1

5. DIMENSIONAL ANALYSIS
5.1 Dimensional Analysis
Since the theoretical analysis of a process is sometimes very difficult and mostly one has to depend on
experimental work, therefore it is necessary to carry dimensional analysis so as to list the parameters on which
the phenomenon depends. From the knowledge of the convective heat transfer, we can broadly classify the
parameters on which the heat transfer coefficient basically depends as follows:
(i) Geometry : Shape of the surface holding or carrying the fluid, say diameter (d), distance (x, L),
(ii) Flow condition, flow velocity (U) in case of forced and buoyant flow (βgΔT) in case of free convection.
(iii) Fluid properties: Properties involved in the process, like ρ, µ, Cp, k, etc.
In case of two phase heat transfer, latent heat or surface tension also come into picture.
The pertinent parameters influencing the phenomenon of convective heat transfer are listed in Table
5.1. The physical quantities along with their dimensional formulas are given in the Table 5.1.

Table 5.1 Primary dimensions of some parameters involved in the convective heat transfer
S .No. Parameter Symbol Unit Primary dimensions
1 Heat transfer coefficient h W/(m2K M t-3T-1
2 Significant Length/diameter L or d m L
3 Fluid density ρ kg/m3 ML-3
4 Fluid viscosity µ kg /m-s ML-1t-1
5 Fluid velocity U m/s Lt-1
6 Thermal Conductivity of fluid k W/mK MLt-3T-1
7 Specific heat of fluid Cp J/kgK L2t-2T-1
8 Temperature difference ∆T o
C T
9 Coefficient of volume expansion β 1/K T-1
10 Acceleration due to gravity g m/s2 Lt-2
11 Buoyant flow βg∆T m/s 2
Lt-2
12 Latent heat hfg J/kg L2t-2

We shall now illustrate the application of Buckingham’s theorem to the cases of forced and free convection
heat transfer and to the process of condensation. The flow of fluid, in any case may due to some forced velocity
or due to buoyant flow inside or across a tube, over a vertical or horizontal plate. Buckingham’s theorem
states that if there are n number variables affecting the physical phenomenon and they contain 4 fundamental
dimensions (L, M, t and T) then there will (n-4) independent dimensionless groups to correlate experimental
data.

Now let us form the ‘core group’ of four variables that are common to both the forced convection and the free
convection. They are d, ρ, µ and k, which will be considered as the repeating variables. The remaining variables
are,
(a) h ; this is to be correlated,
(b) U or (βgΔT); these are the flow conditions, specific to the forced and free convection, respectively and
(c) Cp ; this seems to be an appropriate property that defines heat.
Then, the different π terms are obtained by combining the core group with each one of the remaining (7-4)
variables.
 2

Forced convection

We shall now consider the velocity term, U in the forced flow and define the three dimensionless  groups as:
1  d a bµc k d h (5.1)
2  d mn  p k q U (5.2)
3  d w x  y k z Cp (5.3)
The exponents of the physical quantities in π-terms are chosen so as to make the π – terms dimensionless. So we
start with π1 and write the dimensional formula of each quantity and apply the requirement of dimensional
homogeneity.

For π1: Considering, 1  d  µ k h and writing the dimensions as:

a b c d

M0 L0 .t 0 .T0  1   L ML3  ML1t 1  MLt 3T 1  Mt 3T 1 

a b c d

Equating the terms of [M, L, t and T] on either side, for dimensional homogeneity as follows:
Exponents of M : 0  b  c  d 1
Exponents of L : 0  a  3b  c  d
Exponents of t : 0  c  3d  3
Exponents of T : 0  d 1
Solving the above equations, we get: a  1, b  0, c  0 and d  1
hd
Therefore, 1  d10µ0 k 1h or 1   Nu (This is Nusselt number) (5.4)
k

For π2: Considering  2  d m n  p k q U and writing the dimensions, we have

M0 L0 .t 0 .T 0  1   L ML3  ML1t 1  MLt 3T 1  Lt 1 
m n p q

Equating the terms of M, L, t and T on either side, for dimensional homogeneity:

Exponents of M : 0 npq
Exponents of L : 0  m  3n  p  q  1
Exponents of t : 0  p  3q  1
Exponents of T : 0  q
Solving the above equations, we get: m  1, n  1, p  1 and q  0
Ud
Therefore, 2  d111k0 U or 2  =Re (This is Reynolds number) (5.5)


