Accepted Manuscript

Title: Thermally developing flow inside a porous-filled channel in the presence

of internal heat generation under local thermal non-equilibrium condition: a
perturbation analysis
Author: Maziar Dehghan, M.S. Valipour, S. Saedodin, Yasser Mahmoudi
PII:
DOI:
Reference:

S1359-4311(16)00024-7
http://dx.doi.org/doi: 10.1016/j.applthermaleng.2015.12.133
ATE 7547

To appear in:

Applied Thermal Engineering

Received date:
Accepted date:

2-8-2015
26-12-2015

Please cite this article as: Maziar Dehghan, M.S. Valipour, S. Saedodin, Yasser Mahmoudi,
Thermally developing flow inside a porous-filled channel in the presence of internal heat
generation under local thermal non-equilibrium condition: a perturbation analysis, Applied
Thermal Engineering (2016), http://dx.doi.org/doi: 10.1016/j.applthermaleng.2015.12.133.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service
to our customers we are providing this early version of the manuscript. The manuscript will
undergo copyediting, typesetting, and review of the resulting proof before it is published in its
final form. Please note that during the production process errors may be discovered which could
affect the content, and all legal disclaimers that apply to the journal pertain.

Thermally developing flow inside a porous-filled channel in the presence of internal

heat generation under local thermal non-equilibrium condition: A perturbation analysis

Maziar Dehghan1,*, M.S. Valipour1, S. Saedodin1, Yasser Mahmoudi2

Faculty of Mechanical Engineering, Semnan University, Semnan, Iran;

Department of Engineering, University of Cambridge, Cambridge, United Kingdom.

Highlights

Analytical study of thermally developing forced convection inside porous materials;

Performing a perturbation analysis to avoid utilising thermal boundary condition for

10
11

constant wall heat flux;

12
13

Effects of internal heat generation/absorption on Nusselt number and thermal entry

length;

Presenting the condition of zero-thermal entry length;

14
15
16

Abstract

17

In the present work a perturbation analysis is performed to study the thermally developing

18

forced convection heat transfer inside a channel filled with a porous material under local thermal

19

non-equilibrium (LTNE) condition. Internal heat generations within the solid and fluid phases

20

are considered. Channel walls are subjected to a constant heat flux. It is assumed that there is a

21

small temperature difference between the fluid and solid phases of the porous material. So that,

22

performing a perturbation analysis enables us to avoid utilizing models for the constant wall heat
* Corresponding author:
m-dehghan@semnan.ac.ir, Tel/Fax: +98-2333654122.

Page 1 of 33

23

flux boundary condition to investigate the hydrothermal behavior of the system. Therefore,

24

analytical solutions are developed for temperature difference between the solid and fluid phases

25

as well as the local Nusselt number in the porous medium. Effects of pertinent parameters such

26

as dimensionless axial length, Biot number, effective thermal conductivity ratio and

27

dimensionless heat generation parameters on the Nusselt number are discussed. To further clarify

28

the validity of the solution provided, the obtained results are compared with the solutions for two

29

primary approaches (Models A and B) for the constant wall heat flux boundary condition.

30

Results show that both the Nusselt number and the thermal entry length increase with the

31

increase of thermal conductivity ratio. The Nusselt number and the thermal entry length are

32

found to decrease with the increase of the internal heat generation of the solid phase. It is further

33

observed that the Nusselt number as well as the thermal entry length is less sensitive to the solid

34

internal heat generation at high Biot numbers. Finally, it is found that when the effective thermal

35

conductivity ratio tends to infinity, the thermal entry length tends to zero for high Biot numbers.

36
37

Keywords: thermally developing flow; porous medium; local thermal non-equilibrium;

internal heat generation; perturbation analysis.

38
39

Nomenclature:

40

asf

specific surface area (m-1)

41

a constant used in Eqs. (21) and (26)

42

Bi

Biot number (= h sf sf H 2 / k s ,ef f )

43

cp

specific heat at constant pressure (J.kg-1.K-1)

44

dp

pore diameter (m)

2
Page 2 of 33

45

Dh

hydraulic diameter (m)

46

a function of y used in Eq. (15)

47

a function of x and y used in Eq. (15)

48

half of the channel gap (m)

49

hsf

fluid-to-solid heat transfer coefficient (W.m-2.K-1)

50

effective thermal conductivity ratio (= kf,eff /ks,eff)

51

kf

thermal conductivity of the fluid phase (W.m-1.K-1)

52

kf,eff

effective thermal conductivity of the fluid phase (W.m-1.K-1)

53

km

effective thermal conductivity of the porous medium (= kf,eff +ks,eff) (W.m-

54

.K-1)

55

ks

thermal conductivity of the solid phase (W.m-1.K-1)

56

ks,eff

effective thermal conductivity of the solid phase (W.m-1.K-1)

57

counter

58

Nu

Nusselt number

59

order of magnitude

60

Pe

Peclet number

61

Pr

Prandtl number

62

q"w

heat flux at the channel walls (W.m-2)

63

Re

Reynolds number

64

Temperature (K)
3
Page 3 of 33

65

Ti

inlet temperature (K)

66

Darcian velocity (m.s-1)

67

x*, y*

dimensional coordinates (m)

68

x, y

dimensionless coordinates

69

Greek letters

70

NE

LTNE intensity defined as (= s - f)

71

small parameter (= 1/hsfasf, W-1.m3.K)

72

dimensionless temperature

73

fluid density (kg.m-3)

74

parameter used in Eq. (28)

75

porosity of the porous medium

76

Subscripts

77

dev

developing

78

fluid phase

79

fd

fully developed component

80

inlet

81

bulk-mean value

82

counter

83

solid phase

4
Page 4 of 33

84
85

1. Introduction

86

Many analytical studies have investigated the problem of forced convection heat transfer in

87

porous materials due to its high academic and industrial significance [1, 2] such as electronic

88

cooling, heat pipes, nuclear reactors, drying technology and multiphase catalytic reactors. In

89

majority of this applications a porous material is inserted between the two parallel plates while

90

the walls of the plates are subjected to constant heat flux or constant temperature. In such case,

91

the undeveloped thermal boundary layer leads to a larger heat transfer rate in the thermal

92

entrance region compared to that of the corresponding thermally fully developed region [1]. One

93

of the main characteristics appearing due to the existence of a porous medium is that the local

94

temperature difference between the solid and fluid phases can substantially influence the heat

95

transfer process [3] and hence a proper analysis of the heat transport through the porous medium

96

is required.

