359

Viscous dissipation effect on mixed convection

flow of a micropolar fluid over an exponentially
stretching sheet
Mohamed Abd El-Aziz

Abstract: Micropolar boundary-layer flow and heat transfer characteristics associated with a heated exponential stretching
continuous sheet being cooled by a mixed convection flow are examined. The relevant heat transfer mechanisms are of in-
terest in a wide variety of practical applications such as hot rolling, continuous casting, extrusion, and drawing. The wall
temperature and stretching velocity are assumed to vary according to specific exponential forms. The contributions of
buoyancy along with viscous dissipation on the convective transport in the boundary-layer region is analyzed in the oppos-
ing and assisting flow situations. Local similarity solutions are obtained for the boundary-layer equations governing the
problem. A parametric study of the mixed convection parameter x, the micropolar parameter D, the Eckert number Ec, the
parameter of temperature distribution n, and Prandtl number Pr is conducted and a representative set of numerical results
for the velocity, angular velocity, temperature profiles, local skin friction coefficient, wall couple stress parameter, and lo-
cal Nusselt number are illustrated graphically to show typical trends of the solutions.
PACS No: 44.20.+b
Résumé : Nous avons étudié l’écoulement d’une couche limite micropolaire et le transfert de chaleur caractéristique asso-
cié à une feuille continue chauffée en étirement exponentiel et refroidie par écoulement convectif mixte. Les mécanismes
de transfert de chaleur sont d’intérêt dans une vaste gamme d’applications pratiques, comme le laminage à chaud, le mou-
lage en continu, l’extrusion et l’étirage. La température du mur et la vitesse d’étirement sont présumées varier selon cer-
taines lois exponentielles. Nous analysons les contributions de la flottabilité et de la dissipation visqueuse sur le transport
convectif dans la couche limite en écoulement stimulé. Une transformation de similarité permet de réduire les équations
gouvernantes en un ensemble d’équations différentielles ordinaires qui sont alors solutionnées par la méthode de tir. Nous
complétons une étude paramétrique du paramètre de convection x, du paramètre micropolaire D, du nombre de Eckert Ec,
du paramètre de la distribution en température n et du nombre de Prandtl Pr. Nous en étudions numériquement les effets
sur la vitesse, la vitesse angulaire, les profils de température, le coefficient de friction de peau, le paramètre de contrainte
sur le mur et le nombre de Nusselt et les résultats sont affichés graphiquement, afin de mettre en évidence les tendances
typiques des solutions.
[Traduit par la Rédaction]

1. Introduction hand, Peddiesen and McNitt [5] and Wilson [6] initiated the
boundary-layer approximation in the study of micropolar
Micropolar fluids are non-Newtonian fluids with micro- fluids. Also, the continuous surface concept was introduced
structures such as polymeric additives, colloidal suspensions, by Sakiadis [7]. Continuous surfaces are surfaces such as
liquid crystals, etc. Eringen developed the theory of micro- those of polymer sheets or filaments continuously drawn
polar fluids [1], in which the microscopic effects arising from a die. The boundary-layer flow on continuous surfaces
from the local structure and the micromotions of the fluids is an important type of flow occurring in a number of tech-
elements are taken into account, and extended it into the nical processes. Examples may be found in continuous cast-
theory of thermomicropolar fluids [2]. In engineering, appli- ing, glass fiber production, metal extrusion, hot rolling, the
cations of micropolar fluids include solidification of liquid cooling and (or) drying of paper and textiles, and wire draw-
crystals, lubricants, and colloidals suspensions, while in bi- ing (see refs. 8–10). The study of heat transfer and the flow
ology the animal blood may be modelled as a micropolar field is necessary for determining the quality of the final
fluid. A comprehensive review of the subject and the vari- products of these processes as explained by Karwe and
ous applications of micropolar fluids has been done by Jaluria [11, 12]. A broad effort has been made to gain infor-
Ariman et al. [3]. A recent book by Lukaszewicz [4] focuses mation regarding the stretching flow problems in various sit-
on the theory and applications of such fluids. On the other uations. Such situations include consideration of porous

Received 30 December 2008. Accepted 27 April 2009. Published on the NRC Research Press Web site at cjp.nrc.ca on 22 June 2009.
M. AbdEl-Aziz.1 Helwan University, Faculty of Science, Mathematics Department, P.O.Box, 11795, Cairo, Egypt.
1Present address: King Khalid University, Faculty of Science, Mathematics Department, Abha 9004, Saudi Arabia
(email: [email protected]).

