.022
1290-0729/© 2016 Elsevier Masson SAS. All rights reserved.
 S. Corasaniti, F. Gori / International Journal of Thermal Sciences 112 (2017) 460e469 461

by Lord Rayleigh to smaller porosities [8], later on corrected spherical unit composed of a composite particle surrounded by a
numerically [9], and applied to the case of a cubic array of uniform layer of the continuous medium with a contact resistance between
size spheres [10], with the proposal of a different potential function particle and the continuum negligible, and an exact series solution
and a different final equation. The thermal resistance approach has of the heat conduction equation was obtained for the temperature
been employed in Ref. [11] to calculate the effective thermal con- distribution in each phase. The effect of solid thermal conductivity
ductivity of porous materials taking into account the effects of and particle-particle contact on thermo-diffusion processes was
temperature, porosity, material skeleton thermal conductivity, gas investigated in Ref. [38], showing that for non-consolidates porous
in the pore, pore size and other kinds of heat transfer mechanisms. material, composed of spherical particles, the thermal conductivity
The empirical approach can be resumed by the weighted mean. ratio does not influence the thermal diffusion process. The
Since the series distribution of the two phase in the medium cor- approach of [22] was applied to three-phase media in Ref. [39], to
responds to the weighted harmonic mean and the parallel distri- composite materials [40e42], and to three-phase soils [43]. Mini-
bution to the weighted arithmetic mean, the approach of the mization of thermal resistance in an air cooled porous matrix, made
weighted geometric mean, for the effective medium approxima- of solid spheres with heat generation, was investigated in Ref. [44].
tion, can be considered empirical because it does not make refer- A theoretical model was developed in Ref. [45] to take into account
ence to a physical principle, but simply gives results comprised the decrease of the lattice thermal conductivity in a porous me-
between the other two means. The weighted geometric mean dium with inhomogeneous porosity. Porous media with various
approach has been used in Ref. [12] to compare their experimental pore groups and different diameters of pores leads to the significant
results. The empirical model of Lichtenecker [5] has been applied to decrease in the lattice thermal conductivity, compared, not only to
the effective thermal conductivity of three-component composites, bulk materials with zero porosity but also to materials with ho-
expressed by the volumetric content of each one, [13]. The mogeneous porosity. The approach of [22] has been extended
macroscopic properties of randomly inhomogeneous materials, recently to composites [46,47].
having specific geometry of the constituents, was investigated in The aim of the present paper is to present a new theoretical
Ref. [14] to take into account the different shapes of the inclusions model to evaluate the effective thermal conductivity of two and
embedded in the matrix, while the effective media formation and three-phase media, made of solid spherical particles of the same
conduction through unsaturated granular materials was studied in diameter, and with porosity greater than 0.4764, without any
Ref. [15]. Models of weighted effective medium approximations empirical constant, by extending the previous approach [23]. The
were developed to estimate the thermal conductivity of random paper compares the theoretical results with some experimental
composites with different materials and shapes [16]. Inter-particle measurements carried on by the authors for two-phase media.
contact heat transfer in soil systems at moderate temperatures was
studied in Ref. [17]. An empirical model for the investigation of a
packed bed of spherical particles, with points of contacts in the heat 2. Theoretical model
flow direction was employed in Ref. [18].
Many papers of the literature have solved the heat conduction 2.1. Two-phase media with solid spherical particles
equation in multiple phase systems taking into account also other
phenomena, as thermal radiation. Kunii and Smith [19] predicted The unit cell of the medium, for porosities greater than 0.4764,
the effective thermal conductivity of beds of unconsolidated par- has a rhombic base and contains a solid spherical particle of radius
ticles containing stagnant fluid and neglecting radiation. The effect R1 in the center. Fig. 1 presents the unit cell in three-dimensions.
of air pressure on the effective thermal conductivity of a bed of The angle b of Fig. 1a is variable with the porosity and, for
spherical particles was investigated in Ref. [20], while the thermal b ¼ 90 , the porosity of the cell is 0.4764 according to its definition:
conductivity of granular materials was studied in Ref. [21]. A model
for the evaluation of the effective thermal conductivity of unsatu- Vcell  Vsolid p
ε¼ ¼ 1  ¼ 0:4764 (1)
rated frozen soils was proposed in Ref. [22]. Two-phase media with Vcell 6
the solid particles of the same diameter and porosities smaller than
0.4764 were modelled in Ref. [23]. The theoretical approach of [22] where Vsolid ¼ 43 pR31 , and Vcell ¼ 8R31 . If b < 90 the volume of the cell
was later on extended to bricks [24]. Heat transfer in a packed bed becomes:
with wall effect was studied in Ref. [25], showing that the heat .
conduction characteristics are function of the ratio of the particle Vcell ¼ 2R1 $2R1 $2R1 =sin b ¼ 8R31 sin b; (2)
diameter to the characteristic length of the geometry and to the
ratio of the thermal conductivity of the fluid to that of the solid
where the particle volume is Vsolid ¼ 43 pR31 and the porosity is equal
phase. The approach of [25] was modified for stagnant thermal
to the ratio between the volume of the voids and the total volume
conductivity of porous media [26,27], where a unit cell model was
of the cell
used to determine the effective thermal conductivity of bi-
dispersed porous media, based on the lumped parameter method. Vcell  Vsolid p
The theoretical prediction of the thermal conductivity of soils at ε¼ ¼ 1  sin b > 0:4764 (3)
Vcell 6
moderately high temperatures was done in Ref. [28]. A model to
evaluate the radiation conductivity tensor was developed in The rhombic base of the cell of Fig. 1a has the side equal to a
Ref. [29] for porous media composed by spheres or cylinders, and
the numerical results showed that the radiation contribution can 2R1
a¼ (4)
influence the temperature distribution across the particle surface. sinðbÞ
Theoretical predictions of the thermal conductivity of two-and
The area of the rhombic base, S, is
three-phase water/olivine systems were compared to experi-
ments in Ref. [30]. The approach of [22] was extended to the .
investigation of extraterrestrial soil analogues, as in planets and S ¼ a2 sinðbÞ ¼ ð2R1 Þ2 sinðbÞ (5)
comets [31e34], frozen meats [35], soils at elevated temperatures
[36]. Felske [37] considered the spheres as homogeneous and the while the total volume of the cell is
462 S. Corasaniti, F. Gori / International Journal of Thermal Sciences 112 (2017) 460e469