97

There are two ways of modeling the energy equation in a porous medium: local thermal

98

equilibrium (LTE) and local thermal non-equilibrium (LTNE) models [3]. The LTE model holds

99

when the temperature difference between the solid and fluid phases is negligible. For some

100

applications of porous materials, however, this temperature difference is considerably large.

101

Hence, utilizing LTNE model is required. However, the use of LTNE model under the constant

102

wall heat flux thermal boundary condition requires additional information to account energy

103

transport between the two phases at the channel wall [4, 5]. This information is usually provided

104

in the form of models related to the constant wall heat flux boundary conditions [6, 7]. This, in

105

turn, makes the thermal behavior of the system dependent upon the applied model. Extra levels

106

of complexity are hence added to the problem, which involve devising the proper models and
5
Page 5 of 33

107

including them in the analysis. The applicability region of two common models of the constant

108

heat flux thermal boundary condition was mentioned by Dehghan [8] and Kim and Kim [9]. In

109

the present analysis, a perturbation analysis is performed to avoid utilizing the extra models for

110

the constant heat flux boundary condition.

111

While the forced convective heat transfer is naturally a developing phenomenon (i.e. the

112

temperature profile pattern varies in the thermal entry region as well as the heat transfer rate),

113

only few analytical studies considered the developing section listed in the following. Nield et al.

114

[10] and Kuznetsov et al. [11] used modified Graetz methodology to investigate the thermally

115

developing heat transfer through a porous material sandwiched in a channel or a tube with walls

116

at constant temperature. Nield and Kuznetsov [12] developed study of Refs. [10, 11] to the non-

117

Newtonian fluid flows. Haji-Sheikh et al. [13] found the temperature profile as well as the

118

Nusselt number in the thermal entrance region of rectangular porous passages. Kuznetsov and

119

Nield [14] studied thermally developing forced convection through a porous material in the slip-

120

flow regime. Ouyang et al. [15] utilized three different models for the constant heat flux

121

boundary condition (the so called Models A, B and C) to find the Nusselt number of thermally

122

developing forced convection through porous materials. Dehghan et al. [16] discussed the

123

validity of three models used by Ref. [15] in the entrance region. They showed that these models

124

may yield results that cannot be interpreted in terms of physics of the problem.

125

The role of internal heat generations within both the solid and fluid phases, on the

126

temperature field within the porous medium, has been given attention recently [7-9, 17, 18]. The

127

internal heat generation and absorption may arise from different sources, such as magnetic

128

heating, endothermic/exothermic reactions and viscous heating in both the solid and fluid phases

129

of a porous material. In addition, using porous materials as catalysts are usual in chemical
6
Page 6 of 33

130

engineering applications from burners to refiners. For example, a clean fuel such as hydrogen

131

can be obtained from endothermic reactions of methane reforming inside a porous catalyst

132

coated with Palladium nanoparticles [19, 20]. Meanwhile, a most recent application of porous

133

materials with heat generation/absorption is artificial porous media obtained by grids of PCMs

134

(Phase-Change Materials). These PCMs can act as endothermic or exothermic media. To analyze

135

the problem analytically, the heat generations/absorptions are assumed constant volume-

136

averaged values to find the overall thermal response of such systems. Understanding the thermal

137

response of such systems is mandatory in designing these systems and in selecting the materials

138

involved. Heat generations/absorptions appear as source/sink terms in the energy conversation

139

equations of the solid and fluid phases of a porous material and hence make these equations non-

140

homogeneous as well. Subsequently, obtaining an analytical solution of such problems is

141

accompanied with extra difficulties. The effect of constant internal heat generation on the

142

temperature field in the porous region under LTNE was addresses for the first time by Yang and

143

Vafai [17] in the fully developed region of a channel filled with a porous material. They

144

introduced the heat flux bifurcation phenomenon (i.e. different heat flux directions for the two

145

phases at the walls). Their analysis was further extended to the micro-scale by Mahmoudi [7] and

146

Dehghan et al. [18] to find the thermal response and the Nusselt number in the slip-flow regime

147

wherein the temperature jump phenomenon also govern at the walls. Dehghan et al. [18] used the

148

two-equation energy model [21] to define conditions at which the heat flux bifurcation occurs in

149

the slip-flow regime. They also discussed on the validity region of the one-equation energy

150

model (i.e. the LTE assumption). Dehghan [8] introduced the heat flux bifurcation (splitting)

151

phenomenon in a channel partially filled with a non-Darcian porous material. He discussed on

7
Page 7 of 33

152

the thermal response of such partially porous-filled channels under the LTNE condition using the

153

two common thermal boundary conditions (models A and B).