Can. J. Phys. 87: 359–368 (2009) doi:10.1139/P09-047 Published by NRC Research Press
360 Can. J. Phys. Vol. 87, 2009

surface, non-Newtonian fluids, magnetohydrodynamic fluid, toward the fluid (see Fig. 1). Under the above assumptions
heat transfer, mass transfer, porous medium, slip effects, and invoking the Boussinesq approximation, the boundary-
etc. A vast body of literature is now available on the topic. layer equations governing the flow and heat transfer of a mi-
Some very recent attempts in this direction have been made cropolar fluid over a moving vertical surface can be ex-
in the investigations described in refs. 13–18. All the above pressed as
mentioned investigations were limited to a contiuous surface
moving with a constant or a linear velocity. The steady free @u @v
þ ¼0 ð1Þ
convection flow and heat transfer from an exponentially @x @y
stretching vertical surface with an exponential temperature  
distribution is analyzed by Magyari and Keller [19] both an- @u @u m þ k @ 2 u k @N
u þv ¼ þ þ gbðT  T1 Þ ð2Þ
alytically and numerically. Elbashbeshy [20] examined the @x @y r @y2 r @y
flow and heat transfer characteristics by considering an ex-    
ponentially stretching continuous surface. Recently, viscoe- @N @N @2N @u
rj u þv ¼ g 2  k 2N þ ð3Þ
lastic boundary-layer flow over an exponential stretching @x @y @y @y
continuous sheet have been examined by [21, 22]. In actual
   
practice, the flow over a continuous material moving @T @T @2T m þ k @u 2
through a quiescent fluid is induced by the movement of u þv ¼a 2 þ ð4Þ
@x @y @y rcp @y
the solid material and by thermal buoyancy. Therefore, these
two mechanisms, the surface motion and buoyancy force,
will determine the momentum and thermal transport proc- Subject to the boundary conditions
esses. The thermal buoyancy force arising due to the heating
u ¼ Uw ðxÞ; v ¼ 0; N ¼ 0; T ¼ Tw ðxÞ at y¼0
or cooling of a continuously moving surface, under some ð5Þ
circumstances, may significantly alter the flow and thermal u ! 0; N ! 0; T ! T1 at y ! 1
fields and therby the heat transfer behavior in the manufac- where u and v are the velocity components along the x-and
turing process. Also, on account of the recent advances in y-axes, respectively, T is the fluid temperature, N is the
electronics, nuclear energy, and space technology, the study component of the microrotation vector normal to the x–y
of buoyancy-induced flows of non-Newtonian fluid contin- plane, r is the density, j is the micro-inertia density, m is
ues to be a major area of interest. Some important practical the dynamic viscosity, k is the gyro-viscosity (or vortex
applications involve the predictions of environmental pollu- viscosity), g is the spin-gradient viscosity, að¼ k=rcp Þ is the
tion in the ocean and atmosphere, thermal stratification in thermal diffusivity, k is the fluid thermal conductivity, and
lakes, as well as development and design of chemical proc- cp is the heat capacity at constant pressure. The stretching
essing equipment. However, in all of the papers cited earlier velocity Uw ðxÞ is assumed to be of the form [20]
the effect of the buoyancy force was relaxed and the follow- x
ing papers have taken it into consideration. Such papers are Uw ðxÞ ¼ U0 exp ð6Þ
those of Abo-Eldahab and Abd El-Aziz [23], Abd El-Aziz L
and Salem [24], Salem and Abd El-Aziz [25], Abd El-Aziz
[26] and Partha et al. [27]. The purpose of the present study Here, U0 is a constant and L is the reference length. The
is to extend the work of Magyari and Keller [19] to a micro- above exponential form is valid only when x << L, which oc-
polar fluid flow and heat transfer taking into account the ef- curs very close to the slit. Also the surface temperature
fect of the buoyancy force and viscous dissipation. This Tw ðxÞ of the stretching sheet is assumed to be in the form
study may be applicable to polymer technology involving  nx 
the stretching of plastic sheets. Many metallurgical proc- Tw ðxÞ ¼ T1 þ T0 exp ð7Þ
2L
esses involve the cooling of continuous strips or filaments
by drawing them through a quiescent fluid. During the proc- where n and T0 are the parameters of temperature distribu-
ess of drawing, the strips are sometimes stretched. In such tion on the stretching surface and T? is the temperature far
situations the rate of cooling has a great effect on the prop- away from the stretching sheet.
erties of the final product. By drawing them in a micropolar Introducing the dimensionless variables h, f, h, and q as
fluid the rate of cooling may be controlled, thereby giving rffiffiffiffiffiffiffiffi
the final product some desired characteristics. U0 x
h¼y exp ð8Þ
2nL 2L
2. Analysis pffiffiffiffiffiffiffiffiffiffiffiffiffi x
Consider the mixed convection boundary-layer flow over jðx; yÞ ¼ 2nLU0 exp f ðx; hÞ ð9Þ
2L
an impermeable plane vertical wall stretching with velocity rffiffiffiffiffiffiffiffiffi  
Uw and a given temperature distribution Tw moving through U03 3x
a quiescent micropolar fluid with constant temperature T1 . N¼ exp hðx; hÞ ð10Þ
2nL 2L
The induced motion of the fluid is assumed to be laminar,
steady, and two-dimensional with a thermally active incom- T ¼ T1 þ ðTw  T1 Þqðx; hÞ ð11Þ
pressible viscous fluid with constant properties. The x-axis is
directed along the continuous stretching surface and points in which j(x, y) is the physical stream function that auto-
in the direction of motion with the slot as the origin and the matically assures mass conservation (1), n is the kinematic
y-axis is perpendicular to the plate in the outward direction viscosity of the fluid and ðTw  T1 Þ is given by (7).

