Chapter 6 Fundamentals of Convection

Chapter 6
FUNDAMENTALS OF CONVECTION
Physical Mechanisms of Forced Convection

6-1C In forced convection, the fluid is forced to flow over a surface or in a tube by external means such as
a pump or a fan. In natural convection, any fluid motion is caused by natural means such as the buoyancy
effect that manifests itself as the rise of the warmer fluid and the fall of the cooler fluid. The convection
caused by winds is natural convection for the earth, but it is forced convection for bodies subjected to the
winds since for the body it makes no difference whether the air motion is caused by a fan or by the winds.

6-2C If the fluid is forced to flow over a surface, it is called external forced convection. If it is forced to
flow in a tube, it is called internal forced convection. A heat transfer system can involve both internal and
external convection simultaneously. Example: A pipe transporting a fluid in a windy area.

6-3C The convection heat transfer coefficient will usually be higher in forced convection since heat
transfer coefficient depends on the fluid velocity, and forced convection involves higher fluid velocities.

6-4C The potato will normally cool faster by blowing warm air to it despite the smaller temperature
difference in this case since the fluid motion caused by blowing enhances the heat transfer coefficient
considerably.

6-5C Nusselt number is the dimensionless convection heat transfer coefficient, and it represents the
enhancement of heat transfer through a fluid layer as a result of convection relative to conduction across
hL
the same fluid layer. It is defined as Nu = where L is the characteristic length of the surface and k is
k
the thermal conductivity of the fluid.

6-6C Heat transfer through a fluid is conduction in the absence of bulk fluid motion, and convection in the
presence of it. The rate of heat transfer is higher in convection because of fluid motion. The value of the
convection heat transfer coefficient depends on the fluid motion as well as the fluid properties. Thermal
conductivity is a fluid property, and its value does not depend on the flow.

6-7C A fluid flow during which the density of the fluid remains nearly constant is called incompressible
flow. A fluid whose density is practically independent of pressure (such as a liquid) is called an
incompressible fluid. The flow of compressible fluid (such as air) is not necessarily compressible since the
density of a compressible fluid may still remain constant during flow.

6-1
 Chapter 6 Fundamentals of Convection

6-8 Heat transfer coefficients at different air velocities are given during air cooling of potatoes. The initial
rate of heat transfer from a potato and the temperature gradient at the potato surface are to be determined.
Assumptions 1 Steady operating conditions exist. 2 Potato is spherical in shape. 3 Convection heat transfer
coefficient is constant over the entire surface.
Properties The thermal conductivity of the potato is given to be k = 0.49 W/m.°C.
Analysis The initial rate of heat transfer from a potato is
As = πD 2 = π (0.10 m) 2 = 0.03142 m 2
Air
Q& = hAs (Ts − T∞ ) = (19.1 W/m 2 .°C)(0.03142 m 2 )(20 − 5)°C = 9.0 W V∞ = 1 m/s
T∞ = 5°C
where the heat transfer coefficient is obtained from the table at 1 m/s Potato
velocity. The initial value of the temperature gradient at the potato Ti = 20°C
surface is

 ∂T 
q& conv = q& cond = − k   = h(Ts − T∞ )
 ∂r  r = R
∂T h(Ts − T∞ ) (19.1 W/m 2 .°C)(20 − 5)°C
=− =− = −585 °C/m
∂r r=R k (0.49 W/m.°C)

6-9 The rate of heat loss from an average man walking in still air is to be determined at different walking
velocities.
Assumptions 1 Steady operating conditions exist. 2 Convection heat transfer coefficient is constant over
the entire surface.
Analysis The convection heat transfer coefficients and the rate of heat losses at different walking velocities
are

(a) h = 8.6V 0.53 = 8.6(0.5 m/s) 0.53 = 5.956 W/m 2 .°C

Ts = 30°C
Q& = hAs (Ts − T∞ ) = (5.956 W/m 2 .°C)(1.8 m 2 )(30 − 10)°C = 214.4 W Air
V∞
(b) h = 8.6V 0.53 = 8.6(1.0 m/s) 0.53 = 8.60 W/m 2 .°C T∞ = 10°C
Q& = hAs (Ts − T∞ ) = (8.60 W/m 2 .°C)(1.8 m 2 )(30 − 10)°C = 309.6 W