For π3: Considering 3  d   k Cp and writing the dimensions,

w x y z

M0 L0 .t 0 .T0  1   L ML3  ML1t 1  MLt 3T 1  L2 t 2T 1 

w x y z

Equating the terms of M, L, t and T on either side, for dimensional homogeneity:

Exponents of M : 0 xyz
Exponents of L : 0  w  3x  y  z  2
Exponents of t : 0   y  3z  2
Exponents of T : 0  z 1
Solving the above equations, we get: w  0, x  0, y  1 and z  1
Cp
Therefore, 3  d 001k 1Cp or 3   Pr (This is Prandtl number) (5.6)
k
 3

Then, according to the Buckingham π theorem,

3  F  1 , 2  (5.7)

Or, Nu  C  Rem Pr n  (5.8)

Where C, m and n are constants evaluated experimentally.

Above relation is the desired relation among the various physical quantities affecting forced convection across a
tube, expressed in terms of dimensionless numbers Nu, Re and Pr.
Note that if, instead, we choose (U, µ, ρ, Cp) as the core group then the dimensionless terms obtained will be

Ud µCp h
Re  , Pr  , and St   Stanton number
µ k UCp
The another way of expressing heat transfer correlation is therefore,

St  F  RePr  (5.9)

Natural Convection

In natural convection heat transfer, flow is induced by the density differences caused as a result of temperature
differences. In the gravitational field, the density differences induce a buoyancy force (βg∆T). We shall now
consider the buoyancy forces term (βg∆T) as a single variable and define the three dimensionless  groups as:

1  d a bµc k d h (5.10)
2  d   k gT 
m n p q
(5.11)
3  d   k Cp
w x y z
(5.12)

Again, the exponents of the quantities in π-terms are chosen so as to make the π – terms dimensionless. We start
with π1 and write the dimensional formulas of each quantity and apply the requirement of dimensional
homogeneity.

For π1: Considering, 1  d  µ k h and writing the dimensions, we have

a b c d

M0 L0 .t 0 .T0  1   L ML3  ML1t 1  MLt 3T 1  Mt 3T 1 

a b c d

Equating the terms of M, L, t and Ton either side, for dimensional homogeneity:
Exponents of M : 0  b  c  d 1
Exponents of L : 0  a  3b  c  d
Exponents of t : 0  c  3d  3
Exponents of T : 0  d 1
Solving the above equations, we get: a  1, b  0, c  0 and d  1
hd
Therefore, 1  d10µ0 k 1h or 1   Nu (This is Nusselt number) (5.13)
k

For π2: Considering, 2  d mn p k q gT  and writing the dimensions, we get
M0 L0 .t 0 .T 0  1   L ML3  ML1t 1  MLt 3T 1  Lt 2 
m n p q

Equating the terms of M, L, t and T on either side, for dimensional homogeneity:

Exponents of M : 0 npq
Exponents of L : 0  m  3n  p  q  1
Exponents of t : 0  p  3q  2
 4

Exponents of T : 0  q
Solving the above equations, we get: m  3, n  2, p  2 and q  0
Therefore, 2  d322 k 0 gT 
2gTd3
Or 2  =Gr (This is Grashof number) (5.14)
2
For π3: Considering, 3  d w x  y k z Cp and writing the dimensions,

M0 L0 .t 0 .T0  1   L ML3  ML1t 1  MLt 3T 1  L2 t 2T 1 

w x y z

Equating the terms of M, L, t and T on either side, for dimensional homogeneity:

Exponents of M : 0 xyz
Exponents of L : 0  w  3x  y  z  2
Exponents of t : 0   y  3z  2
Exponents of T : 0  z 1
Solving the above equations, we get: w  0, x  0, y  1 and z  1
Therefore, 3  d 001k 1Cp
Cp
Or 3   Pr (This is Prandtl number) (5.15)
k
Then, according to the Buckingham π theorem,
3  F  1 , 2  (5.16)

Nu  C  Gr Pr 
m
Or (5.17)

Of course, the exact form of equation with associated constants must be determined from experiments.
As we shall see later, in most of empirical relations, product of Gr and Pr is taken together and the relations are
presented in the form:
Nu  CRa m (5.18)
Where
Ra   GrPr   Rayleigh number and ‘C’ and ‘m’ are constants determined from experiments.

