Anda di halaman 1dari 14

Introduction to Convection:

Flow and Thermal Considerations

Chapter Six: Sections 6.1.1 through 6.1.3, 6.2.1,6.2.3 and 6.3 – 6.6

Convection Heat Transfer

Convection

-heat transfer through a fluid in the presence of bulk fluid motion.

Forced- fluid moved by fan or pump.

Natural (free)- fluid moved by natural means, e.g., buoyancy effect

- can be classified as internal or external

(we will discuss natural convection later)

Convection Heat Transfer

Forced convection is achieved by subjecting the fluid to a pressure gradient (e.g., by a fan or pump), thereby forcing motion to occur according to the laws of fluid mechanics.

Convective heat transfer rate is calculated from Neuton’s Law of Cooling

h is called the convective heat transfer coefficient and has units of W/m 2 K or (often W/m 2 o C) or Btu/(hr ft 2 o F).

Convection Heat Transfer

Typical values of h are

Natural convection of air Natural conv. of water around a pipe Forced conv. of air over plate@30 m/s Water @ 2 m/s over plate, DT=15 K Liquid sodium at 5 m/s in 1.3cm pipe

= 5 W/m 2 K = 570 = 80 = 590 = 75,000 (@ 370 C)

Note: Phase change occurring at the surface during convection may also greatly affect the values of “h”, (e.g., boiling or condensing processes such as occur in a thermal power plant)

The heat transfer coefficient contains all the parameters which influence convection heat transfer.

It depends on the conditions in the boundary layer which are influenced by the surface geometry, the nature of the gas motion, and the properties of the fluid.

In fact, the question of convection heat transfer comes down to defining the right heat transfer coefficient.

This depends on the velocity and thermal Boundary Layers.

Velocity and Thermal Boundary Layers

Velocity and Thermal Boundary Layers

τ
=
s

μ

u

u
y

y = 0

Boundary Layer Features

Boundary Layers: Physical Features

• Velocity Boundary Layer

A consequence of viscous effects associated with relative motion between a fluid and a surface.

A region of the flow characterized by shear stresses and velocity gradients.

A region between the surface and the free stream whose

thickness

the flow direction.

δ

increases in

δ

u

()

y

u

=

0.99

Why does

the viscous effects penetrate further into the free stream along the plate and

δ increase in the flow direction?

δ increases

Manifested by a surface shear

stress

force,

that provides a drag .

τ

s

F

D

How does

τ

s

vary in the flow

direction? Why?

τ

s

F

D

u
= μ
y = 0
y
=
τ dA
s
s
A
s

Boundary Layer Features (cont.)

• Thermal Boundary Layer

A consequence of heat transfer between the surface and fluid.

A region of the flow characterized by temperature gradients and heat fluxes.

A region between the surface and the free stream whose thickness increases in the flow direction.

Why does

δ

t

δ

t

increase in the

flow direction?

Manifested by a surface heat

flux

transfer coefficient h .

q

′′

s

and a convection heat

If

(

T

s

T

)

is constant, how do

q′′

s

and

h vary in the flow direction?

δ

t

T

s

()

Ty

=

0.99

T
− T
s
T
q
′′ = −
k
s
f
y = 0
y
−∂ ∂
kT
/
y
f
y = 0
h
T
− T
s

- The temperature gradient at the wall, h and

q ′′

s

decrease with increasing x.

Local and Average Coefficients

Distinction between Local and Average Heat Transfer Coefficients

• Local Heat Flux and Coefficient:

q

′′ =

s

(

hT

s

T

)

• Average Heat Flux and Coefficient for a Uniform Surface Temperature:

q
=
hA
(
T
)
T ∞
s
s
q
= q ′′ dA
= (
T
− T
)
A
s
s
∫ A
s
s
1
h
=
hdA
A
s
A
s
s

hdA

s

• For a flat plate in parallel flow:

1
h
=

L

L hdx

o

Transition

Boundary Layer Transition

• How would you characterize conditions in the laminar region of boundary layer development? In the turbulent region?