(c) h = 8.6V 0.53 = 8.6(1.5 m/s) 0.53 = 10.66 W/m 2 .°C

Q& = hAs (Ts − T∞ ) = (10.66 W/m 2 .°C)(1.8 m 2 )(30 − 10)°C = 383.8 W

(d) h = 8.6V 0.53 = 8.6(2.0 m/s) 0.53 = 12.42 W/m 2 .°C

Q& = hAs (Ts − T∞ ) = (12.42 W/m 2 .°C)(1.8 m 2 )(30 − 10)°C = 447.0 W

6-2
 Chapter 6 Fundamentals of Convection

6-10 The rate of heat loss from an average man walking in windy air is to be determined at different wind
velocities.
Assumptions 1 Steady operating conditions exist. 2 Convection heat transfer coefficient is constant over
the entire surface.
Analysis The convection heat transfer coefficients and the rate of heat losses at different wind velocities
are
Ts = 29°C
(a) h = 14.8V 0.53 = 14.8(0.5 m/s) 0.69 = 9.174 W/m 2 .°C Air
V∞
Q& = hAs (Ts − T∞ ) = (9.174 W/m 2 .°C)(1.7 m 2 )(29 − 10)°C = 296.3 W T∞ = 10°C
(b) h = 14.8V 0.53 = 14.8(1.0 m/s) 0.69 = 14.8 W/m 2 .°C

Q& = hAs (Ts − T∞ ) = (14.8 W/m 2 .°C)(1.7 m 2 )(29 − 10)°C = 478.0 W

(c) h = 14.8V 0.53 = 14.8(1.5 m/s) 0.69 = 19.58 W/m 2 .°C

Q& = hAs (Ts − T∞ ) = (19.58 W/m 2 .°C)(1.7 m 2 )(29 − 10)°C = 632.4 W

6-11 The expression for the heat transfer coefficient for air cooling of some fruits is given. The initial rate
of heat transfer from an orange, the temperature gradient at the orange surface, and the value of the Nusselt
number are to be determined.
Assumptions 1 Steady operating conditions exist. 2 Orange is spherical in shape. 3 Convection heat
transfer coefficient is constant over the entire surface. 4 Properties of water is used for orange.
Properties The thermal conductivity of the orange is given to be k = 0.50 W/m.°C. The thermal
conductivity and the kinematic viscosity of air at the film temperature of (Ts + T∞)/2 = (15+5)/2 = 10°C are
(Table A-15)
k = 0.02439 W/m.°C, υ = 1.426 × 10 -5 m 2 /s
Analysis (a) The Reynolds number, the heat transfer coefficient, and the initial rate of heat transfer from
an orange are

As = πD 2 = π (0.07 m) 2 = 0.01539 m 2
Air
V D (0.5 m/s)(0.07 m) V∞=0.5 m/s
Re = ∞ = = 2454 T∞ = 5°C
υ 1.426 × 10 −5 m 2 /s
Orange
5.05k air Re 1 / 3 5.05(0.02439 W/m.°C)(2454)1 / 3 Ti = 15°C
h= = = 23.73 W/m 2 .°C
D 0.07 m

Q& = hAs (Ts − T∞ ) = (23.73 W/m 2 .°C)(0.01539 m 2 )(15 − 5)°C = 3.65 W

(b) The temperature gradient at the orange surface is determined from

∂T 
q& conv = q& cond = −k   = h(Ts − T∞ )
 ∂r  r = R
∂T h(Ts − T∞ ) (23.73 W/m 2 .°C)(15 − 5)°C
=− =− = −475 °C/m
∂r r =R k (0.50 W/m.°C)

hD (23.73 W/m 2 .°C)(0.07 m)

(c) The Nusselt number is Re = = = 68.11
k 0.02439 W/m.°C

6-3
 Chapter 6 Fundamentals of Convection

Velocity and Thermal Boundary Layers

6-12C Viscosity is a measure of the “stickiness” or “resistance to deformation” of a fluid. It is due to the
internal frictional force that develops between different layers of fluids as they are forced to move relative
to each other. Viscosity is caused by the cohesive forces between the molecules in liquids, and by the
molecular collisions in gases. Liquids have higher dynamic viscosities than gases.

