ME 324

HEAT AND MASS TRANSFER

By
Prof. P. Muthukumar
Department of Mechanical Engineering
IIT Guwahati
 Approximate solutions of boundary layer equations
Consider flow over heated flat plate as shown in Figure below.

Boundary layer equations: U∞ T∞

u v
+ =0 ( Eq.1)
x y
u u 1 p    2u 
u +v =− +  2 ( Eq.2)
x y  x   y 
T T k   2T 
u +v =   ( Eq.3)
x y  C p  y 2 
Tw
Boundary conditions:
Fig. Velocity and thermal boundary layers
@ y = 0; u = v = 0 and T = Tw
@ y =  ; u = U  and T = T
Consider the momentum equation

u u 1 p    2u 
u +v =− +  2
x y  x   y 

Using Integral method, Von Karman and Pohlhausen have written the above equation as follows.
   
u u 1 p    2u 
0 u x dy + 0 v y dy = 0 −  x dy + 0   y 2  dy ( Eqn.4 )

I II III IV
Term III is zero for flow over flat plate (because dp/dx = 0)
 
u v  
0 y  0 0 y udy
 b b

Term II: v dy = uv − 
dv b du
Integration byparts :  u dx = uv a −  vdx 
 a
dx a
dx 

u
= vu  − vu 0  +  u dy
  ( from continuity equation )
0
x u v
=−
u  = u ; u 0 = 0 x y
 
   2u 
Again from continuity Term IV: 0   y 2  dy
u v 
=− 
   2u      u    u 
x y 0   y 2  dy =
 0 y  y 
dy =
  y 0

  u u   u  u 
 
=  − =−  = 0
v u   y y y =0 
  y  y 
0 y dy = − 0 x dy
y = y =0
 y = 

   2u   u
0   y 2  dy = −
 y y =0

u
v y=  = −  dy Eqn. 4 becomes
0
x 
u u

 u
2  u dy − U   dy = −
0
x 0
x  y y =0
II term finally becomes
 
d d  u
( Eq.5)
dx 0 dx 0
 
@ y = 0; u = v = 0  u 2
dy − U udy = −
u u  
 y
−U   dy +  u dy   y =0

0
x 0
x  @ y =  ; u =U  
Eqn. 5 is known as Momentum Integral equation.
Assume the following velocity profile for zero
From Eqn. 5,
pressure gradient flat plate boundary layer flow:  
d d  u
dx 0 dx 0
u 2
dy − U  udy = −
 y
u = A + By + Cy 2 + Dy 3 where A, B, C , D are constants y =0

Boundary conditions: 
U2

  y 2  y 6  y 
4

0 u dy = 4 0 9    +    − 6     dy
2
@ y = 0; u = 0  A = 0
 

 2u
@ y = 0; = 0C = 0 U2   6 
y 2 =  3 + −
4 7 5 

@ y =  ; u = U   B + C 2 + D 3 = U   B + D 3 = U 
U  2 68
=
4 35
u
@ y = ; = 0  B + 3D 2 = 0
y 
U

 y  y 3 
3U
On solving , B =  and D = 3
−U 0 udy =
2 0 3  −    dy
2 2  

3
u 3 y 1 y 5 U 
= −   ( Eqn.6 ) =
U 2  2    4 2
  u    3 y 1 y 
3

− =− U  

−     @ x = 0;  = 0  C = 0
 y 
 y   2  2    
y =0   y =0
Finally

3 U 
x
=−  = 4.64
2  U 
Eqn. 5 becomes
U  2 d   68 5  3 U On substituting Re x = U  x
− = −
4 dx  35 2  2  

x

d  140 
=  = 4.64
dx 13 U  Re x

On Integration this gives

1 13 2  x
 = +C
2 140 U 
 Similarly, casting the integral form of thermal boundary layer equation :

T T k   2T  T 
 T 
+v =  T u  u
T
u  2 ( Eq.3)
0  x x 
+ −   = −
k
x y  C p  y 
u T dy T
x
dy
C p  
 y  y = 0
0
T  T
T T
T k   2T  T
 k  T 
0 u x dy + 0 v y dy = 
0
C p
 2
 y
 dy

0
x
u (T − T ) dy = −
C p 
 y

 y =0
or
T
d k  T 
 u (T − T ) dy = −
I II III
  ( Eqn.6)
dx 0
C p  y  y =0
Integrating II term by parts and applying continuity, we get
T T T
T v  k  T  
dy +  vT 0T

u
0
x
−  T dy = 
0
y   C p

 y

  0
This equation is called Energy Integral Equation
T 
T T
u k  T 
0 x
u dy + 0 x
T dy + T v y = T
= −
C p  
 y  y = 0
T
u
where v y = = −  dy and T v y = 0 = 0
T
0
x
In order to solve Eqn. 6, let us assume the temperature profile
T − Tw y T − Tw T − T
= = C0 + C1 + C2 2 + C3 3 where  =   −1 = − 1=
T − Tw T T − Tw T − Tw
ζ is a similarity variable T − T = ( − 1)(T − Tw )
First term of the energy integral (Eqn. 6) is:
Apply thermal boundary conditions T 𝛿𝑇
d
(T − Tw ) u ( − 1) dy
dx 0
@ y = 0; T = Tw,  @  = 0; = 0  C0 = 0 න 𝑢(𝑇 − 𝑇∞ ሻ 𝑑𝑦
0
 2T d 2 d
T
3 1 3 3 1 3 
@ y = 0; 2 = 0,  @  = 0; = 0  C2 = 0 U  (T − Tw ) 0  2 2   2 2  − 1 dy
 −   −
y d dx
@ y =  T ; T = T ,  @  = 1; = 1  C1 + C3 = 1 d 3  y  1  y   3  y
T 3
 1 y 
3