Published by NRC Research Press

Abd El-Aziz 361

Making use of the transformations (8)–(11) in (1)–(4) we In addition to the velocity and temperature fields, further
obtain the following partial differential equations: quantities of engineering interest are the local skin friction
s
coefficient
ð1 þ DÞfhhh þ f fhh  2fh2 þ Dhh þ 2xq ¼ 2Lðfh fhx  fhh fx Þ ð12Þ "   #
2 @u
Cf x ¼ ðm þ kÞ þ kðNÞy¼0
lhhh þ f hh  3fh h  2DBð2h þ fhh Þ ¼ 2Lðfh hx  fx hh Þ ð13Þ rUw2 @y y¼0
pffiffiffi
1 ¼ 2ð1 þ DÞRe1=2 x f 00 ð0Þ ð20Þ
qhh þ f qh  nfh q þ ð1 þ DÞEc fh2 ¼ 2Lðfh qx  fx qh Þ ð14Þ
Pr the couple stress at the wall
where D ¼ k=m is the micropolar parameter, x ¼ Grx =Re2x is  
the mixed convection or buoyancy parameter @N gUw2 0
Mw ¼ g ¼ h ð0Þ ð21Þ
with Grx ¼ gbðTw  T1 ÞL3 =n2 being the local Grashof @y y¼0 2nL
number, Rex ¼ Uw L=n is the local Reynolds number, and the local Nusselt number
Pr ¼ n=a is the Prandtl number, Ec ¼ Uw2 =cp ðTw  T1 Þ is  
the Eckert number, and l ¼ g=mj and B ¼ nL=ðjUw Þ are the L @T 1 0
Nux ¼  ¼  pffiffiffi Re1=2
x q ð0Þ ð22Þ
dimensionless parameters. ðTw  T1 Þ @y y¼0 2
In our present problem, the presence of a buoyancy term
in the momentum equation (2), the dimensionless parameter
B in the angular momentum equation (13), and the dissipa- 3. Numerical solution
tion term in the energy equation (4) do not allow the prob-
lem to have self-similar solutions. Hence, we turn our The set of nonlinear coupled differential equations (15)–
attention to obtain a local similarity solution of the problem (17) subject to the boundary conditions (18) and (19) consti-
by introducing a pseudo-similarity variable. To obtain a lo- tute a two-point boundary-value problem. To solve these
cal similarity solution, we must have the condition that de- equations numerically, we follow most efficient numerical
rivatives of f, h, and q with respect to x do not exist. shooting technique with fifth-order Runge–Kutta–Fehlberg
Therefore, f(x, h), h(x, h), and q(x, h) should be assumed to integration scheme. In this method it is most important to
be f(h), h(h), and q(h), i.e., f ðx; hÞ ¼ f ðhÞ; hðx; hÞ ¼ hðhÞ, choose the appropriate finite values of h ! 1. To select
and qðx; hÞ ¼ qðhÞ [22]. With this assumptions the right- h1 , we begin with some initially estimated value and solve
hand sides of (12)–(14) vanish and we obtain the following the problem with some particular set of parameters to obtain
third-order nonlinear coupled quasi-ordinary differential fhh ð0Þ, hh ð0Þ, and qh ð0Þ . The solution process is repeated
equation of the form: with other large values of h1 until two successive values of
fhh ð0Þ, hh ð0Þ, and qh ð0Þ differ only after the desired digit
ð1 þ DÞfhhh þ f fhh  2fh2 þ Dhh þ 2xq ¼ 0 ð15Þ signifying the limit of the boundary along h. The last value
of h1 is chosen as an appropriate value of the limit h ! 1
lhhh þ f hh  3fh h  2DBð2h þ fhh Þ ¼ 0 ð16Þ for that particular set of parameters.
The three ordinary differential equations (15)–(17), were
1 2 first formulated as a set of seven first-order simultaneous
qhh þ f qh  nfh q þ ð1 þ DÞEc fhh ¼0 ð17Þ equations of seven unknowns following the method of
Pr
superposition [28]. To solve this system we require seven in-
fh ð0Þ ¼ 0; f ð0Þ ¼ 0; hð0Þ; qð0Þ ¼ 1 ð18Þ itial conditions whilst we have only two initial conditions
f(0) and fh ð0Þ on f, one initial condition h(0) on h and one
fh ð1Þ ¼ 0; hð1Þ ¼ 0; qð1Þ ¼ 0 ð19Þ initial condition q(0) on q. Still there are three initial condi-
tions fhh ð0Þ, hh ð0Þ, and qh ð0Þ that are not prescribed. How-
where subscript h stands for differentiation with respect to h. ever the values of fh ðhÞ, hðhÞ, and qðhÞ are known at
We notice that when x = 0, (15)–(17) are uncoupled and a h ! 1. Now we employ the numerical shooting technique
purely forced convection situation results. In this case, the where these three ending boundary conditions are utilized to
flow field is not affected by the thermal field. The sign of x produce three known initial conditions at h ¼ 0. Finally, the
characterizes the nature of the departure from this situation. problem has been solved numerically using fifth-order
For x > 0 (Tw > T1 ), the buoyancy forces act in the direc- Runge–Kutta integration scheme.
tion of the mainstream and the fluid is accelerated in the To verify the validty and accuracy of the present analysis,
manner of a favorable pressure gradient (assisting flow). results for the heat transfer rate qh ð0Þ were compared with
When x < 0 (Tw < T1 ), the buoyancy forces oppose the those reported by Magyari and Keller [19] for forced con-
stretching induced flow, retarding the fluid in the boundary vective flow (x = 0) of a Newtonian fluid (D = 0) over an
layer, acting as an adverse pressure gradient (opposing exponentially streching sheet in the absence of viscous dissi-
flow). It is worth mentioning that, according to the defini- pation (Ec = 0). The comparison is found to be in good
tion of the viscous dissipation parameter, Ec > 0 in the case agreement, as shown in Table 1.
of assisting flow and Ec < 0 in the case of opposing flow.
We also notice that when D = 0 (Newtonian fluid), (15)–
(17) reduce to those of Magyari and Keller [19]. Thus, the 4. Results and discussion
reported results can be used for comparison and support the Figures 2–4 illustrate the velocity, angular velocity, and
accuracy of the present results. temperature fields for different values of buoyancy parame-