154

The above literature review demonstrate that solving the heat transfer problem in a channel

155

or pipe filled with a porous martial subject to constant wall heat flux boundary condition under

156

LTNE model in the developing region or fully developed region requires utilizing thermal

157

boundary condition models (i.e. Model A, B or C (e.g. [5, 15, 16])) at the solid wall. To date, it is

158

not clear which two boundary conditions are required at the solid walls since all the existing

159

models are validated against experiment. Hence, the previous numerical (e.g. [6]) and analytical

160

(e.g. [5]) studies provided solutions for the temperature fields in the porous region that are model

161

dependent.

162

Thus, in the present study, we revisit the problem and examine the thermally developing

163

forced convection heat transfer inside a channel filled with a porous material in the presence of

164

internal heat generations under LTNE condition. It is assumed that the temperature difference

165

between the two phases in the porous region is small such that a linear perturbation analysis can

166

be performed to avoid using models for the constant heat flux thermal boundary conditions at the

167

channel wall (i.e. A, B and C). Therefore, analytic solutions are obtained for the temperature

168

difference between the solid and fluid phases as well as the Nusselt number in the developing

169

region of the channel. Effects of pertinent parameters including the effective conductivity ratio,

170

k, the Biot number, Bi, and the internal heat generation parameter, , on the Nusselt number, Nu,

171

and the thermal entry length, xdev, are discussed.

172
173

2. Mathematical modeling

8
Page 8 of 33

174

The problem studied is schematically shown in Fig. 1. A channel with height of 2H and

175

length of L is filled with a porous material. Channel walls are subjected to a constant wall heat

176

flux. Laminar flow with a uniform temperature (Ti) enters the channel. Assumptions in the

177

present analysis are as following:

178

179
180

2D channel with impermeable parallel-plates to ensure the two-dimensionality of the

problem (e.g. [5-8]),

the Darcys law of motion for the fluid flow through the porous media which is valid

181

at low pore-scale Reynolds number, i.e. Redp < 1 and when the hydraulic diameter of

182

the channel is much higher than the pore-diameter, i.e. H >> dp (e.g. [15-18]),

183

184
185

through the media (e.g. [1, 2]),

186
187

incompressible flow due to much low velocities compared to the sound velocity

the thermally developing forced convective flow to capture the inlet effects [10, 12,
13, 15],

188

a rigid, isotropic and homogenous porous medium applicable for metal foams (e.g.
[2, 15, 17, 26]),

189

constant thermo-physical properties because of low temperature variations [1-3],

190

the local thermal non-equilibrium condition (LTNE) between the two phases in the

191
192
193

porous medium to consider a wide range of porous media (e.g. [15, 17]),

high Peclet number and hence negligible axial heat conduction which is practical in
common heat exchangers (e.g. [1-3]).

9
Page 9 of 33

194
195

According to these assumptions, the energy conservation equations for the fluid and solid
phases of the porous material respectively are [2]:

c pu

T f
x *

Ts
2

0 k s , eff

Tf
2

k f , eff

y *

y *

h sf a sf T s T f S f ,

(1)

h sf a sf T s T f S s .

(2)

196

The axial heat conduction term is negligible when the Peclet number is high, i.e. Pe >> 1 [2,

197

15, 22-25]. In these equations the subscripts s and f denote the solid and fluid phases,

198

respectively. Eqs. (1) and (2) are derived by a volume-averaging based on a representative

199

elementary volume. They consider the conduction heat transfer within the fluid and solid phases

200

of the porous material using effective conductivities (kf,eff and ks,eff), the convective heat transfer

201

through the fluid phase, the interphase convective heat transfer between the two phases using

202

fluid-to-solid heat transfer coefficient (hsf) and specific surface per unit area (asf), and the internal

203

heat generations/absorptions within the two phases. Subsequently, all these important

204

phenomena are represented by corresponding dimensionless numbers or dimensionless

205

parameters. In these equations, T is the temperature, cp is the specific heat of the fluid phase, u is

206

the Darcian velocity, S is the internal heat generation/absorption, ks,eff and kf,eff are the effective

207

thermal conductivities of the solid and fluid phases written as:

208
209

k s , eff 1 k s ,

(3)

k f , eff k f .

(4)

is the porosity of the medium, and hsf and asf are the fluid-to-solid heat transfer coefficient

and specific surface area [2]:

10
Page 10 of 33

a sf

h sf

210

6 1

(5)

dp

kf

2 1 . 1 Pr
dp

1
3

ud

0 .6

(6)

dp is pore diameter and Pr is the Prandtl number.

211
212
213

3. Analysis and solution

If one divides Eq. (2) by hsf asf arrives at the following relation:

T s

Ts
2

k s ,ef f

Tf

h sf a sf y *

1
~O
.
h sf a sf

Ss
h sf a sf

(7)

214

To find more information about perturbation analysis over energy conservation equations,

215

one can refer to Refs. [26-28]. Since the product of hsf asf is a large value (i.e. hsf asf > 1) [2,

216

26-29] Eq. (7) gives:

1
Ts T f O

h sf a sf

217
218

(8)

where is a small parameter. Now, we define the following dimensionless variables and
parameters:

T i k s ,ef f

(9)

q "w H
x

219

x *kf

c p H u
2

x *

H ,
Pe

y *
H

, k

kf

,ef f

k s ,ef f

SH
q w

Bi

h sf sf H

k s ,ef f

(10)

where Ti is the inlet temperature of the fluid phase. Combining Eqs. (7)-(10) yields:

f NE

1 f
2

Bi y

1
Bi

s O Bi

(11)

11
Page 11 of 33

220

where N E ( s f ) (Ts T f ) k s , eff / q " w H . Consequently, finding the temperature

221

difference between the fluid and solid phases, called the LTNE intensity (NE), is reduced to

222

find

f
2

. Adding Eq. (1) to Eq. (2) and using Eqs. (8)-(10) result in:

k 1

f
2

f s O B i

(12)

223

For the sake of brevity, the subscript f is dropped from f in the remainder of the text.