Fig. 1. Unit cell of the two-phase media with the rhombic base and a solid spherical particle of radius R1 in the center. (a) View from above. (b) Front view.

2.2. Three-phase media with solid spherical particles (3P)

.
Vt ¼ a3 sin2 ðbÞ ¼ ð2R1 Þ3 sinðbÞ (6)
The effective thermal conductivity can be evaluated for a three-
The angle b of the rhombic base is related to porosity by phase media, with solid spherical particles partially saturated with
air and water, as presented in Figs. 2 and 3. The water saturation
  degree of the medium, j, spanning in the range 0e1, is the ratio
6
b ¼ asin $ð1  εÞ (7) between the volume of water and that of the voids
p
Heat conduction by Fourier law is applied to the unit cell of Fig. 1 Vw
j¼ (11)
with the thermal assumption of parallel heat fluxes along the x Vv
direction. The total heat transfer across the cell is the sum of the The present paper considers the following two cases.
heat, Qw, crossing only the fluid (e.g., water or air for fully saturated
medium) and the heat, Qw, s, crossing water and solid particle.
The heat transfer, Qw, crossing the water of the cell throughout 2.2.1. Case 1 e water is surrounding the solid spherical particle
the area ðS  pR21 Þ, is with radius R2  a·sin(b/2)
The radius of the water surrounding the solid spherical particle,
kw $DT   R2, is smaller than a$sin(b/2) and the unit cell is shown in Fig. 2.
Qw ¼ $ S  pR21 (8) According to the symbols of Fig. 2 the saturation degree j is given by
2R1
  h i
while the heat, Qw, s, crossing water and solid particle through the
4 p$
3 R32  R31  6$ p$ðR2  R1 Þ2 $ð2R23þR1 Þ
area pR21 , is j¼ (12)
2$R1 $S  43 p$R31

ZR1 where R1  R2  a$sin(b/2).

2p$r$DT
Qw;s ¼ $dr The effective thermal conductivity is calculated with the ther-
2½ðR1  rÞ=kw þ r=ks 
0
mal assumption of parallel heat fluxes. The heat transfer, Q1,
ZR1 crossing water and solid particle through an area equal to pr22 , with
p$y$DT qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
¼ 
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi  qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi  $dy r2 ¼ R22  R21 , is
0 R1  R21  y2 kw þ R21  y2 ks
   Zr2
4p$R1 $kw lnðkw =ks Þ ks p$y$DT
¼ $  1 þ ln  1 $DT (9) Q1 ¼  qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi  qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi  $dy (13)
1  kw =ks 1  kw =ks kw 0 R1  R21  y2 kw þ R21  y2 ks
The effective thermal conductivity of the two-phase medium is
finally The heat transfer, Q2, crossing air, water and solid particle
through an area equal to pðR21  r22 Þ, is