• What conditions are associated with transition from laminar to turbulent flow?

• Why is the Reynolds number an appropriate parameter for quantifying transition from laminar to turbulent flow?

Transition criterion for a flat plate in parallel flow:

x

c

ρ u

Re

,

x c

x

c

μ

critical Reynolds number

location at which transition to turbulence begins

10

5

< Re

~

x c

,

< 3 x 10

~

6

Transition

Boundary Layer Transition

Transition (cont.)

What may be said about transition if Re L < Re x,c ?

If Re L > Re x,c ?

• Effect of transition on boundary layer thickness and local convection coefficient:

Why does transition provide a significant increase in the boundary layer thickness?

Why does the convection coefficient decay in the laminar region? Why does it increase significantly with transition to turbulence, despite the increase in the boundary layer thickness? Why does the convection coefficient decay in the turbulent region?

Boundary Layer Equations The Boundary Layer Equations

• Consider concurrent velocity and thermal boundary layer development for steady,

two-dimensional, incompressible flow with constant fluid properties negligible body forces.

(

μ c

,

p ,

k

)

and

• Apply conservation of mass, Newton’s 2 nd Law of Motion and conservation of energy to a differential control volume and invoke the boundary layer approximations.

Velocity Boundary Layer:

2

u

∂∂

u

p

2

dp

x

2

∂∂

y

x

2

,

dx

Thermal Boundary Layer:

2

T

x

2

2

T

y

2

Boundary Layer Equations (cont.)

• Conservation of Mass:

u

x

+

v

y

=

0

In the context of flow through a differential control volume, what is the physical

significance of the foregoing terms, if each is multiplied by the mass density of the fluid?

• Newton’s Second Law of Motion:

x-direction :

u

u

x

u

∂∂

+

∂∂ y

v

=−

1 dp

ρ

dx

+

ν

2

u

y

2

What is the physical significance of each term in the foregoing equation?

LHS represents the rate at which x-momentum leaves the CV due to fluid motion across the boundaries. 1 st term on RHS represents the net pressure force. 2 nd term on RHS represents the net force due to viscous shear stresses.

Why can we express the pressure gradient as dp /dx instead of p/x?

The BL is assumed to be thin and incompressible so pressure in free stream dominates

Boundary Layer Equations (cont.)

• Conservation of Energy:

∂∂ TT

u

+= v ∂∂ x y

α

2

T

⎛∂ u ⎞ ⎟

y

+

ν

2

p

yc

2

What is the physical significance of each term in the foregoing equation?

Terms on LHS represent the net rate that thermal energy leaves the CV due to the bulk fluid motion (advection). 1 st term on RHS accounts for the net inflow of thermal energy due to conduction into the CV in the Y direction.

What is the second term on the right-hand side called and under what conditions may it be neglected?

2 nd term on RHS represents energy addition due to viscous dissipation. It may be neglected in most cases, however not in high speed flows, e.g., supersonic flows.

Similarity Considerations

Boundary Layer Similarity

• As applied to the boundary layers, the principle of similarity is based on determining similarity parameters that facilitate application of results obtained for a surface experiencing one set of conditions to geometrically similar surfaces experiencing different conditions. (Recall how introduction of the similarity parameters Bi and Fo permitted generalization of results for transient, one- dimensional condition).

Dependent boundary layer variables of interest are:

τ

s

and

q

′′

or h

• For a prescribed geometry, the corresponding independent variables are:

Geometrical: Size (L), Location (x,y) Hydrodynamic: Velocity (V) Fluid Properties:

 Hydrodynamic: ρμ , Hence, Thermal : c p , k u = f ( xyLV ,,, ρμ ,, )

τ

s

= f

(

x LV

,,,,

ρμ

)

Similarity Considerations (cont.)

and

T = f

h =

f

(

(

x

y LV

,,,,,,

ρμ

x LV

,,,,,

ρμ

c

p

,

c

p

k

,

)

k

)

• Key similarity parameters may be inferred by non-dimensionalizing the momentum and energy equations.