224

Now, the solution is similar to solve a fluid-saturated porous medium under the LTE condition

225

(i.e. Eq. 12). The boundary conditions of the porous-filled channel under a constant heat flux

226

imposed at the walls are:

1
k 1

227

at y 0,

(13-a)

at y 1,

(13-b)

at x 0.

(14)

To solve Eq. (12), we define the following intermediate functions:

(15)

( x , y ) f ( y ) g ( x , y ).

228

Combining Eqs. (12)-(15) gives:

g
x

g ( x , 0)
y

g ( x ,1)
y

g
2

( k 1)

( k 1) f f s O ( B i

f (0)

f (1) 0

1
k 1

),

(16)

at y 0,

(17-a)

at y 1,

(17-b)

12
Page 12 of 33

g (0, y ) f ( y ) 0

(18)

at x 0.

229

To obtain a PDE with homogeneous boundary conditions, we put the non-homogeneity of

230

the PDE (16) along with that of the boundary condition (17-a) into an ordinary differential

231

equation (ODE) as follows:

(19)

( k 1) f f s f 1 ( y ),

f (0)

k 1

(20-a)

(20-b)

f (1) 0.

232
233

f 1 ( y ) is another intermediate function. The simplest solution procedure (and maybe the

only feasible one) corresponds to f 1 ( y ) f s 1 , hence:

f (y )

234

2y

(21)

b,

2 ( k 1)

where b is an unknown intermediate constant. Combining Eqs. (16)-(21) yields:

g
x

( k 1)

g ( x , 0)
y
g ( x ,1)
y

g
y

f s 1 O ( B i

(22)

),

0,

(23-a)

0,

(23-b)

g (0, y )

2y

2( k 1)

(24)

b.

235

Eq. (22) along with the boundary condition (Eqs. 23 and 24) can be solved by separation of

236

variables. Since the boundary conditions in the y-direction are homogenous and the first partial

237

derivative, one can show that the solution would be a Fourier cosine series as follows:

13
Page 13 of 33

g (x , y ) c 0 (x )

(25)

( x ) cos( n y ) .

n 1

238

After some mathematical operations, one can find:

g (x , y )

( f f 1)

239

241

243

( f s 1)

where n

(k
n 1

2
1)( n )

exp( n x ) cos( n y ) O ( B i

(26)

).

3 y 1 1
2

6 k 1

( k 1)( n ) / k
2

n 1

exp n x cos n y O ( B i

kn

),

(27)

. Combining Eqs. (11) and (27) obtains the dimensionless

temperature difference between the fluid and solid phases (i.e. LTNE intensity) as follows:

s f
242

3( k 1)

Combining Eqs. (15), (21) and (26) gives the dimensionless temperature distribution:

x , y

240

N E

1+ s ( k 1)
B i ( k 1)
f ully dev eloped com ponent
1

2
exp

x
cos
n

n
O B i
n 1

dev eloping com ponent

. (28)

In addition, after some mathematical manipulations, one can find the local Nusselt number
based on the effective conductivity of the medium (km) as follows:
Nu

q w D h
k m (T w T f

,m

4
1
1

3 Bi
k 1

n 1 ( n ) 2 exp n x

O (B i

),

(29)

244

where Dh is the hydraulic diameter of the channel.

245

Equation (29) implicitly reveals that the Nusselt number for the developing region is

246

independent of the internal heat generation parameter within the fluid phase (f). Similar results

247

are obtained for the thermally fully developed region of the Darcian regime investigated by Refs.

248

[7, 8, 17, 18]. Moreover, Eq. (29) reveals that semi-local thermal equilibrium situation governs

249

on the thermally developing heat transfer between the solid and fluid phases when s = -(k+1)-1

250

with no respect to the Biot number (Bi) value. This phenomenon was seen in the thermally fully

14
Page 14 of 33

251

developed heat transfers inside porous-filled channels [17], in microchannels filled with porous

252

media [7, 18], and in partially porous-filled channels [8]. Meanwhile, Eq. (17) predicts a

253

discontinuity in the Nusselt number corresponding to a sign change in the denominator. Since,

254

255

change or discontinuity of the Nusselt number may results in an inversion in the heat transfer

256

direction. A similar inversion in the heat transfer direction was predicted in Refs. [18, 30-32].

n 1

/ 6 , the sign change in the denominator happens when s < -(k+1)-1. This sign
2

257

In addition, the denominator of Eq. (29) exhibits four phenomena influencing the Nusselt

258

number. The first term (1/3) obtains the fully developed Nusselt number for a channel flow at

259

LTE condition. The second term (s/Bi) corresponds to the internal heat generation

260

phenomenon. Neglecting this term gives the Nusselt number for a porous medium without

261

internal heat generation. The third term (Bi-1/(k+1)) shows the role of the internal heat exchange

262

between the fluid and solid phases of the porous medium. While the three first terms shows the

263

2
fully developed terms of the Nusselt number, the forth term ( n 1 2 ( n ) exp n x )

264

corresponds to the developing heat transfer phenomena.

265

To summarise, Eqs. (28) and (29) represent analytical expressions for the intensity of LTNE

266

condition (i.e. NE) and the developing (local) Nusselt number based on the analytical approach

267

assuming a small temperature difference between the solid and fluid phases in the porous

268

material. It should be noted that no model for considering the constant wall heat flux thermal

269

boundary condition has been used to obtain the solution. Equations (28) and (29) imply that the

270

error associated with the obtained solution has the order of magnitude equal to Bi-1. In other

271

words, the solution is almost accurate when O (Bi) > 1. Meanwhile, it is assumed that Bi is the

272

large parameter of the problem (i.e. the order of magnitude of Bi is higher than that of k and s)

15
Page 15 of 33

273

to find the solution. Subsequently, the solution is accurate when Bi is the dominant parameter

274

(i.e. O (Bi) > O (k) and O (Bi) > O |s|). Parametrical studies and also comparisons with similar

275

results of other researchers for especial cases are presented in the next section.