Qw þ Qw;s $2R1
k2P ¼ (10)
S$DT The heat transfer, Q3, crossing water and air through an area
equal to pðR22  R21 Þ, is

ZR1
p$y$DT
Q2 ¼  qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi  qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi  qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi  $dy (14)
2
r2 R1  R2  y2 ka þ R2  y2  R21  y2
2
kw þ R21  y2 ks
 S. Corasaniti, F. Gori / International Journal of Thermal Sciences 112 (2017) 460e469 463

Fig. 2. Unit cell of a three-phase media with water disposed around the spherical solid particle with R2  a$sin(b/2).

h  i 2.2.2. Case 2: water is surrounding the solid spherical particle with

ZR2  
y$DT$ p  4$acos Ry1 pffiffiffi
 radius R2 in the range a$sin 2b < R2  2$R1
Q3 ¼ qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi  qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi  $dy
R1 R1  R22  y2 ka þ R22  y2 kw The radius of the water, surrounding p the
ffiffiffi solid particle, R2, is
greater than a$sin(b/2) but smaller than 2$R1 and the water is
(15) distributed around the solid spherical particle according to Fig. 3.
Eqs. (11)e(12) are solved numerically through a Simpson The effective thermal conductivity is calculated with the ther-
procedure. mal assumption of parallel heat fluxes and the heat transfers,
 The heat transfer, Q4, crossing
 only air through an area equal to similarly to Case 1, are the followings. The heat transfer, Q1, crossing
S  p$R22 þ 4$R22 $acos RR12  4$r2 $R1 , is water and solid particle through an area equal to pr22 , is

   
ka $DT R
Q4 ¼ $ S  p$R22 þ 4$R22 $acos 1  4$r2 $R1 (16) Zr2
2$R1 R2 p$y$DT
Q1 ¼  qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi  qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi  $dy (18)
The effective thermal conductivity of the three-phase media is 0 R1  R21  y2 kw þ R21  y2 ks
finally
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
ðQ þ Q2 þ Q3 þ Q4 Þ$2R1 with r2 ¼ R22  R21 .
k3P ¼ 1 (17) The heat transfer, Q2, crossing air, water and solid particle
S$DT
through an area equal to pðR21  r22 Þ, is

ZR1
p$y$DT
Q2 ¼  qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi  qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi  qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi  $dy (19)
2
r2 R1  R2  y2 ka þ R2  y2  R21  y2
2
kw þ R21  y2 ks

The heat transfer, Q3, crossing water and air through an area
equal to pðR22star  R21 Þ, is

h  i
RZ2 star
p  4$acos R1
y $y$DT
Q3 ¼  qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi  qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi  $dy
R1 R1  R22  y2 ka þ R22  y2 kw

(20)

with R2 star ¼ a$sin(b/2).

Fig. 3. Base of the unit cell for a three-phase media with water
 disposed around the The heat transfer, Q4, crossing water and air through an area
pffiffiffi
spherical solid particle with radius R2 in the range a$sin 2b < R2  2$R1 . equal to pðR22  R22 star Þ, is
464 S. Corasaniti, F. Gori / International Journal of Thermal Sciences 112 (2017) 460e469

3. Experiments in two-phase media

h  i
ZR2 p  b  acos R1 $2$y$DT
2 y 3.1. Experimental set-up
Q4 ¼  qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi  qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi  $dy
R2 star
R1  R22  y2 ka þ R22  y2 kw The effective thermal conductivity of the two-phase medium is
measured by using the thermal probe method. The probe used in
(21)
the present experiments is laboratory-made, with a stainless steel
The heat transfer, Q5, crossing air through the area Sstar, equal to tube of external diameter 1.5 mm and length 150 mm, containing a
thermocouple and a Platinum wire heater in its inner part [48,49].
 The thermocouple, type T, is made with a copper wire of diameter
a$cos b 0.35 and a constantan wire of diameter 0.08 mm, soldered together
2
Z    in the middle of the probe. The Platinum wire heater is a platinum
p R
Sstar ¼ 4$y$  b  acos 1 $dy (22) wire spiraled around the entire length of the probe. A special wax is
2 y used to hold the wires inside the tube, while the electric connec-
R2
tions are brought to a plastic handle with external wires.
The thermal probe is positioned on the axis of a cylindrical
is
container, high 155 mm and with diameter 300 mm, where
spherical glass beads of the same diameter, 3 mm, are present. A
ka $DT second thermocouple is set on the internal wall of the container, in
Q5 ¼ $Sstar (23)
2R1 order to control the global heating of the sample. A Dewar flask
contains the cold junctions of the thermal probe and the wall
Eqs. (16)e(20) are solved numerically through a Simpson
thermocouple. After the heater is powered, the temperature in-
procedure.
crease is recorded versus time. The experimental set-up, with the
The effective thermal conductivity of the three-phase media is
sample containing the glass beads, the thermal probe and the
finally
various instruments, is sketched in Fig. 4.
The measurements are carried out with a step heating of the
ðQ1 þ Q2 þ Q3 þ Q4 þ Q5 Þ$2$R1 platinum wire, by circulating direct electric current with a c.c.
k3P ¼ (24)
S$DT feeder (KEPCO bipolar operational power supply/amplifier) at time
zero and by monitoring the thermocouple inside the thermal probe