6-13C The fluids whose shear stress is proportional to the velocity gradient are called Newtonian fluids.
Most common fluids such as water, air, gasoline, and oils are Newtonian fluids.

6-14C A fluid in direct contact with a solid surface sticks to the surface and there is no slip. This is known
as the no-slip condition, and it is due to the viscosity of the fluid.

6-15C For the same cruising speed, the submarine will consume much less power in air than it does in
water because of the much lower viscosity of air relative to water.

6-16C (a) The dynamic viscosity of liquids decreases with temperature. (b) The dynamic viscosity of gases
increases with temperature.

6-17C The fluid viscosity is responsible for the development of the velocity boundary layer. For the
idealized inviscid fluids (fluids with zero viscosity), there will be no velocity boundary layer.

6-18C The Prandtl number Pr = ν / α is a measure of the relative magnitudes of the diffusivity of
momentum (and thus the development of the velocity boundary layer) and the diffusivity of heat (and thus
the development of the thermal boundary layer). The Pr is a fluid property, and thus its value is
independent of the type of flow and flow geometry. The Pr changes with temperature, but not pressure.

6-19C A thermal boundary layer will not develop in flow over a surface if both the fluid and the surface
are at the same temperature since there will be no heat transfer in that case.

Laminar and Turbulent Flows

6-20C A fluid motion is laminar when it involves smooth streamlines and highly ordered motion of
molecules, and turbulent when it involves velocity fluctuations and highly disordered motion. The heat
transfer coefficient is higher in turbulent flow.

6-21C Reynolds number is the ratio of the inertial forces to viscous forces, and it serves as a criteria for
determining the flow regime. For flow over a plate of length L it is defined as Re = VL/υ where V is flow
velocity and υ is the kinematic viscosity of the fluid.

6-22C The friction coefficient represents the resistance to fluid flow over a flat plate. It is proportional to
the drag force acting on the plate. The drag coefficient for a flat surface is equivalent to the mean friction
coefficient.

6-23C In turbulent flow, it is the turbulent eddies due to enhanced mixing that cause the friction factor to
be larger.

6-24C Turbulent viscosity µt is caused by turbulent eddies, and it accounts for momentum transport by
∂u
turbulent eddies. It is expressed as τ t = − ρ u ′v ′ = µ t where u is the mean value of velocity in the
∂y
flow direction and u ′ and u ′ are the fluctuating components of velocity.

6-4
 Chapter 6 Fundamentals of Convection

6-25C Turbulent thermal conductivity kt is caused by turbulent eddies, and it accounts for thermal energy
∂T
transport by turbulent eddies. It is expressed as q& t = ρC p v ′T ′ = − k t where T ′ is the eddy
∂y
temperature relative to the mean value, and q& t = ρC p v ′T ′ the rate of thermal energy transport by turbulent
eddies.

Convection Equations and Similarity Solutions

6-26C A curved surface can be treated as a flat surface if there is no flow separation and the curvature
effects are negligible.

∂u ∂v
6-27C The continuity equation for steady two-dimensional flow is expressed as + = 0 . When
∂x ∂y
multiplied by density, the first and the second terms represent net mass fluxes in the x and y directions,
respectively.

6-28C Steady simply means no change with time at a specified location (and thus ∂u / ∂t = 0 ), but the
value of a quantity may change from one location to another (and thus ∂u / ∂x and ∂u / ∂y may be
different from zero). Even in steady flow and thus constant mass flow rate, a fluid may accelerate. In the
case of a water nozzle, for example, the velocity of water remains constant at a specified point, but it
changes from inlet to the exit (water accelerates along the nozzle).

6-29C In a boundary layer during steady two-dimensional flow, the velocity component in the flow
direction is much larger than that in the normal direction, and thus u >> v, and ∂v / ∂x and ∂v / ∂y are
negligible. Also, u varies greatly with y in the normal direction from zero at the wall surface to nearly the
free-stream value across the relatively thin boundary layer, while the variation of u with x along the flow is
typically small. Therefore, ∂u / ∂y >> ∂u / ∂x . Similarly, if the fluid and the wall are at different
temperatures and the fluid is heated or cooled during flow, heat conduction will occur primarily in the
direction normal to the surface, and thus ∂T / ∂y >> ∂T / ∂x . That is, the velocity and temperature
gradients normal to the surface are much greater than those along the surface. These simplifications are
known as the boundary layer approximations.

