0% found this document useful (0 votes)
48 views4 pages

International Journal of Heat and Mass Transfer: D.A. Nield, A.V. Kuznetsov

This technical note revises a previous model of thermal instability in a porous medium layer saturated by a nanofluid. The previous model imposed boundary conditions for both temperature and nanoparticle fraction that are not physically realistic. The revised model assumes the temperature can be imposed at boundaries but the nanoparticle fraction adjusts so that nanoparticle flux is zero. With this new boundary condition, oscillatory convection cannot occur. The presence of nanoparticles still has a destabilizing effect on non-oscillatory convection.

Uploaded by

Reshu Tyagi
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
48 views4 pages

International Journal of Heat and Mass Transfer: D.A. Nield, A.V. Kuznetsov

This technical note revises a previous model of thermal instability in a porous medium layer saturated by a nanofluid. The previous model imposed boundary conditions for both temperature and nanoparticle fraction that are not physically realistic. The revised model assumes the temperature can be imposed at boundaries but the nanoparticle fraction adjusts so that nanoparticle flux is zero. With this new boundary condition, oscillatory convection cannot occur. The presence of nanoparticles still has a destabilizing effect on non-oscillatory convection.

Uploaded by

Reshu Tyagi
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 4

International Journal of Heat and Mass Transfer 68 (2014) 211–214

Contents lists available at ScienceDirect

International Journal of Heat and Mass Transfer


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

Technical Note

Thermal instability in a porous medium layer saturated by a nanofluid:


A revised model
D.A. Nield a, A.V. Kuznetsov b,⇑
a
Department of Engineering Science, University of Auckland, Private Bag 92019, Auckland 1142, New Zealand
b
Department of Mechanical and Aerospace Engineering, North Carolina State University, Campus Box 7910, Raleigh, NC 27695-7910, USA

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

Article history: We develop an extension of our previous thermal instability analysis of a nanofluid-saturated porous
Received 22 August 2013 layer. The extension is based on a new boundary condition for the nanoparticle fraction, which is phys-
Accepted 13 September 2013 ically more realistic. In the previous model we imposed both temperature and nanoparticle volume frac-
Available online 9 October 2013
tions at the boundaries of the layer. It is now assumed that the value of the temperature can be imposed
on the boundaries, but the nanoparticle fraction adjusts so that the nanoparticle flux is zero on the
Keywords: boundaries. The new boundary condition on the nanoparticle volume fraction is made possible by
Nanofluid
accounting for the contributions of the effect of thermophoresis to the nanoparticle flux. It is shown that,
Porous medium
Instability
with the new boundary conditions, oscillatory convection cannot occur. The effect of the nanoparticles on
Natural convection non-oscillatory convection is destabilizing.
Ó 2013 Elsevier Ltd. All rights reserved.

1. Introduction We have submitted a companion paper revising our paper on


the corresponding problem in a fluid clear of solid material [4].
The problem of the onset of convection in a horizontal layer
occupied by a porous medium uniformly heated from below is a 2. Analysis
classical problem commonly known as the Horton–Rogers–Lap-
wood problem or Bénard–Darcy problem. Work on this problem The analysis follows closely that in [3] and so it is abbreviated
has been surveyed in Chapter 6 of the book by Nield and Bejan [1]. as far as possible here.
An extension to the case of a nanofluid, based on a model pre- We select a coordinate frame in which the z⁄-axis is aligned ver-
sented by Buongiorno [2], in which the effects of Brownian motion tically upwards. We consider a horizontal layer of a porous med-
and thermophoresis are taken into account, was made by Nield and ium of porosity e and permeability K confined between the
Kuznetsov [3]. In their paper it was assumed that one could control planes z⁄ = 0 and z⁄ = H. Asterisks are used to denote dimensional
the value of the nanoparticle fraction at the boundary in the same variables. Each boundary wall is assumed to be impermeable and
way as the temperature there could be controlled. However, since perfectly thermally conducting. The temperatures at the lower
the development of the model presented in [3], it became apparent and upper walls are taken to be T h and T c , respectively, T h being
that in practice controlling the nanoparticle volume fraction on the greater than T c . For simplicity, Darcy’s law is assumed to hold
boundaries may be difficult. Thus it is advisable to replace the and the Oberbeck–Boussinesq approximation is employed. Homo-
boundary conditions used in [3] by a set that are more realistic geneity and local thermal equilibrium in the porous medium is as-
physically. In the present paper we revisit this problem, and we sumed. The reference temperature is taken to be T c . In the linear
now assume that there is no nanoparticle flux at the plate and that theory being applied here the temperature change in the fluid is
the particle fraction value there adjusts accordingly. This change assumed to be small in comparison with T c . The Darcy velocity is
has important consequences. The scaling of dimensionless param- denoted by v⁄. The field equations for total mass, momentum, ther-
eters needs changing. The basic solution for the nanoparticle vol- mal energy, and nanoparticles, respectively, take the form
ume fraction is changed, and conclusions are changed. No longer
are there two opposing agencies affecting instability and hence
r  v  ¼ 0; ð1Þ
oscillatory instability is ruled out. h i
l
0 ¼ r p  v þ / qp þ ð1  / Þfqð1  bðT   T c ÞÞg g; ð2Þ
K
⇑ Corresponding author. Tel.: +1 919 515 5292; fax: +1 919 515 7968.
E-mail addresses: [email protected] (D.A. Nield), [email protected]
(A.V. Kuznetsov).

