INTERNATIONAL JOURNAL OF HEAT AND TECHNOLOGY

Vol.33, No.1, 2015

ANALYSIS OF CONDUCTION-RADIATION HEAT TRANSFER WITH VARIABLE

THERMAL CONDUCTIVITY AND VARIABLE REFRACTIVE INDEX:
APPLICATION OF THE LATTICE BOLTZMANN METHOD
Ahmed Mahmoudi 1 and Imen Mejri 1
1 Unitéde Recherche Matériaux, Energie et Energies Renouvelables (MEER), Facultédes Sciences de Gafsa,
B.P.19, Zarroug, Gafsa, 2112, Tunisie

ABSTRACT

The effect of variable thermal conductivity and variable refractive index on transient conduction and radiation heat transfer
in a planar medium is investigated. Thermal conductivity of the medium is assumed to vary linearly with temperature,
while the other thermo-physical properties are assumed constant. The radiative transfer equation and the nonlinear energy
equation are solved using lattice Boltzmann method (LBM). The effects of various parameters are studied. The LBM
results are compared with those available in literature and a good agreement is found.

Keywords: Lattice Boltzmann method, Conduction, Radiation, Planar medium.

1. INTRODUCTION an open T-shaped cavity cooled by a steady stream of

convection that intrudes into a solid conducting wall subject
Combined conduction-radiation heat transfer has to an area fraction. Lorenzini et al. [12] Lorenzini et al.
numerous engineering applications such as heat transfer optimized the shape of cavities that intrude into a cylindrical
through semitransparent and porous materials, multilayered solid body. The objective is to minimize the global thermal
insulations, burning of coal in furnaces, fluidized beds, resistance between the solid body and the cavities. The
nuclear engineering, and so on. The cases of constant results indicate that the optimal distribution of the hot spots
thermal conductivity and unity refractive index in combined is affected not only by the complexity of the configuration
conduction–radiation heat transfer problems have been (larger N) but also by the area of cavities fraction. Mahmoudi
studied in detail by many investigators [1-5]. Because of the et al. [13] studied MHD natural convection in a nanofluid-
mathematical complexities, a limited literature is available filled open cavity with non uniform boundary condition in
that individually deal with the effects of variable thermal the presence of uniform heat generation/absorption. The
conductivity [6] and constant and/or variable refractive index found results show that the heat transfer rate decreases with
[7]. The case of variable thermal conductivity and variable the rise of the Hartmann number and increases with the
refractive index finds application in the thermal analysis of augmentation of the Rayleigh number. Mahmoudi et al. [14]
graded index medium [8]. The present work is, therefore, studied the effect of the magnetic field intensity and direction
aimed at the analysis of conduction and radiation heat on natural convection in a square enclosure filled with
transfer in a participating medium, by lattice Boltzmann nanofluid. The found results show that the heat transfer and
method. The effect of variable thermal conductivity and fluid flow depends strongly upon the direction of magnetic
constant and/or variable refractive index are considered. In field. Mejri et al. [15] studied magnetic field effect on
recent years, use of the lattice Boltzmann method (LBM) as a natural convection in a nanofluid filled enclosure with non-
potential computational fluid dynamics (CFD) tool for the uniform heating on both side walls. The authors used lattice
solution of a large class of problems in science and Boltzmann method to solve the coupled equations of flow
engineering has gained a momentum. Chaabane et al. [9] and temperature fields. The found results show that the heat
used the lattice boltzmann method to solve the energy transfer rate increases with the increase of the Rayleigh
equation in two dimensional enclosure of a problem number but it decreases with the increase of the Hartmann
involving a variety of boundary conditions, they found that number.
the LBM results agree very well with the finite volume In this paper, the 1-D conduction-radiation problem with
results. Mejri et al. [10] studied 1-D conduction-radiation variable thermal conductivity and variable refractive index is
problem by lattice Boltzmann method. LBM is used to solve solved by the lattice Boltzmann method. The effects of
the energy equation and the radiative transfer equation. The various parameters such as the scattering albedo, the
found results are compared with those available in the conduction–radiation parameter, the wall emissivity and the
literature and a very good agreement was found. Lorenzini et thermal conductivity parameter are studied. In order to check
al. [11] optimized the dimensionless excess of temperature of on the accuracy of the numerical technique employed for the