6-30C For flows with low velocity and for fluids with low viscosity the viscous dissipation term in the
energy equation is likely to be negligible.

6-31C For steady two-dimensional flow over an isothermal flat plate in the x-direction, the boundary
conditions for the velocity components u and v, and the temperature T at the plate surface and at the edge
of the boundary layer are expressed as follows:
u∞, T∞ T∞
At y = 0: u(x, 0) = 0, v(x, 0) = 0, T(x, 0) = Ts
y
As y → ∞ : u(x, ∞) = u∞, T(x, ∞) = T∞

6-32C An independent variable that makes it possible to transforming a setx of partial differential equations
into a single ordinary differential equation is called a similarity variable. A similarity solution is likely to
exist for a set of partial differential equations if there is a function that remains unchanged (such as the
non-dimensional velocity profile on a flat plate).

6-33C During steady, laminar, two-dimensional flow over an isothermal plate, the thickness of the
velocity boundary layer (a) increase with distance from the leading edge, (b) decrease with free-stream
velocity, and (c) and increase with kinematic viscosity

6-34C During steady, laminar, two-dimensional flow over an isothermal plate, the wall shear stress
decreases with distance from the leading edge

6-5
 Chapter 6 Fundamentals of Convection

6-35C A major advantage of nondimensionalizing the convection equations is the significant reduction in
the number of parameters [the original problem involves 6 parameters (L, V , T∞, Ts, ν, α), but the
nondimensionalized problem involves just 2 parameters (ReL and Pr)]. Nondimensionalization also results
in similarity parameters (such as Reynolds and Prandtl numbers) that enable us to group the results of a
large number of experiments and to report them conveniently in terms of such parameters.

6-36C For steady, laminar, two-dimensional, incompressible flow with constant properties and a Prandtl
number of unity and a given geometry, yes, it is correct to say that both the average friction and heat
transfer coefficients depend on the Reynolds number only since C f = f 4 (Re L ) and Nu = g 3 (Re L , Pr)
from non-dimensionalized momentum and energy equations.

6-6
 Chapter 6 Fundamentals of Convection

6-37 Parallel flow of oil between two plates is considered. The velocity and temperature distributions, the
maximum temperature, and the heat flux are to be determined.
Assumptions 1 Steady operating conditions exist. 2 Oil is an incompressible substance with constant
properties. 3 Body forces such as gravity are negligible. 4 The plates are large so that there is no variation
in z direction.
Properties The properties of oil at the average temperature of (40+15)/2 = 27.5°C are (Table A-13):
k = 0.145 W/m-K and µ = 0.580 kg/m-s = 0.580 N-s/m2
Analysis (a) We take the x-axis to be the flow direction, and y to be the normal direction. This is parallel
flow between two plates, and thus v = 0. Then the continuity equation reduces to
∂u ∂v ∂u
Continuity: + = 0 → =0 → u = u(y)
∂x ∂y ∂x
12 m/s
Therefore, the x-component of velocity does not change in the
flow direction (i.e., the velocity profile remains unchanged).
Noting that u = u(y), v = 0, and ∂P / ∂x = 0 (flow is maintained T2=40°C
by the motion of the upper plate rather than the pressure
gradient), the x-momentum equation (Eq. 6-28) reduces to L=0.7 mm Oil
 ∂u ∂u  ∂ 2 u ∂P d 2u
x-momentum: ρ u + v  = µ 2 − → 2 = 0 T1=25°C
 ∂x ∂y  ∂y ∂x dy
This is a second-order ordinary differential equation, and integrating it twice gives
u ( y ) = C1 y + C 2