Condensation

The condensation process involves condensate flow downward due to gravity, equal volume of the
displaced vapour, conduction and viscous work at the wall surface. Let the heat transfer coefficient of the
condensate film forming on the wall surface involve variables such that h=f (d, ρ, µ, k, g, β, hfg)
Let us form the ‘core group’ of four variables d, ρ, µ and k, which will be considered as the repeating variables.
The remaining variables are h, g, β and h fg.

Then, the different π terms are obtained by combining the core group with each one of the remaining (8-4=4)
variables:
1  d a bµc k d h (5.19)
2  d mn  p k q g (5.20)
3  d   k 
w x y z
(5.21)
4  d   k h fg
i j r s
(5.22)

The exponents of physical parameters in π-terms are chosen so as to make the π – terms dimensionless. Again
we start with π1, write the dimensional formulas of each quantity and apply the requirement of dimensional
homogeneity:
 5

For π1: Considering, 1  d  µ k h and writing the dimensions as:

a b c d

M0 L0 .t 0 .T0  1   L ML3  ML1t 1  MLt 3T 1  Mt 3T 1 

a b c d

Equating the terms of M, L, t and Ton either side, for dimensional homogeneity:
Exponents of M : 0  b  c  d 1
Exponents of L : 0  a  3b  c  d
Exponents of t : 0  c  3d  3
Exponents of T : 0  d 1
Solving the above equations, we get: a  1, b  0, c  0 and d  1
hd
Therefore, 1  d10µ0 k 1h or 1   Nu (This is Nusselt number) (5.23)
k

For π2: Considering,  2  d m  n  p k q g and writing the dimensions,

M0 L0 .t 0 .T 0  1   L ML3  ML1t 1  MLt 3T 1  Lt 2 
m n p q

Equating the terms of M, L, t and T on either side, for dimensional homogeneity:

Exponents of M : 0 npq
Exponents of L : 0  m  3n  p  q  1
Exponents of t : 0  p  3q  2
Exponents of T : 0  q
Solving the above equations, we get: m  3, n  2, p  2 and q  0
2 gd 3
Therefore, 2  d322k0g Or 2  (5.24)
2
For π3: Considering, 3  d w  x  y k z and writing the dimensions, we get

M0 L0 .t 0 .T 0  1   L ML3  ML1t 1  MLt 3T 1  T 1 

w x y z

Equating the terms of M, L, t and T on either side, for dimensional homogeneity:

Exponents of M : 0 xyz
Exponents of L : 0  w  3x  y  z
Exponents of t : 0   y  3z
Exponents of T : 0  z 1
Solving the above equations, we get: w  2, x  2, y  3 and z  1
 3
Therefore, 3  d 223 k 1 Or 3  (5.25)
d 22 k
For π4: Considering, 4  di j r k s h fg and writing the dimensions, we have

M0 L0 .t 0 .T0  1   L ML3   ML1t 1   MLt 3T 1  L2 t 2 

i j r s

Equating the terms of M, L, t and T on either side, for dimensional homogeneity:

Exponents of M : 0  j r  s
Exponents of L : 0  i  3j  r  s  2
Exponents of t : 0  r  3s  2
Exponents of T : 0  s
Solving the above equations, we get: i  2, j  2, r  2 and s  0
d 22 h fg
Therefore, 4  d 22 2 k 0 h fg  (5.26)
2
Then, according to the Buckingham π theorem,
 6

1  F  2 , 3 , 4  (5.27)
m
 2gd3   3   d 22 h fg  
m
 2 gd3     
Or, Nu  C     
 
   C       h   (5.28)
    d  k        k 
2 2 2 2 2 fg
  
m
 2gd3T   Cp   h fg    Gr Pr 
m

Or, Nu  C       Or Nu  C   (5.29)
 2  k 
  p  
C T  Ja 
 2 gd3T   Cp 
where, Grashof number, Gr=   , Prandtl number, Pr=  
 2   k 
 Cp T 
and Jacob number, Ja= 
 h fg 
 
The correlations shown can be explained by the principle of similarity which states that the behaviour
of two systems will be similar if the ratios of their linear dimensions, velocities, forces, etc. are the same. If the
ratio of inertia forces to viscous forces, that is Reynolds number, is same in the two fluids, the velocity fields
will be similar in geometrically similar systems. Thus, similar flow conditions will prevail in two systems for a
given value of Reynolds number. In the same way, the values of Nusselt number in geometrically similar
systems during forced convection will be the same for the given values of Reynolds number and Prandtl
number. And, during free convection, the Nusselt number will be the same for the given values of Grashof
number and Prandtl number in geometrically similar systems.