1
solution of the considered problem, the present numerical For a linear anisotropic phase
code was validated with the published study. function p  1  a cos  cos  ' , where a is the anisotropy
factor (1  a  1) , Eq. (6) can be written as:
2. MATHEMATICAL FORMULATION

Consider a 1-D planar conducting–radiating medium of S( x , s , t )  (1  ω) I b ( x , t )

length L, with variable thermal conductivity k. Other
thermo-physical properties such as density ρ, specific heat cp, ω (7)
4π  '4
 I ( x, s , t )(1  a cos  cos  ')d '
and optical properties such as extinction coefficient β and
scattering albedo ω are assumed constant. The system is
initially at temperature TE and for time t > 0, its west
ω
boundary is raised to temperature T W. The variation of S( x, s , t )  (1  ω) I b ( x , t )  G( x, t )  a cos  q R ( x, t ) (8)
thermal conductivity with temperature is taken as: 4π

k  k0   '(T - Tw )  is the polar angle.

(1)

Where k0 is the reference thermal conductivity and δ′ is
the coefficient of thermal conductivity variation. The
G  2 
 0
I ( ) sin( )d 
refractive index n of the medium assumes either a constant (9)
value (n ≥ 1) or varies linearly with
distance n( x)  nW  (nW  nE ) x / L , where nW and nE are 

the refractive indices on the west and the east faces of the
qR  2 
 0
I ( ) cos( )sin( )d 
(10)
medium, respectively. For the problem under consideration,
the energy equation is given by:
For a diffuse-gray boundary with temperature T E/W and
T  T q emissivity εE/W, the boundary intensity is given by:
 cP   (k ) R
t x x x (2)
(1  ε E )
π  '2 
I ( xE , s )  ε E I b ( xE )   I ( xE , s ) n.s  n.s > 0d ' (11)
Where qR is the radiative heat flux and qR for a
x
medium with a variable refractive index is given by:
(1  ε W ) (12)
I ( xW , s )  ε W I b ( xW ) 
π   I ( xW , s ) n.s  n.s < 0d '
q R T 4  ' 2
 β(1  ω)(4n2 π  G)
x π (3) For solving the considered problem, the following
dimensionless numbers are defined:
G is the incident radiation. To solve for G at any location
x, information about the intensity I distribution is required k0   ' Tw N
which for any direction s is obtained from the following N ,    2t ,  =
radiative transfer equation: 4 Tw 3
k0 (13)

1 I ( x , s , t ) I ( x , s , t ) N is the conduction–radiation parameter, ξ is the

  βI ( x , s , t )
c t s dimensionless time and δ is the variable thermal conductivity
βω (4) parameter.
4π  '4
 β(1  ω) I b ( x , t )  I ( x , s , t ) p( s '  s )d '

2.1 Energy equation by LBM

Where c is the speed of light in the medium,
Ib  n2 T 4 / π is the Planck’s black body intensity, d is the For a one-dimensional planar geometry, in the LBM with
solid angle and p(s '  s ) is the anisotropic scattering phase a D1Q2 lattice, the discrete Boltzmann equation with
Bhatanagar- Gross-Krook (BGK) approximation is given by
function. For a given direction s , if the upstream point lies [8]:
on the boundary, then its values have to be computed from
the radiative boundary condition. Eq. (4) can be recast as:
fi ( x, t ) 1
 ei .fi ( x , t )    fi ( x, t )  f eq i ( x, t )  i = 1 and 2 (14)
t 
1 I ( x , s , t )
 s .I ( x, s , t )  βI ( x , s , t )  βS( x , s , t )
c t (5) Where f i is the particle distribution function denoting the
number of particles at the lattice node x and time t moving
Where S is the radiative source term given as: in direction i with velocity ei along the lattice
x  ei t connecting the neighbors,  is the relaxation time,
ω (6)
4π  '4
S( x, s , t )  (1  ω) I b ( x , t )  I ( x , s , t ) p( s '  s )d '