0017-9310/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.ijheatmasstransfer.2013.09.026
212 D.A. Nield, A.V. Kuznetsov / International Journal of Heat and Mass Transfer 68 (2014) 211–214

Nomenclature

c nanofluid specific heat at constant pressure (x,y,z) dimensionless Cartesian coordinates, (x⁄, y⁄, z⁄)/H; z is
DB Brownian diffusion coefficient the vertically-upward coordinate
DT thermophoretic diffusion coefficient (x⁄,y⁄,z⁄) Cartesian coordinates
H dimensional layer depth
km effective thermal conductivity of the porous medium Greek symbols
Le Lewis number, defined by Eq. (15) a dimensionless wavenumber
NA modified diffusivity ratio, defined by Eq. (19) am thermal diffusivity of the porous medium, ðqkcmP Þ
f
NB modified particle-density increment, defined by Eq. (20) l viscosity of the fluid
p⁄ pressure q fluid density
p dimensionless pressure, p K=lam (qc)m effective heat capacity of the porous medium
Ra thermal Rayleigh-Darcy number, defined by Eq. (16) qp nanoparticle mass density
Rm basic-density Rayleigh number, defined by Eq. (17) r parameter defined by Eq. (8)
Rn concentration Rayleigh number, defined by Eq. (18) /⁄ nanoparticle volume fraction
t⁄ time /0 reference value for nanoparticle volume fraction
t dimensionless time, t  am =rH2 /
/ /
relative nanoparticle volume fraction, / 0
T⁄ nanofluid temperature 0

T  T 
T dimensionless temperature, T  T c Superscripts
h c
T c temperature at the upper wall
⁄ dimensional variable
T h temperature at the lower wall 0
perturbation variable
(u,v,w) dimensionless Darcy velocity components, (u⁄, v⁄ ,w⁄)H/
am
v Darcy velocity, ev Subscript
v⁄ dimensional Darcy velocity, (u⁄, v⁄, w⁄) b basic solution

@T  @/ @T
ðqcÞm þ ðqcÞf v   r T  ¼ km r2 T  w ¼ 0; T ¼ 1; þ NA ¼ 0 at z ¼ 0; ð13Þ
@t @z @z
þ eðqcÞp ½DB r /  r T 
 @/ @T
þðDT =T c Þr T   r T  ; ð3Þ w ¼ 0; T ¼ 0; þ NA ¼ 0 at z ¼ 1: ð14Þ
@z @z
@/ 1    Here
þ v  r / ¼ DB r2 / þ DT =T c r2 T  : ð4Þ
@t e
am
Le ¼ ; ð15Þ
We write v⁄ = (u⁄, v⁄, w⁄). We assume that the temperature is con- DB
stant and the nanoparticle flux is zero on the boundaries. Thus
the boundary conditions are qgbKHðT h  T c Þ
Ra ¼ ; ð16Þ
@/ DB @T 
lam