Example 5.1 Air at 2 atmospheric pressure and 475K is heated as it flows at a velocity of 8m/s through a tube
having 30 mm diameter. Calculate the heat transfer per unit length of the tube if a constant heat flux condition is
maintained at the wall so that the wall temperature is 20oC above the air temperature all along the length of the
tube. Also, estimate increase in the bulk temperature after 5m length of the tube.

Solution: Properties of air at the bulk temperature of 475K:

2  1.0132 105
p   1.486kg / m3
RT 287  475
5
  2.57 10 kg/m-s, k = 0.0386 W/mK, Cp = 1.025 kJ/kgK and Pr=0.7
Ud 1.486  8 0.03
Red    13,881 (flow is turbulent)  2300
 2.57 105

L 5 
Since    166  , using Dittus Boelter equation:
 d 0.03 
 7

 L 
Nu d  0.023Re0.8
d Pr
0.4
,  for 0.7  Pr  160, Re d  10000 and  60 
 d 
k 0.0386
Nud  0.02313881  0.7  40.1  52.86 W/m 2K
0.8 0.4
 40.1 or h  Nu d 
d 0.03

Heat flow per unit length, for  Tw  Tb   25 C

o

q
 hd  Tw  Tb    52.86  0.03  25   124.56 W
L m

Making energy balance on the element x in the tube shown in Fig. Ex. 5.1.

q
Tb
Tb1 Tb2

. x Tb .
m m
Tw

x
dq
L
Fig. Ex 5.1 Heat supplied at the tube wall and then extracted
by the flowing fluid

. .
dq  m C p Tb  hd  Tw  Tb  x or q  m Cp  Tb2  Tb1   hdL  Tw  Tb 

.
q m Cp
or,   Tb2  Tb1   hd  Tw  Tb   124.56 ;
L L

.  2
Therefore, m  AU   d U  8.4 x 103 kg/s
4

124.56  L 124.56  5  72.33 o C

 Tb2  Tb1   
8.4 x 103 1025
or, .
m Cp

Tb1  Tb2
But, Tb  or  Tb2  Tb1   2  475  950K
2

(950  72.33)
Thus, we have Tb2   511.16o C and hence, Tb1  438.83o C
2
 8

Example 5.2 A one kW electric plate heater 0.5 x 0.5 m in size is placed in an atmospheric air stream flowing
at a velocity of 6m/s and 300 K temperature. If the heater produces a constant heat flux at the surface facing the
atmosphere, calculate the average temperature difference along the plate and the temperature difference at the
trailing edge.
Solution: Since the wall temperature, Tw is not known, therefore the properties are taken at T  300K .
6
Properties of air at 300 K are as follows:   16.84  10 m / s , Pr  0.708 , k  0.02624 W/mK.
2

Thus, Re L  6  0.5 / 16.84  106  1.78  105  5  105 (Flow is laminar) and

1000
q"   4000 W/m 2 .
0.5  0.5

For constant heat flux on the flat plate wall, the heat transfer relation is,

hx q
 0.453Re x 2 Pr 3 . Since, the heat flux is given as: q"   h  Tw  T  ,
1 1
Nu x 
k A

q" x 1 1
therefore, Nu x   0.453Re x 2 Pr 3
k  Tw  T 

Under the conditions of constant heat flux at the plate, the wall temperature T w will vary along the flow
direction.