276
277

4. Results and discussions

278

4.1. Validation

279

For the validation purpose of the solution presented in section 3, the local Nusselt number

280

obtained using Eq. (29) with zero internal heat generation (i.e. s = 0) is compared to those

281

presented in Ref. [15] and the results are shown in Fig 2. It should be pointed out that Ouyang et

282

al. [15] utilized different thermal boundary conditions, so called Model A, B and C to model the

283

constant wall heat flux thermal boundary condition under LTNE condition. For further

284

comparison, the Nusselt numbers obtained by Yang and Vafai [17] (based on model A) for the

285

thermally fully developed region of the channel are also plotted in Fig. 2. It should be pointed

286

out that the Nusselt number in Refs. [15, 17] was obtained based on the effective thermal

287

conductivity of the fluid phase (kf,eff) rather than the effective thermal conductivity of the

288

medium (km) used in the present study. Hence, the Nusselt of Refs. [15, 17] should be multiplied

289

by a factor of kf,eff / km = k / (k + 1) to be converted to the present form as plotted. Comparing

290

figures 2(a)-2(c) shows that as we expect, increasing the Biot number augments the Nusselt

291

number. This is due to a higher heat exchange between the solid and fluid phases of the porous

292

material. As discussed in section 3, the accuracy of the presented perturbation analysis increases

293

with the increase of the Biot number (Bi). As the Biot number increase the difference between

294

the results of three models decreases and tends to the results of the present perturbation analysis.

295

Predictions of these models are discussed in terms of the physics of the problem in Ref. [16]. In
16
Page 16 of 33

296

addition, it is seen that for relatively high Biot numbers the results of the present work are closer

297

to those predicted by Model A. For high Biot number of 10 (Fig. 2c), the present results are

298

exactly the same as those predicted by Model A in [15, 17]. Even for relatively low Biot number

299

of 1 (Fig. 2b) the present results differ 20% from those obtained based on Model A. Model A

300

assumes that the imposed wall heat flux is divided between the two phases based on the effective

301

thermal conductivity and the temperature gradient at the wall. In addition, the temperature of

302

each phase at the wall is considered to be equal.

303

To study the effect of the internal heat generation or absorption on the hydrothermal

304

behavior of the system, Fig. 3 is plotted showing the variation of Nusselt number as a function of

305

dimensionless axial distance (x) for two different internal heat generation parameters s = 1 and

306

s = -1. Furthermore, a numerical simulation has been performed using the finite difference

307

method [33, 34] based on a non-uniform structured 301201 grid [35, 36] with a mesh-span

308

increase-factor of 1.02 [37, 38] in the flow direction to check the accuracy of the obtained

309

analytical solution (Eq. 17) in the presence of internal heat generation or absorption. The

310

boundary condition of model A [15] has been used for the numerical simulation. A good

311

agreement between both the numerical and analytical results in the developing region is seen. In

312

addition both the numerical and analytical Nusselt numbers asymptotically approach to the fully

313

developed results of Yang and Vafai [17]. It is also seen that, the Nusselt number decreases with

314

the increase of the internal heat generation parameter (s) for moderate values of |s|. This

315

observation is consistent with those obtained by Refs. [7, 18]. The Nusselt number dramatically

316

decreases when |s| has a large value due to the semi-conduction heat transfer instead of the

317

convection heat transfer [18]. As we expect, it is seen that the Nusselt number decreases along

318

the flow direction until reach the asymptotic value of the fully developed Nusselt number. This

17
Page 17 of 33

319

reduction in the Nusselt number arises from the growth of the thermal boundary layer in the flow

320

direction [1, 16].

321
322

4.2. Effect of thermal conductivity ratio

323

Figure 4 shows the influence of the effective conductivity ratio (k) on the developing

324

Nusselt number in the presence of internal heat generation or absorption. It is seen that the

325

Nusselt number increases with the increase of k. An increase in k obtains a porous material with

326

a more enhanced effective thermal conductivity of the fluid phase compared to the solid phase.

327

Consequently, the Nusselt number of the porous material increases since the Nusselt number

328

corresponds to the convection heat transfer of the fluid phase. In other words, the Nusselt

329

number increases when the fluid phase has more ability to convect out the heat. Comparing Fig.

330

4(a) with Fig. 4(b) shows that the Nusselt number increases with decreased s as shown by

331

Dehghan [8]. Furthermore, Fig. 4 shows that for a fixed s, as k decreases the Nusselt number

332

tends to be independent of the x position. This is further clarified by inspection of Eq. (17) where

333

for low k, and hence

334

obtained as N u 4 1 / 3 (1 / B i ) s 1 / ( k 1) , which is independent of the axial position

335

x.

n 1

2 ( n ) 2 ex p n x 0 . Therefore, the Nusselt number is

336
337

4.3. Effect of Biot number

338

Figure 5 shows effects of the Biot number (Bi) on the local Nusselt number. It is seen that

339

the Nusselt number increases with the increase of the Biot number when s > 0. A higher Biot

340

number represents a more intense internal heat exchange between the solid and fluid phases of
18
Page 18 of 33

341

the porous material which leads to reduce the temperature difference between the two phases (i.e.

342

the LTE condition). By improving the internal heat exchange between the two phases, the heat

343

can be transferred from the solid phase to the fluid phase more easily and subsequently can be

344

convected out more easily. It is seen that the LTE condition almost governs when Bi > 1000.