Fig. 4. Experimental set-up.

 S. Corasaniti, F. Gori / International Journal of Thermal Sciences 112 (2017) 460e469 465

q lnt2 =t1
k¼ $ (27)
4 p T2  T1
The presence of a linear increase of the probe temperature, re-
ported in a semi-logarithmic plot, and a constant temperature at
the wall, shown in Fig. 5, confirms the good performance of the
thermal probe. The slope of the curve is determined through a
regression least-squares analysis. The experiments are carried out
at ambient temperature and with different electric powers of the
heater, e.g. from 6.5 W/m to 28 W/m. The sample is made with glass
beads of the same diameter, 3 mm, and is saturated with water
(j ¼ 1).
The assumptions of the problem are:

- Water and glass beads are in local thermal equilibrium every-

where (Tf ¼ Ts).
- The medium is isotropic and homogeneous.
Fig. 5. Typical trend of wall and thermal probe temperatures.

Fig. 5 shows that the temperature of the wall is constant during

and that on the container wall. The digital multimeter, Keytley the test and the hypothesis of infinite sample can be applied,
2700, is controlled by a personal computer. because the ratio between the diameter sample (300 mm) and the
external probe diameter (1.5 mm) is greater than 100, [51].
Fig. 6 reports the experimental effective thermal conductivity
3.2. Experimental analysis versus the mean temperature of the test during heating with
different power, i.e. with electric current ranging from 70 mA to
The effective thermal conductivity of the medium is evaluated 140 mA. Four tests, for each electric current, are carried out and
according to the perfect line source theory [50], which assumes the each experimental result, shown in Fig. 6, represents the average
following relation between temperature and time value.
According to the ISO GUM Standard [52], two different uncer-
 
q 4at tainty sources can be attributed to these tests: type A (statistical)
TðtÞ ¼ ln 2  0:5772 (25) and type B (non-statistical). Data processing produces two kinds of
4pk R
type A uncertainties: the uncertainty due to the least square
where R is the external probe radius. At a fixed radius, the tem- regression procedure, and the standard uncertainty deriving from
peratures T2 and T1, at the times t2 and t1, are related by the test repetitions. The second uncertainty is three times higher
than the first one, which can be neglected, and the type A uncer-
q t tainty is finally about 0.7%. The type B uncertainty derives from
T2  T1 ¼ $ln 2 (26)
4 p k t1 other non-statistical sources, e.g. from the knowledge of the
instrumentation and phenomena. The total uncertainty attributed
and the thermal conductivity is given by to the tests is about 1.2%.

Fig. 6. Effective thermal conductivity of the two-phase media at different mean Fig. 7. Comparison between experiments and the present theoretical model for a two-
temperatures. phase media.
466 S. Corasaniti, F. Gori / International Journal of Thermal Sciences 112 (2017) 460e469

3.3. Comparison between theoretical and experimental results effective thermal conductivity is
" #
The experimental effective thermal conductivity, measured in 2$j$kf þ ð3  2$jÞks
the two-phase media, is compared to the theoretical results of the keff ¼ kf $ (30)
ð3  jÞkf þ j$ks
present model. The porosity of the medium, composed of spherical
glass beads, is measured as 0.395, but it requires a correction, close where kf is the thermal conductivity of the continuous phase and ks
to the probe wall, because the diameter of the cylinder is smaller the thermal conductivity of the solid sphere. For a two-phase media
than the diameter of the glass beads. Indeed, the effect of the wall is kf is due to air and water, and is given by
to increase the local porosity near the cylindrical wall of the probe
where the measurements are carried on. kf ¼ ka $ε$ð1  jÞ þ kw $ε$j (31)
Two empirical models are used to calculate the local porosity of
the glass beads near the cylindrical wall of the thermal probe. Ac- Woodside and Messner [12] investigated a medium composed
cording to Bortolozzi and Deiber [53] the porosity must be cor- by two or more phases, and proposed an effective thermal con-
rected according to the following empirical relation ductivity depending on the volumetric content of each components
according to
  
