January Semester 2023-2024 1

THERMODYNAMICS B (B59TC)
Part 2: Convection
January Semester 2023-2024 2

Convection
January Semester 2023-2024 3

Introduction to Convection
• Is a heat transfer mechanism that occurs in a fluid
due to motion.
• The fluid pocket adjacent to the hot wall is heated
and replaced by a cold pocket.
• The movement of the fluid governed by the fluid
properties which are temperature dependent.
January Semester 2023-2024 4

Introduction to Convection
• Convection occurs in liquids and gases.
• Fluid at a high temperature physically moves from
one region to another, while cooler fluid replaces it.
• Energy is thus transferred from one region to
another.
• We are primarily concerned with convection from a
solid surface to the bulk of the fluid, or from a fluid
to a solid wall.
January Semester 2023-2024 5

Introduction to Convection
• The study of convection is subdivided: if the
movement of the fluid is induced by a blower, fan
or pump then it is known as forced convection.
• Natural (free) convection occurs when the fluid
motion is caused by buoyancy forces that are
induced by density differences due to the
variation of temperature in the fluid.

• Boiling and condensation

are special cases of convection.
 January Semester 2023-2024 6

Newton’s law of cooling

h convection heat transfer coefficient, W/m2°C (W/m2K)

As the surface area through which heat transfer takes place
Ts the surface temperature
T the temperature of the fluid sufficiently far from the surface

The convection heat transfer coefficient h or α is not

a property of the fluid.
January Semester 2023-2024 7

Newton’s law of cooling

h or α is an experimentally determined parameter

whose value depends on all the variables influencing
convection such as:
▪ the surface geometry
▪ the nature of fluid motion
▪ the properties of the fluid
▪ the bulk fluid velocity

NOT a property of the fluid

 January Semester 2023-2024 8

Consider a layer with convection between the

surfaces and a fluid on either side:

T1 and T4 are the fluid

temperatures on either
side of the wall

Methods of determining
the convective or
surface heat transfer
coefficients, h, will be
discussed later
 January Semester 2023-2024 9

Consider a layer with convection between the

surfaces and a fluid on either side:
 January Semester 2023-2024 10

Consider a layer with convection between the

surfaces and a fluid on either side:

Adding these equations gives:

 January Semester 2023-2024 11

Consider a layer with convection between the

surfaces and a fluid on either side:

Where U is the overall heat transfer coefficient between the

fluids on either side of the wall.
January Semester 2023-2024 12

For the general case of multiple layers, thickness xi

conductivity ki and convection with coefficient ha and hb
on each side of a wall with fluids at Ta and Tb:
January Semester 2023-2024 13

The resistance approach

• For our two layers considered earlier

 xa xb 
T1 − T3 = Q  +  = Q ( Ra + Rb ) = Q ( Roverall )
 k a kb 
January Semester 2023-2024 14

The resistance approach

• For a layer separating two fluids considered
earlier
January Semester 2023-2024 15

The resistance approach

• In general, for multiple layers:

• For convective boundary layers:

• For conductive layers:

January Semester 2023-2024 16

The resistance approach

• For convective boundaries:

• For conductive layers:

January Semester 2023-2024 17

Notes
• We can apply the analysis discussed over any
number of layers.

• The U or R must be calculated for the relevant layers

using the appropriate temperatures (e.g. surface-
surface, fluid-fluid).

• Some terms may be negligible - if the resistances are

very different it may be possible to neglect the lowest.
• e.g. Thin metal sheet supporting thermal insulation.
• e.g. Condensing heat transfer coefficient on
insulated surface
January Semester 2023-2024 18

Notes
• If thermal resistance is negligible there is no
temperature difference across that boundary or layer.

• By selecting appropriate layers intermediate T can be

calculated.