2
and f eq i is the equilibrium distribution function. The Discretizing Eq. (23), we obtain:
relaxation time  for the D1Q2 lattice is computed from:
Ii ( x  ei t , t  t )  Ii ( x, t )  teiβ(Si  Ii ) i =1,...,M (24)
k t
 
 c p ei
2
2 Clearly in Eq. (24), the term on the right hand side can be
(15)
seen as the collision term in the LBM, where Ii is the
intensity particle distribution function. Using the standard
For this lattice, the two velocities e1 and e2 , and their
LBM terminology, Eq. (24) can be written as:
corresponding weights w1 and w 2 , are given by:

t
x x I i ( x  ei t , t  t )  I i ( x, t )   I eq ( x, t )  I i ( x, t )  (25)
e1 
t
e2  
t
R  i
(16)
Where  R is the relaxation time for the collision process
1
w1  w 2  and I ieq is the equilibrium particle distribution function.
2 (17)

After discretization, Eq. (14) is written as: 1

R  and I ieq  Si
eiβ (26)
1
fi ( x  ei t , t  t )  fi ( x , t )   fi ( x , t )  f eq i ( x , t ) 
 (18) The irradiation G and the heat flux qR due to diffuse
radiation, are computed from the following:
The temperature is obtained after summing f i overall
direction:
 i
G  4π  I i sin  i sin( ) (27)
2
 f ( x, t )
i 1, M
T( x, t )  i
i 1,2
(19)
q R  2π  Ii sin  i cos  i sin( i ) (28)
To process Eq. (18), an equilibrium distribution function i 1, M
is required, which for a conduction-radiation problem is
given by: 3. RESULTS AND DISCUSSION

f eq i ( x, t )  w i T( x, t ) In this paper, the energy equation of a 1-D transient

(20)
conduction- radiation problem with variable thermal
To account for the volumetric radiation, the energy conductivity k, is solved with LBM. Initially the medium is
equation in the LBM formulation, Eq. (18) is modified to: at temperature TE . For t  0 , the west boundary temperature
is maintained at TW  2TE . The medium is absorbing,
f i ( x  ei t , t  t )  f i ( x , t ) emitting and isotropically scattering. The non dimensional
1 t w i q R time step   104 was considered and steady state condition
  f ( x, t )  f eq i ( x , t )  
  i ρcP x (21) was assumed to have been achieved when the maximum
variation in temperature at any location between two
2.2 Radiative information by LMB consecutive time levels did not exceed 105 . First the effect
of the grid size to the non-dimensional temperature results
Multiplying Eq.(5) throughout by the speed of light c, the (T/Tw) is studied by comparing the steady state (SS) results
radiative transfer equation along any lattice link designated at three locations in the medium for several grid sizes for n =
by the index i can be written as: 1,δ = 0, a = 0, β=1.0, N=0.1, TE=0.0, Tw=1.0, ω=0.5 and εw
= εE = 1.0. The results are listed in table.1 and show that the
DI i I non-dimensional temperature is practically independent of
( x , s , t )  i  c .I i  c ( I i  Si ) i =1,...,M
Dt t (22) the grid size. In table.2, for ξ = 0.05, n = 1, δ = 0, a = 0,
β=1.0, N=0.1, TE=0.0, Tw=1.0, ω=0.5, εw =1.0 and εE = 1.0
Let ei be the velocity of propagation along the ith lattice or 0.0, the non-dimensional temperature results (T/T w) are
compared with those reported in the literature [1-2] at three
link of the D1QM lattice structure. If the velocity of light locations in the medium, It is observed that the LBM results
c is fictitiously made equal to the velocity of particle are in good agreements with the published results.
propagation in the LBM, c  e a convenient tool would be Fig.1a-c shows the effect of the conduction-radiation
obtained to solve the radiative transfer equation using the parameter by comparing the LBM results (T/T w) with those
LBM approach published at several times ξ, for n = 1, δ = 0, a = 0, β=1.0, ω
= 0.0 and N = 0.01, 0.1 and 1.0. The LBM results are in
I i good agreements with the published results.
 ei .I i  eiβ( Si  I i ) i =1,...,M
t (23)