Bn
εloc ¼ ε$ 1 þ A exp  ¼ 0:5135 (28) f f
keff ¼ ka a $kww $ks s
f
(32)
dp

where ε is the porosity far from the probe, n the distance from the where fa, fw and fs are the volumetric contents of air, water and
probe, dp the diameter of the glass beads, A ¼ 0.3 and B ¼ 7.5 are solid.
empirical constants. Brailsford and Major [7] developed a model for a two-phase
According to Fand and Yamamoto [54] the local porosity, due to media, extended to three-phases mixture. The effective thermal
the wall effect, is given by conductivity is

   fa $ka þ fs $ks $2k3kþk

a
þ fw $kw $2k3kþk
a
dp
εloc ¼ ε 2:778  1:778 exp  0:1 ¼ 0:5223 (29) keff ¼ a s a w
(33)
2R fa þ fs $2k3kþk
a
þ fw $2k3kþk
a
a s a w

where dp is the glass bead diameter and R is the cylinder/probe Felske [37] considered the spherical unit composed of a com-
radius. An average porosity equal to 0.51 is used in the predictions posite particle surrounded by a layer of the continuous medium
of the theoretical model. with a negligible contact resistance between particle and contin-
Fig. 7 reports the comparison between the experimental mea- uum. The effective thermal conductivity is

h     i h     i
j j j j
ð1  jÞ$ 2 1 þ 1ε  2 1  1ε $kks þ ð1 þ 2jÞ$ 2 1 þ 1ε ks
ka
 2 1  1ε $kkw
keff ¼ ka $ h    iw
h    ia (34)
j j j j
ð2 þ jÞ$ 1 þ 2 1ε  1  1ε $kks þ ð1  jÞ$ 2 þ 1ε $kks  2 1  1ε ,kkw
w a a

surements and the theoretical predictions of the present model

versus the mean temperature of the probe. The thermal conduc-
tivity used for the glass beads is ks ¼ 1.05 W/(mK), constant in the Kunii and Smith [19] derived a model for predicting the effective
temperature range considered, while that of water is thermal conductivity of beds of unconsolidated particles containing
kw ¼ 0.569 þ 1.8$103$T  7.72$107$T2 W/(m K), variable with the
temperature T, in  C.
Fig. 7 presents also the deviations of ±2.5% from the effective
thermal conductivity, in order to give a visual inspection of the
uncertainty. The agreement is quite acceptable.

4. Comparison with experiments of the literature for three-

phase media

The predictions of the present theoretical model, along to other

well-known theoretical models of the literature, are compared to
the experimental results of the literature for three-phase media
with different degrees of saturation. The theoretical models of the
literature employed are reported in the following section.

4.1. Theoretical models of the literature

The Maxwell approach [3] has been developed for a random

distribution of solid spheres in a continuous medium, under the Fig. 8. Comparison between the experiments for three-phase media [55], and the
assumption of sufficiently dilute dispersion of spheres. The theoretical results of the present, k(3P), and other models of the literature.
 S. Corasaniti, F. Gori / International Journal of Thermal Sciences 112 (2017) 460e469 467

models the values of kw ¼ 0.61 W/(m K) for water and ka ¼ 0.026 W/

(m K) for air are assumed. The continuous line of Fig. 8 represents
the results of the present theoretical model, k (3P), obtained by
using Eqs. (12)e(17), Case 1, in the range of saturation degree
0 < j  0.53, and Eqs. (18)e(24), Case 2, in the range of saturation
degree 0.53 < j  0.85. The others curves represent the theoretical
results of the other models discussed in Section 4.1.
The agreement between the results of the present model, k (3P),
and those of Maxwell [3], with the experiments are quite good,
while the other models are not in fairly good agreement, especially
at the intermediate saturation degrees.
Fig. 9 presents the comparison with the experimental results of
[56], for a three-phase medium with porosity equal to 0.52 and
solid thermal conductivity 2.85 W/(m K). The continuous line of
Fig. 9 represents the results of the present theoretical model, k (3P),
obtained by using Eqs. (12)e(17), Case 1, in the range of saturation
degree 0 < j  0.4764, and Eqs. (18)e(24), Case 2, in the range of
saturation degree 0.4764 < j  0.81. The agreement between the
results of the present model, k (3P), and those of Maxwell [3], with
Fig. 9. Comparison between the experiments for three-phase media [56], and the
the experiments are reasonably good, while the other models are
theoretical results of the present, k(3P), and other models of the literature. not in fair agreement, especially at the intermediate saturation
degrees.
Fig. 10 presents the comparison with the experimental results of
[55] for a New Brunswick (NB01) silty loam or clay loam, with
porosity equal to 0.54 and solid thermal conductivity equal to
4.33 W/(m K). The continuous line of Fig. 10 represents the results
of the present theoretical model, k (3P), obtained by using Eqs.
(12)e(17), Case 1, in the range of saturation degree 0 < j  0.37, and
Eqs. (18)e(24), Case 2, in the range of saturation degree
0.37 < j  0.75. The agreement between the results of the present
model, k (3P), and the experiments is quite good, while the model
of Maxwell [3] predict smaller values, and the other models are not
in agreement, especially at the intermediate saturation degrees.