U  (T − Tw )     −       −   − 1 dy
T d dx 0  2    2      2   T  2  T  
@ y = T ; = 0,  @  = 1; = 0  C1 + 3C3 = 0 
y d T
Substituting  = and rearranging, we get
3
Solving for C1 and C3 ; C1 = and C3 = −
1 
2 2 d   3 2 3 4 3 2 1 4 3 2 1 4 
U  (T − Tw )    −  −  +  −  +  
T − Tw 3 1 dx   4 20 20 28 4 8 
 =  − 3
T − Tw 2 2 d   3 2 3 4 
U  (T − Tw )   −  +  
dx   20 280 
The RHS of energy integral equation (Eqn. 6) is evaluated in the following way :
3  y  1 y  
3

(T − Tw ) = (T − Tw )   −   
 2   T  2   T  
T  3 1 3 y2 
= (T − Tw )  − 3
y  T
2  2  T 

Pr >1
k T 3 (T − Tw )
− = −
 C p y y =0
2 T Fig. Different situations with Pr <1 and Pr >1

Now Eqn.6 becomes

d   3 2 3 4  3 (T − Tw )
U  (T − Tw )  −  +   =− ( Eqn.7)
dx   20 280  2 T
T
For Pr  1,    T ,  =
becomes small, and so  4   2

For Pr >1, Eqn. 7 becomes
d  2  
 U    =
dx  20  2
10 d 2 d This linear differential equation for which the integrating
=  + 2  2
U  dx dx 3
factor can be seen to be x 4
d  140  280  x d  3  43 13  3 1 − 14
where  = , and  2 =
3
x4 + x = x
dx 13 U  13 U  dx 4 x 14  4 Pr

dx
( ) 56 Pr
( )
Finally, on substitution of above two relations Eqn.7 d x 34  = 39 1 x − 14  d x 43  = 39 1 x − 14 dx ( On Integrating )
  56 Pr
 3 + 2 x ( )
d 2 13 1
 = ( Eqn.8 ) 39 1 x − 14 +1

x =
3
dx 14 Pr 4
  + b ; b = Integration constant
56 Pr  −41 + 1 
13 1 −3
= + bx 4
3 d 2 d 14 Pr
put  =  ; = 3
dx dx
Substituting in Eqn.8
4 x d  13 1
+ =
3 dx 14 Pr
d 3 13 1 3
+ =
dx 4 x 14 Pr 4 x

Fig. The flat plate situation with unheated initial length

Apply boundary conditions as described in the above Figure, we get
T
On substituting for we get
@ x = x0 ,  T = 0,  = 0 and so,@ x = x0 ,  = 0 y y =0

Substitutingi in Eqn.8, we get

−k
( T − Tw ) 3
13 1 3 4 T 2
b=− x0 h=
14 Pr (Tw − T )
After substituting in Eqn.8for b, we get or
3 1.026
( Pr ) ( Re x ) 2
3 1 1
13 1 13 1  x0  4
h =k 3

3 =  = −   2 4.64 x
14 Pr 14 Pr  x  hx
= 0.332 ( Re x ) ( Pr ) 3
1 1
2

For x0 → 0, we finally have k

 The local Nusselt number can be expressed as
( Pr )
−1
T = 3

Nu x = 0.332 ( Re x ) ( Pr )
1 1
2 3
1.026
L

 hx dx
Again we know that : Now, the average heat transfer coefficient, h = 0
L and so
 dx
T
0

h (Tw − T ) = −k The Average Nusselt number is

y y =0 Nu L = 0.664 ( Re L )
1
2
( Pr )
1
3
 Reynold Analogy
➢ If two dimensionless equations take the same form the equations are analogous

➢ The momentum and energy boundary layer equations are analogous if the pressure gradient is equal to

zero and Prandtl Number (Pr) ~𝟏

 2 U 2 Nu u  u u *
Cf = = = ;sin ce  =  = u  u  1   2u  
1
V 2 Re L Y y = 0 Re L y L y * u +v =  
2
y =0 y =0
x  y  Re  y 2 
Defining a new non dimensional number

   1   2 
Stanton number, St =
h
=
Nu
=
heat transferred u 
+v 
=  2 
U  C p Re Pr thermal capacity x y Re  y 
The Reynolds Analogy is defined as
Cf
Cf
= St , C f = Fanning Friction factor Re = Nu = Sh
2 2

The Reynolds Analogy implies that under certain conditions (no pressure gradient, Pr = 1) if the velocity
parameters are known then the heat transfer parameters can be determined (and vice versa).
➢ The advantage of the Reynolds analogy lies in that the ‘h’ may not be available for certain
geometries/situations however, for which ‘f’ value may be available as it is easier to perform momentum
Cf
transport experiments and then to calculate the ‘f’. Thus by using = St the ‘h’ may be found out
2
without involving into the exhaustive and difficult heat transfer experiments.

➢ The accuracy of Reynolds analogy depends on having Pr = 1 and dp/dx = 0.

➢ It has been shown that the analogy may be applied over a wide range of Pr, if certain corrections are added.
➢ The empirically modified Reynolds’ analogy is known as Chilton-Colburn analogy and is given

Modified Reynolds Analogy: Chilton-Colburn

Cf
= St Pr 3  jH 0.6  Pr  60
2

2
jH = Colburn j factor for heat transfer

For laminar flow the above equation is appropriate only when dp/dx≈ 0