• In general it is up to you whether you work in terms of

U or R – the conclusions should be the same!!
January Semester 2023-2024 19

Boundary Layer Concept

• When a flowing fluid comes into contact with the
leading edge of a thin plate, a boundary layer
develops between the fluid and the plate. It is
disturbed by the viscous and heat transfer effects at
the surface.

Development of the boundary thermal boundary layer

layer over a flat plate
January Semester 2023-2024 20

Boundary Layer Concept

• If the fluid particles flow parallel to one another (low
velocities), the flow is called laminar flow.

• If the particles flow at high velocities, the flow pattern

becomes unpredictable. This is called turbulent flow.
January Semester 2023-2024 21

Boundary Layer Concept

• For flow over a flat plate, we may determine whether the
flow in the boundary layer is likely to be laminar or
turbulent by applying the following conditions:

 V  x 
Re x  =   10
5
Laminar
  

 V x 
Re x  =   10
6
Turbulent
  

where x is the distance from the leading edge of the plate.

January Semester 2023-2024 22

Thermal boundary layer

• A thermal boundary layer is the region which
facilitates the convective heat transfer.
• The convection heat transfer coefficient h or α then
describes the conductance of heat in the thermal
boundary layer in the equation
January Semester 2023-2024 23

Dimensional Analysis
• The fact that many factors are involved in convective
heat transfer suggests that we may use dimensional
analysis to derive dimensionless groups which assist
in presenting and using convective heat transfer data.

• Three dimensionless groups can be used to

represent forced convective heat transfer data.
January Semester 2023-2024 24

Nusselt Number, Nu
• The dimensionless Nusselt number Nu, is the ratio of
convection to conduction heat transfer in a fluid.

• h - heat transfer coefficient W/m2K

• L - characteristic dimension m

• k - thermal conductivity W/mK

January Semester 2023-2024 25

Prandtl Number, Pr
• is often described as the ratio of viscous (momentum
diffusion) to thermal diffusion.

• α - thermal diffusivity m2/s (Velocity boundary

layer vs Thermal
• k - thermal conductivity W/mK boundary layer)

• μ - dynamic viscosity kg/ms

• ν - μ / ρ - kinematic viscosity m2/s

• cp – specific heat capacity J/kgK

January Semester 2023-2024 26

Reynolds number, Re
• is the ratio of inertia force to the viscous force.

• 𝛒 the density of the fluid (kg/m3),

• u the mean fluid velocity (m/s),
• D Characteristic dimension (m),
• μ dynamic viscosity of the fluid (Pa.s), (kg/ms)
• 𝛎 is the kinematic velocity of the fluid m2/s
January Semester 2023-2024 27

Convective heat transfer correlations

• Empirical correlations are based upon
experimental observations.
• The form of the correlations may have some
theoretical or conceptual justification.
• Their accuracy relies upon the reliability of the
experimental observations and the similarity of
the experimental conditions to those to which the
correlation is to be applied.
January Semester 2023-2024 28

Convective heat transfer correlations

• It is obvious that correlations are geometry
dependent (e.g. applying to flow inside tubes,
crossflow outside tubes or over flat plates)
• There are other, less obvious transitions, the
most important being between laminar and
turbulent flow and the occurrence or otherwise of
a phase change.
• From the designer’s or engineer’s point of view it
is essential that an appropriate correlation is
applied.
January Semester 2023-2024 29

Convective heat transfer correlations

• Use of a correlation outside the range of
conditions for which it has been experimentally
verified is very dangerous.
• You are not expected to learn correlations – but
you are expected to understand their use and
their limitations.
January Semester 2023-2024 30

Reynolds Analogy
• This is a highly simplified model which serves to
demonstrate the form which we would expect for
heat transfer correlations .
• Consider a turbulent flow past a wall.
• Assume that in unit time, over an area A, a mass
m, of fluid moves from the free stream to the wall
and the same mass moves away from the wall.
Free stream at V, Tf
m