345

Furthermore, Fig. 5 shows that the Nusselt number changes more significantly with x for higher

346

Biot number. This is because as Bi, the second term in denominator of Eq. (17) vanishes and

347

hence the Nusselt number is mainly affected by the third term evolving x. While for low Bi

348

number

349

N u 4 1 / 3 (1 / B i ) s 1 / ( k 1)

350

(i.e.

Bi0)

the

term
1

1
1
s

Bi
k 1

dominates

and

hence

, which is independent of x.

4.4. Thermal entry length

351

Figure 6 shows the thermal entry length (i.e. an axial location where the flow becomes

352

thermally fully developed). The thermal entry length (xdeveloping) is a position where the local

353

Nusselt number obtained by Eq. (17) is almost equal to that of the fully developed Nusselt

354

number, i.e. N u

355

the maximum value for s = -1. According to the discussions given by Refs. [8, 17], a reduction

356

in the internal heat generation parameter increases the convection heat transfer rate represented

357

by the Nusselt number. Subsequently, the thermal entry length would be more considerable when

358

the convective heat transfer is augmented.

N u ( x developing ) N u fd
N u fd

0.01 [39-41]. Figure 6(a) shows that xdeveloping yields

359

To investigate effects of a relatively high Biot number, Fig. 6(b) is plotted. It reveals that the

360

dependency of the thermal entrance length on the internal heat generation/absorption parameter

361

decreases by an increase in the Biot number. According to Eq. (17) for finite s values, the
19
Page 19 of 33

1
1
s
0
Bi
k 1

362

second term in the denominator tends to zero, i.e.

, as the Bi tends to

363

infinity. Hence, the Nusselt number is less sensitive to the internal heat generation and

364

consequently the thermally developing length is also has negligible dependency on small and

365

moderate values of s. However, relatively high values of the internal heat generation/absorption

366

parameter may influence the thermal entrance length even at high Biot numbers as seen it Fig.

367

6(c).

368

In addition, in Fig. 6 it is seen that the thermal entry length increases with increased

369

effective thermal conductivity ratio (k). As it has been discussed earlier, a porous medium with a

370

lower k translates to a porous medium with lower convective heat transfer effects compared to

371

the conductive heat transfer ones. In addition, the thermal entry length is meaningful for the

372

convective heat transfer. Consequently, the thermal entry length decreases when the role of

373

convective heat transfer decreases compared to the conductive heat transfer (within both the

374

solid and fluid phases). At the limiting case when the effective thermal conductivity tends to zero

375

( k 0 ), n tends to infinity ( n ). Subsequently, the term containing

376

denominator of Eq. (17), corresponding to the developing heat transfer, would be almost equal to

377

zero. Consequently, the developing heat transfer would be meaningless and hence the thermal

378

entry length would be zero as well. The thermally fully developed flow assumption would be

379

closer to the reality when the thermal entry length is smaller.

exp n x

in the

380
381

5. Conclusion

382

A thermally developing forced convective heat transfer inside a porous-filled channel in the

383

presence of internal heat generations within both the solid and fluid phases of the porous material
20
Page 20 of 33

384

has analytically been investigated under the iso-flux thermal boundary condition by assuming a

385

small temperature difference between the two phases. To examine the validity of the present

386

model, the present solution is compared with the results obtained by utilizing two existing

387

models (A and B) describing the constant wall heat flux boundary condition at the channel wall.

388

The major remarks on the behavior of the investigated system can be summarized as follows:

389

An exact solution has been provided for the Nusselt number in the entrance region of

390

a porous medium in the presence of internal heat generation under the Darcy law of

391

motion and LTNE condition.

392

To close the equations under LTNE condition, we have found a new extra analytic

393

relation between the temperatures of the fluid and solid phases based on a small

394

temperature difference between the two phases. Hence, no thermal boundary

395

condition has been used to solve the energy transport equations.

396

The analytical solution has been obtained by assuming a relatively high heat transfer

397

rate between fluid and solid phase at pore scale. Hence, the solution is accurate for

398

high Biot numbers (i.e. O (Bi) > 1) and when the order of magnitude of Bi is higher

399

than that of other parameters of the problem (i.e. k and s).

400

number has a relatively large value (O (Bi) > 102).

401
402

403
404
405

The local Nu is less sensitive to the internal heat generation parameter when Biot

The local Nu increases with increased Biot number (Bi) and increased conductivity
ratio (k).

Results obtained by model A are found to be closer to the results of the present
analysis using no model for the constant heat flux boundary condition.

21
Page 21 of 33

406

407
408
409

The thermal entry length (xdeveloping) increases with both increased conductivity ratio
(k) and decreased internal heat generation parameter (s).

At the limiting case of k 0 , the thermal entry length (xdeveloping) would be zero and
would be meaningless.

410
411

References

412

[1] A. Bejan, Convection Heat Transfer, John Wiley & Sons, New York, 2013.

413

[2] D.A. Nield, A. Bejan, Convection in Porous Media, Springer, New York, 2013.

414

[3] K. Vafai, Handbook of Porous Media, Marcel Dekker, Ohio, 2000.

415

[4] A. Amiri, K. Vafai, T.M. Kuzay, Effects of boundary conditions on non-Darcian heat

416

transfer through porous media and experimental comparisons, Numer. Heat Transfer,

417

Part A 27 (1995) 651664.

418

[5] Y. Mahmoudi, N. Karimi, K. Mazaheri, Analytical investigation of heat transfer

419

enhancement in a channel partially filled with a porous material under local thermal

420

non-equilibrium condition: effects of different thermal boundary conditions at the

421

porous-fluid interface, Int. J. Heat Mass Transfer 70 (2014) 875891.