• Recast the boundary layer equations by introducing dimensionless forms of the independent and dependent variables.

*

x

*

u

x

L

u

V

*

y

*

v

y

L

v

V

T

*

T

T

s

T

T

s

• Neglecting viscous dissipation, the following normalized forms of the x-momentum and energy equations are obtained:

u

2*

u

∂∂ u

+

∂∂ x y

*

*

dp

*

1

*

u

*

v

*

*

=−

dx

*

1

+

Re

L

2*

T

y

*2

*

*

∂∂ TT

*

*

u

*

+

v

∂∂ xy

*

Re

L

Pr

y

*2

=

Similarity Considerations (cont.)

We see there are two important dimensionless similarity parameters associated with the heat transfer in this case. These are the Reynolds and Prandtl numbers.

The Reynolds number represents the ratio of the inertia to the viscous forces

The Prandtl number is the ratio of the momentum and thermal diffusivities. These are calculated as

Re

Pr

L

VL

ρ

μ

=→ VL

v

the Reynolds Numbe r

c

p

μ

k

v

=→

α

the Prandtl Number

The dimensionless parameters allow us to apply results obtained for a surface experiencing convective heat transfer conditions, to geometrically similar surfaces experiencing different conditions, e.g., those with different fluids and fluid velocities, or differences in size as described by the characteristic length.

Therefore, as long as the similarity parameters and dimensionless boundary conditions are the same, the solution of the differential equations for non-dimension velocity and temperature will be the same

Table 6.1 in the text summarises the boundary layer equations and their assumed boundary conditions.

We see that the dependence of these equations can be simplified by grouping variables into non-dimensional parameters, e.g., Reynolds and Prandtl number.

• thus for a prescribed geometry, the dimensionless velocity has the following dependencies

*

u

= fxy

,

(

**

,Re L

)

and the shear stress that the surface can be expressed as

*
u
μ
Vu ⎞∂
τ
=
μ
s
*
y
= ⎜ ⎝
Ly ⎠ ⎟
*
y = 0
y
= 0

Similarity Considerations (cont.)

The local friction coefficient, is then

C

f

*
2
∂ u
τ s
=
2
*
ρ
V
/ 2
Re
∂ y
L

*

y

= 0

What is the functional dependence of the average friction coefficient?

Since

*
u
*
y

*

y

=

0

= f

(

x

* ,Re

L )

Then for a prescribed geometry, the friction coefficient may be expressed exclusively in terms of a dimensionless space coordinate and Reynolds number. This result should be universally applicable

2
C
=
fx
(
* ,Re
)
f
L
Re
L

Similarity Considerations (cont.)

• thus, for a prescribed geometry,

*
T
= fxy
(
**
,
,Re ,Pr
)
L
−∂ ∂
kT
/
y
k
( T
)
*
k
*
f
T
∂ T
T
yf = 0
s
f
h =
=−
=+
*
T
T
LT
(
−∂
T
)
*
y
L
y
s
s
*
y
= 0
The dimensionless local convection coefficient is then
*
hL
∂ T
Nu ≡=
* = f
(
*
x
,Re ,Pr
)
L
k
∂ y
f
*
y
= 0

Nu local Nusselt number

*

y

= 0

What is the functional dependence of the average Nusselt number?

How does the Nusselt number differ from the Biot number?

Problem 6.19:

Determination of heat transfer rate for prescribed turbine blade operating conditions from wind tunnel data obtained for a geometrically similar but smaller blade. The blade surface area may be assumed to be directly proportional to its characteristic length( A L

s

)

.

ASSUMPTIONS: (1) Steady-state conditions, (2) Constant properties, (3) Surface area A is directly proportional to characteristic length L, (4) Negligible radiation, (5) Blade shapes are geometrically similar.

ANALYSIS: For a prescribed geometry,

=

k

f

(

Re

L

,Pr

)

.