Wall at Tw
January Semester 2023-2024 31

Reynolds Analogy
• Assuming that the fluid at the wall is stationary
and reaches thermal equilibrium with the wall we
can say:
force = rate of change of momentum

Free stream at V, Tf
m

Wall at Tw
January Semester 2023-2024 32

Reynolds Analogy
• Transfer of momentum from the fluid to the wall in
flow direction = m(V − 0)
• Transfer of heat from the wall to the fluid =
(
− mcp Tf − Tw )
• shear stress on
Free stream at V, Tf
the wall is equal to m

the force exerted

by the fluid in the
flow direction per
Wall at Tw
unit area
January Semester 2023-2024 33

Reynolds Analogy

• Rearranging and putting ( )

T = Tw − Tf gives:

• showing that the heat transfer coefficient and wall

shear stress are closely related
January Semester 2023-2024 34

Reynolds Analogy

• May be rearranged in dimensionless form to give:

• which may be expressed:

January Semester 2023-2024 35

Reynolds Analogy
• Finally, remembering the empirical Blasius
equation:

• This equation is not used in practice!!! But

demonstrates that:

• Is an appropriate form for turbulent convective

heat transfer correlations
January Semester 2023-2024 36

Reynolds Analogy

where m and n are constant exponents, and

constant C depends on geometry and flow.

• Note that Pr for air/gas does not vary much with

temperature and often correlation for gas does
not have a Pr term and a different constant than
liquid.
January Semester 2023-2024 37

Correlation for forced convection from

a Flat Plate
• Transition from Laminar to turbulent at Re ~ 5 x 105

• Laminar:
Re  5 105
Pr  0.6

• Turbulent:

5 105  Re  107
0.6  Pr  60
January Semester 2023-2024 38

Correlation for forced convection from a

Flat Plate
• These correlations give the average heat transfer
coefficient over the plate
• Fluid properties are evaluated at the mean film
temperature i.e. the average of the fluid and the
plate temperatures.
• A general equation for flat plate:
January Semester 2023-2024 39

Flow over cylinder (External flow)

• Flow across cylinders and spheres is frequently
found in many heat transfer.
• Applications:
❑ Shell-and-tube heat exchanger
❑ Pin fin heat sinks for electronic cooling
• The characteristic length for a cylinder and sphere is
taken to be the external diameter D.

Re cr  2 10 5
January Semester 2023-2024 40

Flow over cylinder (External flow)

• For air:

Note that Pr for air/gas dose not vary much with temperature and often
correlation for gas not have a Pr term and a different constant than
liquid.
January Semester 2023-2024 41

Flow over cylinder (External flow)

Nud = c Re d Pr m n

Red c m 1
1-4 0.989 0.33
n=0.333 =
3
4-40 0.911 0.385
40-4000 0.683 0.466 Note these values differ slightly
from those in Dr Houston’s
4000-40,000 0.193 0.618 printed notes – different workers
40,000-250,000 0.027 0.805 have developed slightly different
correlations

It is recommended that the use of these constants be restricted to

the range 0.5<Pr<10.
January Semester 2023-2024 42

Flow over cylinder (External flow)

January Semester 2023-2024 43

Flow inside pipes/tubes

Developing vs Developed region
• In the thermally fully developed region of a tube, the
local convection coefficient is constant (does not vary
with x)
• Both local friction factor fx and local convection
coefficient hx remain constant in hydrodynamically and
thermally fully developed regions, respectively.