422

[6] B. Alazmi, K. Vafai, Constant wall heat flux boundary conditions in porous media

423

under local thermal non-equilibrium conditions, International Journal of Heat and

424

Mass Transfer 45 (2002) 30713087.

425

[7] Y. Mahmoudi, Constant wall heat flux boundary condition in micro-channels filled

426

with a porous medium with internal heat generation under local thermal non-

427

equilibrium condition, International Journal of Heat and Mass Transfer 85 (2015)

428

524542.
22
Page 22 of 33

429

[8] M. Dehghan, Effects of heat generations on the thermal response of channels partially

430

filled with non-Darcian porous materials, Transport in Porous Media (2015) DOI:

431

10.1007/s11242-015-0567-9.

432
433

[9] S.J. Kim, D. and Kim, Thermal Interaction at the Interface Between a Porous Medium
and an Impermeable Wall, Journal of Heat Transfer 123 (2001) 527-533.

434

[10] D.A. Nield, A.V. Kuznetsov, M. Xiong, Thermally developing forced convection in a

435

porous medium: parallel plate channel with walls at uniform temperature, with axial

436

conduction and viscous dissipation effects, International Journal of Heat and Mass

437

Transfer 46 (2003) 643651.

438

[11] A.V. Kuznetsov, M. Xiong, D.A. Nield, Thermally Developing Forced Convection in

439

a Porous Medium: Circular Duct with Walls at Constant Temperature, with

440

Longitudinal Conduction and Viscous Dissipation Effects, Transport in Porous Media

441

53 (2003) 331345.

442

[12] D.A. Nield, A.V. Kuznetsov, Thermally developing forced convection in a channel

443

occupied by a porous medium saturated by a non-Newtonian fluid, International

444

Journal of Heat and Mass Transfer 48 (2005) 12141218.

445

[13] A. Haji-Sheikh, D.A. Nield, K. Hooman, Heat transfer in the thermal entrance region

446

for flow through rectangular porous passages, International Journal of Heat and Mass

447

Transfer 49 (2006) 30043015.

448

[14] A.V. Kuznetsov, D.A. Nield, Thermally developing forced convection in a porous

449

medium occupied by a rarefied gas: parallel plate channel or circular tube with walls

450

at constant heat flux, Transport in Porous Media 76 (2009) 345362.

23
Page 23 of 33

451

[15] X. Ouyang, K. Vafai, P. Jiang, Analysis of thermally developing flow in porous

452

media under local thermal non-equilibrium conditions, Int. J. Heat Mass Transfer 67

453

(2013) 768775.

454

[16] M. Dehghan, M.S. Valipour, A. Keshmiri, S. Saedodin, N. Shokri, On the thermally

455

developing forced convection through a porous material under the local thermal non-

456

equilibrium condition: an analytical study, International Journal of Heat and Mass

457

Transfer 92 (2016) 815-823.

458

[17] K. Yang, K. Vafai, Analysis of temperature gradient bifurcation in porous media

459

An exact solution, International Journal of Heat and Mass Transfer 53 (2010) 4316

460

4325.

461

[18] M. Dehghan, M.S. Valipour, S. Saedodin, Analytical study of heat flux splitting in

462

micro-channels filled with porous media, Transport in Porous Media 109 (2015) 571-

463

587.

464
465

[19] B. Eilers, Microchannel steam-methane reforming under constant and variable

surface temperature distributions, M.S. Thesis, Oregon State University, 2010.

466

[20] V. Kuznetsov, S. Kozlov, Modeling of methane steam reforming in a microchannel

467

with a heat flow distributed in length, Journal of Engineering Thermophysics 17(1)

468

(2008) 53-59.

469

[21] A. White, J. McTigue, C. Markides, Wave propagation and thermodynamic losses in

470

packed-bed thermal reservoirs for energy storage, Applied Energy 130 (2014) 648

471

657.

24
Page 24 of 33

472

[22] Y. Mahmoudi, M. Maerefat, Analytical investigation of heat transfer enhancement in

473

a channel partially filled with a porous material under local thermal non-equilibrium

474

condition, International Journal of Thermal Sciences 50 (12) (2011) 2386-2401.

475

[23] M. Dehghan, M.S. Valipour, S. Saedodin, Temperature-dependent conductivity in

476

forced convection of heat exchangers filled with porous media: a perturbation

477

solution, Energy Conversion and Management 91 (2015) 259266.

478

[24] K. Wang, F. Tavakkoli, S. Wang, K. Vafai, Forced convection gaseous slip flow in a

479

porous circular microtube: An exact solution, International Journal of Thermal

480

Sciences 97 (2015) 152-162.

481

[25] K. Wang, F. Tavakkoli, K. Vafai, Analysis of gaseous slip flow in a porous micro-

482

annulus under local thermal non-equilibrium condition An exact solution,

483

International Journal of Heat and Mass Transfer 89 (2015) 13311341.

484

[26] A.V. Kuznetsov, Thermal Nonequilibrium, Non-Darcian Forced Convection in a

485

Channel Filled with a Fluid Saturated Porous Medium- A Perturbation Solution,

486

Applied Scientific Research 57 (1997) 119-131.

487

[27] M. Dehghan, M.S. Valipour, S. Saedodin, Perturbation analysis of the local thermal

488

non-equilibrium condition in a fluid saturated porous medium bounded by an iso-

489

thermal channel, Transport in Porous Media 102 (2) (2014) 139-152.

490

[28] M. Dehghan, M.T. Jamal-Abad, S. Rashidi, Analytical interpretation of the local

491

thermal non-equilibrium condition of porous media imbedded in tube heat

492

exchangers, Energy Conversion Management 85 (2014) 264-271.