Published by NRC Research Press

362 Can. J. Phys. Vol. 87, 2009

Table 1. Comparison of the values of –q 0 (0) for x = D = l = B = Ec = 0 and various values of Pr and n with those of
Magyari and Keller [19].

n = –0.5 n = 0.0 n=1 n=3

Pr Ref. 19 Present Ref. 19 Present Ref. 19 Present Ref. 19 Present
0.5 0.175 815 0.176 182 0.330 493 0.330 764 0.594 338 0.594 493 1.008 405 1.008 465
1 0.299 876 0.299 877 0.549 643 0.549 644 0.954 782 0.954 785 1.560 294 1.560 295
3 0.634 113 0.634 110 1.122 188 1.122 087 1.869 075 1.869 074 2.938 535 2.938 534
5 0.870 431 0.870 427 1.521 243 1.521 239 2.500 135 2.500 132 3.886 555 3.886 556
10 1.308 613 1.308 608 2.257 429 2.257 424 3.660 379 3.660 372 5.628 198 5.628 196

Fig. 1. Schematic representation of the physical model and coordi- Fig. 2. Velocity profiles for various values of x and D with Ec =
nates system. 0.2, n = 1, l = 0.3, B = 0.1, and Pr = 0.72.

ter x and micropolar parameter D in the absence of viscous Fig. 3. Angular velocity profiles for various values of x and D with
dissipation (Ec = 0) for Pr = 0.72, n = 1, l = 0.3, and B = Ec = 0.2, n = 1, l = 0.3, B = 0.1, and Pr = 0.72.
0.1. It can be seen from Fig. 2 that for assisting flow,
namely, for positive values of x, the velocity gradient at the
sheet increases as the buoyancy increases, with an accompa-
nying increase in the velocity f 0 (h) for small and large values
of D. For the opposing flows (x < 0), on the other hand, the
effect of buoyancy is to reduce the velocity compared to
those for pure forced convection (x = 0). This is due to the
fact that a positive x induces a favorable pressure gradient
that enhances the fluid flow in the boundary layer, while a
negative x produces an adverse pressure gradient that slows
the fluid motion down. Also, it can be seen that for large val-
ues of the buoyancy parameter (x = 10) the velocity near the
wall overshoots the moving speed of the sheet. Further, f 0 (h)
increases as D increases for opposing flow (x = –0.2) and
small values of x (x = 0 and 1) in the assisting flow while
as x increases (x = 10) the velocity f 0 (h) is greatly decreased
as D increases near the sheet where 0  h < h0 ffi 1:1 but the
opposite result is observed for h > h0 .
Figure 3 shows for assisting flow that close to the surface
of the sheet, an increase in the values of x leads to the de- the angular velocity away from the surface. From Fig. 3, it
crease in the value of the angular velocity h(h) while for is noteworthy that the direction of the angular velocity is
most of the boundary layer at a fixed position h the angular negative close to the wall for large values of the buoyancy
velocity h(h) increases with the increase in the buoyancy pa- parameter (x = 10) and both values of D, which may be a
rameter x. For the opposing flows, on the other hand, the ef- direct result of the velocity overshoot of f 0 (h) near the wall
fect of buoyancy parameter on h(h) is opposite to that of the for x = 10. Also, as D increases the magnitude of the angu-
assisting flow. Also, increasing the value of the buoyancy lar velocity increases with an increase in the boundary-layer
parameter x moves the location of the maximum value of thickness for fixed values of buoyancy parameter x.