The Reynolds numbers for the blades are

Re

L,1

()

==V L /ν

1

1

1

15m

2

/ s ν

1

Hence, with constant properties (

Therefore,

v

1

= v

2

) ,

Nu
= Nu
2
1
(
h
L
/k
)
=
(
h L
/k
)
2
2
2
11
1
LL
q
11
1
h
=
h
=
2
1
L
L
AT
(
)
2
2
1
s,1
T ∞

Re

L,2

Re

L,1

2
/νν= 15m
)
/ s
.
2
2
2
.
Also, Pr
= Pr
.
1
2

=

(

V

2

L

= Re

L,2

The heat rate for the second blade is then

q

2

=

hA

2

2

(

T

s,2

q 2

q

=

T s,2

T s,1

T

T

2 = 2066 W.

 ) ( L 1 A 2 ( T s,2 − T ∞ ) L 400 2 − A 1 35 ) ( ( T s,1 − 1500 W T ∞ ) ) ( 300 − 35 )

−= T

q

1

=

q

1

COMMENTS: (i) The variation in ν from Case 1 to Case 2 would cause Re L,2 to differ from Re L,1 . However, for air and the prescribed temperatures, this non-constant property effect is

small. (ii) If the Reynolds numbers were not equal

(

Re

L

,1

Re

L

2

)

,

knowledge of the specific form of

f

(

Re

L

Pr

,

)

would be needed to determine h 2.

Problem: Nusselt Number

Problem 6.26:

Use of a local Nusselt number correlation to estimate the surface temperature of a chip on a circuit board.

KNOWN: Expression for the local heat transfer coefficient of air at prescribed velocity and temperature flowing over electronic elements on a circuit board and heat dissipation rate for a 4 × 4 mm chip located 120mm from the leading edge.

FIND: Surface temperature of the chip surface, T s .

Problem: Nusselt Number (cont.)

ASSUMPTIONS: (1) Steady-state conditions, (2) Power dissipated within chip is lost by convection across the upper surface only, (3) Chip surface is isothermal, (4) The average heat transfer coefficient for the chip surface is equivalent to the local value at x = L, (5) Negligible radiation.

PROPERTIES: Table A-4, Air (Evaluate properties at the average temperature of air in the boundary layer. Assuming T s = 45°C, T ave = (45 + 25)/2 = 35°C = 308K. Also, p = 1atm): ν = 16.69 ×

10 -6 m 2 /s, k = 26.9 × 10 -3 W/mK, Pr = 0.703.

ANALYSIS: From an energy balance on the chip,

q

conv

= E

g

= 30mW.

Newton’s law of cooling for the upper chip surface can be written as

chip

where

T

s

=

A chip

T

=

+

2

.

q

/ h A

conv

(2)

Assuming that the average heat transfer coefficient ()

coefficient evaluated at x = L, that is,

applying the prescribed correlation at x = L.

h

over the chip surface is equivalent to the local

)

, the local coefficient can be evaluated by

h

chip

(

hL

x

Nu

x

=

h

x

x

k

=

0.04

⎣ ⎢

Vx

ν

⎦ ⎥

0.85

Pr

1/3

h

L

=

0.04

k

⎢ ⎣

VL

L

ν

⎥ ⎦

0.85

Pr

1/3

Problem: Nusselt Number (cont.)

h

L

=

0.04

 0.0269 W/m K ⋅ ⎤ ⎡ 10 m/s × 0.120 m ⎤ ⎡ ⎢ ⎣ 0.120 m ⎦ ⎥ ⎢ ⎣ 16.69 × 10 -6 m 2 ⎥ / s ⎦

0.85

(

0.703

)

1/3

107 W/m

=⋅

2

K.

From Eq. (2), the surface temperature of the chip is

T

s

= 25 C + 30×10

-3

W/107 W/m

2

⋅×K

(

0.004m

)

2

= 42.5 C.

COMMENTS: (1) The estimated value of T ave used to evaluate the air properties is reasonable.

(2) How else could

h

chip

have been evaluated? Is the assumption of

h = h

L

reasonable?