The fluid velocities at the plate surfaces must be equal to the velocities of the plates because of the no-slip
condition. Therefore, the boundary conditions are u(0) = 0 and u(L) = V , and applying them gives the
velocity distribution to be
y
u( y) = V
L
Frictional heating due to viscous dissipation in this case is significant because of the high
viscosity of oil and the large plate velocity. The plates are isothermal and there is no change in the flow
direction, and thus the temperature depends on y only, T = T(y). Also, u = u(y) and v = 0. Then the energy
equation with dissipation (Eqs. 6-36 and 6-37) reduce to
2 2
∂ 2T  ∂u  d 2T V
Energy: 0=k + µ   → k = −µ 
 ∂y 
2 2
∂y dy L
since ∂u / ∂y = V / L . Dividing both sides by k and integrating twice give
2
µ y 
T ( y) = −  V + C3 y + C 4
2k  L 

Applying the boundary conditions T(0) = T1 and T(L) = T2 gives the temperature distribution to be

T2 − T1 µV 2  y y2 
T ( y) = y + T1 +  − 
L 2k  L L2 
 
(b) The temperature gradient is determined by differentiating T(y) with respect to y,

dT T2 − T1 µV 2  y
= + 1 − 2 
dy L 2kL  L

The location of maximum temperature is determined by setting dT/dy = 0 and solving for y,
dT T2 − T1 µV 2  y  T −T 1 
= + 1 − 2  = 0 → y = L k 2 2 1 + 
dy L 2kL  L  µV 2 


6-7
 Chapter 6 Fundamentals of Convection

The maximum temperature is the value of temperature at this y, whose numeric value is
 T −T 1   ( 40 − 15)°C 1
y = L k 2 2 1 +  = (0.0007 m) (0.145 W/m.°C) + 
 µV 
2 
2
(0.580 N.s/m )(12 m/s) 2 2 

= 0.0003804 m = 0.3804 mm
Then

T2 − T1 µV 2  y y2 
Tmax = T (0.0003804 ) = y + T1 +  − 
L 2k  L L2 
 
(40 − 15)°C (0.58 N ⋅ s/m 2 )(12 m/s) 2  0.0003804 m (0.0003804 m) 2 
= (0.0003804 m) + 15°C +  
0.0007 m 2(0.145 W/m ⋅ °C)  0.0007 m − (0.0007 m) 2 
 
= 100.0°C
(c) Heat flux at the plates is determined from the definition of heat flux,

dT T2 − T1 µV 2 T − T µV 2
q& 0 = − k = −k −k (1 − 0) = −k 2 1 −
dy y =0
L 2kL L 2L
(40 − 15)°C (0.58 N ⋅ s/m 2 )(12 m/s) 2  1 W  4 2
= −(0.145 W/m.°C) −   = −6.48 × 10 W/m
0.0007 m 2(0.0007 m )  1 N ⋅ m/s 

dT T2 − T1 µV 2 T −T µV 2
q& L = − k = −k −k (1 − 2) = −k 2 1 +
dy y=L
L 2kL L 2L
(40 − 15)°C (0.58 N ⋅ s/m 2 )(12 m/s) 2  1 W  4 2
= −(0.145 W/m.°C) +   = 5.45 × 10 W/m
0.0007 m 2(0.0007 m )  1 N ⋅ m/s 

Discussion A temperature rise of about 72.5°C confirms our suspicion that viscous dissipation is very
significant. Calculations are done using oil properties at 27.5°C, but the oil temperature turned out to be
much higher. Therefore, knowing the strong dependence of viscosity on temperature, calculations should
be repeated using properties at the average temperature of about 64°C to improve accuracy.

6-8
 Chapter 6 Fundamentals of Convection

6-38 Parallel flow of oil between two plates is considered. The velocity and temperature distributions, the
maximum temperature, and the heat flux are to be determined.
Assumptions 1 Steady operating conditions exist. 2 Oil is an incompressible substance with constant
properties. 3 Body forces such as gravity are negligible. 4 The plates are large so that there is no variation
in z direction.
Properties The properties of oil at the average temperature of (40+15)/2 = 27.5°C are (Table A-13):
k = 0.145 W/m-K and µ = 0.580 kg/m-s = 0.580 N-s/m2
Analysis (a) We take the x-axis to be the flow direction, and y to be the normal direction. This is parallel
flow between two plates, and thus v = 0. Then the continuity equation (Eq. 6-21) reduces to
∂u ∂v ∂u
Continuity: + = 0 → =0 → u = u(y)
∂x ∂y ∂x
12 m/s
Therefore, the x-component of velocity does not change in the
flow direction (i.e., the velocity profile remains unchanged).
Noting that u = u(y), v = 0, and ∂P / ∂x = 0 (flow is maintained T2=40°C
by the motion of the upper plate rather than the pressure
gradient), the x-momentum equation reduces to L=0.4 mm Oil
 ∂u ∂u  ∂ 2 u ∂P d 2u
x-momentum: ρ u + v  = µ 2 − → 2 = 0 T1=25°C
 ∂x ∂y  ∂y ∂x dy
This is a second-order ordinary differential equation, and integrating it twice gives
u ( y ) = C1 y + C 2