Published by NRC Research Press

Abd El-Aziz 363

Fig. 4. Temperature profiles for various values of x and D with Ec Fig. 6. Angular velocity profiles for various values of x and Ec with
= 0.2, n = 1, l = 0.3, B = 0.1, and Pr = 0.72. D = 1, n = 1, l = 0.3, B = 0.1, and Pr = 0.72.

Fig. 5. Velocity profiles for various values of x and Ec with D = 1, Fig. 7. Temperature profiles for various values of x and Ec with D
n = 1, l = 0.3, B = 0.1, and Pr = 0.72. = 1, n = 1, l = 0.3, B = 0.1, and Pr = 0.72.

From Fig. 4, we conclude that, for both values of the mi- definition of Eckert number, a positive Ec corresponds to
cropolar parameter D, the effect of buoyancy parameter x is fluid heating (heat is being supplied across the wall into the
to increase the temperature q(h) in the opposing flow and fluid) case (Tw > T1 Þ so that the fluid is being heated
decrease it in the case of the assisting flow. Further, it is whereas a negative Ec means that the fluid is being cooled.
noted from Fig. 4 that the effect of D is to decrease the tem- From Fig. 7, it can be seen that the dimensionless tempera-
perature q(h) significantly for the opposing flow and forced ture increases when the fluid is being heated (Ec > 0) but de-
convection regime (x = 0) and this effect is imperceptible creases when the fluid is being cooled (Ec < 0). For Ec < 0
for x = 1 but as x increases more (x = 10), the relation be- the dimensionless fluid temperature q(h) decreases monotoni-
tween the temperature q(h) and D explained earlier is cally with h, from unity at the wall towards its free-stream
switched. In other words, the micropolar fluid is better value. It is noted from the definition of q that this behavior
suited for effective cooling of the stretching sheet only implies the monotonous decrease in the actual fluid temper-
when the flow is forced convective. ature in the horizontal direction from the sheet temperature
Figures 5–7 illustrate the effect of viscous dissipation on Tw to the free-stream temperature. The increase in the fluid
the velocity, microrotation, and temperature fields in the temperature due to viscous heating is observed to be more
case of assisting and opposing flow for Pr = 0.72, n = 1, l = pronounced for a higher value of Ec, as expected. On the
0.3, B = 0.1, and D = 1. From Fig. 5 it can be seen that the other hand, for Ec < 0 (i.e., Tw < T1 ) the dimensionless fluid
viscous dissipation increases the velocity f 0 (h) in both oppos- temperature q decreases with h rapidly at first, arriving at a
ing (x < 0 and hence Ec < 0 according its definition) and as- negative minimum value, for Ec < –2 and then increases
sisting (x > 0 and hence Ec > 0) flow. According to the more gradually to its free surface value. Correspondingly,

Published by NRC Research Press

364 Can. J. Phys. Vol. 87, 2009

Fig. 8. Velocity profiles for various values of D and Ec with x = 1, Fig. 10. Temperature profiles for various values of D and Ec with x
n = 1, l = 0.3, B = 0.1, and Pr = 0.72. = 1, n = 1, l = 0.3, B = 0.1, and Pr = 0.72.

Fig. 9. Angular velocity profiles for various values of D and Ec Fig. 11. Velocity profiles for various values of n with D = 1, x = 1,
with x = 1, n = 1, l = 0.3, B = 0.1, and Pr = 0.72. Ec = 0.2, n = 1, l = 0.3, B = 0.1, and Pr = 0.72.

the actual fluid temperature in the horizontal direction in- angular velocity h(h) first decreases near the sheet surface
creases at first from the surface temperature Tw to a maxi- where 0 £ h < 2.1 but the situation is completely reversed in
mum value and then decreases to its free-stream value. It the other part of the boundary layer where h > 2.1 and this is
should be noted that for the fluid cooling case (Ec < 0) a neg- true for opposing and assisting flow.
ative q indicates the excess of actual fluid temperature T over Representative velocity f 0 (h), angular velocity h(h), and
that at the plate because of the viscous dissipation effect. It is temperature q(h) distributions in the boundary layer for vari-
known that viscous dissipation produces heat due to drag be- ous values of a micropolar parameter D and Eckert number
tween the fluid particles and this extra heat causes an in- Ec for Pr = 0.72, n = 1, l = 0.3, B = 0.1, and x = 1, are pre-
crease of the initial fluid temperature (see Fig. 7). This sented in Figs. 8–10, respectively. Figures 8–10 show that
increase in temperature causes an increase in the buoyant the velocity and temperature profiles of a micropolar fluid
force. There is a continuous interaction between the viscous (D = 3) exhibit similar fashions to those of a Newtonian
heating and the buoyant force. This mechanism produces dif- fluid (D = 0). Also, viscous dissipation demonstrates a more
ferent results in the assisting (upward) and opposing (down- pronounced influence on the velocity and temperature distri-
ward) flow. In the assisting (opposing) flow, the increase in bution of a micropolar fluid than on those of a Newtonian
the values of positive (negative) Ec will increase the buoyant fluid. This may be explained by the fact that the micropolar
force in the upward (downward) direction, which results in parameter D is proportional to the spin gradient viscosity of
an increase in the fluid velocity as shown in Fig. 5. The pos- the fluid microstructure and increasing it results in an
itive and negative Eckert numbers assist the upward and increase in the viscosity of the micropolar fluid, which in
downward flow, respectively. It is noted from Fig. 6 that the turn increases the viscous dissipation. On the other hand, the