3
 Fig.2a-c shows for several times the effect of the a
scattering albedo ω=0.0, 0.5 and 0.9 for n = 1, δ = 0, a = 0,
β =1.0 and N=0.01 by comparing the LBM results (T/T w)
with those published. Excellent agreement is also found.

Table 1. Effect of grid size on non-dimensional temperature

steady state

x/L = 0.25 x/L = 0.50 x/L = 0.75

Nx=20 M=4 0.8265 0.6076 0.3339
M=8 0.8356 0.6204 0.3438
M=16 0.8389 0.6254 0.3479
M=32 0.8400 0.6270 0.3492 b

M=32 Nx=20 0.8400 0.6270 0.3492

Nx=30 0.8438 0.6210 0.3365
Nx=40 0.8269 0.6181 0.3441
Nx=60 0.8227 0.6152 0.3425

Table 2. Comparison of transient temperature

x/L = 0.25 x/L = 0.5 x/L = 0.75

ε W  1.0 [1] 0.4888 0.1778 0.0591
[2] 0.4889 0.1773 0.0588
ε E  1.0 LBM 0.4893 0.1787 0.05724 c
ε W  1.0 [1] 0.5030 0.2005 0.0833
[2] 0.5031 0.2001 0.0830
ε E  0.0 LBM 0.5037 0.1993 0.0841

Figure 2. Comparison of the non dimensional temperature

(T/Tw) at different ξ for (a) ω=0.1, (b) 0.5 and (c) 0.9

Fig.3a-c show the effect of the emissivity εW = 0.1, 0.5 and

0.9 by comparing the LBM results (T/T w) with those
b published for n = 1, δ = 0, a = 0, β = 1.0, ω=0.0, N=0.01 and
εE=1.0. It is shown that the LBM results are in good
agreements with the published results.
Fig.4 shows the effect of the extinction coefficient by
comparing the steady-state (SS) results obtained by the LBM
and the published results, for n = 1, δ = 0, a = 0, ω=0.0 and
N = 0.1. These comparisons are shown for β = 0.1, 1.0 and
2.0. Excellent agreement is also found.
Fig.5a-b shows the effect of the variable thermal
conductivity parameter for different extinction coefficients
(β = 0.1 and 1.0) by comparing the steady-state results
c obtained by the LBM and the published results, for n = 1, a =
0, ω=0.5, N = 0.5, εw = 1.0 and εE = 0.5.The LBM results are
in good agreement with those published.

Figure 1. Comparison of the non dimensional temperature

(T/Tw) at different instants ξ for (a) N = 0.01, (b) 0.1 and (c)
1.0

4
 a a

b b

c
Figure 5. Comparison of the non dimensional temperature
(T/Tw) for different δ for (a) β = 0.1 and (b) β = 1.0