493
494

[29] P.X. Jiang, R.N. Xu, W. Gong, Particle-to-fluid heat transfer coefficients in
miniporous media, Chem. Eng. Sci. 61 (2006) 72137222.

25
Page 25 of 33

495
496

[30] A.A. Kornhauser, J.L. Smith, Effects of heat transfer on gas spring performance,
Journal of Energy Resources Technology 115 (1993) 70-75.

497

[31] A.A. Kornhauser, J.L. Smith, Application of a Complex Nusselt Number to Heat

498

Transfer During Compression and Expansion, Journal of Heat Transfer 116 (1994)

499

536-542.

500

[32] R. Mathie, C.N. Markides, A.J. White, A Framework for the Analysis of Thermal

501

Losses in Reciprocating Compressors and Expanders, Heat Transfer Engineering 35

502

(2014) 1435-1449.

503

[33] M. Dehghan, Y. Rahmani, D.D. Ganji, S. Saedodin, M.S. Valipour, S. Rashidi,

504

Convectionradiation heat transfer in solar heat exchangers filled with a porous

505

medium: Homotopy perturbation method versus numerical analysis, Renewable

506

Energy 74 (2015) 448-455.

507

[34] M. Dehghan, Y. Mahmoudi, M.S. Valipour, S. Saedodin, Combined Conduction-

508

Convection-Radiation Heat Transfer of Slip Flow inside a Micro-channel Filled with

509

a Porous Material, Transport in Porous Media 108 (2015) 413-436.

510

[35] M. Dehghan, H. Basirat Tabrizi, On near-wall behavior of particles in a dilute

511

turbulent gassolid flow using kinetic theory of granular flows, Powder Technology

512

224 (2012) 273280.

513

[36] M. Dehghan, H. Basirat Tabrizi, Turbulence effects on the granular model of particle

514

motion in a boundary layer flow, Canadian Journal of Chemical Engineering 92

515

(2014) 189195.

26
Page 26 of 33

516

[37] M. Mirzaei, M. Dehghan, Investigation of flow and heat transfer of nanofluid in

517

microchannel with variable property approach, Heat Mass Transfer 49 (2013) 1803-

518

1811.

519

[38] M. Dehghan, M. Daneshipour, M.S. Valipour, R. Rafee, S. Saedodin, Enhancing heat

520

transfer in microchannel heat sinks using converging flow passages, Energy

521

Conversion and Management 92 (2015) 244-250.

522

[39] M. Mital, Analytical analysis of heat transfer and pumping power of laminar

523

nanofluid developing flow in microchannels, Applied Thermal Engineering 50 (2013)

524

429436.

525

[40] M. Ahmeda, M. Eslamian, Laminar forced convection of a nanofluid in a

526

microchannel: Effect of flow inertia and external forces on heat transfer and fluid

527

flow characteristics, Applied Thermal Engineering 78 (2015) 326338.

528

[41] J.A. Esfahani, S.M. Vahidhosseini, E. Barati, Three-dimensional analytical solution

529

for transport problem during convection drying using Green's function method

530

(GFM), Applied Thermal Engineering 85 (2015) 264277.

531
532

27
Page 27 of 33

533
534

Fig. 1. Schematic diagram of the problem.

535
5

Present analysis
Yang and Vafai [17] Model A

Ouyang et al. [15] Model A

Ouyang et al. [15] Model B

Nu

Ouyang et al. [15] Model C

0
0.01
(a)

0.1

28
Page 28 of 33

14

Present analysis
Yang and Vafai [17] Model A

12

Ouyang et al. [15] Model A

10

Ouyang et al. [15] Model B

Nu

Ouyang et al. [15] Model C

8
6

4
2
0
0.01
(b)

0.1

20

Present analysis
Yang and Vafai [17] Model A
Ouyang et al. [15] Model A

15

Ouyang et al. [15] Model B

Nu

Ouyang et al. [15] Model C

10

0
0.01
(c)

0.1

536

Fig. 2. The present analysis compared to Refs. [15, 17] for s = 0 and k = 0.1; (a) Bi = 0.1,

537

(b) Bi = 1 and (c) Bi = 10.

538

29
Page 29 of 33

20

Analytical results
Numerical resutls

Nu

15

Yang and Vafai [17]

s = -1

10

s = 1

0
0.01

0.1

539

Fig. 3. Effects of the internal heat generation (s) on the developing Nusselt number with k

540

= 0.1 and Bi = 10.

541
30

Bi=10
Bi=100

25

Bi=1000
LTE

Nu

20
15
10
5
0
0.01

0.1

542

Fig. 5. Effects of the Biot number (Bi) on the developing Nusselt number for s = 1 and k =

543

1.
30
Page 30 of 33

Nu

100

10
k=1
k=0.1

1
0.01
(a)

0.1

Nu

10

k=1
k = 0.1

1
0.01
(b)

0.1

544

Fig. 4. Effects of the conductivity ratio (k) on the developing Nusselt number for Bi = 10;

545

(a) s = -1 and (b) s = 1.

546

31
Page 31 of 33

0.5

xdeveloping

0.4
0.3
0.2
s =-1
-1
_s=
s =00
_s=
s =11
_s=

0.1
0

0.1

(a)

10

100

0.5

xdeveloping

0.4
0.3
0.2
s =-1
-1
_s=
s =00
_s=

0.1

s =11
_s=

0.1

(b)

10

100

32
Page 32 of 33

0.5

xdeveloping

0.4
0.3
0.2
s = -10
_s=-10
s = 0
_s=0

0.1

s = 10
_s=10

0
0.1

(c)

10

100

547

Fig. 6. Thermal entry length (xdeveloping) versus the conductivity ratio (k); (a) Bi = 10, (b) Bi =

548

100 and (c) Bi = 100.

549

33
Page 33 of 33