5. Conclusions

Heat transfer by conduction in two and three-phase media, with

solid spherical particles of the same diameter, and porosity greater
than 0.4764 is investigated theoretically. The effective thermal
conductivity is evaluated across a unit cell made with a rhomboidal
base in order to be able to calculate porosity greater than 0.4764.
The measurements are carried on with a thermal probe, hand-
Fig. 10. Comparison between the experiments for three-phase media [55], and the
made in the laboratory. The two-phase media is composed of
theoretical results of the present, k(3P), and other models of the literature.
glass beads of the same diameter, 3 mm, and is water saturated. The
tests are carried out at different powers of the heater. The experi-
stagnant fluid and neglecting radiation. The following equation has mental thermal conductivity of the two-phase media is measured
been obtained with the perfect line source theory. The experimental results of the
" two-phase media are compared favorably with the present theo-
  #
ka 1 retical model with a deviation smaller than 2.5%. The theoretical
keff ¼ ka $ ε þ d$ð1  εÞ$ 4 þ g$ (35) model employs a sample porosity corrected because of the wall
ks
effect. The theoretical model is extended to three-phase media,
composed of air, water and spherical solid particle. The experi-
with d ¼ 0.895, g ¼ 2/3 and 4 a parameter depending on the
mental results of the literature for three-phase soils are compared
packing type of the spheres.
with the present theoretical model, and some theoretical models
available in the literature, with a reasonably good agreement for the
4.2. Comparison between theoretical models and experimental present model and the Maxwell one.
results
Nomenclature
Figs. 8e10 show the comparison between the experimental
results of [55,56] for three-phase media, the predictions of the
present theoretical model, k (3P), and of the models of the literature Latin symbol
reported in Sec. 4.1, versus the degree of saturation. a cell dimension, m
Fig. 8 reports the results for the samples investigated in Ref. [55], d diameter, m
Ontario (ON02) sand to silt loam, with porosity equal to 0.51 and k thermal conductivity, W/(m K)
solid thermal conductivity equal to 2.55 W/(m K). In the theoretical I electric current, A
468 S. Corasaniti, F. Gori / International Journal of Thermal Sciences 112 (2017) 460e469