Published by NRC Research Press

Abd El-Aziz 365

Fig. 12. Angular velocity profiles for various values of n with D = Fig. 14. Local skin friction coefficient f 00 (0) versus x for various
1, x = 1, Ec = 0.2, n = 1, l = 0.3, B = 0.1, and Pr = 0.72. values of D and Ec with n = 1, l = 0.3, B = 0.1, and Pr = 0.72.

Fig. 13. Temperature profiles for various values of n with D = 1, x Fig. 15. Wall couple stress h 0 (0) versus x for various values of D
= 1, Ec = 0.2, n = 1, l = 0.3, B = 0.1, and Pr = 0.72. and Ec with n = 1, l = 0.3, B = 0.1, and Pr = 0.72.

angular velocity h(h) decreases in the vicinity of the plate the fluid as depicted in Fig. 11. On the other hand, Fig. 12
but increases as one moves away from it, on increasing the shows that increasing the values of n results in increasing
Eckert number Ec but this trend is found to be more notice- the angular velocity h(h) close to the plate along with shift-
able at a larger micropolar parameter as shown in Fig. 9. ing the maximum away from the plate while the opposite
Figures 11–13 depict the respective changes in the profiles tendency is observed as one moves away from the plate.
of f 0 (h), h(h), and q(h) as the temperaure parameter n is Typical variations of the local skin friction coefficient in
changed for Pr = 0.72, n = 1, l = 0.3, B = 0.1, Ec = 0.2, terms of f 00 (0), wall couple stress in terms of h 0 (0), and the
and D = x = 1. From Fig. 13, it can be seen that the temper- local Nusselt number in terms of –q 0 (0) as a function of the
ature q(h) decreases as n increases. It is also observed that a buoyancy parameter x for various values of a micropolar pa-
distinctive peak in the temperature profile occurs in the fluid rameter D and Eckert number Ec are presented in Figs. 14–
adjacent to the wall for some negative values of n. The pres- 16. It is clear from Fig. 14 that for fixed x, the local skin
ence of the peak indicates that the maximum temperature friction coefficient f 00 (0) increases as Ec increases for both
value occurs in the body of the fluid close to the surface Newtonian (D = 0) and micropolar (D = 3) fluids. This is
and not at the surface and consequently heat is expected to true in opposing (x < 0) and assisting (x > 0) flow. In addi-
transfer to the wall from the ambient fluid. Also the decrease tion, the effect of viscous dissipation on f 00 (0) is more pro-
in temperature has a direct effect on decreasing the thermal nounced for higher values of jxj. It is also obseved that the
buoyancy forces, which in turn decrease the velocty f 0 (h) of local skin friction coefficient f 00 (0) of a micropolar fluid is

Published by NRC Research Press

366 Can. J. Phys. Vol. 87, 2009

Fig. 16. Local Nusselt number –q 0 (0) versus x for various values of Fig. 18. Wall couple stress h 0 (0) versus x for various values of Pr
D and Ec with n = 1, l = 0.3, B = 0.1, and Pr = 0.72. and n with D = 1, Ec = 0.2, l = 0.3, and B = 0.1.

Fig. 17. Local skin friction coefficient f 00 (0) versus x for various Fig. 19. Local Nusselt number –q 0 (0) versus x for various values of
values of Pr and n with D = 1, Ec = 0.2, l = 0.3, and B = 0.1. Pr and n with D = 1, Ec = 0.2, l = 0.3, and B = 0.1.