• Velocity distribution and variation of local heat transfer

coefficient for turbulent flow occur near the entrance of
a uniformly heated tube.
January Semester 2023-2024 44

Flow inside pipes/tubes

January Semester 2023-2024 45

Correlation for Flow in Tubes and

Ducts

Laminar (Re<2300)
Developing Developed

Thermal Lt = 0.05 Re Pr d
Velocity Lh = 0.05 Re d
0.14
 Re Pr d    
1
3

Nu = 1.86    
 L   s 
January Semester 2023-2024 46

Correlation for Flow in Tubes and Ducts

The Dittus-Boelter
Equation

Developed
Turbulent (Re > 2300)
Developing
L  10d
L  10d 𝑁𝑢𝑑 = 0.023𝑅𝑒𝑑0.8 𝑃𝑟 𝑛

n = 0.4 for the fluid being heated (Ts>Tm)

Since most tubes and ducts n = 0.3 for the fluid being cooled (Ts < Tm)
used in practice are Properties evaluated at
significantly longer than 10d mean bulk temperature
Dittus-Boelter equation can be T +T
Tm = inlet outlet
used. 2
January Semester 2023-2024 47

Final Comments
• The simple correlations given here are accurate to within
20%
• More accurate but more complex correlations are
available in the literature
• Correlations are available for more complex geometries
• Tube banks
• Finned tubes
• Corrugated surfaces ...........
• Non-circular ducts use equivalent or hydraulic diameter
January Semester 2023-2024 48

Hydraulic Diameter Example

• Circular Tube or Pipe:

• Annulus Circular Cylinder Tube or Pipe:

• Rectangular Tube or Pipe:

January Semester 2023-2024 49

Summary: Helpful Tips

• For heat convection, instead of determining h, we
determine Nu. Nu correlation equations are different
for different conditions, and they are derived from
experiments.
Proper selection of correlations
• Identify the type of fluid flows (Internal or external?)
• Check the Re number. Is the flow laminar or
turbulent?
• Identify the classification of the heat transfer process.
Is it developing or developed region?
• Identify surface boundary condition. Is it uniform
temperature or uniform flux? Is it heating or cooling?
• Use a consistent set of units when doing calculations.
January Semester 2023-2024 50

Example 4
Engine oil at 60°C flows over the upper surface of
a 5-m-long flat plate whose temperature is 20°C
with a velocity of 2 m/s. Determine the rate of heat
transfer per unit width of the entire plate.
January Semester 2023-2024 51

Example 5
A long 10 cm diameter steam pipe whose external
surface temperature is 110°C passes through
some open area that is not protected against the
winds. Determine the rate of heat loss (1093 W)
from the pipe per unit length when the air is at 1
atm pressure and 10°C and the wind is blowing
across the pipe at a velocity of 8 m/s.

@Tf = 60°C
k = 0.02808 W/m°C
Pr = 0.7202
ν = 1.896 x 10-5 m2/s
January Semester 2023-2024 52

Example 6

(a) h = 262.3 W/m2K

(b) 22.8 °C
January Semester 2023-2024 53

Example 7
Water flows in a 10mm ID tube with an average velocity
of 1m/s. The temperature of the water at a particular
location is 60oC and an electric heater maintains the
heat flux at 200 kW/m2. Estimate the local heat transfer
coefficient (6719.26 W/m2K) and surface temperature
(89.77 °C) of the tube at a location where the water
temperature is 60oC. The flow may be assumed to be
fully developed at this location.
Water properties at 60 oC
• Viscosity 463 x 10-6 kg/ms
• Density 983 kg/m3
• Prandtl No 2.97
• Conductivity 653 x 10-3 W/mK
January Semester 2023-2024 54

Example 8
A commercial grade cubical freezer, 3m on a side, has
a composite wall consisting of an exterior sheet of
6.35-mm-thick plain carbon steel, an intermediate
layer of 100-mm-thick cork insulation, and an inner
sheet of 6.35-mm-thick aluminium alloy (2024).
Adhesive interfaces between insulation and the
metallic strips are each characterised by a thermal
′′
contact resistance of 𝑅𝑡,𝑐 = 2.5 × 10−4 m2K/W. What is
the steady-state cooling load that must be maintained
by the refrigerator under conditions for which the outer
and inner surface temperatures are 22 °C and -6 °C,
respectively (590 W)?
kal = 173 W/mK, kst = 64 W/mK, kins = 0.039 W/mK.