Figure 3. Comparison of the non dimensional temperature

(T/Tw) at different ξ for (a) εW = 0.1, (b) 0.5 and (c) 0.9

Figure 6. Comparison of the non dimensional temperature

(T/Tw) for different δ for (a) N = 0.01 and (b) N = 0.1
Figure 4. Comparison of the non dimensional temperature
(T/Tw) for different β
Fig.6a-b shows the effect of the variable thermal
conductivity parameter for different conduction– radiation
parameters (N = 0.01 and 0.1) by comparing the steady-state
results obtained by the LBM and the published results, for
n = 1, a = 0, ω=0.0, β = 1.0, εw = 1.0 and εE = 0.5. The LBM
results are in good agreement with those published.
Fig.7a-c shows the effect of the variable thermal
conductivity parameter for different scattering albedo (ω =
0.0, 0.5 and 0.1) by comparing the steady-state results
obtained by the LBM and the published results, for a = 0,
N=0.5, β = 1.0,

5
εw = 1.0 and εE = 0.5. The LBM results are in good a = 0, ω=0.0, β = 10.0, εw = εE = 1.0, δ=0.0 and TW =3TE.
agreement with those published. The LBM results are in good agreement with those published.

Figure 9. Comparison of the non dimensional temperature

(T/Tw) for different refractive index

Figure 7. Comparison of the non dimensional temperature

(T/Tw) for different δ for (a) ω = 0.0, (b) 0.5 and (c) 0.9 Figure 10. Comparison of the non dimensional temperature
(T/Tw) for variable refractive index

Figure 8. Comparison of the non dimensional temperature Figure 11. Comparison of the non dimensional temperature
(T/Tw) for different N (T/Tw) for variable refractive index

Fig.8 shows the effect of the conduction– radiation Fig.9 shows the effect of the refractive index by comparing
parameter for n E =nW=1.5 by comparing the steady-state the steady-state results obtained by the LBM and the
results obtained by the LBM and the published results, for published results, for a = 0, ω=0.0, β = 10.0, εw = εE = 1.0,