l length, m 1960;6(1):71e8.
[20] Masamune S, Smith JM. Thermal conductivity of beds of spherical particles.
q heat transfer for unit length, W/m
Ind Eng Chem Fund 1963;2(2):136e43.
Q heat transfer, W [21] Zehner P, Schlunder EU. Thermal conductivity of granular materials at mod-
R radius, m erate temperatures. Chem Ingr Techn 1970;42:933e41 (in German).
S surface, m2 [22] Gori F. A theoretical model for predicting the effective thermal conductivity of
unsaturated frozen soils. In: Proceedings of the 4th international conference
t time, s on permafrost, Fairbanks, USA; 1983. p. 363e8.
T temperature, K or  C [23] Gori F. Prediction of effective thermal conductivity of two-phase media
TC thermocouple (including saturated soils). In: XVI international congress of refrigeration,
Paris, vol. B1e26; 1983. p. 525e31.
TP thermal probe [24] Gori F. On the theoretical prediction of the effective thermal conductivity of
V volume, m3 bricks. In: The eight international heat transfer conference, S. Francisco
x,y,z cartesian coordinates, m (U.S.A.), vol. 2; 1986. p. 627e32.
[25] Cheng P. Heat conduction in a packed bed with wall effects. Int Commun Heat
Mass Transf 1986;13:11e21.
Greek symbol [26] Hsu CT, Cheng P, Wong KW. Modified Zehner-Schlunder models for stagnant
a thermal diffusivity, m2/s thermal conductivity of porous media. Int J Heat Mass Transf 1994;37(17):
2751e9.
b rhombic angle, rad [27] Chen ZQ, Cheng P, Hsu CT. A theoretical and experimental study on stagnant
D difference thermal conductivity of bi-bispersed porous media. Int Comm Heat Mass
ε porosity Transf 2000;27(5):601e10.
[28] Gori F, Corasaniti S. Theoretical prediction of the thermal conductivity of soils
f volumetric content at moderately high temperatures. J Heat Transf ASME 2002;124(6):1001e8.
j saturation degree [29] Rubiolo P, Gatt J. Modeling of the radiative contribution to heat transfer in
porous media composed of spheres or cylinders. Int J Therm Sci 2002;41:
401e11.
Subscripts [30] Gori F, Corasaniti S. Experimental measurements and theoretical prediction of
a air the thermal conductivity of two-and three-phase water/olivine systems. Int J
eff effective Thermophys 2003;24(5):1339e53.
[31] F. Gori and S. Corasaniti, Temperature variations inside dry and partially
exp experimental data frozen mars soils, IEEE aerospace conference, Big Sky, MT, USA, March 8-15,
f fluid 2003.
loc local [32] Gori F, Corasaniti S. Theoretical prediction of the thermal conductivity and
temperature variation inside mars soil analogues. Planet Space Sci
p particle 2004;52(1e3):91e9.
s solid [33] Gori F, Corasaniti S. Detection of a dryefrozen boundary inside Martian
t total regolith. Planet Space Sci 2008;56:1093e102.
[34] Gori F, Corasaniti S. Thermal properties and temperature variations in martian
v voids
soil analogues. In: Space science research developments. Nova Science Pub-
w water lishers, Inc.; 2011. p. 203e33.
∞ undisturbed fluid [35] Tarnawski VR, Cleland DJ, Corasaniti S, Gori F, Mascheroni RH. Extension of
soil thermal conductivity models to frozen meats with low and high fat
contents. Int J Refrig 2005;28(6):840e50.
References [36] Leong WH, Tarnawski VR, Gori F, Buchan GD, Sundberg J. Inter-particle con-
tact heat transfer models: an extension to soils at elevated temperatures. Int J
[1] Crane RA, Vachon RI. A prediction of the bounds on the effective thermal Energy Res 2005;29(2):131e44.
conductivity of granular materials. Int J Heat Mass Transf 1977;20:711e23. [37] Felske JD. Effective thermal conductivity of composite spheres in a continuous
[2] Prakouras AG, Vachon RI, Crane RA, Khader MS. Thermal conductivity of medium with contact resistance. J Heat Mass Transf 2004;47:3453e61.
heterogeneous mixtures. Int J Heat Mass Transf 1978;21:1157e66. [38] Davarzani H, Marcoux M, Quintard M. Effect of solid thermal conductivity and
[3] Maxwell JC. A treatise on electricity and magnetism. third ed. New York: particle-particle contact on effective thermo-diffusion coefficient in porous
Dover; 1954. media. Int J Therm Sci 2011;50:2328e39.
[4] Eucken A. Forsh. Gebiete Ingenieur, vol. B3. VDIeForschungsheft; 1932. 353. [39] F. Gori, S. Corasaniti. Effective thermal conductivity of three-phase soils. In:
[5] Lichtenecker K. Die dielektrizitatskonstante naturlicher und kunstlicher mis- ASME international mechanical engineering conference and exposition, pro-
chkorper. Phys Z 1926;27:115e8. ceedings (IMECE), Houston, Texas, USA, 7 (D), 2369e2374, November 9e15,
[6] de Vries DA. In: Van Wijk WR, editor. Physics of plant environment. New York: 2012.
John Wiley; 1963. p. 210e35. [40] F. Gori, J.F. Ciparisse and S. Corasaniti. Thermal anisotropic properties of