The fluid velocities at the plate surfaces must be equal to the velocities of the plates because of the no-slip
condition. Therefore, the boundary conditions are u(0) = 0 and u(L) = V , and applying them gives the
velocity distribution to be
y
u( y) = V
L
Frictional heating due to viscous dissipation in this case is significant because of the high
viscosity of oil and the large plate velocity. The plates are isothermal and there is no change in the flow
direction, and thus the temperature depends on y only, T = T(y). Also, u = u(y) and v = 0. Then the energy
equation with dissipation reduces to
2 2
∂ 2T  ∂u  d 2T V
Energy: 0=k + µ   → k = −µ 
 ∂y 
2 2
∂y dy L
since ∂u / ∂y = V / L . Dividing both sides by k and integrating twice give
2
µ y 
T ( y) = −  V + C3 y + C 4
2k  L 

Applying the boundary conditions T(0) = T1 and T(L) = T2 gives the temperature distribution to be

T2 − T1 µV 2  y y2 
T ( y) = y + T1 +  − 
L 2k  L L2 
 
(b) The temperature gradient is determined by differentiating T(y) with respect to y,

dT T2 − T1 µV 2  y
= + 1 − 2 
dy L 2kL  L

The location of maximum temperature is determined by setting dT/dy = 0 and solving for y,
dT T2 − T1 µV 2  y  T −T 1 
= + 1 − 2  = 0 → y = L k 2 2 1 + 
dy L 2kL  L  µV 2 


6-9
 Chapter 6 Fundamentals of Convection

The maximum temperature is the value of temperature at this y, whose numeric value is
 T −T 1   (40 − 15)°C 1
y = L k 2 2 1 +  = (0.0004 m) (0.145 W/m.°C) + 
 µV 
2 
2
(0.580 N.s/m )(12 m/s) 2 2 

= 0.0002174 m = 0.2174 mm
Then

T2 − T1 µV 2  y y2 
Tmax = T (0.0002174 ) = y + T1 +  − 
L 2k  L L2 
 
(40 − 15)°C (0.58 N ⋅ s/m 2 )(12 m/s) 2  0.0002174 m (0.0002174 m) 2 
= (0.0002174 m) + 15°C +  
0.0004 m 2(0.145 W/m ⋅ °C)  0.0004 m − (0.0004 m) 2 
 
= 100.0°C
(c) Heat flux at the plates is determined from the definition of heat flux,

dT T2 − T1 µV 2 T − T µV 2
q& 0 = − k = −k −k (1 − 0) = −k 2 1 −
dy y =0
L 2kL L 2L
(40 − 15)°C (0.58 N ⋅ s/m 2 )(12 m/s) 2  1 W  5 2
= −(0.145 W/m.°C) −   = −1.135 × 10 W/m
0.0004 m 2(0.0004 m )  1 N ⋅ m/s 

dT T2 − T1 µV 2 T −T µV 2
q& L = − k = −k −k (1 − 2) = −k 2 1 +
dy y=L
L 2kL L 2L
(40 − 15)°C (0.58 N ⋅ s/m 2 )(12 m/s) 2  1 W  4 2
= −(0.145 W/m.°C) +   = 9.53 × 10 W/m
0.0004 m 2(0.0004 m )  1 N ⋅ m/s 

Discussion A temperature rise of about 72.5°C confirms our suspicion that viscous dissipation is very
significant. Calculations are done using oil properties at 27.5°C, but the oil temperature turned out to be
much higher. Therefore, knowing the strong dependence of viscosity on temperature, calculations should
be repeated using properties at the average temperature of about 64°C to improve accuracy.

6-10