Thus, the average temperature difference for such conditions can be obtained as:

1  
L L L 1
1 1 q"x q" x 2
Tw  T    Tw  T  dx  
L 0  0.453k  U   12 Pr 13 
dx  dx
L0 L 0 k  0.453 Re 12 Pr 13
x  

3
1  2  q" L2 q "L
Tw  T    
L  3  0.453k  U   1 2 Pr 13 0.6795k Re 1 2 Pr 13
L

q" x q"L
But,  Tw  T    .
k  0.453 Re x 2 Pr k  0.453 Re L2 Pr
1 1 1 1
3 3
x L

Tw  T 0.453 2
Therefore,  
Tw  T x  L 0.6795 3
Substituting the respective values, we have

4000  0.5
 Tw  T    298.26K
 0.02624  0.6795  1.78 105  0.708 1
3
 9

And,  Tw  T  
3
2
 3

Tw  T   298.26  447.4K. Thus, Tw  447.4  300  747.4K
2

Now the calculations are repeated by taking the properties of air at the film temperatures, Tf   Tw  T  / 2

6
Properties of air at Tf = 523.7 K:   40.95  10 m /s , Pr  0.687 , k  0.04189 W / mK
2

6  0.5
Re L   0.7326 105 , Tw  T  295K and  TW  T   1.5  295K  442K
40.95  106

Example 5.3 Water enters a tube of 30 mm diameter at a mean flow velocity of 3 cm/s and temperature 500C.
Calculate the exit temperature of water if the tube is 2 m long and its wall is maintained at a constant
temperature of 90oC.
Solution: Since the bulk temperature of water only at the inlet is given, therefore the properties are taken at the

inlet bulk temperature, Tb1  50 o C which are as follows:

  985kg / m3 ,   5.5 104 kg/m-s , Cp  4.175kJ / kgo C, k  0.645W / mo C , Pr  3.4

Ud  985   3.0 10  30 10   1611.82 (Laminar flow) (Laminar flow)
2 3

ReD  
 5.5 104

1 0.14
 d 3  
Nu d  1.86  Red Pr     8.79 (at 90o C, w  3.0 104 kg / ms)
 L   w 

kNu d
Thus, h  189W / m 2o C
d

  T  Tb2   . .  2
   m C p  Tb2  Tb1  and m   d U  20.88  10 kg / s
3
Now, q  hdL Tw   b1
  2  4

.
  T  T   m Cp
or Tw   b1 b2     Tb2  Tb1  .
  2   hdL

Substituting the respective values, we have

 50  Tb2    20.88x10   4175 

3

90      T  50 
  2   189    0.03 2  b2
  T 
or 90  25   b2    2.447  Tb2  50  or Tb2  63.57 o C
  2 
 10

Tb1  Tb2 50  63.57

Thus, Tb    56.785o C and
2 2
q  2 189   30 103 [90  56.785]  1183.3W

Now, the calculation is repeated by taking properties at Tb=56.785oC:

  984kg / m3 , C p  4.179kJ / kgK ,

  4.9 104 kg / m  s, k =0.651 W / m oC and Pr  3.15 . Thus, Red  1807.35 ,

Nud  8.76, h  190.1 W/m2o C and Tb2  63.65o Cand

.
q  1189.92 W and m  20.86x10 3 kg/s . These results being close to those obtained previously may be
treated as correct.

Example 5.4 The gate of a baking oven in a bakery is made of a vertical metallic plate 1m high and 0.5 wide. It
receives a net radiant energy flux of 1 kW/m 2 from inside the oven while its other side, which is painted black,
is exposed to air at 25oC in a large room. Assuming that all the radiation received by the plate is lost by free
convection to the room air, estimate the average temperature that the plate will attained.

Solution:

This problem may be treated as a surface with constant heat flux.

To evaluate properties of the air, let us determine an approximate value of the film temperature. As a 1 st
approximation, let use the simplified equations for free convection in air.

If the free convection is turbulent then the heat transfer coefficient is given by, h  0.95T 
1
3

Also, the heat flux cane written as: q  1000  hT  0.95  T 
4
" 3

This gives, T  184.8 C and h  5.4 W/m C

0 2o

 T 
0.25

Now if the free convection is laminar then, h  1.42   hence,

 L 

0.25
1
q ''  1000  1.42    T 
1.25
, then T  189.75 C and h  5.27 W/m C since, L= 1m.
o 2o