[7] Brailsford AD, Major KG. The thermal conductivity of aggregates of several composite materials. In: Proceedings of the IEEE aerospace conference, Big
phases, including porous materials. Br J Appl Phys 1964;15:313e9. Sky Montana, March 3-10, 2012.
[8] Rayleigh Lord. On the influence of obstacles in rectangular order upon the [41] Gori F, Corasaniti S, Worek WM, Minkowycz WJ. Theoretical prediction of
properties of a medium. Philos Mag 1892;56:481e502. thermal conductivity for thermal protection systems. Appl Therm Eng
[9] Runge I. Zur electrischen Leitfahigkeit metallisher Aggregate. Z Tech Phys 2012;49:124e30.
1925;6:61e8. [42] Gori F, Corasaniti S, Ciparisse JF. Theoretical prediction of the anisotropic
[10] Meredith RE, Tobias CW. Resistance to potential flow through a cubical array effective thermal conductivity of composite materials. In: Proceedings of the
of spheres. J Appl Phys 1969;31:1270e3. ASME international mechanical engineering conference and exposition,
[11] Luikov AV, Shashkov AG, Vasiliev LL, Fraiman YUE. Thermal conductivity of IMECE, November 9-15, Houston, Texas, USA; 2012. p. 91e6. IMECE2012-
porous systems. Int J Heat Mass Transf 1968;11:117e40. 86114.
[12] Woodside W, Messmer JH. Thermal conductivity of porous media. I uncon- [43] Gori F, Corasaniti S. New model to evaluate the effective thermal conductivity
solidated sands. J Appl Phys 1961;32(9):1688e99. of three-phase soils. Int Commun Heat Mass Transf 2013;47:1e6.
[13] Zakri T, Laurent JP, Vauclin M. Theoretical evidence for “Lichtenecker's [44] Baloyi J, Bello-Ochende T, Meyer JP. Minimization of thermal resistance in an
mixture formulae” based on the effective medium theory. J Phys D Appl Phys air cooled porous matrix made up of solid spheres with heat generation. Int
1998;31:1589e94. Commun Heat Mass Transf 2012;39:966e70.
[14] Krupiczka R. Analysis of thermal conductivity in granular materials. Int Chem [45] Tarkhanyan RH, Niarchos DG. Reduction in lattice thermal conductivity of
Eng 1967;7:122e44. porous materials due to inhomogeneous porosity. Int J Therm Sci 2013;67:
[15] Pande RN, Gori F. Effective media formation and conduction through unsat- 107e12.
urated granular materials. Int J Heat Mass Transf 1987;30(5):993e1000. [46] Gori F, Corasaniti S. Effective thermal conductivity of composites. Int J Heat
[16] Pham DC. Weighted effective medium approximations for conductivity of Mass Transf 2014;77:653e61.
random composites. Int J Heat Mass Transf 2008;51:3355e61. [47] Corasaniti S, Gori F. Further considerations on anisotropic thermal efficiency
[17] Tarnawski VR, Leong WH, Gori F, Buchan GD, Sundberg J. Inter-particle con- of symmetric composite. Int J Heat Mass Transf 2015;88:836e43.
tact heat transfer in soil systems at moderate temperatures. Int J Energy Res [48] Gori F, Pietrafesa M. On the construction method of the thermal probe used
2002;26:1345e58. for the measurements of thermal properties of insulating materials. In:
[18] Lu Y, Lu S, Horton R, Ren T. An empirical model for estimating soil thermal Beasley DE, Chen JLS, editors. ASME - WAM, HT - 7A - 5, Atlanta, December
conductivity from texture, water content, and bulk density. Soil Sci Soc Am J 1991, fundamental experimental measurements in heat transfervol. 179.
2014;78:1859e68. ASME, HTD; 1991. p. 83e9. Book N. H00663-1991.
[19] Kunii D, Smith JM. Heat transfer characteristic of porous rocks. AIChE J [49] Gori F, Marino C, Pietrafesa M. Experimental measurements and theoretical
 S. Corasaniti, F. Gori / International Journal of Thermal Sciences 112 (2017) 460e469 469

predictions of the effective thermal conductivity of porous saturated two natural convection in a porous media. Chem Eng Sci 2001;56:157e72.
phases media. Int Commun Heat Mass Transf 2001;28(8):1091e102. [54] Fand RM, Yamamoto RH. In: Proc. 9th int. heat transfer conf., Jerusalem, 15 e
[50] Carslaw HS, Jaeger JC. Conduction of heat in solids. London: Oxford University PB e 07; 1990.
Press; 1959. €nenberger J, Momose T, Wagner B, Leong WH, Tarnawski VR. Canadian
[55] Scho
[51] Wechsler AE. Compendium of thermophysical properties. Specific methods, field soils I. Mineral composition by XRD/XRF measurements. Int J Thermo-
vol. II. NY and London: Plenum Press; 1984. phys 2012;33(2):342e62.
[52] ISO GUM. Guide to expression of uncertainties in measurements. 2008. ISO/ [56] Singh AK, Singh R, Chaudhary DR. Prediction of effective thermal conductivity
IEC Guide 98e3. of moist porous materials. J Phys D Appl Phys 1990;23:698e702.
[53] Bortolozzi RA, Deiber JA. Comparison between two- and one-field models for

View publication stats