w ¼ 0; T ¼ T h ; DB  þ ¼ 0 at z ¼ 0; ð5Þ
@z T 1 @z h i
qp /1 þ qð1  /1 Þ gKH
@/ DB @T  Rm ¼ ; ð17Þ
w ¼ 0; T  ¼ T c ; DB þ ¼ 0 at z ¼ H: ð6Þ lam
@z T 1 @z
We introduce dimensionless variables as follows. We define ðqp  qÞ/0 gKH
Rn ¼ ; ð18Þ
ðx; y; zÞ ¼ ðx ; y ; z Þ=H; t ¼ t am =rH2 ;
lam
ðu; v ; wÞ ¼ ðu ; v  ; w ÞH=am ;
DT ðT h  T c Þ
/  /0 T   T c NA ¼ ; ð19Þ
p ¼ p K=lam ; /¼ ; T¼  ; ð7Þ DB T c /0
/0 T h  T c
where /0 is a reference scale for the nanopartical fraction and eðqcÞp 
NB ¼ / : ð20Þ
ðqcÞf 0
k ðqc Þ
am ¼ m ; r¼ P m: ð8Þ
ðqcP Þf ðqcP Þf The parameter Le is a Lewis number and Ra is the familiar ther-
mal Rayleigh–Darcy number. The parameters Rm and Rn may be
Then Eqs. (1)–(6) take the form:
regarded as a basic-density Rayleigh number and a concentration
r  v ¼ 0; ð9Þ Rayleigh number, respectively. The parameter NA is a modified dif-
fusivity ratio and is somewhat similar to the Soret parameter that
0 ¼ rp  v  Rme
^z þ RaT e
^z  Rn/e
^z ; ð10Þ arises in cross-diffusion phenomena in solutions, while NB is a
modified particle-density increment.
@T NB N N By extending the Oberbeck–Boussinesq approximation, Eq. (10)
þ v  rT ¼ r2 T þ r/  rT þ A B rT  rT; ð11Þ
@t Le Le has been linearized by the neglect of a term proportional to the
product of / and T. We believe that this assumption will be valid
1 @/ 1 1 2 N in the case of small temperature gradients in a dilute suspension
þ v  r/ ¼ r / þ A r2 T; ð12Þ
r @t e Le Le of nanoparticles.
D.A. Nield, A.V. Kuznetsov / International Journal of Heat and Mass Transfer 68 (2014) 211–214 213

2.1. Basic solution and substitute into the differential equations to obtain

We seek a time-independent quiescent solution of Eqs. (9)–(14)


ðD2  a2 ÞW þ Ra a2 H  Rn a2 U ¼ 0; ð35Þ
with temperature and nanoparticle volume fraction varying in the  
NB 2NA NB NB
z-direction only, that is a solution of the form v = 0, T = Tb(z), W þ D2 þ D D  a2  s H  DU ¼ 0; ð36Þ
/ = /b(z). Le Le Le
Eqs. (11) and (12) reduce to  
 2 1 NA 2 1 2 s
2
d T b NB d/b dT b NA NB dT b W ðD  a2 ÞH  ðD  a2 Þ  U ¼ 0; ð37Þ
þ þ ¼0 ð21Þ e Le Le r
2 Le dz dz Le dz
dz
W ¼ 0; H ¼ 0; D/ þ NA DH ¼ 0 at z ¼ 0 and at z ¼ 1; ð38Þ
2 2
d /b d Tb
2
þ NA 2
¼0 ð22Þ where
dz dz
d 2 1=2
Using the boundary conditions (13) and (14), Eq. (22) may be D and a ¼ ðl þ m2 Þ : ð39Þ
dz
integrated to give
d/b dT b Here a is a dimensionless horizontal wavenumber.
þ NA ¼ 0; ð23Þ For neutral stability the real part of s is zero. Hence we now
dz dz
write s = ix, where x is real and is a dimensionless frequency.
and substitution of this into Eq. (21) gives An approximate solution to the system of Eqs. (35)–(38) is ob-
2
d Tb tained by using a Galerkin-type weighted residuals method. As
2
¼ 0: ð24Þ trial functions (satisfying the boundary conditions) we choose
dz
The solution of Eq. (24) satisfying Eqs. (13) and (14) is W p ¼ Hp ¼ zp1 ð1  zÞ; p ¼ 1; 2; 3; . . . ; ð40Þ
T b ¼ 1  z: ð25Þ
1
U1 ¼ NA ðz2  zÞ and Up ¼ NA z2 ; p ¼ 2; 3; 4; . . . : ð41Þ
Equation (23) then integrates, when Eqs. (13) and (14) are applied, to 2
/b ¼ /0 þ NA z: ð26Þ We then write
X
N X
N X
N
W¼ Ap W p ; H¼ Bp Hp ; U¼ C p Up ; ð42Þ
2.2. Perturbation solution p¼1 p¼1 p¼1

substitute into Eqs. (35)–(37), and make the expressions on the


We now superimpose perturbations on the basic solution. We
left-hand sides of those equations (the residuals) orthogonal to
write
the trial functions, thereby obtaining a system of 3N linear
v ¼ v0 ; p ¼ pb þ p0 ; T ¼ Tb þ T0; / ¼ /b þ /0 ; ð27Þ algebraic equations in the 3N unknowns Ap, Bp, Cp; p = 1, 2,...,N.
substitute in Eqs. (9)–(14), and linearize by neglecting products The vanishing of the determinant of coefficients produces the
of primed quantities. The following equations are obtained when eigenvalue equation for the system. One can regard Ra as the
Eqs. (25) and (26) are used. eigenvalue. Thus Ra is found in terms of the other parameters.