L

Let us consider the turbulent convection for which, T  184.8 C , Where, T   Tw  T 

o

With T  25o C, the wall temperature will be Tw  209.8o C

 11

Tw  T
Thus, Tf   117.4o C(390.5K)
2

Taking properties at 390 K,   2.48  10 m /s ,   1

5
390.5 k  0.0328 W / m C
2 o
, and Pr = 0.693

gq"x 4
Using the equation for constant wall heat flux, Nu x  0.17  Grx Pr  , where  Gr Pr   Grx 
1
4
k 2

Nu x k
For x=L=1m, Nu x  163.85 and hence, h x   5.37 W/m 2 K
x

q" 1000
Re-calculating, T    186.20 C
h 5.37

These values of h and T are quite close to the values obtained from the turbulent flow equation,

Example 5.5 Calculate the rate of heat transfer from a large vertical plate 5m high and 10 m wide, which is
maintained at 84oC and exposed to the atmospheric air at 20oC.

84  20
Solution: Properties of air at the film temperature, Tf   52o C  325K
2

1
 , k  0.02813W / m o C,   18.8  106 m 2 /s and Pr  0.7025
325

 9.81  84  20  5
3

Gr Pr   0.7025  4.799 x 1011  turbulent flow 

18.8 10  6 2

h  0.95T 
1
Using, 3 = 3.8 W/m oC

Nu  0.021 Gr.Pr   for 10  GrPr  103  ;

2
Using 5 9
Nu  987.87

h
 987.87    0.02813  5.557 W/m 2 o C
5

q  hA  Tw  T   5.557 x  5 x 1  84 – 20   1778.5 W

Example 5.6 Determine the heat loss from a hot square plate of size 0.6m x 0.6m which is maintained at a
constant temperature of 95 oC and is exposed to an atmospheric air at 25oC. Estimate the percentage decrease in
the heat loss when the plate position is changed from vertical to the horizontal.
 12

95  25
Solution: Properties of air at the film temperature Tf   60o C  333.15K are given below:
2

1
 ,   18.97x106 m2 /s, k=0.02894 W/mo C, Cp  1007 J/kgK and
333.15

6
  2.05x10 kg/m-s and   1.06 kg / m , Pr= 3
Cp

 2.05x10  1007   0.07133
6

k 0.02894

For vertical position of the plate,

gTL3  9.81 0.003 70  0.6 

3

Ra L  GrL Pr   0.07133  8.82x107

 18.97x10 
2 6 2

Using,

Nu L 
hL
 0.13  GrL Pr  3  0.13 8.82x107  3 =57.86, h=
1 1  57.86  0.02894   2.79 W/m2 K
k 0.6

Heat loss from both surfaces of the vertical plate,

q  h  2A  Tw  T    2.79  2x0.6 x 0.6 95-25  140.655 W

For horizontal position of the plate, characteristic length, L=(A/P)=(0.6x0.6)/(0.6x4)=0.15

gTL3  9.81 0.003 70  0.15 

3

Ra L  GrL Pr   0.07133  1378125

2 18.97x10 
6 2

 0.54  Gr Pr  4 for  2x104  Ra L  8x106 

hL 1
For horizontal plane (heated upper surface): Nu L 
k
Therefore for the upper surface:

Nu L 
hL
 0.54 1378125 4 =18.5, h u =
1 18.5 0.02894   3.57 W/m2 K
k 0.15

 0.27  Gr Pr  4 for 105  Ra L 1011 

hL 1
For horizontal plane (heated lower surface): Nu L 
k

Therefore for the lower surface:

Nu L 
hL
 0.27 1378125 4 =9.25, h l =
1  9.25 0.02894   1.78 W/m2 K
k 0.15
Heat loss from top and bottom surfaces of the horizontal plate,

q   h u  h l  A  Tw  T    3.57  1.78 0.6 x 0.6 95-25   134.94 W

140.655  134.94
Percentage decrease in heat transfer  x100  4.06%
140.655

Questions for discussion

 13

1. Explain the importance of dimensional Analysis.

2. What is prototype system?
3. Explain the similarity criteria required for two systems to be identical.
4. Define the dimensionless number which identifies the free, forced and mixed convection.
5. What is the difference between analytical and empirical relations?
6. What are the dominating parameters in the free and forced convections?