higher than that of a Newtonian fluid for all x and Ec but this are not affected by the thermal field when buoyancy force
behavior is more pronounced in opposing flow. For all Ec is zero. Further, viscous dissipation demonstrates a more
and D, the buoyancy force enhances the local skin friction pronounced influence on the local skin friction coefficient
coefficient in assisting flow, which is a consequence of the f 00 (0) in the opposing flow than that of assisting flow.
existence of a favorable pressure gradient along the plate Figure 15 reveals that for a given x, the couple stress
due to the buoyancy effects. On the other hand, the buoy- h 0 (0) decreases with viscous dissipation in both fluid heating
ancy force in opposing flow reduces f 00 (0) of a Newtonian (Ec > 0) and cooling (Ec < 0) cases and for both D values
fluid (D = 0) for all Ec values, but for a micropolar fluid (D but this trend is found to be more noticeable at larger values
= 3) the local skin friction coefficient decreases with buoy- of jxj and D. Also, the couple stress for assisting flow is
ancy force for Ec = –0.5 approximately constant for Ec = –4 greatly decreased as x is increased for all D values but this
and increases slightly for Ec = –7. Further, it is noted from trend is found to be more noticeable at larger values of D.
Fig. 14 that a single value of f 00 (0) = –1.281 86 for a New- In opposing flow, on the other hand, when D = 1 the couple
tonian fluid and –0.611 838 for a micropolar fluid is ob- stress h 0 (0) decreases slightly with the buoyancy force for
tained for all |Ec| values when x = 0 (the forced Ec = –0.5, nearly constant for Ec = –4 and increases greatly
convection flow). This is because that (15) and (17) are for Ec = –7 while, h 0 (0) is greatly decreased for Ec = –0.5, –4
uncoupled when x = 0, i.e., the solution to the flow field and is greatly increased for Ec = –7. Moreover, for forced

Published by NRC Research Press

Abd El-Aziz 367

convection flow (x = 0), the wall couple stress has a tinent parameters is discussed. The following conclusions
unique value of 0.162 001 for D = 1 despite the magnitude can be drawn as a result of the computations:
of Ec while the corresponding unique value of h 0 (0) is (1) An increase in the micropolar parameter leads to a fas-
0.319 919 for D = 3 for the same reason mentioned above. ter rate of cooling of the sheet for forced convective
For given D and x, Fig. 16 reveals that the local Nusselt flow only.
number –q 0 (0) of assisting flow is reduced due to an in- (2) The local Nusselt number decreases for negative va-
crease in the Eckert number Ec. Also, the influence of vis- lues of n as Pr increases, while the opposite trend is
cous dissipation on –q 0 (0) is seen to be more noticeable for true for positive values of n.
lower values of x and higher values of D. The opposite is (3) When n is increased, the heat transfer rate is enhanced
true in the case of the opposing flow. Further, it worth not- for all Prandtl numbers and the buoyancy parameter.
ing here that an adiabatic (–q 0 (0) = 0) case where the heat
(4) The local Nusselt number is increased monotonously
transfer process is stopped occurred for Newtonian (D = 0)
with increase in the buoyancy parameter for all Pr
fluid when Ec = 0.4 and x = 0.28. This can be explained by
and positive values of n.
the fact that the heat transfer process is reversed at some
values of Ec and x, i.e., the flow of heat is directed from (5) For negative values of n and all Pr, the local Nusselt
the surface to the ambient fluid (positive values of –q 0 (0) in number is increased for small and moderate values of
Fig. 16) and then reversed from the ambient fluid to the sur- x and then decreases for higher values of x.
face (negative values of –q 0 (0)). (6) For all values of the buoyancy parameter in buoyancy-
assisting flow, the local Nusselt number is increased
Variation of the local skin friction coefficient f 00 (0), wall
for positive n as Pr increases while the opposite trend
couple stress h 0 (0), and the local Nusselt number –q 0 (0) as a
is true for negative n.
function of x for various values of n and Pr are presented in
Figs. 17–19. As shown in Fig. 17, an increase in the value (7) Viscous dissipation effect reduces the heat transfer
of n leads to a decrease in the values of f 00 (0) for fixed x rate for all x and D in the buoyancy-assisting flow but
and both values of Pr. Also, it is obvious that for given n this effect is seen to be more noticeable for small va-
and x, the local skin friction coefficient f 00 (0) is decreased lues of x while the opposite is true in the case of buoy-
with an increase in Pr. In addition, the effect of x on f 00 (0) ancy-opposing flow.
becomes more significant for a smaller Pr. This behavior (8) The influence of viscouss dissipation in both assisting
implies that fluids having a smaller Pr are much responsive and opposing flow on the heat transfer of a micropolar
to thermal buoyancy than fluids having larger Pr. fluid is seen to be more noticeable than that of a New-
tonian fluid.
From Fig. 18, it is seen that the effect of n, Pr, and x on
the wall couple stress h 0 (0) is opposite to that on f 00 (0). In (9) A rapid growth in the local skin friction coefficient of
addition, for forced convection flow (x = 0), f 00 (0), and a micropolar fluid is noticed with the mixed convec-
tion parameter for all n and Pr.
h 0 (0) have a unique value of –0.896 048 and 0.162 001, re-
spectively, despite the values of n and Pr. (10) The local skin friction coefficient of a micropolar fluid
decreases with the buoyancy force of the opposing
The curves in Fig. 19 show that the local Nusselt nimber –
flow for small values of Ec, is nearly constant for
q 0 (0) is increased with increasing the value of n for all Pr
moderate values of Ec, and increases with higher va-
and x, in other words, the heat transfer rate can be enhanced lues of Ec.
by enlarging the surface temperature variation. It is noted
that negative heat transfer rates are obtained for n = –0.8.
References
Negative values of –q 0 (0) indicate that heat is transferred
from the fluid to the moving surface in spite of the excess 1. A.C. Eringen. J. Math. Mech. 6, 1 (1960).
of surface temperature over that of the free stream fluid. 2. A.C. Eringen. J. Math. Anal. Appl. 38, 480 (1972). doi:10.
This phenomenon can be explained as a fluid particle heated 1016/0022-247X(72)90106-0.
to nearly the wall temperature being convected downstream 3. T. Ariman, M.A. Turk, and N.D. Sylvester. Int. J. Eng. Sci.
to a location where the wall temperature is lower. Also, it is 12, 273 (1974). doi:10.1016/0020-7225(74)90059-7.
interesting to note that there are two adiabatic (–q 0 (0) = 0) 4. G. Lukaszewicz. Micropolar fluids: theory and applications.
Birkhauser, Basel. 1999.
cases where the heat transfer process is stopped for x = 0
5. J. Peddiesen and R.P. McNitt. Recent Adv. Eng. 5, 405
and 3.6 with n = –0.8 and Pr = 0.72. Finally, it is observed
(1970).
that for all x the local Nusselt number decreases for negative
6. A.J. Wilson. Math. Proc. Camb. Philos. Soc. 67, 469 (1970).
n and increases for positive n as Pr increases. doi:10.1017/S0305004100045746.
7. B.C. Sakiadis. Am. Inst. Chem. Eng. J. 7, 221 (1961).
5. Conclusion doi:10.1002/aic.690070211.
8. T. Altan, S. Oh, and H. Gegel. Metal forming fundamentals
The problem of mixed convective flow of a steady, in- and applications. American Society of Metals, Metals Park,
compressible micropolar fluid adjacent to an exponentially Ohio. 1979.
stretching sheet taking into account the effect of viscous dis- 9. E.G. Fisher. Extrusion of plastics. Wiley, New York. 1976.
sipation has been studied. This situation may arise in poly- 10. Z. Tadmor and I. Klein. Engineering principles of plasticat-
mer technology involving the stretching of plastics sheets. ing extrusion. Van nostrand Reinhold, New York. 1970.
Numerical solutions are presented for the fluid flow and 11. M.V. Karwe and Y. Jaluria. J. Heat Transfer, 110, 655
heat transfer characteristics and their dependence on the per- (1988).