r  v 0 ¼ 0; ð28Þ 3. Results and discussion

0 ¼ rp0  v 0 þ RaT 0 e
^z  Rn/0 e
^z ; ð29Þ Since oscillatory convection has been ruled out, because of the
absence of two opposing buoyancy forces, we need to consider
 
@T 0 NB @T 0 @/0 2NA NB @T 0 only the case of non-oscillatory instability, when x = 0.
 w0 ¼ r2 T 0 þ NA   ; ð30Þ For a first approximation we take N = 1. This produces the result
@t Le @z @z Le @z
  2
Le ð10 þ a2 Þ
1 @/0 1 0 1 NA 2 0 Ra þ N A þ Rn ¼ : ð43Þ
þ w ¼ r2 /0 þ r T; ð31Þ e 2 a
r @t e Le Le
The minimum of the right-hand side of this equation as a varies is
pffiffiffiffiffiffi
@/0 @T 0 attained when a ¼ 10; and the minimum value is 40.pWe
w0 ¼ 0; T 0 ¼ 0; þ NA ¼ 0 at z ¼ 0 and at z ¼ 1: ð32Þ ffiffiffiffiffiffi conclude
@z @z that the critical wavenumber ac is approximately 10 or about
We reduced the six unknowns u0 , v0 , w0 , p0 , T0 , /0 to three by operat- 3.16, and the critical Rayleigh number Rac is given approximately by
ing on Eq. (29) with e ^z  curl curl and using the identity  
Le
curl curl  grad div  r2 together with Eq. (28). Rac ¼ 40  NA þ Rn: ð44Þ
e
The result is
We recognize that in the absence of nanoparticles we recover,
r2 w0 ¼ Rar2H T 0 þ Rnr2H /0 : ð33Þ
approximately, the well-known result that the critical Rayleigh
Here r2H is the two-dimensional Laplacian operator on the horizon- number is equal to 4p2. Usually when one employs a single-term
tal plane. Galerkin approximation in this context one gets an overestimate
The differential Eqs. (23), (30), (31) and the boundary condi- by about 3% (e.g. 1750 instead of 1708 in the case of the standard
tions (32) constitute a linear boundary-value problem that can Bénard problem). In this case the approximation happens to give
be solved using the method of normal modes. an overestimate by about 1%. For a non-zero value of Rn we expect
We write that the approximation will be less accurate, but nevertheless we
believe that Eq. (44) gives a useful upper bound to the value of
ðw0 ; T 0 ; /0 Þ ¼ ½WðzÞ; HðzÞ; UðzÞ expðst þ ilx þ imyÞ; ð34Þ the critical Rayleigh number.
214 D.A. Nield, A.V. Kuznetsov / International Journal of Heat and Mass Transfer 68 (2014) 211–214

We observe that the parameter NB does not appear to this order a different scaling (a typical nanofluid fraction rather than the dif-
of approximation. Terms in NB drop out because of an orthogonal ference of two fractions). A major difference is that the sign of Rn
property of the first-order trial functions and their first derivatives. cannot be negative, and hence oscillatory convection cannot occur.
This means that, at this approximation, the average contribution of Also, it should be noted that in deriving the result in [3] it was as-
the nanoparticle flux in the thermal energy conservation Eq. (36) is sumed that the Lewis number Le was large compared with unity,
zero. but in the present paper we have imposed no restriction on the
magnitude of Le.
4. Conclusions
References
Except for the approximation of 4p2 by 40, Eq. (44) is formally
[1] D.A. Nield, A. Bejan, Convection in Porous Media, fourth ed., Springer, New York,
the same as Eq. (63) of [3]. Thus the conclusions in [3] concerning 2013.
non-oscillatory instability are unchanged by the change of bound- [2] J. Buongiorno, Convective transport in nanofluids, ASME J. Heat Transfer 128
ary conditions. These conclusions are that the parameter NB is not (2006) 240–250.
[3] D.A. Nield, A.V. Kuznetsov, Thermal instability in a porous medium layer
significant at first order, and the effect of the nanofluid Rayleigh saturated by a nanofluid, Int. J. Heat Mass Transfer 52 (2009) 5796–5801.
number Rn is destabilizing by an amount that depends of the [4] D.A. Nield, A.V. Kuznetsov, The onset of convection in a horizontal nanofluid
porosity, the Lewis number, and the modified diffusion ratio NA de- layer of finite depth, Eur. J. Mech. B/Fluids 29 (2010) 217–223.
fined by Eq. (19). However, it should be noted that Rn now involves

You might also like