Unsolved Problems

1. Air flow through a long air conditioning duct, rectangular shape (0.3 m height x 0.6 m width) runs
horizontally through a room in which the air temperature is at 15 oC. If the duct is un-insulated and its
outer surface temperature is maintained at 15 oC, calculate the heat gained by the duct per meter length.
Following relations may be used.

For vertical position of the plate, Nu L  0.13  GrL Pr 

1
3

For horizontal plane (heated upper surface): Nu L  0.54  Gr Pr  for  2x104  Ra L  8x106 
1
4

For horizontal plane (heated lower surface): Nu L  0.27  Gr Pr  for 105  Ra L 1011 
1
4

2. Liquid metal while flowing at a rate of 2.0 kg/s through an electrically heated tube of 25mm diameter
gets heated from 120oC to 150oC. If a constant heat flux is maintained on its wall surface calculate the
minimum length of the tube if its wall temperature is not to exceed 180 oC. Properties of sodium at 135 oC
are:

 916kg / m3 ,   0.594x106 m 2 / s, Cp 1.3565 kJ / kg oC, k  84.9 W / moC and Pr  0.0087

The relation for Nusselt number that can be used is, Nu  4.82  0.0185Re Pr0.827

3. Water flowing at an average velocity of 2m/s through a long tube with its wall maintained at a constant
temperature of 95oC gets heated from 15oC to 60oC. If the tube is of 20 mm diameter find the average
heat transfer coefficient and length of the tube. The well known equations which can be used are:
Dittus-Boelter equation: Nud  0.023 Red0.8 Pr0.4 and

Sieder and Tate equation: Nu d  0.027 Red0.8 Pr 0.33  b / w 

0.14

4. A tube placed at the focal line of parabolic solar collector heats water from 20˚C to 90˚C when it flows at
the rate of 0.01 kg/s. Assuming a uniform heat flux of 1500 W/m2 on the tube wall surface, determine
length of the tube if its diameter is 50mm. Also estimate the surface temperature at outlet of the tube.
Properties of water at the mean bulk temperature of 55oC are:
 985.5 kg / m3 , Cp  4.18kJ / kg K,  =0.517x106 m2 / s, k  0.654 W / mK, Pr  3.26
 14

5. Water flowing at a velocity of 1.2 m/s through a 15 mm x 30 mm rectangular cross section tube gets
heated from 30oC to 70oC. If the tube walls are maintained at 85 oC, determine length of the tube. The
properties of water at 50oC are,

  985kg / m3 ,   5.5 104 kg/m-s , C  4.175kJ / kg o C , k  0.645W / mo C , Pr  3.4

p

Questions of multiple choice

1. The dimensionless heat transfer Stanton Number, St is defined as:

a.  Re Pr  Nu
b. Nu  Re Pr  Ans.
c. Gr  Re Nu 
d. GrNu  Re

2. The dimensionless heat transfer, Peclet Number, Pe is defined as:

a. Re Pr Ans.
b. Re Gr
c. Gr Pr
d. NuGr

3. For film condensation, the non dimensional Nusselt number is related as:
Nu  C  Gr Pr/ Re  4
1
a.
Nu  C  Re Pr/ Ja 
1
b. 4

Nu  C  Gr Pr/ Ja 
1
c. 4 Ans.
Nu  C  Gr Pr/ Ja 
1
d. 3

4. The dimensionless heat transfer Graetz Number, Gz is defined as:

a.
 
 Re Pr d 
L   Ans.

b.
 
Gr Pr d 
L  
c.
 
 Nu Pr d 
L  
d.
 
 Re G r d 
L  

5. The dimensionless heat transfer, Rayleigh Number, Ra is defined as:

a. Re Pr
b. Re Gr
c. Gr Pr Ans.
d. NuGr

6. The dimensionless Richardson Number, Ri is defined as:

a. Gr / Re 2 Ans.
b. Re/ Gr 2
c. Gr Pr/ Re 2
 15

d. Nu 2 / Gr

7. For a given situation, heat flowing through a solid body due to conduction is dissipating to the
surrounding fluid by convection. For this situation, if the expression for Biot number and Nusselt
number is same given by (hL/k), which of the following statement is correct
a. h and k, both are for the surrounding fluid
b. h and k, both are for the solid body
c. h is for the fluid and k is for the solid
d. In Biot number, k is for the solid while in Nusselt number, k is for the fluid Ans