6
δ=0.5, N=0.2 and TW =3TE. The LBM results are in good 9. R. Chaabane, F. askri and S.B. Nasrallah, numerical
agreement with those published. modeling of boundary conditions for two dimansional
Fig.10 and 11 show the effect of the variable refractive conduction heat transfer equation using lattica
index for different conduction– radiation parameters and boltzmann method, International Journal of Heat and
different variable thermal conductivity parameters by Technology, vol. 28(2), pp. 51–57, 2010.
comparing the steady-state results obtained by the LBM and 10. I. Mejri, A. Mahmoudi, M.A Abbassi , A. Omri, Lattice
the published results, for a = 0, ω=0.0, β = 10.0, εw = εE = 0.0 Boltzmann simulation of conduction-radiation heat
and TW =3TE. The LBM results are in good agreement with transfer in a planar medium, International Journal of
those published. Heat and Technology, vol. 32, pp. 213–218, 2014.
11. G. Lorenzini, C. Biserni, F.L. Garcia, L.A.O. Rocha,
4. CONCLUSIONS Geometric optimization of a convective T-shaped cavity
on the basisof constructal theory, Int J of Heat and Mass
The LBM is used to analyze combined conduction– Transfer, vol. 55, 6951–6958, 2012.
radiation heat transfer in a planar absorbing, emitting and 12. G. Lorenzini, L. A. O. Rocha, C. Biserni, E. D. dos
scattering medium with variable thermal conductivity and Santos, L.A. Isoldi, Constructal Design of Cavities
variable refractive index. The radiative information and the Inserted Into a Cylindrical Solid Body, ASME Journal
energy equation are solved using the LBM. The results of the of Heat Transfer, Vol. 134 (7), pp. 0713011 - 0713016,
LBM-LBM formulation are compared with those available in 2012.
the literature. A very good agreement was found. 13. A. Mahmoudi, I. Mejri, M.A. Abbassi, A. Omri,
Analysis of MHD natural convection in a nanofluid-
filled open cavity with non uniform boundary condition
REFERENCES in the presence of uniform heat generation/absorption,
Powder Technology, vol.269, 275–289, 2015.
14. A. Mahmoudi, I. Mejri, M.A. Abbassi, A. Omri, Lattice
1. S.C. Mishra, A. Lankadasu, Analysis of Transient Boltzmann simulation of magnetic field direction effect
Conduction and Radiation Heat Transfer Using the on natural convection of nanofluid-filled cavity.
Lattice Boltzmann Method and the Discrete Transfer International Journal of Heat and Technology, vol. 32,
Method, Numerical Heat Transfer, Part A, vol.47, pp. pp. 9–4, 2014.
935–954, 2005. 15. I. Mejri, A. Mahmoudi, M.A Abbassi , A. Omri,
2. R. Raj , A. Prasad , P.R. Parida, S.C. Mishra, Analysis magnetic field effect on natural convection in a
of Solidification of a Semi transparent Planar Layer nanofluidfilled enclosure with non-uniform heating on
Using the Lattice Boltzmann Method and the Discrete both side walls, International Journal of Heat and
Transfer Method, Numerical Heat Transfer, Part A, Technology, vol. 32, pp. 127–133, 2014.
vol.49, pp. 279–299, 2006.
3. B. Mondal, S.C. Mishra, Application of the lattice
boltzmann method and the discrete ordinates method for
solving transient conduction and radiation heat transfer
problems, Numerical Heat Transfer, Part A, vol.52, pp. NOMENCLATURE
757–775, 2007.
4. S.C. Mishra, P. Talukdar, D. Trimis, F. Durst, a Anisotropy factor
Computational efficiency improvements of the radiative c Speed of light (m/s)
transfer problems with or without conduction-a cp Specific heat at constant pressure (JKg-1K-1)
comparison of the collapsed dimension method and the ei Lattice speed (m/s)
discrete transfer method, Int J of Heat and Mass f Internal energy distribution functions (K)
Transfer, vol. 46, 3083–3095, 2003. f eq Equilibrium internal energy distribution (K)
5. S.C. Mishra, R.R Vernekar, Analysis of transport of G Incident radiation (W/m²)
collimated radiation in a participating media using the I Intensity of radiation (W/m²)
lattice Boltzmann method, Journal of Quantitative I eq Equilibrium intensity (W/m²)
Spectroscopy & Radiative Transfer, vol. 113, pp. 2088– k Thermal conductivity (Wm-1K-1)
2099, 2012. L Length of the planar geometry (m)
6. P. Talukdar, S.C. Mishra, Transient conduction and M Total number of discrete directions
radiation heat transfer with variable thermal N Conduction-radiation parameter(=kβ/4σTw3)
conductivity, Numerical Heat Transfer, Part A, vol.41, Nx Total number of node
pp.851–867, 2002. p Scattering phase function
7. X.L. Xia, Y. Huang, H.P. Tan, X.B. Zhang, qR Heat flux (W/m²)
Simultaneous radiation and conduction heat transfer in S Source term (W/m²)
a graded index semitransparent slab with gray t Time (s)
boundaries, Int. J. Heat Mass Transfer, vol.45, pp. T Temperature (K)
2673–2688, 2002. w Weight in the LBM
8. S.C. Mishra, N.A. Krishna, N. Gupta, G.R. Chaitanya,
Combined conduction and radiation heat transfer with
variable thermal conductivity and variable refractive
index, Int. J. Heat Mass Transfer, vol.51, pp. 83–90,
2008.

7
GREEK SYMBOLS

Δx Lattice spacing (m)

Δt Time increment (s)
α Thermal diffusivity (m2/s)
β Extinction coefficient (1/m)
δ Thermal conductivity parameter
ε Emissivity
γ Polar angle (rad)
σ Stefan–Boltzmann constant,5.67 10-8 (W/m2 K4)
ρ Density (Kg/m3)
ξ Non dimensional time (  2t )
τ Relaxation time for temperature (s)
τR Relaxation time for radiation (s)
ω Scattering albedo

SUBSCRIPT

E East
w West