Published by NRC Research Press

368 Can. J. Phys. Vol. 87, 2009

12. M.V. Karwe and Y. Jaluria. J. Heat Transfer, 113, 655 21. S.K. Khan and E. Sanjayanand. Int. J. Heat Mass Transfer,
(1988). 48, 1534 (2005). doi:10.1016/j.ijheatmasstransfer.2004.10.
13. F.K. Tsou, E.M. Sparrow, and R.J. Goldstein. Int. J. Heat 032.
Mass Transfer, 10, 219 (1967). doi:10.1016/0017-9310(67) 22. E. Sanjayanand and S.K. Khan. Int. J. Therm. Sci. 45, 819
90100-7. (2006). doi:10.1016/j.ijthermalsci.2005.11.002.
14. L.G. Grubka and K.M. Bobba. J. Heat Transfer, 107, 248 23. E.M. Abo-Eldahab and M. Abd El-Aziz. Int. J. Therm. Sci.
(1985). 43, 709 (2004). doi:10.1016/j.ijthermalsci.2004.01.005.
15. L.E. Erickson, L.T. Fan, and V.G. Fox. Ind. Eng. Chem. 24. M. Abd El-Aziz and A.M. Salem. Can. J. Phys. 85, 359
Fund. 5, 19 (1966). doi:10.1021/i160017a004. (2007). doi:10.1139/P07-048.
16. R. Cortell. Fluid Dyn. Res. 37, 231 (2005). doi:10.1016/j. 25. A.M. Salem and M. Abd El-Aziz. Appl. Math. Model. 32,
fluiddyn.2005.05.001. 1236 (2008). doi:10.1016/j.apm.2007.03.008.
17. A. Ali and A. Mehmood. Commun. Nonlinear Sci. Numer. 26. M. Abd El-Aziz. Phys. Lett. 372A, 263 (2008). doi:10.1016/
Simul. 13, 340 (2008). doi:10.1016/j.cnsns.2006.03.008. j.physleta.2007.07.015.
18. I. Liu; I–Chung Liu. Int. J. Non-linear Mech. 40, 465 (2005). 27. M.K. Partha, P.V.S.N. Murthy, and G.P. Rajasekhar. Heat
doi:10.1016/j.ijnonlinmec.2004.07.008. Mass Transfer, 41, 360 (2005). doi:10.1007/s00231-004-
19. E. Magyari and B. Keller. J. Phys. D: Appl. Phys. 32, 577 0552-2.
(1999). doi:10.1088/0022-3727/32/5/012. 28. T.Y. Na. Computational methods in engineering boundary
20. E.M.A. Elbashbeshy. Arch. Mech. 53, 643 (2001). value problems. Academic Press, New York. 1979.

Published by NRC Research Press